diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,65144 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9999462510077936,
+  "eval_steps": 500,
+  "global_step": 9302,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.52533925718378,
+      "learning_rate": 3.5714285714285714e-06,
+      "loss": 2.1829,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8917807391691538,
+      "learning_rate": 7.142857142857143e-06,
+      "loss": 1.9838,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.0748449554888,
+      "learning_rate": 1.0714285714285714e-05,
+      "loss": 1.9405,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6850439438507572,
+      "learning_rate": 1.4285714285714285e-05,
+      "loss": 2.0572,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.233502306710884,
+      "learning_rate": 1.7857142857142855e-05,
+      "loss": 2.017,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.1190437540863045,
+      "learning_rate": 2.1428571428571428e-05,
+      "loss": 2.1505,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 78.75075566973746,
+      "learning_rate": 2.5e-05,
+      "loss": 1.9595,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.491814243878795,
+      "learning_rate": 2.857142857142857e-05,
+      "loss": 1.9188,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.0583872857696868,
+      "learning_rate": 3.214285714285714e-05,
+      "loss": 2.2103,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.885577656564717,
+      "learning_rate": 3.571428571428571e-05,
+      "loss": 2.0415,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.4388687007921046,
+      "learning_rate": 3.928571428571428e-05,
+      "loss": 2.0554,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5874308565723774,
+      "learning_rate": 4.2857142857142856e-05,
+      "loss": 1.9838,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7501093864144527,
+      "learning_rate": 4.642857142857143e-05,
+      "loss": 2.103,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7408455109981018,
+      "learning_rate": 5e-05,
+      "loss": 1.8546,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.0611197819252878,
+      "learning_rate": 5.357142857142857e-05,
+      "loss": 1.9603,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9561526566868136,
+      "learning_rate": 5.714285714285714e-05,
+      "loss": 1.9864,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8141552130974743,
+      "learning_rate": 6.0714285714285715e-05,
+      "loss": 1.8761,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9098149261285031,
+      "learning_rate": 6.428571428571427e-05,
+      "loss": 1.767,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8866212598000296,
+      "learning_rate": 6.785714285714285e-05,
+      "loss": 2.1052,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.1741636005388276,
+      "learning_rate": 7.142857142857142e-05,
+      "loss": 1.9065,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3333604730387096,
+      "learning_rate": 7.5e-05,
+      "loss": 2.1443,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3112663647673257,
+      "learning_rate": 7.857142857142857e-05,
+      "loss": 1.9296,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4304082960837077,
+      "learning_rate": 8.214285714285714e-05,
+      "loss": 1.8629,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5095743351765647,
+      "learning_rate": 8.571428571428571e-05,
+      "loss": 2.096,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.5458230630716185,
+      "learning_rate": 8.928571428571429e-05,
+      "loss": 2.0096,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.761767871209414,
+      "learning_rate": 9.285714285714286e-05,
+      "loss": 1.9721,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.81257268388485,
+      "learning_rate": 9.642857142857143e-05,
+      "loss": 2.0577,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.486947255640102,
+      "learning_rate": 0.0001,
+      "loss": 1.916,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.0720761542391175,
+      "learning_rate": 0.00010357142857142858,
+      "loss": 1.9207,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6505238857303257,
+      "learning_rate": 0.00010714285714285714,
+      "loss": 2.0371,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.291313495395931,
+      "learning_rate": 0.00011071428571428571,
+      "loss": 2.0557,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7313893024574583,
+      "learning_rate": 0.00011428571428571428,
+      "loss": 1.9889,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5935351470037225,
+      "learning_rate": 0.00011785714285714286,
+      "loss": 1.9542,
+      "step": 33
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.066218796348364,
+      "learning_rate": 0.00012142857142857143,
+      "loss": 1.9928,
+      "step": 34
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3713446928171047,
+      "learning_rate": 0.000125,
+      "loss": 1.9124,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3523693102135053,
+      "learning_rate": 0.00012857142857142855,
+      "loss": 2.0267,
+      "step": 36
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9515514517206934,
+      "learning_rate": 0.00013214285714285715,
+      "loss": 1.8685,
+      "step": 37
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9729557725179657,
+      "learning_rate": 0.0001357142857142857,
+      "loss": 1.7549,
+      "step": 38
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.564147690807114,
+      "learning_rate": 0.0001392857142857143,
+      "loss": 1.9658,
+      "step": 39
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.002970099447507,
+      "learning_rate": 0.00014285714285714284,
+      "loss": 2.0966,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.0123033933136292,
+      "learning_rate": 0.00014642857142857144,
+      "loss": 1.9739,
+      "step": 41
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7633814778573,
+      "learning_rate": 0.00015,
+      "loss": 1.847,
+      "step": 42
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.1686064851334461,
+      "learning_rate": 0.0001535714285714286,
+      "loss": 1.9012,
+      "step": 43
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9361446756319182,
+      "learning_rate": 0.00015714285714285713,
+      "loss": 1.9118,
+      "step": 44
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5200884270670607,
+      "learning_rate": 0.00016071428571428573,
+      "loss": 1.9476,
+      "step": 45
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.450100365815453,
+      "learning_rate": 0.00016428571428571428,
+      "loss": 1.7931,
+      "step": 46
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.288950921055788,
+      "learning_rate": 0.00016785714285714285,
+      "loss": 1.9861,
+      "step": 47
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.803610469800681,
+      "learning_rate": 0.00017142857142857143,
+      "loss": 1.9061,
+      "step": 48
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4356333181084613,
+      "learning_rate": 0.000175,
+      "loss": 1.8542,
+      "step": 49
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2358575078876701,
+      "learning_rate": 0.00017857142857142857,
+      "loss": 1.9251,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0864304720316573,
+      "learning_rate": 0.00018214285714285714,
+      "loss": 1.8825,
+      "step": 51
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1309351656853563,
+      "learning_rate": 0.00018571428571428572,
+      "loss": 1.846,
+      "step": 52
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9981874490118918,
+      "learning_rate": 0.0001892857142857143,
+      "loss": 1.7579,
+      "step": 53
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.065892064141929,
+      "learning_rate": 0.00019285714285714286,
+      "loss": 1.8208,
+      "step": 54
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8816359706176079,
+      "learning_rate": 0.00019642857142857144,
+      "loss": 1.7919,
+      "step": 55
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0578385928836473,
+      "learning_rate": 0.0002,
+      "loss": 1.8116,
+      "step": 56
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9421173828756452,
+      "learning_rate": 0.00020357142857142858,
+      "loss": 1.7467,
+      "step": 57
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6289823592081432,
+      "learning_rate": 0.00020714285714285716,
+      "loss": 1.655,
+      "step": 58
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4741147658119709,
+      "learning_rate": 0.00021071428571428573,
+      "loss": 1.7672,
+      "step": 59
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.57246021644073,
+      "learning_rate": 0.00021428571428571427,
+      "loss": 1.6553,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1375051314898432,
+      "learning_rate": 0.00021785714285714287,
+      "loss": 1.7275,
+      "step": 61
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3522322434105656,
+      "learning_rate": 0.00022142857142857142,
+      "loss": 1.6659,
+      "step": 62
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3498064220242347,
+      "learning_rate": 0.00022500000000000002,
+      "loss": 1.7152,
+      "step": 63
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4251392957299767,
+      "learning_rate": 0.00022857142857142857,
+      "loss": 1.5624,
+      "step": 64
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.119320800018593,
+      "learning_rate": 0.00023214285714285717,
+      "loss": 1.667,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0918414494076893,
+      "learning_rate": 0.0002357142857142857,
+      "loss": 1.5963,
+      "step": 66
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9746103076509893,
+      "learning_rate": 0.0002392857142857143,
+      "loss": 1.504,
+      "step": 67
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9492418576791255,
+      "learning_rate": 0.00024285714285714286,
+      "loss": 1.6585,
+      "step": 68
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8174319326574438,
+      "learning_rate": 0.00024642857142857143,
+      "loss": 1.5783,
+      "step": 69
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.184127183820156,
+      "learning_rate": 0.00025,
+      "loss": 1.6589,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6662930221337083,
+      "learning_rate": 0.0002535714285714286,
+      "loss": 1.5221,
+      "step": 71
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2301561957715066,
+      "learning_rate": 0.0002571428571428571,
+      "loss": 1.7142,
+      "step": 72
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7534991533961916,
+      "learning_rate": 0.0002607142857142857,
+      "loss": 1.5588,
+      "step": 73
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5609113126608429,
+      "learning_rate": 0.0002642857142857143,
+      "loss": 1.5584,
+      "step": 74
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.958306688822885,
+      "learning_rate": 0.00026785714285714287,
+      "loss": 1.523,
+      "step": 75
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6114223761007516,
+      "learning_rate": 0.0002714285714285714,
+      "loss": 1.6552,
+      "step": 76
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6799693126275542,
+      "learning_rate": 0.000275,
+      "loss": 1.6153,
+      "step": 77
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.846263726121709,
+      "learning_rate": 0.0002785714285714286,
+      "loss": 1.5276,
+      "step": 78
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5330196945414835,
+      "learning_rate": 0.00028214285714285716,
+      "loss": 1.5136,
+      "step": 79
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6961419376772497,
+      "learning_rate": 0.0002857142857142857,
+      "loss": 1.4979,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.44617572572355957,
+      "learning_rate": 0.0002892857142857143,
+      "loss": 1.5388,
+      "step": 81
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5404931034315101,
+      "learning_rate": 0.0002928571428571429,
+      "loss": 1.53,
+      "step": 82
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.625345659039276,
+      "learning_rate": 0.00029642857142857145,
+      "loss": 1.5604,
+      "step": 83
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3973305228158473,
+      "learning_rate": 0.0003,
+      "loss": 1.5368,
+      "step": 84
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5150470529051457,
+      "learning_rate": 0.00030357142857142855,
+      "loss": 1.5111,
+      "step": 85
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5165971204137647,
+      "learning_rate": 0.0003071428571428572,
+      "loss": 1.6526,
+      "step": 86
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5061721865580466,
+      "learning_rate": 0.00031071428571428575,
+      "loss": 1.5447,
+      "step": 87
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7102067251701049,
+      "learning_rate": 0.00031428571428571427,
+      "loss": 1.4482,
+      "step": 88
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5244806059205487,
+      "learning_rate": 0.00031785714285714284,
+      "loss": 1.6764,
+      "step": 89
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5221727806541643,
+      "learning_rate": 0.00032142857142857147,
+      "loss": 1.6043,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7848075737761744,
+      "learning_rate": 0.00032500000000000004,
+      "loss": 1.5074,
+      "step": 91
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4290039738082572,
+      "learning_rate": 0.00032857142857142856,
+      "loss": 1.5482,
+      "step": 92
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5232736843606722,
+      "learning_rate": 0.00033214285714285713,
+      "loss": 1.5139,
+      "step": 93
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.44862425899364705,
+      "learning_rate": 0.0003357142857142857,
+      "loss": 1.5531,
+      "step": 94
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4858044328676179,
+      "learning_rate": 0.00033928571428571433,
+      "loss": 1.5697,
+      "step": 95
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.40783462592279174,
+      "learning_rate": 0.00034285714285714285,
+      "loss": 1.4921,
+      "step": 96
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4833210430418762,
+      "learning_rate": 0.0003464285714285714,
+      "loss": 1.6326,
+      "step": 97
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3349518955445,
+      "learning_rate": 0.00035,
+      "loss": 1.481,
+      "step": 98
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5119405076678242,
+      "learning_rate": 0.0003535714285714286,
+      "loss": 1.5565,
+      "step": 99
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4644397775386018,
+      "learning_rate": 0.00035714285714285714,
+      "loss": 1.5632,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3798960509359317,
+      "learning_rate": 0.0003607142857142857,
+      "loss": 1.6424,
+      "step": 101
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.38836469924030503,
+      "learning_rate": 0.0003642857142857143,
+      "loss": 1.5311,
+      "step": 102
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.40155819248605457,
+      "learning_rate": 0.0003678571428571429,
+      "loss": 1.4503,
+      "step": 103
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5577097012658978,
+      "learning_rate": 0.00037142857142857143,
+      "loss": 1.4728,
+      "step": 104
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3302891581452391,
+      "learning_rate": 0.000375,
+      "loss": 1.4451,
+      "step": 105
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5105057674282057,
+      "learning_rate": 0.0003785714285714286,
+      "loss": 1.4823,
+      "step": 106
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.34139026979805664,
+      "learning_rate": 0.0003821428571428571,
+      "loss": 1.5526,
+      "step": 107
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3526373961195756,
+      "learning_rate": 0.0003857142857142857,
+      "loss": 1.6479,
+      "step": 108
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.36739582888447553,
+      "learning_rate": 0.0003892857142857143,
+      "loss": 1.5424,
+      "step": 109
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.31509373847267547,
+      "learning_rate": 0.0003928571428571429,
+      "loss": 1.7238,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7305039688024733,
+      "learning_rate": 0.0003964285714285714,
+      "loss": 1.5271,
+      "step": 111
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3581584781956524,
+      "learning_rate": 0.0004,
+      "loss": 1.6029,
+      "step": 112
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3224198059244875,
+      "learning_rate": 0.0004035714285714286,
+      "loss": 1.5386,
+      "step": 113
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3471545607154616,
+      "learning_rate": 0.00040714285714285717,
+      "loss": 1.468,
+      "step": 114
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3904049444253878,
+      "learning_rate": 0.0004107142857142857,
+      "loss": 1.6968,
+      "step": 115
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3121284899151761,
+      "learning_rate": 0.0004142857142857143,
+      "loss": 1.7429,
+      "step": 116
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3244004142779458,
+      "learning_rate": 0.0004178571428571429,
+      "loss": 1.5028,
+      "step": 117
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5252291278727379,
+      "learning_rate": 0.00042142857142857146,
+      "loss": 1.4616,
+      "step": 118
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3378902942457399,
+      "learning_rate": 0.000425,
+      "loss": 1.4332,
+      "step": 119
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.28884442583413833,
+      "learning_rate": 0.00042857142857142855,
+      "loss": 1.3443,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.35609856553313685,
+      "learning_rate": 0.0004321428571428572,
+      "loss": 1.5219,
+      "step": 121
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.35554901725913257,
+      "learning_rate": 0.00043571428571428575,
+      "loss": 1.5723,
+      "step": 122
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3440599128217032,
+      "learning_rate": 0.00043928571428571427,
+      "loss": 1.3977,
+      "step": 123
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.31192531792308886,
+      "learning_rate": 0.00044285714285714284,
+      "loss": 1.5228,
+      "step": 124
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.32046127353392895,
+      "learning_rate": 0.00044642857142857147,
+      "loss": 1.4763,
+      "step": 125
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3101840268627974,
+      "learning_rate": 0.00045000000000000004,
+      "loss": 1.4582,
+      "step": 126
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4288094131576573,
+      "learning_rate": 0.00045357142857142856,
+      "loss": 1.5662,
+      "step": 127
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.44158522598033534,
+      "learning_rate": 0.00045714285714285713,
+      "loss": 1.4745,
+      "step": 128
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3239538132892933,
+      "learning_rate": 0.0004607142857142857,
+      "loss": 1.4899,
+      "step": 129
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2871588061072913,
+      "learning_rate": 0.00046428571428571433,
+      "loss": 1.6439,
+      "step": 130
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.38892543132373336,
+      "learning_rate": 0.00046785714285714285,
+      "loss": 1.4552,
+      "step": 131
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3825473937891277,
+      "learning_rate": 0.0004714285714285714,
+      "loss": 1.5597,
+      "step": 132
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4309140645349847,
+      "learning_rate": 0.000475,
+      "loss": 1.5602,
+      "step": 133
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4017136556479656,
+      "learning_rate": 0.0004785714285714286,
+      "loss": 1.5129,
+      "step": 134
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.28489478417342257,
+      "learning_rate": 0.00048214285714285715,
+      "loss": 1.4758,
+      "step": 135
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3099684276288773,
+      "learning_rate": 0.0004857142857142857,
+      "loss": 1.5706,
+      "step": 136
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.29120739116594524,
+      "learning_rate": 0.0004892857142857142,
+      "loss": 1.5954,
+      "step": 137
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.36657854067868006,
+      "learning_rate": 0.0004928571428571429,
+      "loss": 1.5739,
+      "step": 138
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4317725139749333,
+      "learning_rate": 0.0004964285714285715,
+      "loss": 1.4752,
+      "step": 139
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.31764086869505503,
+      "learning_rate": 0.0005,
+      "loss": 1.6824,
+      "step": 140
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3385632495847182,
+      "learning_rate": 0.0005035714285714285,
+      "loss": 1.5375,
+      "step": 141
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.37461770956336343,
+      "learning_rate": 0.0005071428571428572,
+      "loss": 1.412,
+      "step": 142
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3399190654827032,
+      "learning_rate": 0.0005107142857142857,
+      "loss": 1.4806,
+      "step": 143
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4216971640700005,
+      "learning_rate": 0.0005142857142857142,
+      "loss": 1.5229,
+      "step": 144
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3746707384688907,
+      "learning_rate": 0.0005178571428571429,
+      "loss": 1.5051,
+      "step": 145
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.360108642576056,
+      "learning_rate": 0.0005214285714285714,
+      "loss": 1.4608,
+      "step": 146
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3508119200146609,
+      "learning_rate": 0.0005250000000000001,
+      "loss": 1.5631,
+      "step": 147
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3979211809075355,
+      "learning_rate": 0.0005285714285714286,
+      "loss": 1.7342,
+      "step": 148
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2965752820354549,
+      "learning_rate": 0.0005321428571428571,
+      "loss": 1.4366,
+      "step": 149
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.36692441991683633,
+      "learning_rate": 0.0005357142857142857,
+      "loss": 1.6711,
+      "step": 150
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3433996604400371,
+      "learning_rate": 0.0005392857142857143,
+      "loss": 1.5016,
+      "step": 151
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3136388055796795,
+      "learning_rate": 0.0005428571428571428,
+      "loss": 1.4356,
+      "step": 152
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.30192687104297594,
+      "learning_rate": 0.0005464285714285714,
+      "loss": 1.4473,
+      "step": 153
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.33129198015257594,
+      "learning_rate": 0.00055,
+      "loss": 1.4901,
+      "step": 154
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3165111191353476,
+      "learning_rate": 0.0005535714285714287,
+      "loss": 1.4939,
+      "step": 155
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.29059641041159867,
+      "learning_rate": 0.0005571428571428572,
+      "loss": 1.5284,
+      "step": 156
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.30213428826667454,
+      "learning_rate": 0.0005607142857142857,
+      "loss": 1.6462,
+      "step": 157
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2845219847958041,
+      "learning_rate": 0.0005642857142857143,
+      "loss": 1.526,
+      "step": 158
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.35986971493563835,
+      "learning_rate": 0.0005678571428571428,
+      "loss": 1.4807,
+      "step": 159
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.43297526779692946,
+      "learning_rate": 0.0005714285714285714,
+      "loss": 1.4925,
+      "step": 160
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.309817795851931,
+      "learning_rate": 0.000575,
+      "loss": 1.4316,
+      "step": 161
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3532850398670643,
+      "learning_rate": 0.0005785714285714286,
+      "loss": 1.3771,
+      "step": 162
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2921544010366211,
+      "learning_rate": 0.0005821428571428572,
+      "loss": 1.3419,
+      "step": 163
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.29158256657962844,
+      "learning_rate": 0.0005857142857142858,
+      "loss": 1.3892,
+      "step": 164
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.26959751530851456,
+      "learning_rate": 0.0005892857142857143,
+      "loss": 1.5987,
+      "step": 165
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.27858915227839154,
+      "learning_rate": 0.0005928571428571429,
+      "loss": 1.6159,
+      "step": 166
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.27887679232979334,
+      "learning_rate": 0.0005964285714285714,
+      "loss": 1.6343,
+      "step": 167
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2902063856837706,
+      "learning_rate": 0.0006,
+      "loss": 1.5628,
+      "step": 168
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.26242952479313464,
+      "learning_rate": 0.0006035714285714286,
+      "loss": 1.6507,
+      "step": 169
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2593759252347874,
+      "learning_rate": 0.0006071428571428571,
+      "loss": 1.5475,
+      "step": 170
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.31011666182187436,
+      "learning_rate": 0.0006107142857142858,
+      "loss": 1.5178,
+      "step": 171
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.25639111105134527,
+      "learning_rate": 0.0006142857142857143,
+      "loss": 1.498,
+      "step": 172
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.31652842035082357,
+      "learning_rate": 0.0006178571428571429,
+      "loss": 1.4955,
+      "step": 173
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2987038846515858,
+      "learning_rate": 0.0006214285714285715,
+      "loss": 1.4423,
+      "step": 174
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3263219499540308,
+      "learning_rate": 0.000625,
+      "loss": 1.7098,
+      "step": 175
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.32379973578168375,
+      "learning_rate": 0.0006285714285714285,
+      "loss": 1.5546,
+      "step": 176
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.30673266201950766,
+      "learning_rate": 0.0006321428571428572,
+      "loss": 1.5595,
+      "step": 177
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.31025981831403515,
+      "learning_rate": 0.0006357142857142857,
+      "loss": 1.5682,
+      "step": 178
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2631500803146406,
+      "learning_rate": 0.0006392857142857142,
+      "loss": 1.4778,
+      "step": 179
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.263772189519905,
+      "learning_rate": 0.0006428571428571429,
+      "loss": 1.5322,
+      "step": 180
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3170171832408717,
+      "learning_rate": 0.0006464285714285715,
+      "loss": 1.5399,
+      "step": 181
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.25839825036928027,
+      "learning_rate": 0.0006500000000000001,
+      "loss": 1.5088,
+      "step": 182
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2693335388497259,
+      "learning_rate": 0.0006535714285714286,
+      "loss": 1.3628,
+      "step": 183
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3374590349944099,
+      "learning_rate": 0.0006571428571428571,
+      "loss": 1.5421,
+      "step": 184
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2880039486353611,
+      "learning_rate": 0.0006607142857142857,
+      "loss": 1.5958,
+      "step": 185
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2532226088788996,
+      "learning_rate": 0.0006642857142857143,
+      "loss": 1.544,
+      "step": 186
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.31236526445694107,
+      "learning_rate": 0.0006678571428571428,
+      "loss": 1.4988,
+      "step": 187
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.296432786908887,
+      "learning_rate": 0.0006714285714285714,
+      "loss": 1.5167,
+      "step": 188
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3319120080791588,
+      "learning_rate": 0.000675,
+      "loss": 1.4906,
+      "step": 189
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2671474694497889,
+      "learning_rate": 0.0006785714285714287,
+      "loss": 1.506,
+      "step": 190
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.29230866138530726,
+      "learning_rate": 0.0006821428571428572,
+      "loss": 1.5605,
+      "step": 191
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.29237517306757743,
+      "learning_rate": 0.0006857142857142857,
+      "loss": 1.5592,
+      "step": 192
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.23998407955562542,
+      "learning_rate": 0.0006892857142857143,
+      "loss": 1.5919,
+      "step": 193
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.33358955469999935,
+      "learning_rate": 0.0006928571428571428,
+      "loss": 1.5007,
+      "step": 194
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2588549835888095,
+      "learning_rate": 0.0006964285714285714,
+      "loss": 1.583,
+      "step": 195
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2985579696888121,
+      "learning_rate": 0.0007,
+      "loss": 1.5594,
+      "step": 196
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2837689469382843,
+      "learning_rate": 0.0007035714285714286,
+      "loss": 1.4313,
+      "step": 197
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.26100830899090055,
+      "learning_rate": 0.0007071428571428572,
+      "loss": 1.5072,
+      "step": 198
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2858482873084629,
+      "learning_rate": 0.0007107142857142858,
+      "loss": 1.6042,
+      "step": 199
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.28657146589590393,
+      "learning_rate": 0.0007142857142857143,
+      "loss": 1.5303,
+      "step": 200
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.29257530177935365,
+      "learning_rate": 0.0007178571428571429,
+      "loss": 1.6107,
+      "step": 201
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3042172997552271,
+      "learning_rate": 0.0007214285714285714,
+      "loss": 1.5416,
+      "step": 202
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.30426651346545824,
+      "learning_rate": 0.000725,
+      "loss": 1.5159,
+      "step": 203
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.26088662022664416,
+      "learning_rate": 0.0007285714285714286,
+      "loss": 1.511,
+      "step": 204
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.22126894553880855,
+      "learning_rate": 0.0007321428571428571,
+      "loss": 1.3603,
+      "step": 205
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.22402688666484616,
+      "learning_rate": 0.0007357142857142858,
+      "loss": 1.4555,
+      "step": 206
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.22146626840818323,
+      "learning_rate": 0.0007392857142857144,
+      "loss": 1.345,
+      "step": 207
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.234618942149512,
+      "learning_rate": 0.0007428571428571429,
+      "loss": 1.4908,
+      "step": 208
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.29634089484886406,
+      "learning_rate": 0.0007464285714285715,
+      "loss": 1.5035,
+      "step": 209
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.25852822888367205,
+      "learning_rate": 0.00075,
+      "loss": 1.6435,
+      "step": 210
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.292756900485729,
+      "learning_rate": 0.0007535714285714285,
+      "loss": 1.5641,
+      "step": 211
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2039367704094928,
+      "learning_rate": 0.0007571428571428572,
+      "loss": 1.4156,
+      "step": 212
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.25648074373805585,
+      "learning_rate": 0.0007607142857142857,
+      "loss": 1.4194,
+      "step": 213
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2552155993199289,
+      "learning_rate": 0.0007642857142857142,
+      "loss": 1.4379,
+      "step": 214
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2585414529627138,
+      "learning_rate": 0.0007678571428571429,
+      "loss": 1.5787,
+      "step": 215
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2546374137167054,
+      "learning_rate": 0.0007714285714285715,
+      "loss": 1.5555,
+      "step": 216
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.27081684679812273,
+      "learning_rate": 0.0007750000000000001,
+      "loss": 1.3571,
+      "step": 217
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.24325223266409077,
+      "learning_rate": 0.0007785714285714286,
+      "loss": 1.5655,
+      "step": 218
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.24276431673429932,
+      "learning_rate": 0.0007821428571428571,
+      "loss": 1.4558,
+      "step": 219
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2366434468516385,
+      "learning_rate": 0.0007857142857142857,
+      "loss": 1.5843,
+      "step": 220
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.23005762850876868,
+      "learning_rate": 0.0007892857142857143,
+      "loss": 1.4706,
+      "step": 221
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.21661794426819167,
+      "learning_rate": 0.0007928571428571428,
+      "loss": 1.5066,
+      "step": 222
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.24364550930674125,
+      "learning_rate": 0.0007964285714285714,
+      "loss": 1.4984,
+      "step": 223
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.24544373452185395,
+      "learning_rate": 0.0008,
+      "loss": 1.5392,
+      "step": 224
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3164181294977973,
+      "learning_rate": 0.0008035714285714287,
+      "loss": 1.4713,
+      "step": 225
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2057971791624359,
+      "learning_rate": 0.0008071428571428572,
+      "loss": 1.6402,
+      "step": 226
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.22646235028301068,
+      "learning_rate": 0.0008107142857142857,
+      "loss": 1.6311,
+      "step": 227
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2434939861749518,
+      "learning_rate": 0.0008142857142857143,
+      "loss": 1.4522,
+      "step": 228
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.25565301421024494,
+      "learning_rate": 0.0008178571428571428,
+      "loss": 1.4731,
+      "step": 229
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2145112325466064,
+      "learning_rate": 0.0008214285714285714,
+      "loss": 1.508,
+      "step": 230
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.24235525781497383,
+      "learning_rate": 0.000825,
+      "loss": 1.5515,
+      "step": 231
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.24808658822607518,
+      "learning_rate": 0.0008285714285714286,
+      "loss": 1.4308,
+      "step": 232
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.25257648565419344,
+      "learning_rate": 0.0008321428571428573,
+      "loss": 1.5598,
+      "step": 233
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2052328252860909,
+      "learning_rate": 0.0008357142857142858,
+      "loss": 1.4058,
+      "step": 234
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.30554429044632603,
+      "learning_rate": 0.0008392857142857143,
+      "loss": 1.4984,
+      "step": 235
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2684656793305302,
+      "learning_rate": 0.0008428571428571429,
+      "loss": 1.5596,
+      "step": 236
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.24919207229849352,
+      "learning_rate": 0.0008464285714285714,
+      "loss": 1.5213,
+      "step": 237
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.23262987526628107,
+      "learning_rate": 0.00085,
+      "loss": 1.6697,
+      "step": 238
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.206144476035226,
+      "learning_rate": 0.0008535714285714286,
+      "loss": 1.6066,
+      "step": 239
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1916608965478554,
+      "learning_rate": 0.0008571428571428571,
+      "loss": 1.5925,
+      "step": 240
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.18887206567344306,
+      "learning_rate": 0.0008607142857142858,
+      "loss": 1.5143,
+      "step": 241
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.18012370414006335,
+      "learning_rate": 0.0008642857142857144,
+      "loss": 1.5059,
+      "step": 242
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.23347982878855053,
+      "learning_rate": 0.0008678571428571429,
+      "loss": 1.5216,
+      "step": 243
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2113463845182561,
+      "learning_rate": 0.0008714285714285715,
+      "loss": 1.6146,
+      "step": 244
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2069102374474599,
+      "learning_rate": 0.000875,
+      "loss": 1.4846,
+      "step": 245
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.20672558766688756,
+      "learning_rate": 0.0008785714285714285,
+      "loss": 1.4937,
+      "step": 246
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.21223510647754198,
+      "learning_rate": 0.0008821428571428572,
+      "loss": 1.5495,
+      "step": 247
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2702576482389292,
+      "learning_rate": 0.0008857142857142857,
+      "loss": 1.4101,
+      "step": 248
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.22179795777881917,
+      "learning_rate": 0.0008892857142857142,
+      "loss": 1.677,
+      "step": 249
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2231004575446933,
+      "learning_rate": 0.0008928571428571429,
+      "loss": 1.6451,
+      "step": 250
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.19345695749788838,
+      "learning_rate": 0.0008964285714285715,
+      "loss": 1.4773,
+      "step": 251
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2164011557010416,
+      "learning_rate": 0.0009000000000000001,
+      "loss": 1.4006,
+      "step": 252
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.20521116028853958,
+      "learning_rate": 0.0009035714285714286,
+      "loss": 1.4699,
+      "step": 253
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.21510699790233587,
+      "learning_rate": 0.0009071428571428571,
+      "loss": 1.3917,
+      "step": 254
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.18317267392579845,
+      "learning_rate": 0.0009107142857142857,
+      "loss": 1.5035,
+      "step": 255
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2783009940687441,
+      "learning_rate": 0.0009142857142857143,
+      "loss": 1.5334,
+      "step": 256
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2054533522968223,
+      "learning_rate": 0.0009178571428571428,
+      "loss": 1.3404,
+      "step": 257
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.21102299105750338,
+      "learning_rate": 0.0009214285714285714,
+      "loss": 1.3101,
+      "step": 258
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.18963226471367908,
+      "learning_rate": 0.000925,
+      "loss": 1.3663,
+      "step": 259
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2184587395865125,
+      "learning_rate": 0.0009285714285714287,
+      "loss": 1.5179,
+      "step": 260
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.20605484569236143,
+      "learning_rate": 0.0009321428571428572,
+      "loss": 1.5511,
+      "step": 261
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2791226050077258,
+      "learning_rate": 0.0009357142857142857,
+      "loss": 1.5101,
+      "step": 262
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.19037886699486145,
+      "learning_rate": 0.0009392857142857143,
+      "loss": 1.4686,
+      "step": 263
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1981693942215429,
+      "learning_rate": 0.0009428571428571429,
+      "loss": 1.3974,
+      "step": 264
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.22826376509236745,
+      "learning_rate": 0.0009464285714285714,
+      "loss": 1.5342,
+      "step": 265
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.19772832728602222,
+      "learning_rate": 0.00095,
+      "loss": 1.4764,
+      "step": 266
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2584142237517251,
+      "learning_rate": 0.0009535714285714286,
+      "loss": 1.4261,
+      "step": 267
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2144015182353289,
+      "learning_rate": 0.0009571428571428573,
+      "loss": 1.5228,
+      "step": 268
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.17217525622544874,
+      "learning_rate": 0.0009607142857142858,
+      "loss": 1.5678,
+      "step": 269
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1996368107538457,
+      "learning_rate": 0.0009642857142857143,
+      "loss": 1.5514,
+      "step": 270
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.20732376738259717,
+      "learning_rate": 0.0009678571428571429,
+      "loss": 1.5184,
+      "step": 271
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.19770305356463294,
+      "learning_rate": 0.0009714285714285714,
+      "loss": 1.6127,
+      "step": 272
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.18981688274012634,
+      "learning_rate": 0.000975,
+      "loss": 1.5789,
+      "step": 273
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.20918579733765114,
+      "learning_rate": 0.0009785714285714285,
+      "loss": 1.408,
+      "step": 274
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.20927524823711904,
+      "learning_rate": 0.0009821428571428572,
+      "loss": 1.4623,
+      "step": 275
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.17053415437999953,
+      "learning_rate": 0.0009857142857142857,
+      "loss": 1.4153,
+      "step": 276
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.17353729231705273,
+      "learning_rate": 0.0009892857142857142,
+      "loss": 1.4628,
+      "step": 277
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.19297924630966062,
+      "learning_rate": 0.000992857142857143,
+      "loss": 1.4706,
+      "step": 278
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.18001888908262587,
+      "learning_rate": 0.0009964285714285715,
+      "loss": 1.5269,
+      "step": 279
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.18830669239089334,
+      "learning_rate": 0.001,
+      "loss": 1.5408,
+      "step": 280
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.17643323360815724,
+      "learning_rate": 0.0009999999696866382,
+      "loss": 1.4809,
+      "step": 281
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2040443477107501,
+      "learning_rate": 0.000999999878746556,
+      "loss": 1.5866,
+      "step": 282
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1793843334457987,
+      "learning_rate": 0.000999999727179765,
+      "loss": 1.4915,
+      "step": 283
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2035576998367501,
+      "learning_rate": 0.0009999995149862835,
+      "loss": 1.5028,
+      "step": 284
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.18822278840465206,
+      "learning_rate": 0.0009999992421661369,
+      "loss": 1.5095,
+      "step": 285
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.22599095980346726,
+      "learning_rate": 0.0009999989087193582,
+      "loss": 1.4853,
+      "step": 286
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.19948992318384096,
+      "learning_rate": 0.0009999985146459881,
+      "loss": 1.4658,
+      "step": 287
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2077750753103998,
+      "learning_rate": 0.0009999980599460746,
+      "loss": 1.4903,
+      "step": 288
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2042826255511603,
+      "learning_rate": 0.0009999975446196726,
+      "loss": 1.5716,
+      "step": 289
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1809148903047096,
+      "learning_rate": 0.0009999969686668442,
+      "loss": 1.5172,
+      "step": 290
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.17323034884698793,
+      "learning_rate": 0.0009999963320876598,
+      "loss": 1.4453,
+      "step": 291
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.20664753532753552,
+      "learning_rate": 0.0009999956348821966,
+      "loss": 1.6475,
+      "step": 292
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5668139338801725,
+      "learning_rate": 0.0009999948770505387,
+      "loss": 1.4983,
+      "step": 293
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.20916989817361614,
+      "learning_rate": 0.0009999940585927781,
+      "loss": 1.5331,
+      "step": 294
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.20045916684556297,
+      "learning_rate": 0.0009999931795090142,
+      "loss": 1.4464,
+      "step": 295
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1991163579687259,
+      "learning_rate": 0.0009999922397993537,
+      "loss": 1.5759,
+      "step": 296
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.26448205876932024,
+      "learning_rate": 0.0009999912394639103,
+      "loss": 1.4446,
+      "step": 297
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.19273562442141076,
+      "learning_rate": 0.0009999901785028053,
+      "loss": 1.5955,
+      "step": 298
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.16933403058061827,
+      "learning_rate": 0.0009999890569161677,
+      "loss": 1.4588,
+      "step": 299
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1830637464672982,
+      "learning_rate": 0.000999987874704133,
+      "loss": 1.6497,
+      "step": 300
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.17282182901114648,
+      "learning_rate": 0.0009999866318668449,
+      "loss": 1.4537,
+      "step": 301
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.18143770203988144,
+      "learning_rate": 0.0009999853284044537,
+      "loss": 1.5288,
+      "step": 302
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.21283164213442005,
+      "learning_rate": 0.000999983964317118,
+      "loss": 1.6833,
+      "step": 303
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.16871865453821608,
+      "learning_rate": 0.0009999825396050028,
+      "loss": 1.4128,
+      "step": 304
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.17788316189586986,
+      "learning_rate": 0.000999981054268281,
+      "loss": 1.4958,
+      "step": 305
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.17628497956862654,
+      "learning_rate": 0.0009999795083071327,
+      "loss": 1.5826,
+      "step": 306
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.17774954930179934,
+      "learning_rate": 0.0009999779017217452,
+      "loss": 1.5999,
+      "step": 307
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.16138411201607197,
+      "learning_rate": 0.0009999762345123135,
+      "loss": 1.448,
+      "step": 308
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.14636359639747132,
+      "learning_rate": 0.0009999745066790397,
+      "loss": 1.5426,
+      "step": 309
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.17395839471064034,
+      "learning_rate": 0.0009999727182221335,
+      "loss": 1.4609,
+      "step": 310
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1596077796811495,
+      "learning_rate": 0.0009999708691418112,
+      "loss": 1.5566,
+      "step": 311
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.17254640291130927,
+      "learning_rate": 0.0009999689594382976,
+      "loss": 1.5635,
+      "step": 312
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.21093878480731898,
+      "learning_rate": 0.0009999669891118238,
+      "loss": 1.4798,
+      "step": 313
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.14639537332716396,
+      "learning_rate": 0.0009999649581626292,
+      "loss": 1.3929,
+      "step": 314
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.18027095555325548,
+      "learning_rate": 0.0009999628665909597,
+      "loss": 1.6339,
+      "step": 315
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.18042986872074054,
+      "learning_rate": 0.0009999607143970689,
+      "loss": 1.547,
+      "step": 316
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1586215202794954,
+      "learning_rate": 0.0009999585015812178,
+      "loss": 1.4079,
+      "step": 317
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.17834825140193727,
+      "learning_rate": 0.0009999562281436747,
+      "loss": 1.4367,
+      "step": 318
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.17982659450306954,
+      "learning_rate": 0.0009999538940847156,
+      "loss": 1.6693,
+      "step": 319
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.164622497777815,
+      "learning_rate": 0.0009999514994046231,
+      "loss": 1.5315,
+      "step": 320
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2006206552979814,
+      "learning_rate": 0.0009999490441036877,
+      "loss": 1.5092,
+      "step": 321
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.16907762296767967,
+      "learning_rate": 0.0009999465281822072,
+      "loss": 1.5472,
+      "step": 322
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.19421538970369667,
+      "learning_rate": 0.0009999439516404863,
+      "loss": 1.5062,
+      "step": 323
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.16267349147204174,
+      "learning_rate": 0.000999941314478838,
+      "loss": 1.4601,
+      "step": 324
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.17687924044685271,
+      "learning_rate": 0.0009999386166975816,
+      "loss": 1.4468,
+      "step": 325
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1658946269536119,
+      "learning_rate": 0.0009999358582970443,
+      "loss": 1.3812,
+      "step": 326
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.19961931592072243,
+      "learning_rate": 0.000999933039277561,
+      "loss": 1.451,
+      "step": 327
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.18247338360131146,
+      "learning_rate": 0.0009999301596394727,
+      "loss": 1.5772,
+      "step": 328
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.18531837950187607,
+      "learning_rate": 0.0009999272193831293,
+      "loss": 1.499,
+      "step": 329
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.17578485273271977,
+      "learning_rate": 0.000999924218508887,
+      "loss": 1.4941,
+      "step": 330
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1716177706116155,
+      "learning_rate": 0.0009999211570171098,
+      "loss": 1.3478,
+      "step": 331
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.15907076461451958,
+      "learning_rate": 0.0009999180349081686,
+      "loss": 1.5074,
+      "step": 332
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.14106955725340992,
+      "learning_rate": 0.0009999148521824424,
+      "loss": 1.6304,
+      "step": 333
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.18473843834209136,
+      "learning_rate": 0.0009999116088403167,
+      "loss": 1.3852,
+      "step": 334
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.12962084897330353,
+      "learning_rate": 0.0009999083048821851,
+      "loss": 1.5796,
+      "step": 335
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.21469389717837936,
+      "learning_rate": 0.000999904940308448,
+      "loss": 1.5622,
+      "step": 336
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.13965253223738205,
+      "learning_rate": 0.0009999015151195135,
+      "loss": 1.4831,
+      "step": 337
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.14771018846987435,
+      "learning_rate": 0.000999898029315797,
+      "loss": 1.4489,
+      "step": 338
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.14126087838350085,
+      "learning_rate": 0.0009998944828977208,
+      "loss": 1.4695,
+      "step": 339
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.13103131840592866,
+      "learning_rate": 0.0009998908758657153,
+      "loss": 1.3719,
+      "step": 340
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.15938251830581013,
+      "learning_rate": 0.0009998872082202176,
+      "loss": 1.6759,
+      "step": 341
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.14062844047046466,
+      "learning_rate": 0.0009998834799616726,
+      "loss": 1.5435,
+      "step": 342
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.14069295274521235,
+      "learning_rate": 0.0009998796910905324,
+      "loss": 1.5798,
+      "step": 343
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.14899985863071655,
+      "learning_rate": 0.0009998758416072562,
+      "loss": 1.4963,
+      "step": 344
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.15403995672516727,
+      "learning_rate": 0.000999871931512311,
+      "loss": 1.5071,
+      "step": 345
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.15965488508555337,
+      "learning_rate": 0.0009998679608061705,
+      "loss": 1.4844,
+      "step": 346
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1397205113440922,
+      "learning_rate": 0.0009998639294893166,
+      "loss": 1.3994,
+      "step": 347
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.16837720499970224,
+      "learning_rate": 0.0009998598375622382,
+      "loss": 1.6294,
+      "step": 348
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.14290960526839144,
+      "learning_rate": 0.0009998556850254307,
+      "loss": 1.393,
+      "step": 349
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1647613660769307,
+      "learning_rate": 0.0009998514718793986,
+      "loss": 1.5986,
+      "step": 350
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.126604301560717,
+      "learning_rate": 0.000999847198124652,
+      "loss": 1.5888,
+      "step": 351
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1625076881295614,
+      "learning_rate": 0.0009998428637617094,
+      "loss": 1.4581,
+      "step": 352
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1354795426114141,
+      "learning_rate": 0.0009998384687910967,
+      "loss": 1.5685,
+      "step": 353
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1960614626166613,
+      "learning_rate": 0.000999834013213346,
+      "loss": 1.4762,
+      "step": 354
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.15410756920154223,
+      "learning_rate": 0.0009998294970289983,
+      "loss": 1.5645,
+      "step": 355
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.18398803112017642,
+      "learning_rate": 0.0009998249202386007,
+      "loss": 1.4068,
+      "step": 356
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.19952868680271008,
+      "learning_rate": 0.0009998202828427085,
+      "loss": 1.5943,
+      "step": 357
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.17194494265665142,
+      "learning_rate": 0.000999815584841884,
+      "loss": 1.5148,
+      "step": 358
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.16781358427269294,
+      "learning_rate": 0.0009998108262366965,
+      "loss": 1.4017,
+      "step": 359
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.17487641147080715,
+      "learning_rate": 0.0009998060070277232,
+      "loss": 1.4302,
+      "step": 360
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.14854649570076,
+      "learning_rate": 0.0009998011272155485,
+      "loss": 1.435,
+      "step": 361
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.13955296740945117,
+      "learning_rate": 0.0009997961868007642,
+      "loss": 1.3177,
+      "step": 362
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.17180197963616256,
+      "learning_rate": 0.0009997911857839688,
+      "loss": 1.4593,
+      "step": 363
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1450351295687952,
+      "learning_rate": 0.0009997861241657694,
+      "loss": 1.4266,
+      "step": 364
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.17459394228876687,
+      "learning_rate": 0.0009997810019467793,
+      "loss": 1.4735,
+      "step": 365
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.21035233323036612,
+      "learning_rate": 0.0009997758191276197,
+      "loss": 1.5312,
+      "step": 366
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1522354572488286,
+      "learning_rate": 0.0009997705757089192,
+      "loss": 1.4431,
+      "step": 367
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.15293858625262405,
+      "learning_rate": 0.0009997652716913134,
+      "loss": 1.6787,
+      "step": 368
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1353046205100757,
+      "learning_rate": 0.0009997599070754454,
+      "loss": 1.4503,
+      "step": 369
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.17227780219049219,
+      "learning_rate": 0.0009997544818619657,
+      "loss": 1.4085,
+      "step": 370
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1437172909509328,
+      "learning_rate": 0.000999748996051532,
+      "loss": 1.5073,
+      "step": 371
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1504855289987324,
+      "learning_rate": 0.0009997434496448099,
+      "loss": 1.5767,
+      "step": 372
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.13326666119325065,
+      "learning_rate": 0.0009997378426424715,
+      "loss": 1.4447,
+      "step": 373
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.12945116710966437,
+      "learning_rate": 0.0009997321750451968,
+      "loss": 1.5107,
+      "step": 374
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1633221149641643,
+      "learning_rate": 0.0009997264468536732,
+      "loss": 1.5965,
+      "step": 375
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.14094383798016874,
+      "learning_rate": 0.000999720658068595,
+      "loss": 1.5529,
+      "step": 376
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1424006539876174,
+      "learning_rate": 0.0009997148086906642,
+      "loss": 1.4084,
+      "step": 377
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1659677155871326,
+      "learning_rate": 0.0009997088987205903,
+      "loss": 1.6574,
+      "step": 378
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.16406857672291258,
+      "learning_rate": 0.0009997029281590892,
+      "loss": 1.4398,
+      "step": 379
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.14181116310622616,
+      "learning_rate": 0.0009996968970068857,
+      "loss": 1.4507,
+      "step": 380
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.14808479870132174,
+      "learning_rate": 0.0009996908052647105,
+      "loss": 1.5897,
+      "step": 381
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.17030159407878712,
+      "learning_rate": 0.0009996846529333027,
+      "loss": 1.5398,
+      "step": 382
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.16969702631552683,
+      "learning_rate": 0.0009996784400134078,
+      "loss": 1.4812,
+      "step": 383
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.17226460579810668,
+      "learning_rate": 0.0009996721665057795,
+      "loss": 1.469,
+      "step": 384
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1721386837516993,
+      "learning_rate": 0.0009996658324111785,
+      "loss": 1.5467,
+      "step": 385
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.14892597511153813,
+      "learning_rate": 0.0009996594377303725,
+      "loss": 1.5863,
+      "step": 386
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1575483529907697,
+      "learning_rate": 0.0009996529824641374,
+      "loss": 1.4743,
+      "step": 387
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1456398333006279,
+      "learning_rate": 0.0009996464666132553,
+      "loss": 1.5276,
+      "step": 388
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.15489743549736446,
+      "learning_rate": 0.0009996398901785167,
+      "loss": 1.482,
+      "step": 389
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1583860129059049,
+      "learning_rate": 0.000999633253160719,
+      "loss": 1.5561,
+      "step": 390
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.17262883033342483,
+      "learning_rate": 0.0009996265555606667,
+      "loss": 1.4898,
+      "step": 391
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.16439961527862196,
+      "learning_rate": 0.000999619797379172,
+      "loss": 1.5272,
+      "step": 392
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.15828301920189175,
+      "learning_rate": 0.0009996129786170546,
+      "loss": 1.4813,
+      "step": 393
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1731170888583976,
+      "learning_rate": 0.000999606099275141,
+      "loss": 1.4245,
+      "step": 394
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1559416559722423,
+      "learning_rate": 0.0009995991593542656,
+      "loss": 1.5365,
+      "step": 395
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.18583660593985893,
+      "learning_rate": 0.0009995921588552695,
+      "loss": 1.4156,
+      "step": 396
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1707241641157939,
+      "learning_rate": 0.000999585097779002,
+      "loss": 1.5137,
+      "step": 397
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1407581665746861,
+      "learning_rate": 0.0009995779761263189,
+      "loss": 1.5203,
+      "step": 398
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1388777158799526,
+      "learning_rate": 0.000999570793898084,
+      "loss": 1.5761,
+      "step": 399
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.16295821219358653,
+      "learning_rate": 0.000999563551095168,
+      "loss": 1.6157,
+      "step": 400
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1539280261207723,
+      "learning_rate": 0.0009995562477184492,
+      "loss": 1.416,
+      "step": 401
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.14246096761905572,
+      "learning_rate": 0.0009995488837688132,
+      "loss": 1.5411,
+      "step": 402
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.15716424737797968,
+      "learning_rate": 0.0009995414592471527,
+      "loss": 1.4782,
+      "step": 403
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.14283758228041246,
+      "learning_rate": 0.000999533974154368,
+      "loss": 1.4147,
+      "step": 404
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.15328695456751443,
+      "learning_rate": 0.000999526428491367,
+      "loss": 1.5153,
+      "step": 405
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1436552033181903,
+      "learning_rate": 0.0009995188222590645,
+      "loss": 1.5375,
+      "step": 406
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1454954430103715,
+      "learning_rate": 0.0009995111554583825,
+      "loss": 1.34,
+      "step": 407
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.16441780739523218,
+      "learning_rate": 0.0009995034280902509,
+      "loss": 1.5478,
+      "step": 408
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.13139245357731225,
+      "learning_rate": 0.0009994956401556065,
+      "loss": 1.3986,
+      "step": 409
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1609687663696494,
+      "learning_rate": 0.0009994877916553937,
+      "loss": 1.651,
+      "step": 410
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1774917612967736,
+      "learning_rate": 0.0009994798825905644,
+      "loss": 1.4559,
+      "step": 411
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.12610286657532213,
+      "learning_rate": 0.000999471912962077,
+      "loss": 1.3864,
+      "step": 412
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.15091836283777837,
+      "learning_rate": 0.0009994638827708986,
+      "loss": 1.3917,
+      "step": 413
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1274380080548309,
+      "learning_rate": 0.0009994557920180024,
+      "loss": 1.4435,
+      "step": 414
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1403056631620941,
+      "learning_rate": 0.0009994476407043694,
+      "loss": 1.5053,
+      "step": 415
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.13960834407799336,
+      "learning_rate": 0.000999439428830988,
+      "loss": 1.4803,
+      "step": 416
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.14244926848342568,
+      "learning_rate": 0.0009994311563988542,
+      "loss": 1.5339,
+      "step": 417
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.14869433694885337,
+      "learning_rate": 0.0009994228234089708,
+      "loss": 1.4244,
+      "step": 418
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.14528003568402328,
+      "learning_rate": 0.0009994144298623485,
+      "loss": 1.4579,
+      "step": 419
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.13406799469371342,
+      "learning_rate": 0.0009994059757600048,
+      "loss": 1.5282,
+      "step": 420
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.14796217619302254,
+      "learning_rate": 0.0009993974611029646,
+      "loss": 1.5673,
+      "step": 421
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1370563358302789,
+      "learning_rate": 0.0009993888858922605,
+      "loss": 1.4697,
+      "step": 422
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.12170722368190066,
+      "learning_rate": 0.0009993802501289326,
+      "loss": 1.4959,
+      "step": 423
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.10003558537131502,
+      "learning_rate": 0.0009993715538140276,
+      "loss": 1.4291,
+      "step": 424
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1334012983526422,
+      "learning_rate": 0.0009993627969486002,
+      "loss": 1.558,
+      "step": 425
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.12602036845602907,
+      "learning_rate": 0.000999353979533712,
+      "loss": 1.3761,
+      "step": 426
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.09994334678040766,
+      "learning_rate": 0.0009993451015704324,
+      "loss": 1.3765,
+      "step": 427
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1429494796596272,
+      "learning_rate": 0.0009993361630598377,
+      "loss": 1.4477,
+      "step": 428
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1459824322844078,
+      "learning_rate": 0.0009993271640030116,
+      "loss": 1.5808,
+      "step": 429
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.13516497528709895,
+      "learning_rate": 0.0009993181044010454,
+      "loss": 1.5225,
+      "step": 430
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.12460002385750782,
+      "learning_rate": 0.000999308984255038,
+      "loss": 1.5176,
+      "step": 431
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.12165124868540639,
+      "learning_rate": 0.0009992998035660945,
+      "loss": 1.4966,
+      "step": 432
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.11682420685092966,
+      "learning_rate": 0.0009992905623353286,
+      "loss": 1.4564,
+      "step": 433
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.13472134097524974,
+      "learning_rate": 0.0009992812605638605,
+      "loss": 1.4921,
+      "step": 434
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.11455037083596988,
+      "learning_rate": 0.0009992718982528186,
+      "loss": 1.4622,
+      "step": 435
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.13050817933155195,
+      "learning_rate": 0.0009992624754033377,
+      "loss": 1.5472,
+      "step": 436
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.15767461304956595,
+      "learning_rate": 0.0009992529920165602,
+      "loss": 1.6125,
+      "step": 437
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.146525866672167,
+      "learning_rate": 0.0009992434480936366,
+      "loss": 1.4863,
+      "step": 438
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1321979814402815,
+      "learning_rate": 0.0009992338436357235,
+      "loss": 1.5751,
+      "step": 439
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1292076236415919,
+      "learning_rate": 0.0009992241786439858,
+      "loss": 1.4775,
+      "step": 440
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.13500666830362648,
+      "learning_rate": 0.0009992144531195955,
+      "loss": 1.5966,
+      "step": 441
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.13695901467453922,
+      "learning_rate": 0.0009992046670637316,
+      "loss": 1.5805,
+      "step": 442
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.13922716501708265,
+      "learning_rate": 0.0009991948204775807,
+      "loss": 1.5072,
+      "step": 443
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.14647658645269154,
+      "learning_rate": 0.000999184913362337,
+      "loss": 1.3592,
+      "step": 444
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.12899480292336285,
+      "learning_rate": 0.0009991749457192013,
+      "loss": 1.4513,
+      "step": 445
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.12393656639381344,
+      "learning_rate": 0.0009991649175493827,
+      "loss": 1.4687,
+      "step": 446
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.16754428326189647,
+      "learning_rate": 0.000999154828854097,
+      "loss": 1.4184,
+      "step": 447
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.14210675981117007,
+      "learning_rate": 0.0009991446796345676,
+      "loss": 1.4588,
+      "step": 448
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1377863515377502,
+      "learning_rate": 0.0009991344698920246,
+      "loss": 1.5647,
+      "step": 449
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.15677195035868116,
+      "learning_rate": 0.0009991241996277068,
+      "loss": 1.4818,
+      "step": 450
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.14291065447977097,
+      "learning_rate": 0.0009991138688428588,
+      "loss": 1.4337,
+      "step": 451
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1500952470396202,
+      "learning_rate": 0.0009991034775387335,
+      "loss": 1.5161,
+      "step": 452
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.14513044258257118,
+      "learning_rate": 0.000999093025716591,
+      "loss": 1.5096,
+      "step": 453
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.16032928911490424,
+      "learning_rate": 0.000999082513377698,
+      "loss": 1.4002,
+      "step": 454
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.17202853259574485,
+      "learning_rate": 0.0009990719405233303,
+      "loss": 1.4792,
+      "step": 455
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.15835946799615688,
+      "learning_rate": 0.000999061307154769,
+      "loss": 1.4615,
+      "step": 456
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.16951071663191195,
+      "learning_rate": 0.0009990506132733037,
+      "loss": 1.4748,
+      "step": 457
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1398963494982491,
+      "learning_rate": 0.000999039858880231,
+      "loss": 1.5173,
+      "step": 458
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.12997095802643854,
+      "learning_rate": 0.000999029043976855,
+      "loss": 1.2695,
+      "step": 459
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.11616314177597527,
+      "learning_rate": 0.000999018168564487,
+      "loss": 1.4087,
+      "step": 460
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.17461980067454186,
+      "learning_rate": 0.0009990072326444455,
+      "loss": 1.5803,
+      "step": 461
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.14272014194613272,
+      "learning_rate": 0.0009989962362180569,
+      "loss": 1.4678,
+      "step": 462
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.16570305570684601,
+      "learning_rate": 0.0009989851792866543,
+      "loss": 1.6578,
+      "step": 463
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.11998225014078424,
+      "learning_rate": 0.0009989740618515787,
+      "loss": 1.4982,
+      "step": 464
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.13537021139680433,
+      "learning_rate": 0.0009989628839141775,
+      "loss": 1.5582,
+      "step": 465
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.14086644816346083,
+      "learning_rate": 0.0009989516454758066,
+      "loss": 1.4891,
+      "step": 466
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1358567824453613,
+      "learning_rate": 0.0009989403465378284,
+      "loss": 1.4645,
+      "step": 467
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1401683725102955,
+      "learning_rate": 0.0009989289871016132,
+      "loss": 1.5034,
+      "step": 468
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.14130289874302648,
+      "learning_rate": 0.0009989175671685383,
+      "loss": 1.6208,
+      "step": 469
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1476505273386164,
+      "learning_rate": 0.0009989060867399884,
+      "loss": 1.4666,
+      "step": 470
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1494861721735433,
+      "learning_rate": 0.0009988945458173552,
+      "loss": 1.458,
+      "step": 471
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.14147325490478396,
+      "learning_rate": 0.0009988829444020385,
+      "loss": 1.6521,
+      "step": 472
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.15133662684402113,
+      "learning_rate": 0.0009988712824954451,
+      "loss": 1.5475,
+      "step": 473
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.15736937108778778,
+      "learning_rate": 0.0009988595600989886,
+      "loss": 1.5526,
+      "step": 474
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.13815053807297753,
+      "learning_rate": 0.0009988477772140908,
+      "loss": 1.4171,
+      "step": 475
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.15170448316189497,
+      "learning_rate": 0.00099883593384218,
+      "loss": 1.4429,
+      "step": 476
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.14523768793650085,
+      "learning_rate": 0.0009988240299846926,
+      "loss": 1.6591,
+      "step": 477
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.15517158286559127,
+      "learning_rate": 0.0009988120656430719,
+      "loss": 1.4886,
+      "step": 478
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.13328382363243924,
+      "learning_rate": 0.0009988000408187685,
+      "loss": 1.4964,
+      "step": 479
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.12208130405531557,
+      "learning_rate": 0.0009987879555132405,
+      "loss": 1.3444,
+      "step": 480
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1371216763034477,
+      "learning_rate": 0.0009987758097279534,
+      "loss": 1.5955,
+      "step": 481
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.11997966531046428,
+      "learning_rate": 0.0009987636034643798,
+      "loss": 1.462,
+      "step": 482
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.15171065892823024,
+      "learning_rate": 0.0009987513367239996,
+      "loss": 1.3923,
+      "step": 483
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.13082361088442068,
+      "learning_rate": 0.0009987390095083004,
+      "loss": 1.4156,
+      "step": 484
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.12932044553418595,
+      "learning_rate": 0.0009987266218187772,
+      "loss": 1.4468,
+      "step": 485
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1475713076816688,
+      "learning_rate": 0.0009987141736569314,
+      "loss": 1.4624,
+      "step": 486
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.14178318017974623,
+      "learning_rate": 0.000998701665024273,
+      "loss": 1.3946,
+      "step": 487
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.16305685355062952,
+      "learning_rate": 0.0009986890959223181,
+      "loss": 1.5231,
+      "step": 488
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1686444442480543,
+      "learning_rate": 0.0009986764663525913,
+      "loss": 1.6783,
+      "step": 489
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.15549050758741176,
+      "learning_rate": 0.0009986637763166237,
+      "loss": 1.62,
+      "step": 490
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.13218940398691503,
+      "learning_rate": 0.0009986510258159541,
+      "loss": 1.4141,
+      "step": 491
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.14336555374093515,
+      "learning_rate": 0.0009986382148521283,
+      "loss": 1.545,
+      "step": 492
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.12407347653401202,
+      "learning_rate": 0.0009986253434267,
+      "loss": 1.4171,
+      "step": 493
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.15672602263906668,
+      "learning_rate": 0.00099861241154123,
+      "loss": 1.5143,
+      "step": 494
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.13339275812986198,
+      "learning_rate": 0.000998599419197286,
+      "loss": 1.4366,
+      "step": 495
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.13385666580192612,
+      "learning_rate": 0.0009985863663964434,
+      "loss": 1.4831,
+      "step": 496
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1704216054223673,
+      "learning_rate": 0.000998573253140285,
+      "loss": 1.365,
+      "step": 497
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1419854321243664,
+      "learning_rate": 0.0009985600794304007,
+      "loss": 1.3812,
+      "step": 498
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1294114515649719,
+      "learning_rate": 0.0009985468452683882,
+      "loss": 1.4495,
+      "step": 499
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.16788358525374333,
+      "learning_rate": 0.0009985335506558519,
+      "loss": 1.6707,
+      "step": 500
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1633606897699836,
+      "learning_rate": 0.0009985201955944039,
+      "loss": 1.6144,
+      "step": 501
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.13574456823443043,
+      "learning_rate": 0.0009985067800856635,
+      "loss": 1.5012,
+      "step": 502
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.12805624518518505,
+      "learning_rate": 0.0009984933041312573,
+      "loss": 1.469,
+      "step": 503
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.14562449161498103,
+      "learning_rate": 0.0009984797677328194,
+      "loss": 1.557,
+      "step": 504
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.13274042544010126,
+      "learning_rate": 0.0009984661708919913,
+      "loss": 1.5591,
+      "step": 505
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.13783829279522242,
+      "learning_rate": 0.0009984525136104215,
+      "loss": 1.4461,
+      "step": 506
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.14150450438484263,
+      "learning_rate": 0.000998438795889766,
+      "loss": 1.4764,
+      "step": 507
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.17191114370741217,
+      "learning_rate": 0.0009984250177316881,
+      "loss": 1.5333,
+      "step": 508
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.16755229866536533,
+      "learning_rate": 0.0009984111791378582,
+      "loss": 1.5334,
+      "step": 509
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.15257417846103322,
+      "learning_rate": 0.0009983972801099548,
+      "loss": 1.4058,
+      "step": 510
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.15478833011778742,
+      "learning_rate": 0.000998383320649663,
+      "loss": 1.5657,
+      "step": 511
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.12880041372659604,
+      "learning_rate": 0.0009983693007586752,
+      "loss": 1.6036,
+      "step": 512
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.14047382096291214,
+      "learning_rate": 0.0009983552204386916,
+      "loss": 1.5327,
+      "step": 513
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.12671081085946684,
+      "learning_rate": 0.0009983410796914197,
+      "loss": 1.5109,
+      "step": 514
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.13315349868182663,
+      "learning_rate": 0.0009983268785185735,
+      "loss": 1.4648,
+      "step": 515
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.12974308859167147,
+      "learning_rate": 0.0009983126169218755,
+      "loss": 1.4275,
+      "step": 516
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.10708978863759408,
+      "learning_rate": 0.0009982982949030546,
+      "loss": 1.4725,
+      "step": 517
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.11232917145575787,
+      "learning_rate": 0.0009982839124638475,
+      "loss": 1.5388,
+      "step": 518
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.13263578439023072,
+      "learning_rate": 0.0009982694696059982,
+      "loss": 1.5457,
+      "step": 519
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1180388204627775,
+      "learning_rate": 0.0009982549663312581,
+      "loss": 1.4521,
+      "step": 520
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1069615275333495,
+      "learning_rate": 0.0009982404026413854,
+      "loss": 1.4397,
+      "step": 521
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.11920448256909753,
+      "learning_rate": 0.0009982257785381464,
+      "loss": 1.3893,
+      "step": 522
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.13802661336336072,
+      "learning_rate": 0.0009982110940233138,
+      "loss": 1.3956,
+      "step": 523
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.12589487672919536,
+      "learning_rate": 0.0009981963490986686,
+      "loss": 1.6111,
+      "step": 524
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1388066167759072,
+      "learning_rate": 0.0009981815437659985,
+      "loss": 1.4834,
+      "step": 525
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.11860878855329496,
+      "learning_rate": 0.0009981666780270989,
+      "loss": 1.5376,
+      "step": 526
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1291593285336083,
+      "learning_rate": 0.000998151751883772,
+      "loss": 1.5099,
+      "step": 527
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.13856146675132416,
+      "learning_rate": 0.0009981367653378278,
+      "loss": 1.4985,
+      "step": 528
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1428411422383799,
+      "learning_rate": 0.0009981217183910834,
+      "loss": 1.4989,
+      "step": 529
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.13401950703669366,
+      "learning_rate": 0.0009981066110453633,
+      "loss": 1.5726,
+      "step": 530
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.11238669635358409,
+      "learning_rate": 0.0009980914433024997,
+      "loss": 1.4765,
+      "step": 531
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.14635913263213093,
+      "learning_rate": 0.0009980762151643313,
+      "loss": 1.5009,
+      "step": 532
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.12077582640456455,
+      "learning_rate": 0.0009980609266327044,
+      "loss": 1.402,
+      "step": 533
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.13303350194920682,
+      "learning_rate": 0.0009980455777094733,
+      "loss": 1.6252,
+      "step": 534
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.13303747826576684,
+      "learning_rate": 0.0009980301683964988,
+      "loss": 1.5057,
+      "step": 535
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.14834784322318495,
+      "learning_rate": 0.0009980146986956495,
+      "loss": 1.4399,
+      "step": 536
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1385305315636288,
+      "learning_rate": 0.000997999168608801,
+      "loss": 1.5656,
+      "step": 537
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1264521765532537,
+      "learning_rate": 0.0009979835781378367,
+      "loss": 1.4,
+      "step": 538
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.15363887930073927,
+      "learning_rate": 0.0009979679272846462,
+      "loss": 1.5321,
+      "step": 539
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.12205084836277039,
+      "learning_rate": 0.0009979522160511282,
+      "loss": 1.4562,
+      "step": 540
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.11622818650530785,
+      "learning_rate": 0.000997936444439187,
+      "loss": 1.4356,
+      "step": 541
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.12491589190127463,
+      "learning_rate": 0.0009979206124507355,
+      "loss": 1.3947,
+      "step": 542
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.12922947991275946,
+      "learning_rate": 0.0009979047200876932,
+      "loss": 1.49,
+      "step": 543
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.13489214169177516,
+      "learning_rate": 0.000997888767351987,
+      "loss": 1.5153,
+      "step": 544
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.12878593154366547,
+      "learning_rate": 0.0009978727542455511,
+      "loss": 1.4892,
+      "step": 545
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.13996023103395688,
+      "learning_rate": 0.0009978566807703274,
+      "loss": 1.5035,
+      "step": 546
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.12336944314011723,
+      "learning_rate": 0.0009978405469282647,
+      "loss": 1.3817,
+      "step": 547
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1420723471733146,
+      "learning_rate": 0.0009978243527213196,
+      "loss": 1.5272,
+      "step": 548
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.14789430461564015,
+      "learning_rate": 0.0009978080981514553,
+      "loss": 1.5767,
+      "step": 549
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1267611684634138,
+      "learning_rate": 0.0009977917832206431,
+      "loss": 1.4002,
+      "step": 550
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.13572961076613538,
+      "learning_rate": 0.0009977754079308608,
+      "loss": 1.3569,
+      "step": 551
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.12869792779865574,
+      "learning_rate": 0.0009977589722840942,
+      "loss": 1.4989,
+      "step": 552
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.12884701104585888,
+      "learning_rate": 0.0009977424762823363,
+      "loss": 1.4594,
+      "step": 553
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1496458530775414,
+      "learning_rate": 0.000997725919927587,
+      "loss": 1.6697,
+      "step": 554
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.16089313194101512,
+      "learning_rate": 0.0009977093032218544,
+      "loss": 1.4965,
+      "step": 555
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.12898716410042255,
+      "learning_rate": 0.0009976926261671523,
+      "loss": 1.5562,
+      "step": 556
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.13909417311407565,
+      "learning_rate": 0.000997675888765504,
+      "loss": 1.4994,
+      "step": 557
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1364769705088755,
+      "learning_rate": 0.0009976590910189382,
+      "loss": 1.6546,
+      "step": 558
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1407938612275754,
+      "learning_rate": 0.0009976422329294919,
+      "loss": 1.6437,
+      "step": 559
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.12342258081030302,
+      "learning_rate": 0.0009976253144992093,
+      "loss": 1.4347,
+      "step": 560
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.11970500471927006,
+      "learning_rate": 0.0009976083357301417,
+      "loss": 1.2459,
+      "step": 561
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.11644855616958355,
+      "learning_rate": 0.0009975912966243478,
+      "loss": 1.4681,
+      "step": 562
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1399132278939777,
+      "learning_rate": 0.0009975741971838938,
+      "loss": 1.4792,
+      "step": 563
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.12385326071720139,
+      "learning_rate": 0.000997557037410853,
+      "loss": 1.4996,
+      "step": 564
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.13958038993004748,
+      "learning_rate": 0.0009975398173073062,
+      "loss": 1.5275,
+      "step": 565
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.11635118724721759,
+      "learning_rate": 0.000997522536875341,
+      "loss": 1.5998,
+      "step": 566
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1135776748032988,
+      "learning_rate": 0.0009975051961170532,
+      "loss": 1.4715,
+      "step": 567
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.14026212563791537,
+      "learning_rate": 0.0009974877950345452,
+      "loss": 1.5512,
+      "step": 568
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.13372265079333248,
+      "learning_rate": 0.000997470333629927,
+      "loss": 1.5538,
+      "step": 569
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.09829929826672491,
+      "learning_rate": 0.0009974528119053156,
+      "loss": 1.404,
+      "step": 570
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1400205748949228,
+      "learning_rate": 0.0009974352298628359,
+      "loss": 1.4348,
+      "step": 571
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.12816133710759073,
+      "learning_rate": 0.0009974175875046196,
+      "loss": 1.5441,
+      "step": 572
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.11826809528085243,
+      "learning_rate": 0.0009973998848328061,
+      "loss": 1.4434,
+      "step": 573
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.11123840372385792,
+      "learning_rate": 0.0009973821218495415,
+      "loss": 1.5314,
+      "step": 574
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1042235432756459,
+      "learning_rate": 0.0009973642985569803,
+      "loss": 1.5121,
+      "step": 575
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.11749153433901713,
+      "learning_rate": 0.0009973464149572828,
+      "loss": 1.5447,
+      "step": 576
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.15188595873537555,
+      "learning_rate": 0.000997328471052618,
+      "loss": 1.5764,
+      "step": 577
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.11940122268611729,
+      "learning_rate": 0.0009973104668451617,
+      "loss": 1.4156,
+      "step": 578
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1230749446399393,
+      "learning_rate": 0.0009972924023370967,
+      "loss": 1.5258,
+      "step": 579
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.11807230862866323,
+      "learning_rate": 0.0009972742775306133,
+      "loss": 1.5498,
+      "step": 580
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.12698953546566902,
+      "learning_rate": 0.0009972560924279097,
+      "loss": 1.5633,
+      "step": 581
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.12048242833327667,
+      "learning_rate": 0.0009972378470311904,
+      "loss": 1.4828,
+      "step": 582
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.12821890991850848,
+      "learning_rate": 0.0009972195413426679,
+      "loss": 1.3623,
+      "step": 583
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1364893555564914,
+      "learning_rate": 0.000997201175364562,
+      "loss": 1.4673,
+      "step": 584
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.10110545251686494,
+      "learning_rate": 0.0009971827490990993,
+      "loss": 1.5114,
+      "step": 585
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.12792269127038894,
+      "learning_rate": 0.0009971642625485144,
+      "loss": 1.3687,
+      "step": 586
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.11988262596625479,
+      "learning_rate": 0.0009971457157150485,
+      "loss": 1.4693,
+      "step": 587
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.14192927684022144,
+      "learning_rate": 0.0009971271086009507,
+      "loss": 1.5541,
+      "step": 588
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1335038186372788,
+      "learning_rate": 0.0009971084412084771,
+      "loss": 1.3357,
+      "step": 589
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.11027670743622948,
+      "learning_rate": 0.0009970897135398913,
+      "loss": 1.4499,
+      "step": 590
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.10736693798136951,
+      "learning_rate": 0.000997070925597464,
+      "loss": 1.3992,
+      "step": 591
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.10885656905907487,
+      "learning_rate": 0.0009970520773834733,
+      "loss": 1.4657,
+      "step": 592
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.09049890929626538,
+      "learning_rate": 0.0009970331689002046,
+      "loss": 1.4588,
+      "step": 593
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.10742491587756407,
+      "learning_rate": 0.0009970142001499505,
+      "loss": 1.3222,
+      "step": 594
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1267830112432258,
+      "learning_rate": 0.0009969951711350114,
+      "loss": 1.5056,
+      "step": 595
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.11243589034871186,
+      "learning_rate": 0.0009969760818576941,
+      "loss": 1.4724,
+      "step": 596
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1280646485857755,
+      "learning_rate": 0.0009969569323203138,
+      "loss": 1.5029,
+      "step": 597
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.13283166676950728,
+      "learning_rate": 0.000996937722525192,
+      "loss": 1.5652,
+      "step": 598
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.11385817819290642,
+      "learning_rate": 0.0009969184524746585,
+      "loss": 1.3798,
+      "step": 599
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1104480250063874,
+      "learning_rate": 0.000996899122171049,
+      "loss": 1.3786,
+      "step": 600
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1153007090104645,
+      "learning_rate": 0.0009968797316167082,
+      "loss": 1.397,
+      "step": 601
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.12277337551873949,
+      "learning_rate": 0.0009968602808139869,
+      "loss": 1.6086,
+      "step": 602
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.12101584666596928,
+      "learning_rate": 0.0009968407697652434,
+      "loss": 1.4888,
+      "step": 603
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.15500860601806254,
+      "learning_rate": 0.000996821198472844,
+      "loss": 1.5664,
+      "step": 604
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1178828194892715,
+      "learning_rate": 0.0009968015669391613,
+      "loss": 1.5685,
+      "step": 605
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.12335417800873578,
+      "learning_rate": 0.000996781875166576,
+      "loss": 1.5528,
+      "step": 606
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.12233699469045906,
+      "learning_rate": 0.0009967621231574753,
+      "loss": 1.3711,
+      "step": 607
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.11591524885822231,
+      "learning_rate": 0.000996742310914255,
+      "loss": 1.4826,
+      "step": 608
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.14044536601206312,
+      "learning_rate": 0.0009967224384393168,
+      "loss": 1.6158,
+      "step": 609
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1381027639030021,
+      "learning_rate": 0.0009967025057350705,
+      "loss": 1.4215,
+      "step": 610
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.125010089541578,
+      "learning_rate": 0.000996682512803933,
+      "loss": 1.4867,
+      "step": 611
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.12207318684054917,
+      "learning_rate": 0.0009966624596483285,
+      "loss": 1.3042,
+      "step": 612
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.10878370098894088,
+      "learning_rate": 0.0009966423462706884,
+      "loss": 1.4636,
+      "step": 613
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.13328193000271166,
+      "learning_rate": 0.0009966221726734517,
+      "loss": 1.3897,
+      "step": 614
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.11689483535778034,
+      "learning_rate": 0.0009966019388590644,
+      "loss": 1.6064,
+      "step": 615
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1301761197932106,
+      "learning_rate": 0.00099658164482998,
+      "loss": 1.6075,
+      "step": 616
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1346747568866436,
+      "learning_rate": 0.0009965612905886592,
+      "loss": 1.3378,
+      "step": 617
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.12025260657555636,
+      "learning_rate": 0.00099654087613757,
+      "loss": 1.3927,
+      "step": 618
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.14311098764679006,
+      "learning_rate": 0.0009965204014791879,
+      "loss": 1.5443,
+      "step": 619
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.11841687436158872,
+      "learning_rate": 0.0009964998666159952,
+      "loss": 1.538,
+      "step": 620
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.11248266795214713,
+      "learning_rate": 0.000996479271550482,
+      "loss": 1.4871,
+      "step": 621
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.13459548022776915,
+      "learning_rate": 0.0009964586162851455,
+      "loss": 1.3075,
+      "step": 622
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.13005288898830575,
+      "learning_rate": 0.0009964379008224901,
+      "loss": 1.4756,
+      "step": 623
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.12150477440478978,
+      "learning_rate": 0.0009964171251650277,
+      "loss": 1.3786,
+      "step": 624
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.11943429960726923,
+      "learning_rate": 0.0009963962893152778,
+      "loss": 1.4501,
+      "step": 625
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.12382509486585465,
+      "learning_rate": 0.0009963753932757662,
+      "loss": 1.5678,
+      "step": 626
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.15440640002770115,
+      "learning_rate": 0.000996354437049027,
+      "loss": 1.5596,
+      "step": 627
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.11091940708776127,
+      "learning_rate": 0.0009963334206376012,
+      "loss": 1.4617,
+      "step": 628
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.11455971413894141,
+      "learning_rate": 0.0009963123440440368,
+      "loss": 1.3403,
+      "step": 629
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.15444295040707226,
+      "learning_rate": 0.0009962912072708897,
+      "loss": 1.5607,
+      "step": 630
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1059896049882209,
+      "learning_rate": 0.0009962700103207228,
+      "loss": 1.4402,
+      "step": 631
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.10979744315974159,
+      "learning_rate": 0.0009962487531961063,
+      "loss": 1.509,
+      "step": 632
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1271977452069575,
+      "learning_rate": 0.0009962274358996178,
+      "loss": 1.5976,
+      "step": 633
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.11723721165067925,
+      "learning_rate": 0.0009962060584338417,
+      "loss": 1.5683,
+      "step": 634
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.11516730530763199,
+      "learning_rate": 0.0009961846208013704,
+      "loss": 1.4661,
+      "step": 635
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1176706269429394,
+      "learning_rate": 0.0009961631230048032,
+      "loss": 1.4224,
+      "step": 636
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1072376383600733,
+      "learning_rate": 0.0009961415650467467,
+      "loss": 1.5348,
+      "step": 637
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.12294971323557731,
+      "learning_rate": 0.000996119946929815,
+      "loss": 1.6004,
+      "step": 638
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.11965915531049516,
+      "learning_rate": 0.0009960982686566294,
+      "loss": 1.3978,
+      "step": 639
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.09157392129523097,
+      "learning_rate": 0.0009960765302298184,
+      "loss": 1.4615,
+      "step": 640
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.11716155321247809,
+      "learning_rate": 0.0009960547316520182,
+      "loss": 1.5232,
+      "step": 641
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1170791482693603,
+      "learning_rate": 0.0009960328729258711,
+      "loss": 1.4715,
+      "step": 642
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.12259710702230697,
+      "learning_rate": 0.0009960109540540284,
+      "loss": 1.5141,
+      "step": 643
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.12107772105788704,
+      "learning_rate": 0.0009959889750391472,
+      "loss": 1.461,
+      "step": 644
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.10557229425276307,
+      "learning_rate": 0.0009959669358838932,
+      "loss": 1.5069,
+      "step": 645
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.09790643592491724,
+      "learning_rate": 0.0009959448365909384,
+      "loss": 1.4338,
+      "step": 646
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.12391777117899554,
+      "learning_rate": 0.0009959226771629622,
+      "loss": 1.4601,
+      "step": 647
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.11025736594706048,
+      "learning_rate": 0.0009959004576026516,
+      "loss": 1.5652,
+      "step": 648
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.10460029117111654,
+      "learning_rate": 0.000995878177912701,
+      "loss": 1.4367,
+      "step": 649
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.10927989562994969,
+      "learning_rate": 0.0009958558380958116,
+      "loss": 1.4751,
+      "step": 650
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.12173129880732504,
+      "learning_rate": 0.0009958334381546927,
+      "loss": 1.5573,
+      "step": 651
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.12838332720662465,
+      "learning_rate": 0.0009958109780920598,
+      "loss": 1.4785,
+      "step": 652
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.12814839246956414,
+      "learning_rate": 0.0009957884579106363,
+      "loss": 1.5043,
+      "step": 653
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.12270162223081243,
+      "learning_rate": 0.0009957658776131536,
+      "loss": 1.6063,
+      "step": 654
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.13428707781101876,
+      "learning_rate": 0.0009957432372023486,
+      "loss": 1.4086,
+      "step": 655
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.13177426626485697,
+      "learning_rate": 0.000995720536680967,
+      "loss": 1.4597,
+      "step": 656
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.12414797214999132,
+      "learning_rate": 0.0009956977760517615,
+      "loss": 1.4479,
+      "step": 657
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.12911162849346897,
+      "learning_rate": 0.0009956749553174918,
+      "loss": 1.6136,
+      "step": 658
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.12716929664557133,
+      "learning_rate": 0.0009956520744809248,
+      "loss": 1.4265,
+      "step": 659
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1179166264201855,
+      "learning_rate": 0.000995629133544835,
+      "loss": 1.4051,
+      "step": 660
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.14156709119896776,
+      "learning_rate": 0.000995606132512004,
+      "loss": 1.4486,
+      "step": 661
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.14670926950068083,
+      "learning_rate": 0.000995583071385221,
+      "loss": 1.6208,
+      "step": 662
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.10603558357744294,
+      "learning_rate": 0.000995559950167282,
+      "loss": 1.4793,
+      "step": 663
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1224055075034946,
+      "learning_rate": 0.0009955367688609905,
+      "loss": 1.4483,
+      "step": 664
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1395496872898413,
+      "learning_rate": 0.0009955135274691573,
+      "loss": 1.5365,
+      "step": 665
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1079570336167769,
+      "learning_rate": 0.0009954902259946008,
+      "loss": 1.4479,
+      "step": 666
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.13238288662712247,
+      "learning_rate": 0.0009954668644401462,
+      "loss": 1.6067,
+      "step": 667
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.10990400407308812,
+      "learning_rate": 0.0009954434428086259,
+      "loss": 1.4747,
+      "step": 668
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.12106482867993244,
+      "learning_rate": 0.0009954199611028802,
+      "loss": 1.4404,
+      "step": 669
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.10078426132952148,
+      "learning_rate": 0.0009953964193257564,
+      "loss": 1.5387,
+      "step": 670
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.09989641475889491,
+      "learning_rate": 0.0009953728174801088,
+      "loss": 1.2998,
+      "step": 671
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1222248510617004,
+      "learning_rate": 0.0009953491555687991,
+      "loss": 1.5839,
+      "step": 672
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.10030915780605011,
+      "learning_rate": 0.0009953254335946969,
+      "loss": 1.6116,
+      "step": 673
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.10147470567669492,
+      "learning_rate": 0.000995301651560678,
+      "loss": 1.4606,
+      "step": 674
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.12189077525868597,
+      "learning_rate": 0.0009952778094696262,
+      "loss": 1.5718,
+      "step": 675
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.12389923995874418,
+      "learning_rate": 0.0009952539073244326,
+      "loss": 1.5397,
+      "step": 676
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1079829259763348,
+      "learning_rate": 0.0009952299451279954,
+      "loss": 1.5822,
+      "step": 677
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.13025108271421548,
+      "learning_rate": 0.00099520592288322,
+      "loss": 1.3588,
+      "step": 678
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.09252175398506655,
+      "learning_rate": 0.0009951818405930194,
+      "loss": 1.3074,
+      "step": 679
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.12190380184039762,
+      "learning_rate": 0.0009951576982603133,
+      "loss": 1.5712,
+      "step": 680
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.12012031597622391,
+      "learning_rate": 0.0009951334958880292,
+      "loss": 1.5322,
+      "step": 681
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.11291624944914505,
+      "learning_rate": 0.0009951092334791016,
+      "loss": 1.4539,
+      "step": 682
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0931588349385203,
+      "learning_rate": 0.0009950849110364729,
+      "loss": 1.3937,
+      "step": 683
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1053379448621117,
+      "learning_rate": 0.0009950605285630916,
+      "loss": 1.4337,
+      "step": 684
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.10645922517410703,
+      "learning_rate": 0.0009950360860619147,
+      "loss": 1.3934,
+      "step": 685
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.10572018719307143,
+      "learning_rate": 0.0009950115835359054,
+      "loss": 1.4284,
+      "step": 686
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.09603500722299851,
+      "learning_rate": 0.0009949870209880354,
+      "loss": 1.4436,
+      "step": 687
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1171822427524059,
+      "learning_rate": 0.0009949623984212824,
+      "loss": 1.4438,
+      "step": 688
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.10775144263060174,
+      "learning_rate": 0.0009949377158386323,
+      "loss": 1.5625,
+      "step": 689
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.11412434036504529,
+      "learning_rate": 0.0009949129732430778,
+      "loss": 1.501,
+      "step": 690
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.10324234472443945,
+      "learning_rate": 0.0009948881706376192,
+      "loss": 1.6314,
+      "step": 691
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.12233091194141885,
+      "learning_rate": 0.0009948633080252636,
+      "loss": 1.5398,
+      "step": 692
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.10981153329077435,
+      "learning_rate": 0.000994838385409026,
+      "loss": 1.5315,
+      "step": 693
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.11124117283144365,
+      "learning_rate": 0.000994813402791928,
+      "loss": 1.468,
+      "step": 694
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.10859983867961821,
+      "learning_rate": 0.000994788360176999,
+      "loss": 1.4983,
+      "step": 695
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.10679836737127918,
+      "learning_rate": 0.0009947632575672757,
+      "loss": 1.509,
+      "step": 696
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1452761412325607,
+      "learning_rate": 0.0009947380949658017,
+      "loss": 1.5244,
+      "step": 697
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.10376320328347177,
+      "learning_rate": 0.0009947128723756281,
+      "loss": 1.4568,
+      "step": 698
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.13737925652052532,
+      "learning_rate": 0.0009946875897998131,
+      "loss": 1.5241,
+      "step": 699
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.12421770725172457,
+      "learning_rate": 0.0009946622472414223,
+      "loss": 1.5322,
+      "step": 700
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.12125159942786252,
+      "learning_rate": 0.000994636844703529,
+      "loss": 1.3998,
+      "step": 701
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.13120907346940802,
+      "learning_rate": 0.0009946113821892128,
+      "loss": 1.4633,
+      "step": 702
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.15391223685243563,
+      "learning_rate": 0.0009945858597015613,
+      "loss": 1.4451,
+      "step": 703
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1360647191323169,
+      "learning_rate": 0.0009945602772436692,
+      "loss": 1.4341,
+      "step": 704
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.10529263459373553,
+      "learning_rate": 0.0009945346348186385,
+      "loss": 1.4727,
+      "step": 705
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.11431211623838632,
+      "learning_rate": 0.0009945089324295785,
+      "loss": 1.4321,
+      "step": 706
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.13598904701131048,
+      "learning_rate": 0.0009944831700796053,
+      "loss": 1.5741,
+      "step": 707
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1355862716894886,
+      "learning_rate": 0.0009944573477718435,
+      "loss": 1.4294,
+      "step": 708
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.12494946367200017,
+      "learning_rate": 0.0009944314655094232,
+      "loss": 1.4831,
+      "step": 709
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.11809725685709474,
+      "learning_rate": 0.0009944055232954832,
+      "loss": 1.4706,
+      "step": 710
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1198085117849122,
+      "learning_rate": 0.0009943795211331693,
+      "loss": 1.579,
+      "step": 711
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.13247770297281755,
+      "learning_rate": 0.0009943534590256338,
+      "loss": 1.5206,
+      "step": 712
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.11403899759015178,
+      "learning_rate": 0.000994327336976037,
+      "loss": 1.5245,
+      "step": 713
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1273405472522352,
+      "learning_rate": 0.0009943011549875466,
+      "loss": 1.4184,
+      "step": 714
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1358059212062907,
+      "learning_rate": 0.000994274913063337,
+      "loss": 1.5701,
+      "step": 715
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.10773623159909808,
+      "learning_rate": 0.0009942486112065902,
+      "loss": 1.5458,
+      "step": 716
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.11792824796737497,
+      "learning_rate": 0.0009942222494204954,
+      "loss": 1.5711,
+      "step": 717
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.11176245858069334,
+      "learning_rate": 0.000994195827708249,
+      "loss": 1.4227,
+      "step": 718
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.11809148806007971,
+      "learning_rate": 0.0009941693460730548,
+      "loss": 1.4618,
+      "step": 719
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.12523379292415296,
+      "learning_rate": 0.0009941428045181235,
+      "loss": 1.4765,
+      "step": 720
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.11416568669285194,
+      "learning_rate": 0.0009941162030466736,
+      "loss": 1.4385,
+      "step": 721
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.11968357238475509,
+      "learning_rate": 0.0009940895416619307,
+      "loss": 1.4373,
+      "step": 722
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.15665080698553804,
+      "learning_rate": 0.0009940628203671275,
+      "loss": 1.5316,
+      "step": 723
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.10064658632776849,
+      "learning_rate": 0.0009940360391655042,
+      "loss": 1.5352,
+      "step": 724
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.10251226529146976,
+      "learning_rate": 0.0009940091980603076,
+      "loss": 1.4791,
+      "step": 725
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.10378073760870751,
+      "learning_rate": 0.000993982297054793,
+      "loss": 1.5029,
+      "step": 726
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.10539285260130456,
+      "learning_rate": 0.0009939553361522217,
+      "loss": 1.4734,
+      "step": 727
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1100171833779912,
+      "learning_rate": 0.0009939283153558628,
+      "loss": 1.353,
+      "step": 728
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.11572299828458692,
+      "learning_rate": 0.000993901234668993,
+      "loss": 1.4491,
+      "step": 729
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.10411765589679892,
+      "learning_rate": 0.0009938740940948958,
+      "loss": 1.4687,
+      "step": 730
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1213664111538291,
+      "learning_rate": 0.0009938468936368618,
+      "loss": 1.4684,
+      "step": 731
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.11803576700547622,
+      "learning_rate": 0.0009938196332981897,
+      "loss": 1.5678,
+      "step": 732
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1189255355811123,
+      "learning_rate": 0.0009937923130821844,
+      "loss": 1.4901,
+      "step": 733
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1140209620430202,
+      "learning_rate": 0.0009937649329921589,
+      "loss": 1.3862,
+      "step": 734
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.12863941643968582,
+      "learning_rate": 0.000993737493031433,
+      "loss": 1.4651,
+      "step": 735
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.11018747815379674,
+      "learning_rate": 0.0009937099932033338,
+      "loss": 1.4941,
+      "step": 736
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.11099330218951524,
+      "learning_rate": 0.000993682433511196,
+      "loss": 1.5038,
+      "step": 737
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.11147184255727886,
+      "learning_rate": 0.000993654813958361,
+      "loss": 1.4039,
+      "step": 738
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.10734348471301973,
+      "learning_rate": 0.000993627134548178,
+      "loss": 1.4491,
+      "step": 739
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.10710535668640941,
+      "learning_rate": 0.0009935993952840032,
+      "loss": 1.39,
+      "step": 740
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.13094386533114225,
+      "learning_rate": 0.0009935715961692,
+      "loss": 1.5734,
+      "step": 741
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.12488034601195075,
+      "learning_rate": 0.000993543737207139,
+      "loss": 1.5388,
+      "step": 742
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.12403785629260212,
+      "learning_rate": 0.0009935158184011985,
+      "loss": 1.2961,
+      "step": 743
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.11652789448505044,
+      "learning_rate": 0.0009934878397547635,
+      "loss": 1.4781,
+      "step": 744
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.11516732956770669,
+      "learning_rate": 0.0009934598012712268,
+      "loss": 1.4467,
+      "step": 745
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.12245986579048859,
+      "learning_rate": 0.000993431702953988,
+      "loss": 1.4653,
+      "step": 746
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.10983200899730307,
+      "learning_rate": 0.0009934035448064538,
+      "loss": 1.5349,
+      "step": 747
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.09167922119798957,
+      "learning_rate": 0.000993375326832039,
+      "loss": 1.4424,
+      "step": 748
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1069490038423996,
+      "learning_rate": 0.0009933470490341649,
+      "loss": 1.447,
+      "step": 749
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.11574964570738946,
+      "learning_rate": 0.0009933187114162603,
+      "loss": 1.4137,
+      "step": 750
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.13373086874940893,
+      "learning_rate": 0.0009932903139817611,
+      "loss": 1.4324,
+      "step": 751
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.12154833008669642,
+      "learning_rate": 0.0009932618567341108,
+      "loss": 1.4614,
+      "step": 752
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.12863675940870534,
+      "learning_rate": 0.0009932333396767596,
+      "loss": 1.6192,
+      "step": 753
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.12602366816139124,
+      "learning_rate": 0.0009932047628131659,
+      "loss": 1.4902,
+      "step": 754
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.11828428253206899,
+      "learning_rate": 0.000993176126146794,
+      "loss": 1.5057,
+      "step": 755
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.11447959551486472,
+      "learning_rate": 0.0009931474296811169,
+      "loss": 1.4539,
+      "step": 756
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.11652175026432009,
+      "learning_rate": 0.0009931186734196136,
+      "loss": 1.5865,
+      "step": 757
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.15953279308615217,
+      "learning_rate": 0.0009930898573657712,
+      "loss": 1.5144,
+      "step": 758
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.11445487382353405,
+      "learning_rate": 0.0009930609815230838,
+      "loss": 1.5207,
+      "step": 759
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.09523137267014015,
+      "learning_rate": 0.0009930320458950523,
+      "loss": 1.4406,
+      "step": 760
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.11500133371745935,
+      "learning_rate": 0.0009930030504851856,
+      "loss": 1.5745,
+      "step": 761
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.11011675099438246,
+      "learning_rate": 0.0009929739952969994,
+      "loss": 1.4388,
+      "step": 762
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.11046935647517428,
+      "learning_rate": 0.0009929448803340166,
+      "loss": 1.5235,
+      "step": 763
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.11679227263335894,
+      "learning_rate": 0.0009929157055997677,
+      "loss": 1.4837,
+      "step": 764
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.10443444082409932,
+      "learning_rate": 0.0009928864710977902,
+      "loss": 1.5236,
+      "step": 765
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.11647164928646948,
+      "learning_rate": 0.0009928571768316288,
+      "loss": 1.4907,
+      "step": 766
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.10566648958207446,
+      "learning_rate": 0.0009928278228048357,
+      "loss": 1.4907,
+      "step": 767
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.09826190297521414,
+      "learning_rate": 0.00099279840902097,
+      "loss": 1.4079,
+      "step": 768
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1017904451693403,
+      "learning_rate": 0.0009927689354835981,
+      "loss": 1.4102,
+      "step": 769
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.09993971482709918,
+      "learning_rate": 0.0009927394021962941,
+      "loss": 1.6505,
+      "step": 770
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.10591663082599534,
+      "learning_rate": 0.0009927098091626388,
+      "loss": 1.4566,
+      "step": 771
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.11498870565123132,
+      "learning_rate": 0.0009926801563862205,
+      "loss": 1.6038,
+      "step": 772
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.11407235573294057,
+      "learning_rate": 0.0009926504438706348,
+      "loss": 1.5839,
+      "step": 773
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.12308677147643597,
+      "learning_rate": 0.0009926206716194842,
+      "loss": 1.5441,
+      "step": 774
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.10318309978776286,
+      "learning_rate": 0.0009925908396363789,
+      "loss": 1.5983,
+      "step": 775
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.10265958152079739,
+      "learning_rate": 0.0009925609479249362,
+      "loss": 1.467,
+      "step": 776
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1041780857939972,
+      "learning_rate": 0.0009925309964887803,
+      "loss": 1.3516,
+      "step": 777
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1127904379754784,
+      "learning_rate": 0.0009925009853315432,
+      "loss": 1.3867,
+      "step": 778
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.10759215513795695,
+      "learning_rate": 0.0009924709144568636,
+      "loss": 1.4878,
+      "step": 779
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.10524365237305855,
+      "learning_rate": 0.0009924407838683878,
+      "loss": 1.4779,
+      "step": 780
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.09420759561249992,
+      "learning_rate": 0.0009924105935697693,
+      "loss": 1.475,
+      "step": 781
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.10835188218183127,
+      "learning_rate": 0.0009923803435646689,
+      "loss": 1.5956,
+      "step": 782
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.10517000343653546,
+      "learning_rate": 0.0009923500338567541,
+      "loss": 1.4006,
+      "step": 783
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.10588943505093504,
+      "learning_rate": 0.0009923196644497003,
+      "loss": 1.4583,
+      "step": 784
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.09840148056749254,
+      "learning_rate": 0.00099228923534719,
+      "loss": 1.5177,
+      "step": 785
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.087873993730566,
+      "learning_rate": 0.0009922587465529128,
+      "loss": 1.4561,
+      "step": 786
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.10730475413220047,
+      "learning_rate": 0.0009922281980705653,
+      "loss": 1.4842,
+      "step": 787
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.09819906057900965,
+      "learning_rate": 0.0009921975899038521,
+      "loss": 1.443,
+      "step": 788
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.10673849388513212,
+      "learning_rate": 0.0009921669220564842,
+      "loss": 1.4846,
+      "step": 789
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.09585259890060156,
+      "learning_rate": 0.00099213619453218,
+      "loss": 1.4982,
+      "step": 790
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.10886473711799136,
+      "learning_rate": 0.0009921054073346659,
+      "loss": 1.5461,
+      "step": 791
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.11529970337773958,
+      "learning_rate": 0.0009920745604676743,
+      "loss": 1.483,
+      "step": 792
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1049014082110331,
+      "learning_rate": 0.0009920436539349459,
+      "loss": 1.428,
+      "step": 793
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.11341000507473294,
+      "learning_rate": 0.0009920126877402283,
+      "loss": 1.4127,
+      "step": 794
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0972092028209167,
+      "learning_rate": 0.000991981661887276,
+      "loss": 1.5701,
+      "step": 795
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.10979701692895823,
+      "learning_rate": 0.0009919505763798509,
+      "loss": 1.4882,
+      "step": 796
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.12347853346750738,
+      "learning_rate": 0.0009919194312217226,
+      "loss": 1.5197,
+      "step": 797
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.12862095545623062,
+      "learning_rate": 0.0009918882264166671,
+      "loss": 1.4307,
+      "step": 798
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.11011053871767552,
+      "learning_rate": 0.0009918569619684686,
+      "loss": 1.5358,
+      "step": 799
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.10923348365641565,
+      "learning_rate": 0.0009918256378809178,
+      "loss": 1.3628,
+      "step": 800
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.10848730542687637,
+      "learning_rate": 0.0009917942541578127,
+      "loss": 1.4926,
+      "step": 801
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.09815586793245867,
+      "learning_rate": 0.0009917628108029588,
+      "loss": 1.4147,
+      "step": 802
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.12150522280091537,
+      "learning_rate": 0.000991731307820169,
+      "loss": 1.3304,
+      "step": 803
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.10961537259089094,
+      "learning_rate": 0.0009916997452132624,
+      "loss": 1.4825,
+      "step": 804
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.11017075427928086,
+      "learning_rate": 0.0009916681229860669,
+      "loss": 1.5064,
+      "step": 805
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.10544774278956268,
+      "learning_rate": 0.0009916364411424164,
+      "loss": 1.4638,
+      "step": 806
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.121938065418309,
+      "learning_rate": 0.0009916046996861522,
+      "loss": 1.4857,
+      "step": 807
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.12330109961033686,
+      "learning_rate": 0.0009915728986211237,
+      "loss": 1.3688,
+      "step": 808
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1371285563688456,
+      "learning_rate": 0.0009915410379511865,
+      "loss": 1.5859,
+      "step": 809
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1053850031308369,
+      "learning_rate": 0.0009915091176802035,
+      "loss": 1.3979,
+      "step": 810
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.10649860867690676,
+      "learning_rate": 0.000991477137812046,
+      "loss": 1.5179,
+      "step": 811
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.13922844265421294,
+      "learning_rate": 0.0009914450983505908,
+      "loss": 1.3672,
+      "step": 812
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1287741960027465,
+      "learning_rate": 0.0009914129992997232,
+      "loss": 1.465,
+      "step": 813
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.10611709028639908,
+      "learning_rate": 0.0009913808406633354,
+      "loss": 1.568,
+      "step": 814
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.10666038719575555,
+      "learning_rate": 0.0009913486224453266,
+      "loss": 1.4613,
+      "step": 815
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.09552780948354576,
+      "learning_rate": 0.0009913163446496032,
+      "loss": 1.3986,
+      "step": 816
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.10458514944773992,
+      "learning_rate": 0.0009912840072800796,
+      "loss": 1.4997,
+      "step": 817
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.10781289805463203,
+      "learning_rate": 0.0009912516103406759,
+      "loss": 1.4163,
+      "step": 818
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.11066849525641222,
+      "learning_rate": 0.0009912191538353212,
+      "loss": 1.3647,
+      "step": 819
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.11037102460141801,
+      "learning_rate": 0.0009911866377679505,
+      "loss": 1.4986,
+      "step": 820
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.10550276331452135,
+      "learning_rate": 0.0009911540621425066,
+      "loss": 1.4142,
+      "step": 821
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.09357877084866194,
+      "learning_rate": 0.0009911214269629394,
+      "loss": 1.4046,
+      "step": 822
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.10821731143868805,
+      "learning_rate": 0.000991088732233206,
+      "loss": 1.4454,
+      "step": 823
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.10885502691894844,
+      "learning_rate": 0.0009910559779572707,
+      "loss": 1.4416,
+      "step": 824
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.09442699393531002,
+      "learning_rate": 0.0009910231641391052,
+      "loss": 1.5394,
+      "step": 825
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.10525545876158422,
+      "learning_rate": 0.0009909902907826883,
+      "loss": 1.546,
+      "step": 826
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.11720992112248244,
+      "learning_rate": 0.000990957357892006,
+      "loss": 1.5321,
+      "step": 827
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.10864424543494355,
+      "learning_rate": 0.0009909243654710514,
+      "loss": 1.4025,
+      "step": 828
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1117918912175246,
+      "learning_rate": 0.000990891313523825,
+      "loss": 1.692,
+      "step": 829
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1313258688345413,
+      "learning_rate": 0.0009908582020543345,
+      "loss": 1.5812,
+      "step": 830
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.11780429486456513,
+      "learning_rate": 0.0009908250310665947,
+      "loss": 1.556,
+      "step": 831
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.11012726750810582,
+      "learning_rate": 0.0009907918005646276,
+      "loss": 1.4705,
+      "step": 832
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.10699556897583314,
+      "learning_rate": 0.000990758510552463,
+      "loss": 1.4487,
+      "step": 833
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1012229677196335,
+      "learning_rate": 0.0009907251610341368,
+      "loss": 1.5403,
+      "step": 834
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1236371026793818,
+      "learning_rate": 0.0009906917520136932,
+      "loss": 1.4242,
+      "step": 835
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.12682607602434304,
+      "learning_rate": 0.000990658283495183,
+      "loss": 1.4958,
+      "step": 836
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1035656380844131,
+      "learning_rate": 0.0009906247554826643,
+      "loss": 1.4714,
+      "step": 837
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.10153729327718629,
+      "learning_rate": 0.0009905911679802027,
+      "loss": 1.4798,
+      "step": 838
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.09357476745648424,
+      "learning_rate": 0.0009905575209918705,
+      "loss": 1.5151,
+      "step": 839
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.10110221465858242,
+      "learning_rate": 0.0009905238145217477,
+      "loss": 1.5721,
+      "step": 840
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.09581386231874159,
+      "learning_rate": 0.0009904900485739213,
+      "loss": 1.5697,
+      "step": 841
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.11136596229104052,
+      "learning_rate": 0.0009904562231524857,
+      "loss": 1.5124,
+      "step": 842
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.10563905529316776,
+      "learning_rate": 0.0009904223382615417,
+      "loss": 1.5153,
+      "step": 843
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.105291268018836,
+      "learning_rate": 0.000990388393905199,
+      "loss": 1.5121,
+      "step": 844
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.13132650708524377,
+      "learning_rate": 0.0009903543900875726,
+      "loss": 1.565,
+      "step": 845
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.11176225233742272,
+      "learning_rate": 0.000990320326812786,
+      "loss": 1.5904,
+      "step": 846
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.09906334204646819,
+      "learning_rate": 0.0009902862040849694,
+      "loss": 1.434,
+      "step": 847
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.10512336498042481,
+      "learning_rate": 0.0009902520219082602,
+      "loss": 1.4771,
+      "step": 848
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.10922470793825259,
+      "learning_rate": 0.0009902177802868033,
+      "loss": 1.5708,
+      "step": 849
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1013422099157301,
+      "learning_rate": 0.0009901834792247503,
+      "loss": 1.543,
+      "step": 850
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.11685308384489977,
+      "learning_rate": 0.0009901491187262609,
+      "loss": 1.5184,
+      "step": 851
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1068500340383022,
+      "learning_rate": 0.0009901146987955007,
+      "loss": 1.3863,
+      "step": 852
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.10535661820687749,
+      "learning_rate": 0.0009900802194366437,
+      "loss": 1.4796,
+      "step": 853
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.11783230529707762,
+      "learning_rate": 0.0009900456806538707,
+      "loss": 1.3684,
+      "step": 854
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.11384519631304521,
+      "learning_rate": 0.0009900110824513691,
+      "loss": 1.4558,
+      "step": 855
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.10264058083146971,
+      "learning_rate": 0.0009899764248333348,
+      "loss": 1.434,
+      "step": 856
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.11519712314754389,
+      "learning_rate": 0.0009899417078039696,
+      "loss": 1.4387,
+      "step": 857
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1123372880638399,
+      "learning_rate": 0.0009899069313674832,
+      "loss": 1.4141,
+      "step": 858
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1097586324100307,
+      "learning_rate": 0.0009898720955280925,
+      "loss": 1.5496,
+      "step": 859
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.11048300266047464,
+      "learning_rate": 0.0009898372002900213,
+      "loss": 1.3956,
+      "step": 860
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.10667135360093603,
+      "learning_rate": 0.000989802245657501,
+      "loss": 1.4431,
+      "step": 861
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.10169688375776761,
+      "learning_rate": 0.0009897672316347696,
+      "loss": 1.4944,
+      "step": 862
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.10805112813479695,
+      "learning_rate": 0.000989732158226073,
+      "loss": 1.5502,
+      "step": 863
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.11383346267574627,
+      "learning_rate": 0.0009896970254356637,
+      "loss": 1.5222,
+      "step": 864
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.10437239928845843,
+      "learning_rate": 0.000989661833267802,
+      "loss": 1.557,
+      "step": 865
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.12880169356158996,
+      "learning_rate": 0.0009896265817267548,
+      "loss": 1.6166,
+      "step": 866
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.12070472393705185,
+      "learning_rate": 0.0009895912708167967,
+      "loss": 1.5124,
+      "step": 867
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.12068593858069836,
+      "learning_rate": 0.000989555900542209,
+      "loss": 1.3678,
+      "step": 868
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.11386637597424645,
+      "learning_rate": 0.0009895204709072806,
+      "loss": 1.3299,
+      "step": 869
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.12175046980633066,
+      "learning_rate": 0.0009894849819163075,
+      "loss": 1.5199,
+      "step": 870
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.10357307243540072,
+      "learning_rate": 0.000989449433573593,
+      "loss": 1.4439,
+      "step": 871
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.11587763922735016,
+      "learning_rate": 0.000989413825883447,
+      "loss": 1.424,
+      "step": 872
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.10361221368654966,
+      "learning_rate": 0.0009893781588501875,
+      "loss": 1.5112,
+      "step": 873
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1307499271286858,
+      "learning_rate": 0.000989342432478139,
+      "loss": 1.467,
+      "step": 874
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.10556346561030903,
+      "learning_rate": 0.0009893066467716336,
+      "loss": 1.6212,
+      "step": 875
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.13185572299133208,
+      "learning_rate": 0.0009892708017350104,
+      "loss": 1.4886,
+      "step": 876
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.11593000129396568,
+      "learning_rate": 0.0009892348973726157,
+      "loss": 1.4242,
+      "step": 877
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.10263973471205634,
+      "learning_rate": 0.0009891989336888033,
+      "loss": 1.4018,
+      "step": 878
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.09388484545466252,
+      "learning_rate": 0.0009891629106879333,
+      "loss": 1.5406,
+      "step": 879
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.10579398130719597,
+      "learning_rate": 0.0009891268283743742,
+      "loss": 1.4365,
+      "step": 880
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.10205073561963973,
+      "learning_rate": 0.0009890906867525008,
+      "loss": 1.36,
+      "step": 881
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0993925966084655,
+      "learning_rate": 0.0009890544858266953,
+      "loss": 1.5417,
+      "step": 882
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.12462405391860587,
+      "learning_rate": 0.0009890182256013476,
+      "loss": 1.4766,
+      "step": 883
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.11871289446812809,
+      "learning_rate": 0.0009889819060808541,
+      "loss": 1.3472,
+      "step": 884
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.10788692034328204,
+      "learning_rate": 0.0009889455272696186,
+      "loss": 1.5064,
+      "step": 885
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.10830663754630337,
+      "learning_rate": 0.0009889090891720524,
+      "loss": 1.4562,
+      "step": 886
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1020377985895941,
+      "learning_rate": 0.0009888725917925735,
+      "loss": 1.4743,
+      "step": 887
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.11402004313995222,
+      "learning_rate": 0.0009888360351356076,
+      "loss": 1.5885,
+      "step": 888
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.10404694253312369,
+      "learning_rate": 0.0009887994192055872,
+      "loss": 1.4231,
+      "step": 889
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.11522811129082161,
+      "learning_rate": 0.0009887627440069518,
+      "loss": 1.4904,
+      "step": 890
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.11120179814086706,
+      "learning_rate": 0.0009887260095441488,
+      "loss": 1.4243,
+      "step": 891
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.11751114935366559,
+      "learning_rate": 0.0009886892158216323,
+      "loss": 1.4856,
+      "step": 892
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.09387632888638782,
+      "learning_rate": 0.0009886523628438635,
+      "loss": 1.4743,
+      "step": 893
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1089212935769675,
+      "learning_rate": 0.0009886154506153113,
+      "loss": 1.4551,
+      "step": 894
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.10504625831674352,
+      "learning_rate": 0.000988578479140451,
+      "loss": 1.5762,
+      "step": 895
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.11992942665130872,
+      "learning_rate": 0.0009885414484237657,
+      "loss": 1.5388,
+      "step": 896
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.12013797851445274,
+      "learning_rate": 0.0009885043584697457,
+      "loss": 1.3869,
+      "step": 897
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1055384732184131,
+      "learning_rate": 0.000988467209282888,
+      "loss": 1.4898,
+      "step": 898
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.10227730583423617,
+      "learning_rate": 0.000988430000867697,
+      "loss": 1.4475,
+      "step": 899
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.11746558731715007,
+      "learning_rate": 0.0009883927332286846,
+      "loss": 1.5964,
+      "step": 900
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.11176430850331567,
+      "learning_rate": 0.0009883554063703697,
+      "loss": 1.4568,
+      "step": 901
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.10878283844133003,
+      "learning_rate": 0.0009883180202972781,
+      "loss": 1.5068,
+      "step": 902
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1241270556094801,
+      "learning_rate": 0.0009882805750139432,
+      "loss": 1.5213,
+      "step": 903
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.11725153365662314,
+      "learning_rate": 0.000988243070524905,
+      "loss": 1.5395,
+      "step": 904
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.11545711343200074,
+      "learning_rate": 0.0009882055068347114,
+      "loss": 1.5103,
+      "step": 905
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.10455006262284446,
+      "learning_rate": 0.000988167883947917,
+      "loss": 1.4557,
+      "step": 906
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.10220852893232839,
+      "learning_rate": 0.0009881302018690833,
+      "loss": 1.4858,
+      "step": 907
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0935560395575124,
+      "learning_rate": 0.0009880924606027802,
+      "loss": 1.5,
+      "step": 908
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.10430753292676732,
+      "learning_rate": 0.0009880546601535834,
+      "loss": 1.4719,
+      "step": 909
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.10808182563791133,
+      "learning_rate": 0.0009880168005260766,
+      "loss": 1.4541,
+      "step": 910
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1250933395522429,
+      "learning_rate": 0.0009879788817248503,
+      "loss": 1.5065,
+      "step": 911
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1072916390151232,
+      "learning_rate": 0.000987940903754502,
+      "loss": 1.456,
+      "step": 912
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.09340424157858447,
+      "learning_rate": 0.0009879028666196373,
+      "loss": 1.5287,
+      "step": 913
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.11313097745359296,
+      "learning_rate": 0.0009878647703248677,
+      "loss": 1.5945,
+      "step": 914
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.09626876557166207,
+      "learning_rate": 0.0009878266148748128,
+      "loss": 1.5226,
+      "step": 915
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1136594247718088,
+      "learning_rate": 0.000987788400274099,
+      "loss": 1.3699,
+      "step": 916
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.09857785705146646,
+      "learning_rate": 0.0009877501265273603,
+      "loss": 1.4187,
+      "step": 917
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.09942441719629294,
+      "learning_rate": 0.000987711793639237,
+      "loss": 1.5443,
+      "step": 918
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.11513364964116303,
+      "learning_rate": 0.0009876734016143773,
+      "loss": 1.5316,
+      "step": 919
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.10223246559180174,
+      "learning_rate": 0.0009876349504574365,
+      "loss": 1.5269,
+      "step": 920
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.09292632355039271,
+      "learning_rate": 0.000987596440173077,
+      "loss": 1.3696,
+      "step": 921
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.11279813746379812,
+      "learning_rate": 0.0009875578707659676,
+      "loss": 1.4942,
+      "step": 922
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.09788112736253952,
+      "learning_rate": 0.0009875192422407859,
+      "loss": 1.418,
+      "step": 923
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.09913455232850231,
+      "learning_rate": 0.0009874805546022153,
+      "loss": 1.5017,
+      "step": 924
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.10018718019093535,
+      "learning_rate": 0.0009874418078549467,
+      "loss": 1.4764,
+      "step": 925
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.10335561243043875,
+      "learning_rate": 0.0009874030020036787,
+      "loss": 1.4936,
+      "step": 926
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1014320149701531,
+      "learning_rate": 0.0009873641370531162,
+      "loss": 1.4366,
+      "step": 927
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.11382615640806051,
+      "learning_rate": 0.0009873252130079718,
+      "loss": 1.527,
+      "step": 928
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.09703283161987435,
+      "learning_rate": 0.0009872862298729653,
+      "loss": 1.3459,
+      "step": 929
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.12078726730680549,
+      "learning_rate": 0.0009872471876528235,
+      "loss": 1.4748,
+      "step": 930
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.13779708131618348,
+      "learning_rate": 0.0009872080863522806,
+      "loss": 1.4424,
+      "step": 931
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.09662842113778519,
+      "learning_rate": 0.0009871689259760771,
+      "loss": 1.506,
+      "step": 932
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.10470466535314994,
+      "learning_rate": 0.0009871297065289623,
+      "loss": 1.4008,
+      "step": 933
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.09825370182641988,
+      "learning_rate": 0.000987090428015691,
+      "loss": 1.457,
+      "step": 934
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.10128986233646643,
+      "learning_rate": 0.000987051090441026,
+      "loss": 1.5307,
+      "step": 935
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.10257667805231513,
+      "learning_rate": 0.0009870116938097374,
+      "loss": 1.5253,
+      "step": 936
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.13212406497778306,
+      "learning_rate": 0.0009869722381266016,
+      "loss": 1.4041,
+      "step": 937
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.09740066939177527,
+      "learning_rate": 0.0009869327233964032,
+      "loss": 1.4791,
+      "step": 938
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.09993232471086905,
+      "learning_rate": 0.0009868931496239334,
+      "loss": 1.3653,
+      "step": 939
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.10008462495414401,
+      "learning_rate": 0.0009868535168139907,
+      "loss": 1.3846,
+      "step": 940
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.13685743246388624,
+      "learning_rate": 0.0009868138249713805,
+      "loss": 1.4776,
+      "step": 941
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.09142847020327265,
+      "learning_rate": 0.0009867740741009159,
+      "loss": 1.3617,
+      "step": 942
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.11429730776068008,
+      "learning_rate": 0.0009867342642074165,
+      "loss": 1.4035,
+      "step": 943
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.12908152202529422,
+      "learning_rate": 0.0009866943952957096,
+      "loss": 1.4352,
+      "step": 944
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.12166464155376931,
+      "learning_rate": 0.0009866544673706294,
+      "loss": 1.3091,
+      "step": 945
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.11394285539064337,
+      "learning_rate": 0.0009866144804370172,
+      "loss": 1.418,
+      "step": 946
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.11662230091559467,
+      "learning_rate": 0.0009865744344997216,
+      "loss": 1.4271,
+      "step": 947
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.10758530149575866,
+      "learning_rate": 0.0009865343295635985,
+      "loss": 1.4356,
+      "step": 948
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.10608808920552162,
+      "learning_rate": 0.0009864941656335105,
+      "loss": 1.4282,
+      "step": 949
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.10309934585985372,
+      "learning_rate": 0.0009864539427143278,
+      "loss": 1.514,
+      "step": 950
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.10917924921838347,
+      "learning_rate": 0.0009864136608109272,
+      "loss": 1.4121,
+      "step": 951
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.11296901752973143,
+      "learning_rate": 0.0009863733199281938,
+      "loss": 1.4354,
+      "step": 952
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.12866370183838677,
+      "learning_rate": 0.0009863329200710182,
+      "loss": 1.5468,
+      "step": 953
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1286966396951818,
+      "learning_rate": 0.0009862924612442994,
+      "loss": 1.4508,
+      "step": 954
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.13960278705237253,
+      "learning_rate": 0.0009862519434529434,
+      "loss": 1.538,
+      "step": 955
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.11652432188490669,
+      "learning_rate": 0.0009862113667018627,
+      "loss": 1.5395,
+      "step": 956
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.10072364916856659,
+      "learning_rate": 0.0009861707309959777,
+      "loss": 1.4169,
+      "step": 957
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.12757051256843824,
+      "learning_rate": 0.0009861300363402153,
+      "loss": 1.4389,
+      "step": 958
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.111629121012831,
+      "learning_rate": 0.0009860892827395103,
+      "loss": 1.598,
+      "step": 959
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.12919976312893156,
+      "learning_rate": 0.0009860484701988037,
+      "loss": 1.4479,
+      "step": 960
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.11279528892079407,
+      "learning_rate": 0.0009860075987230446,
+      "loss": 1.5771,
+      "step": 961
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1278424802958141,
+      "learning_rate": 0.0009859666683171885,
+      "loss": 1.5236,
+      "step": 962
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.11501125985783359,
+      "learning_rate": 0.0009859256789861986,
+      "loss": 1.4961,
+      "step": 963
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.09565466997074817,
+      "learning_rate": 0.000985884630735045,
+      "loss": 1.4487,
+      "step": 964
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.09876986226137081,
+      "learning_rate": 0.0009858435235687044,
+      "loss": 1.617,
+      "step": 965
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.10299888552267045,
+      "learning_rate": 0.000985802357492162,
+      "loss": 1.4978,
+      "step": 966
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.11351964792472877,
+      "learning_rate": 0.0009857611325104088,
+      "loss": 1.4704,
+      "step": 967
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.11014860613660067,
+      "learning_rate": 0.0009857198486284435,
+      "loss": 1.5225,
+      "step": 968
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.10712646048648043,
+      "learning_rate": 0.0009856785058512721,
+      "loss": 1.5689,
+      "step": 969
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.12399232842174956,
+      "learning_rate": 0.0009856371041839075,
+      "loss": 1.5708,
+      "step": 970
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.09681065676026035,
+      "learning_rate": 0.00098559564363137,
+      "loss": 1.6005,
+      "step": 971
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.10178011038545998,
+      "learning_rate": 0.0009855541241986863,
+      "loss": 1.512,
+      "step": 972
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.12471617013219558,
+      "learning_rate": 0.000985512545890891,
+      "loss": 1.6852,
+      "step": 973
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.10686030336674802,
+      "learning_rate": 0.000985470908713026,
+      "loss": 1.4567,
+      "step": 974
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.10697882274944155,
+      "learning_rate": 0.0009854292126701397,
+      "loss": 1.5434,
+      "step": 975
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.10325976456318522,
+      "learning_rate": 0.0009853874577672875,
+      "loss": 1.4225,
+      "step": 976
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.09506887103679255,
+      "learning_rate": 0.0009853456440095327,
+      "loss": 1.4139,
+      "step": 977
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.09578804705306913,
+      "learning_rate": 0.0009853037714019454,
+      "loss": 1.4377,
+      "step": 978
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.09398835151976992,
+      "learning_rate": 0.000985261839949603,
+      "loss": 1.5171,
+      "step": 979
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1143179413310966,
+      "learning_rate": 0.0009852198496575894,
+      "loss": 1.5511,
+      "step": 980
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.10145389270154191,
+      "learning_rate": 0.000985177800530996,
+      "loss": 1.4236,
+      "step": 981
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.08855066022940801,
+      "learning_rate": 0.0009851356925749217,
+      "loss": 1.428,
+      "step": 982
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.10133105738476883,
+      "learning_rate": 0.0009850935257944722,
+      "loss": 1.3685,
+      "step": 983
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.10335410310098274,
+      "learning_rate": 0.0009850513001947604,
+      "loss": 1.464,
+      "step": 984
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.09331951297898144,
+      "learning_rate": 0.0009850090157809061,
+      "loss": 1.4766,
+      "step": 985
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0851802762504621,
+      "learning_rate": 0.0009849666725580367,
+      "loss": 1.6014,
+      "step": 986
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.09555168236454595,
+      "learning_rate": 0.0009849242705312863,
+      "loss": 1.4576,
+      "step": 987
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0921061434635937,
+      "learning_rate": 0.000984881809705796,
+      "loss": 1.4993,
+      "step": 988
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.08929323937431682,
+      "learning_rate": 0.000984839290086715,
+      "loss": 1.5275,
+      "step": 989
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.08698430865853823,
+      "learning_rate": 0.0009847967116791985,
+      "loss": 1.4085,
+      "step": 990
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1106918708421856,
+      "learning_rate": 0.0009847540744884094,
+      "loss": 1.4351,
+      "step": 991
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1198492958950366,
+      "learning_rate": 0.0009847113785195175,
+      "loss": 1.5471,
+      "step": 992
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.09043846175343777,
+      "learning_rate": 0.0009846686237776998,
+      "loss": 1.3394,
+      "step": 993
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.09081924089189229,
+      "learning_rate": 0.0009846258102681406,
+      "loss": 1.4382,
+      "step": 994
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.09804043302232386,
+      "learning_rate": 0.0009845829379960312,
+      "loss": 1.3546,
+      "step": 995
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.10160497776628699,
+      "learning_rate": 0.00098454000696657,
+      "loss": 1.425,
+      "step": 996
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.09855611576341153,
+      "learning_rate": 0.0009844970171849624,
+      "loss": 1.3847,
+      "step": 997
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.08069941782676368,
+      "learning_rate": 0.000984453968656421,
+      "loss": 1.4353,
+      "step": 998
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.10779230448674895,
+      "learning_rate": 0.0009844108613861662,
+      "loss": 1.5072,
+      "step": 999
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.10385625952467904,
+      "learning_rate": 0.000984367695379424,
+      "loss": 1.4946,
+      "step": 1000
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.10086289488325921,
+      "learning_rate": 0.0009843244706414292,
+      "loss": 1.4217,
+      "step": 1001
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1119121824309659,
+      "learning_rate": 0.0009842811871774225,
+      "loss": 1.4483,
+      "step": 1002
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1080440169974913,
+      "learning_rate": 0.000984237844992652,
+      "loss": 1.4309,
+      "step": 1003
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.10987133844288485,
+      "learning_rate": 0.0009841944440923736,
+      "loss": 1.6328,
+      "step": 1004
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.10625514927826236,
+      "learning_rate": 0.0009841509844818496,
+      "loss": 1.4141,
+      "step": 1005
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.10159122308468017,
+      "learning_rate": 0.0009841074661663496,
+      "loss": 1.4201,
+      "step": 1006
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.10218291648959085,
+      "learning_rate": 0.0009840638891511504,
+      "loss": 1.4711,
+      "step": 1007
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.09931046150217804,
+      "learning_rate": 0.0009840202534415357,
+      "loss": 1.5503,
+      "step": 1008
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.10118813706201002,
+      "learning_rate": 0.0009839765590427968,
+      "loss": 1.4787,
+      "step": 1009
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.11350892060283965,
+      "learning_rate": 0.0009839328059602316,
+      "loss": 1.4651,
+      "step": 1010
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.10192194476484162,
+      "learning_rate": 0.000983888994199145,
+      "loss": 1.3939,
+      "step": 1011
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.09806497870255045,
+      "learning_rate": 0.0009838451237648498,
+      "loss": 1.544,
+      "step": 1012
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1076394976182956,
+      "learning_rate": 0.0009838011946626652,
+      "loss": 1.4081,
+      "step": 1013
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.10085057657545801,
+      "learning_rate": 0.0009837572068979179,
+      "loss": 1.4804,
+      "step": 1014
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.10371985397156486,
+      "learning_rate": 0.0009837131604759416,
+      "loss": 1.6341,
+      "step": 1015
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.11205728480425982,
+      "learning_rate": 0.0009836690554020768,
+      "loss": 1.4994,
+      "step": 1016
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0986560095523421,
+      "learning_rate": 0.0009836248916816716,
+      "loss": 1.3913,
+      "step": 1017
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.11832071985582042,
+      "learning_rate": 0.000983580669320081,
+      "loss": 1.4542,
+      "step": 1018
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.11660566131058599,
+      "learning_rate": 0.0009835363883226673,
+      "loss": 1.3678,
+      "step": 1019
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.11717444386866575,
+      "learning_rate": 0.0009834920486947994,
+      "loss": 1.5154,
+      "step": 1020
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.11973506149009075,
+      "learning_rate": 0.0009834476504418535,
+      "loss": 1.5199,
+      "step": 1021
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.12117972897901545,
+      "learning_rate": 0.0009834031935692135,
+      "loss": 1.5541,
+      "step": 1022
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.09979961498083265,
+      "learning_rate": 0.0009833586780822697,
+      "loss": 1.4248,
+      "step": 1023
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.08245349674048885,
+      "learning_rate": 0.0009833141039864198,
+      "loss": 1.3094,
+      "step": 1024
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.09588735228913653,
+      "learning_rate": 0.0009832694712870688,
+      "loss": 1.4229,
+      "step": 1025
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.09828741670587042,
+      "learning_rate": 0.000983224779989628,
+      "loss": 1.569,
+      "step": 1026
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.11780202591373197,
+      "learning_rate": 0.0009831800300995166,
+      "loss": 1.418,
+      "step": 1027
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.09290873549896511,
+      "learning_rate": 0.000983135221622161,
+      "loss": 1.4123,
+      "step": 1028
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.10994217265314823,
+      "learning_rate": 0.000983090354562994,
+      "loss": 1.5884,
+      "step": 1029
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.11691382359685173,
+      "learning_rate": 0.0009830454289274562,
+      "loss": 1.4539,
+      "step": 1030
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1054038050930932,
+      "learning_rate": 0.0009830004447209948,
+      "loss": 1.4914,
+      "step": 1031
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.093228195566531,
+      "learning_rate": 0.0009829554019490643,
+      "loss": 1.6054,
+      "step": 1032
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.09753879636453995,
+      "learning_rate": 0.0009829103006171263,
+      "loss": 1.5552,
+      "step": 1033
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0990335731096939,
+      "learning_rate": 0.0009828651407306494,
+      "loss": 1.4238,
+      "step": 1034
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.10536520343116573,
+      "learning_rate": 0.0009828199222951097,
+      "loss": 1.5421,
+      "step": 1035
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.10416559849628501,
+      "learning_rate": 0.0009827746453159897,
+      "loss": 1.5409,
+      "step": 1036
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.11005172462656163,
+      "learning_rate": 0.0009827293097987798,
+      "loss": 1.4537,
+      "step": 1037
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.11247632552430167,
+      "learning_rate": 0.0009826839157489767,
+      "loss": 1.5894,
+      "step": 1038
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.10756012516652874,
+      "learning_rate": 0.0009826384631720848,
+      "loss": 1.3542,
+      "step": 1039
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.10824115615017481,
+      "learning_rate": 0.0009825929520736155,
+      "loss": 1.4874,
+      "step": 1040
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1300208355066187,
+      "learning_rate": 0.0009825473824590866,
+      "loss": 1.4001,
+      "step": 1041
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.12730737924436886,
+      "learning_rate": 0.0009825017543340245,
+      "loss": 1.344,
+      "step": 1042
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.12974938372532713,
+      "learning_rate": 0.000982456067703961,
+      "loss": 1.507,
+      "step": 1043
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1035502393897516,
+      "learning_rate": 0.0009824103225744359,
+      "loss": 1.5458,
+      "step": 1044
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1141337981032992,
+      "learning_rate": 0.000982364518950996,
+      "loss": 1.5196,
+      "step": 1045
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.12199297553231467,
+      "learning_rate": 0.0009823186568391955,
+      "loss": 1.5455,
+      "step": 1046
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.11478738447697275,
+      "learning_rate": 0.000982272736244595,
+      "loss": 1.4725,
+      "step": 1047
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1262147429809612,
+      "learning_rate": 0.0009822267571727623,
+      "loss": 1.3743,
+      "step": 1048
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.10300865953073311,
+      "learning_rate": 0.000982180719629273,
+      "loss": 1.3518,
+      "step": 1049
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1337500675482,
+      "learning_rate": 0.0009821346236197092,
+      "loss": 1.5529,
+      "step": 1050
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.12392587654224337,
+      "learning_rate": 0.00098208846914966,
+      "loss": 1.3774,
+      "step": 1051
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.15327272162495567,
+      "learning_rate": 0.000982042256224722,
+      "loss": 1.5009,
+      "step": 1052
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.17463158449167768,
+      "learning_rate": 0.0009819959848504985,
+      "loss": 1.3791,
+      "step": 1053
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.10993680189395197,
+      "learning_rate": 0.0009819496550326002,
+      "loss": 1.3499,
+      "step": 1054
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.09452638170222592,
+      "learning_rate": 0.0009819032667766445,
+      "loss": 1.3927,
+      "step": 1055
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.11780194487896538,
+      "learning_rate": 0.0009818568200882566,
+      "loss": 1.5436,
+      "step": 1056
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.12564058560924768,
+      "learning_rate": 0.0009818103149730679,
+      "loss": 1.4736,
+      "step": 1057
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.12738094503349265,
+      "learning_rate": 0.0009817637514367174,
+      "loss": 1.5559,
+      "step": 1058
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.11203298042668797,
+      "learning_rate": 0.0009817171294848514,
+      "loss": 1.5329,
+      "step": 1059
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.09762718006430858,
+      "learning_rate": 0.0009816704491231226,
+      "loss": 1.5596,
+      "step": 1060
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.09700332763521427,
+      "learning_rate": 0.0009816237103571913,
+      "loss": 1.4408,
+      "step": 1061
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1141120906538291,
+      "learning_rate": 0.0009815769131927246,
+      "loss": 1.5046,
+      "step": 1062
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.09951655888040634,
+      "learning_rate": 0.0009815300576353969,
+      "loss": 1.3907,
+      "step": 1063
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.11335129257836248,
+      "learning_rate": 0.0009814831436908897,
+      "loss": 1.46,
+      "step": 1064
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.12462823502348326,
+      "learning_rate": 0.0009814361713648915,
+      "loss": 1.5528,
+      "step": 1065
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.10207058518187534,
+      "learning_rate": 0.0009813891406630975,
+      "loss": 1.5344,
+      "step": 1066
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.11331609691483102,
+      "learning_rate": 0.0009813420515912108,
+      "loss": 1.3399,
+      "step": 1067
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.11008646442048181,
+      "learning_rate": 0.0009812949041549408,
+      "loss": 1.4344,
+      "step": 1068
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.10103273692277252,
+      "learning_rate": 0.0009812476983600046,
+      "loss": 1.4093,
+      "step": 1069
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.10079447646547779,
+      "learning_rate": 0.0009812004342121257,
+      "loss": 1.4924,
+      "step": 1070
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.10309180507533766,
+      "learning_rate": 0.0009811531117170352,
+      "loss": 1.4854,
+      "step": 1071
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.10701374638267942,
+      "learning_rate": 0.000981105730880471,
+      "loss": 1.5011,
+      "step": 1072
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.10438559121872394,
+      "learning_rate": 0.0009810582917081786,
+      "loss": 1.5557,
+      "step": 1073
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.08928725239161178,
+      "learning_rate": 0.0009810107942059096,
+      "loss": 1.5582,
+      "step": 1074
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.10607406200382605,
+      "learning_rate": 0.0009809632383794237,
+      "loss": 1.4577,
+      "step": 1075
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.10140866770455789,
+      "learning_rate": 0.0009809156242344868,
+      "loss": 1.587,
+      "step": 1076
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.11324183419584681,
+      "learning_rate": 0.0009808679517768727,
+      "loss": 1.3907,
+      "step": 1077
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.09985830953044832,
+      "learning_rate": 0.0009808202210123615,
+      "loss": 1.5259,
+      "step": 1078
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.10726443153559254,
+      "learning_rate": 0.000980772431946741,
+      "loss": 1.5433,
+      "step": 1079
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.09371546019627704,
+      "learning_rate": 0.0009807245845858054,
+      "loss": 1.4472,
+      "step": 1080
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.09099857207405426,
+      "learning_rate": 0.000980676678935357,
+      "loss": 1.5692,
+      "step": 1081
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1026646049901164,
+      "learning_rate": 0.000980628715001204,
+      "loss": 1.4125,
+      "step": 1082
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.10668880279133305,
+      "learning_rate": 0.000980580692789162,
+      "loss": 1.3309,
+      "step": 1083
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1119693933886867,
+      "learning_rate": 0.0009805326123050544,
+      "loss": 1.445,
+      "step": 1084
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.11640640857224897,
+      "learning_rate": 0.000980484473554711,
+      "loss": 1.4927,
+      "step": 1085
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.10229689293466915,
+      "learning_rate": 0.0009804362765439688,
+      "loss": 1.4752,
+      "step": 1086
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.10514504374029744,
+      "learning_rate": 0.0009803880212786715,
+      "loss": 1.6104,
+      "step": 1087
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.11079636321731676,
+      "learning_rate": 0.0009803397077646704,
+      "loss": 1.4712,
+      "step": 1088
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.10082080188781722,
+      "learning_rate": 0.000980291336007824,
+      "loss": 1.4789,
+      "step": 1089
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.10675297685692006,
+      "learning_rate": 0.0009802429060139974,
+      "loss": 1.5433,
+      "step": 1090
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.126227264307551,
+      "learning_rate": 0.0009801944177890624,
+      "loss": 1.4064,
+      "step": 1091
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.09845560036111164,
+      "learning_rate": 0.000980145871338899,
+      "loss": 1.4893,
+      "step": 1092
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1273710597482172,
+      "learning_rate": 0.0009800972666693935,
+      "loss": 1.5066,
+      "step": 1093
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.10197482636295606,
+      "learning_rate": 0.000980048603786439,
+      "loss": 1.4119,
+      "step": 1094
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.09923180377874398,
+      "learning_rate": 0.0009799998826959366,
+      "loss": 1.4974,
+      "step": 1095
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.10551711580597985,
+      "learning_rate": 0.0009799511034037933,
+      "loss": 1.5412,
+      "step": 1096
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.10527622567768133,
+      "learning_rate": 0.0009799022659159242,
+      "loss": 1.5318,
+      "step": 1097
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.09689921924266803,
+      "learning_rate": 0.000979853370238251,
+      "loss": 1.5553,
+      "step": 1098
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.09900786363899416,
+      "learning_rate": 0.0009798044163767023,
+      "loss": 1.3221,
+      "step": 1099
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.11176675418735046,
+      "learning_rate": 0.0009797554043372138,
+      "loss": 1.7212,
+      "step": 1100
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.10447474360809943,
+      "learning_rate": 0.000979706334125729,
+      "loss": 1.4666,
+      "step": 1101
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0931764807312025,
+      "learning_rate": 0.0009796572057481968,
+      "loss": 1.5391,
+      "step": 1102
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.09105412831210441,
+      "learning_rate": 0.000979608019210575,
+      "loss": 1.4852,
+      "step": 1103
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.11365343481767007,
+      "learning_rate": 0.0009795587745188275,
+      "loss": 1.5142,
+      "step": 1104
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.09720993573412151,
+      "learning_rate": 0.0009795094716789252,
+      "loss": 1.4296,
+      "step": 1105
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.10445259805972958,
+      "learning_rate": 0.0009794601106968466,
+      "loss": 1.4978,
+      "step": 1106
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.08713210273408448,
+      "learning_rate": 0.0009794106915785763,
+      "loss": 1.3949,
+      "step": 1107
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.09792767712858018,
+      "learning_rate": 0.000979361214330107,
+      "loss": 1.3676,
+      "step": 1108
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.10885611133876523,
+      "learning_rate": 0.0009793116789574379,
+      "loss": 1.372,
+      "step": 1109
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0952070184084006,
+      "learning_rate": 0.0009792620854665753,
+      "loss": 1.475,
+      "step": 1110
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.09550091244632411,
+      "learning_rate": 0.0009792124338635325,
+      "loss": 1.5107,
+      "step": 1111
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.09616490303940983,
+      "learning_rate": 0.00097916272415433,
+      "loss": 1.5391,
+      "step": 1112
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1050502168365546,
+      "learning_rate": 0.0009791129563449952,
+      "loss": 1.445,
+      "step": 1113
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.09273724100814855,
+      "learning_rate": 0.0009790631304415628,
+      "loss": 1.4769,
+      "step": 1114
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.09836197010370934,
+      "learning_rate": 0.000979013246450074,
+      "loss": 1.3792,
+      "step": 1115
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.10609141191664713,
+      "learning_rate": 0.000978963304376578,
+      "loss": 1.4917,
+      "step": 1116
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.08593793619652067,
+      "learning_rate": 0.00097891330422713,
+      "loss": 1.4464,
+      "step": 1117
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.09539697052300453,
+      "learning_rate": 0.0009788632460077927,
+      "loss": 1.399,
+      "step": 1118
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.10822664643859031,
+      "learning_rate": 0.000978813129724636,
+      "loss": 1.4785,
+      "step": 1119
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.10154289276419569,
+      "learning_rate": 0.0009787629553837367,
+      "loss": 1.4826,
+      "step": 1120
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.10862578787342332,
+      "learning_rate": 0.0009787127229911783,
+      "loss": 1.4277,
+      "step": 1121
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.10981816955085401,
+      "learning_rate": 0.000978662432553052,
+      "loss": 1.3717,
+      "step": 1122
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1023940491468338,
+      "learning_rate": 0.0009786120840754556,
+      "loss": 1.6124,
+      "step": 1123
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.11702668759170876,
+      "learning_rate": 0.0009785616775644938,
+      "loss": 1.4212,
+      "step": 1124
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.12158573099045404,
+      "learning_rate": 0.000978511213026279,
+      "loss": 1.3692,
+      "step": 1125
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.10410236569465127,
+      "learning_rate": 0.0009784606904669297,
+      "loss": 1.4622,
+      "step": 1126
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.11484842201707575,
+      "learning_rate": 0.0009784101098925723,
+      "loss": 1.5727,
+      "step": 1127
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.10040496777521027,
+      "learning_rate": 0.0009783594713093397,
+      "loss": 1.4933,
+      "step": 1128
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.11033351716891086,
+      "learning_rate": 0.000978308774723372,
+      "loss": 1.5888,
+      "step": 1129
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.12154411199523346,
+      "learning_rate": 0.0009782580201408164,
+      "loss": 1.4452,
+      "step": 1130
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.11269562719555688,
+      "learning_rate": 0.0009782072075678271,
+      "loss": 1.4995,
+      "step": 1131
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.101394103359456,
+      "learning_rate": 0.0009781563370105653,
+      "loss": 1.5348,
+      "step": 1132
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.12337547846107601,
+      "learning_rate": 0.0009781054084751991,
+      "loss": 1.5912,
+      "step": 1133
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.10895999658008346,
+      "learning_rate": 0.0009780544219679039,
+      "loss": 1.5698,
+      "step": 1134
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.10875009724116637,
+      "learning_rate": 0.000978003377494862,
+      "loss": 1.418,
+      "step": 1135
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.08717663220170035,
+      "learning_rate": 0.0009779522750622625,
+      "loss": 1.5179,
+      "step": 1136
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0901001609158058,
+      "learning_rate": 0.0009779011146763019,
+      "loss": 1.5814,
+      "step": 1137
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.10275116736092334,
+      "learning_rate": 0.0009778498963431837,
+      "loss": 1.4128,
+      "step": 1138
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.09762112462628392,
+      "learning_rate": 0.000977798620069118,
+      "loss": 1.4922,
+      "step": 1139
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.10060723320049322,
+      "learning_rate": 0.0009777472858603226,
+      "loss": 1.5383,
+      "step": 1140
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1164092627557891,
+      "learning_rate": 0.0009776958937230216,
+      "loss": 1.5062,
+      "step": 1141
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.10534018168294584,
+      "learning_rate": 0.0009776444436634466,
+      "loss": 1.4397,
+      "step": 1142
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.10459737964269737,
+      "learning_rate": 0.0009775929356878362,
+      "loss": 1.4836,
+      "step": 1143
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1129596993044548,
+      "learning_rate": 0.0009775413698024358,
+      "loss": 1.5252,
+      "step": 1144
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0977021657766086,
+      "learning_rate": 0.000977489746013498,
+      "loss": 1.3986,
+      "step": 1145
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.10225167190944157,
+      "learning_rate": 0.0009774380643272822,
+      "loss": 1.6317,
+      "step": 1146
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.08850304042119721,
+      "learning_rate": 0.0009773863247500554,
+      "loss": 1.4967,
+      "step": 1147
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.11049263300762383,
+      "learning_rate": 0.0009773345272880906,
+      "loss": 1.5815,
+      "step": 1148
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.09691390291798195,
+      "learning_rate": 0.000977282671947669,
+      "loss": 1.4033,
+      "step": 1149
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.10663292694602702,
+      "learning_rate": 0.000977230758735078,
+      "loss": 1.5099,
+      "step": 1150
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.08971210350781093,
+      "learning_rate": 0.000977178787656612,
+      "loss": 1.4351,
+      "step": 1151
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.12450168045568234,
+      "learning_rate": 0.000977126758718573,
+      "loss": 1.5371,
+      "step": 1152
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0948433794803075,
+      "learning_rate": 0.0009770746719272696,
+      "loss": 1.4678,
+      "step": 1153
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.10085071740419524,
+      "learning_rate": 0.0009770225272890177,
+      "loss": 1.4806,
+      "step": 1154
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4484143809430015,
+      "learning_rate": 0.0009769703248101397,
+      "loss": 1.607,
+      "step": 1155
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.15922980601920514,
+      "learning_rate": 0.0009769180644969653,
+      "loss": 1.5416,
+      "step": 1156
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.11107798130456996,
+      "learning_rate": 0.0009768657463558315,
+      "loss": 1.6267,
+      "step": 1157
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2946033526128502,
+      "learning_rate": 0.0009768133703930819,
+      "loss": 1.4602,
+      "step": 1158
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.15370026664096484,
+      "learning_rate": 0.0009767609366150673,
+      "loss": 1.5199,
+      "step": 1159
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.10375667791821207,
+      "learning_rate": 0.0009767084450281456,
+      "loss": 1.4155,
+      "step": 1160
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.12207675726090313,
+      "learning_rate": 0.0009766558956386814,
+      "loss": 1.3242,
+      "step": 1161
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.22605444819633985,
+      "learning_rate": 0.0009766032884530465,
+      "loss": 1.3588,
+      "step": 1162
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.12489025319914,
+      "learning_rate": 0.00097655062347762,
+      "loss": 1.5016,
+      "step": 1163
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.11854980300003969,
+      "learning_rate": 0.0009764979007187873,
+      "loss": 1.4542,
+      "step": 1164
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.26535399019300804,
+      "learning_rate": 0.0009764451201829414,
+      "loss": 1.5479,
+      "step": 1165
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4479661893821317,
+      "learning_rate": 0.0009763922818764822,
+      "loss": 1.5013,
+      "step": 1166
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.17293052163018774,
+      "learning_rate": 0.0009763393858058164,
+      "loss": 1.4978,
+      "step": 1167
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.12328773343089187,
+      "learning_rate": 0.0009762864319773579,
+      "loss": 1.4241,
+      "step": 1168
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.10961795118064666,
+      "learning_rate": 0.0009762334203975276,
+      "loss": 1.5303,
+      "step": 1169
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1175993556438544,
+      "learning_rate": 0.0009761803510727531,
+      "loss": 1.4783,
+      "step": 1170
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.11106899286073688,
+      "learning_rate": 0.0009761272240094695,
+      "loss": 1.458,
+      "step": 1171
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.13369412952321702,
+      "learning_rate": 0.0009760740392141186,
+      "loss": 1.4819,
+      "step": 1172
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.09580765095849292,
+      "learning_rate": 0.0009760207966931489,
+      "loss": 1.4705,
+      "step": 1173
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.11617828359160304,
+      "learning_rate": 0.0009759674964530167,
+      "loss": 1.4023,
+      "step": 1174
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.10434869245175268,
+      "learning_rate": 0.0009759141385001847,
+      "loss": 1.5545,
+      "step": 1175
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1199786542618325,
+      "learning_rate": 0.0009758607228411225,
+      "loss": 1.5722,
+      "step": 1176
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0964001059293876,
+      "learning_rate": 0.0009758072494823072,
+      "loss": 1.5165,
+      "step": 1177
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.11184263546413122,
+      "learning_rate": 0.0009757537184302225,
+      "loss": 1.4181,
+      "step": 1178
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.11404339323309447,
+      "learning_rate": 0.0009757001296913593,
+      "loss": 1.5288,
+      "step": 1179
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.11956579528844924,
+      "learning_rate": 0.0009756464832722154,
+      "loss": 1.4549,
+      "step": 1180
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0978519457773404,
+      "learning_rate": 0.0009755927791792956,
+      "loss": 1.3303,
+      "step": 1181
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.17000533389215636,
+      "learning_rate": 0.0009755390174191117,
+      "loss": 1.4161,
+      "step": 1182
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.07996186824950127,
+      "learning_rate": 0.0009754851979981826,
+      "loss": 1.3548,
+      "step": 1183
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.09044166304385826,
+      "learning_rate": 0.0009754313209230339,
+      "loss": 1.5155,
+      "step": 1184
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.10621964304981515,
+      "learning_rate": 0.0009753773862001985,
+      "loss": 1.6404,
+      "step": 1185
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.08254025272480696,
+      "learning_rate": 0.0009753233938362161,
+      "loss": 1.4376,
+      "step": 1186
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1106607801453522,
+      "learning_rate": 0.0009752693438376336,
+      "loss": 1.498,
+      "step": 1187
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.09876265807599625,
+      "learning_rate": 0.0009752152362110045,
+      "loss": 1.5122,
+      "step": 1188
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.10039646627078934,
+      "learning_rate": 0.0009751610709628897,
+      "loss": 1.3371,
+      "step": 1189
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1005714762578938,
+      "learning_rate": 0.000975106848099857,
+      "loss": 1.4913,
+      "step": 1190
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.09409073411849112,
+      "learning_rate": 0.0009750525676284811,
+      "loss": 1.4359,
+      "step": 1191
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1114939909639379,
+      "learning_rate": 0.0009749982295553436,
+      "loss": 1.4572,
+      "step": 1192
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.11159452971642979,
+      "learning_rate": 0.0009749438338870331,
+      "loss": 1.4944,
+      "step": 1193
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.10054333256733074,
+      "learning_rate": 0.0009748893806301455,
+      "loss": 1.4464,
+      "step": 1194
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.09895279648206912,
+      "learning_rate": 0.0009748348697912832,
+      "loss": 1.509,
+      "step": 1195
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.10061217401043886,
+      "learning_rate": 0.000974780301377056,
+      "loss": 1.4162,
+      "step": 1196
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.10417574356188845,
+      "learning_rate": 0.0009747256753940803,
+      "loss": 1.3959,
+      "step": 1197
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.10853524853515546,
+      "learning_rate": 0.00097467099184898,
+      "loss": 1.3809,
+      "step": 1198
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.10205080178365547,
+      "learning_rate": 0.0009746162507483854,
+      "loss": 1.5636,
+      "step": 1199
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.11428455096975486,
+      "learning_rate": 0.0009745614520989341,
+      "loss": 1.3596,
+      "step": 1200
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.10288778226541649,
+      "learning_rate": 0.0009745065959072708,
+      "loss": 1.5153,
+      "step": 1201
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.13135365823412376,
+      "learning_rate": 0.0009744516821800469,
+      "loss": 1.5387,
+      "step": 1202
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.10102362449138394,
+      "learning_rate": 0.0009743967109239206,
+      "loss": 1.4004,
+      "step": 1203
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.13512092208553644,
+      "learning_rate": 0.0009743416821455577,
+      "loss": 1.5138,
+      "step": 1204
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1087914935718707,
+      "learning_rate": 0.0009742865958516307,
+      "loss": 1.5475,
+      "step": 1205
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.11166016102754622,
+      "learning_rate": 0.0009742314520488187,
+      "loss": 1.4857,
+      "step": 1206
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.10473240664950298,
+      "learning_rate": 0.0009741762507438083,
+      "loss": 1.4585,
+      "step": 1207
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.09850077005465968,
+      "learning_rate": 0.0009741209919432928,
+      "loss": 1.475,
+      "step": 1208
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.09199217175728568,
+      "learning_rate": 0.0009740656756539723,
+      "loss": 1.5659,
+      "step": 1209
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.09683493951065482,
+      "learning_rate": 0.0009740103018825543,
+      "loss": 1.4967,
+      "step": 1210
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.09121350319498756,
+      "learning_rate": 0.0009739548706357532,
+      "loss": 1.3673,
+      "step": 1211
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.10048743165141734,
+      "learning_rate": 0.0009738993819202901,
+      "loss": 1.5065,
+      "step": 1212
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.09127861055952981,
+      "learning_rate": 0.0009738438357428929,
+      "loss": 1.2944,
+      "step": 1213
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.09363988777597988,
+      "learning_rate": 0.0009737882321102972,
+      "loss": 1.4625,
+      "step": 1214
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.10479854745256931,
+      "learning_rate": 0.000973732571029245,
+      "loss": 1.5235,
+      "step": 1215
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.09677667521333407,
+      "learning_rate": 0.0009736768525064851,
+      "loss": 1.4624,
+      "step": 1216
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.09244710850250462,
+      "learning_rate": 0.0009736210765487741,
+      "loss": 1.5458,
+      "step": 1217
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.10714498510511104,
+      "learning_rate": 0.0009735652431628747,
+      "loss": 1.5138,
+      "step": 1218
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.10730277307980761,
+      "learning_rate": 0.000973509352355557,
+      "loss": 1.3659,
+      "step": 1219
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.11548224356442682,
+      "learning_rate": 0.0009734534041335977,
+      "loss": 1.4155,
+      "step": 1220
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.10400406071826049,
+      "learning_rate": 0.000973397398503781,
+      "loss": 1.4974,
+      "step": 1221
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.10990936584302333,
+      "learning_rate": 0.0009733413354728977,
+      "loss": 1.455,
+      "step": 1222
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.09334736281891305,
+      "learning_rate": 0.0009732852150477456,
+      "loss": 1.4218,
+      "step": 1223
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.10295858393774611,
+      "learning_rate": 0.0009732290372351294,
+      "loss": 1.5732,
+      "step": 1224
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.09349978355010372,
+      "learning_rate": 0.0009731728020418611,
+      "loss": 1.5234,
+      "step": 1225
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.10091993747486702,
+      "learning_rate": 0.0009731165094747593,
+      "loss": 1.4055,
+      "step": 1226
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1048721686348737,
+      "learning_rate": 0.0009730601595406496,
+      "loss": 1.4747,
+      "step": 1227
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.09189084245289222,
+      "learning_rate": 0.0009730037522463647,
+      "loss": 1.4501,
+      "step": 1228
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.09610235865729654,
+      "learning_rate": 0.0009729472875987439,
+      "loss": 1.4709,
+      "step": 1229
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.09998948369758241,
+      "learning_rate": 0.0009728907656046343,
+      "loss": 1.4859,
+      "step": 1230
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.09704555535405772,
+      "learning_rate": 0.0009728341862708889,
+      "loss": 1.5897,
+      "step": 1231
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.11181733909784186,
+      "learning_rate": 0.0009727775496043685,
+      "loss": 1.4797,
+      "step": 1232
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1078281133859294,
+      "learning_rate": 0.0009727208556119401,
+      "loss": 1.5068,
+      "step": 1233
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.12417520601082405,
+      "learning_rate": 0.0009726641043004782,
+      "loss": 1.4862,
+      "step": 1234
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.09808494647507954,
+      "learning_rate": 0.0009726072956768643,
+      "loss": 1.4989,
+      "step": 1235
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.11618660624020795,
+      "learning_rate": 0.0009725504297479864,
+      "loss": 1.3851,
+      "step": 1236
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.10189474305066351,
+      "learning_rate": 0.0009724935065207399,
+      "loss": 1.4341,
+      "step": 1237
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.099810201488847,
+      "learning_rate": 0.0009724365260020267,
+      "loss": 1.3454,
+      "step": 1238
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.10607882500756213,
+      "learning_rate": 0.0009723794881987559,
+      "loss": 1.4519,
+      "step": 1239
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.10124419381866068,
+      "learning_rate": 0.0009723223931178438,
+      "loss": 1.6662,
+      "step": 1240
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.08695360420074442,
+      "learning_rate": 0.0009722652407662129,
+      "loss": 1.4175,
+      "step": 1241
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.09549842078403618,
+      "learning_rate": 0.0009722080311507937,
+      "loss": 1.5201,
+      "step": 1242
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.08806136818436658,
+      "learning_rate": 0.0009721507642785226,
+      "loss": 1.4395,
+      "step": 1243
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.09556559977297367,
+      "learning_rate": 0.0009720934401563437,
+      "loss": 1.5403,
+      "step": 1244
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.08343873926288949,
+      "learning_rate": 0.0009720360587912075,
+      "loss": 1.4373,
+      "step": 1245
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.09441068917886046,
+      "learning_rate": 0.0009719786201900719,
+      "loss": 1.5036,
+      "step": 1246
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.09864797097988,
+      "learning_rate": 0.0009719211243599014,
+      "loss": 1.3718,
+      "step": 1247
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.08882461434600883,
+      "learning_rate": 0.0009718635713076676,
+      "loss": 1.473,
+      "step": 1248
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.08388976795441185,
+      "learning_rate": 0.0009718059610403491,
+      "loss": 1.4691,
+      "step": 1249
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.10270330807860907,
+      "learning_rate": 0.0009717482935649312,
+      "loss": 1.4669,
+      "step": 1250
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.096526175780071,
+      "learning_rate": 0.0009716905688884063,
+      "loss": 1.4292,
+      "step": 1251
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.07826161807267731,
+      "learning_rate": 0.0009716327870177739,
+      "loss": 1.4802,
+      "step": 1252
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.07991878427102167,
+      "learning_rate": 0.00097157494796004,
+      "loss": 1.4538,
+      "step": 1253
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.08512603268481847,
+      "learning_rate": 0.000971517051722218,
+      "loss": 1.3659,
+      "step": 1254
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.08429100786699598,
+      "learning_rate": 0.0009714590983113279,
+      "loss": 1.4721,
+      "step": 1255
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.08571040834333599,
+      "learning_rate": 0.0009714010877343966,
+      "loss": 1.5399,
+      "step": 1256
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.08559073310258722,
+      "learning_rate": 0.0009713430199984583,
+      "loss": 1.3743,
+      "step": 1257
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.09653287635713186,
+      "learning_rate": 0.0009712848951105539,
+      "loss": 1.4327,
+      "step": 1258
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.09277545525213768,
+      "learning_rate": 0.0009712267130777312,
+      "loss": 1.4792,
+      "step": 1259
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.08815163163035124,
+      "learning_rate": 0.000971168473907045,
+      "loss": 1.5024,
+      "step": 1260
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.08767532362959778,
+      "learning_rate": 0.0009711101776055569,
+      "loss": 1.4843,
+      "step": 1261
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.09195990998443754,
+      "learning_rate": 0.0009710518241803356,
+      "loss": 1.5151,
+      "step": 1262
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.09294957831798757,
+      "learning_rate": 0.0009709934136384568,
+      "loss": 1.5475,
+      "step": 1263
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.09036887036299439,
+      "learning_rate": 0.0009709349459870027,
+      "loss": 1.3581,
+      "step": 1264
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.10196282506888943,
+      "learning_rate": 0.0009708764212330629,
+      "loss": 1.3326,
+      "step": 1265
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.10355169189765306,
+      "learning_rate": 0.0009708178393837336,
+      "loss": 1.5029,
+      "step": 1266
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.09318762477278592,
+      "learning_rate": 0.0009707592004461182,
+      "loss": 1.5204,
+      "step": 1267
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.08563077439899368,
+      "learning_rate": 0.0009707005044273267,
+      "loss": 1.3752,
+      "step": 1268
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.08600170949975434,
+      "learning_rate": 0.0009706417513344764,
+      "loss": 1.5756,
+      "step": 1269
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.08164799944411805,
+      "learning_rate": 0.0009705829411746911,
+      "loss": 1.3627,
+      "step": 1270
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.07849971938700012,
+      "learning_rate": 0.000970524073955102,
+      "loss": 1.3703,
+      "step": 1271
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.07646692578878106,
+      "learning_rate": 0.0009704651496828466,
+      "loss": 1.4787,
+      "step": 1272
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.09464792546642692,
+      "learning_rate": 0.00097040616836507,
+      "loss": 1.3937,
+      "step": 1273
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.09949300784748961,
+      "learning_rate": 0.0009703471300089236,
+      "loss": 1.5593,
+      "step": 1274
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.08122148070582941,
+      "learning_rate": 0.0009702880346215664,
+      "loss": 1.3546,
+      "step": 1275
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.10318702598898441,
+      "learning_rate": 0.0009702288822101634,
+      "loss": 1.52,
+      "step": 1276
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.11343488674792428,
+      "learning_rate": 0.0009701696727818874,
+      "loss": 1.3261,
+      "step": 1277
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.10000234811603913,
+      "learning_rate": 0.0009701104063439177,
+      "loss": 1.4701,
+      "step": 1278
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.08002664638456945,
+      "learning_rate": 0.0009700510829034404,
+      "loss": 1.4196,
+      "step": 1279
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.08976865563700746,
+      "learning_rate": 0.0009699917024676488,
+      "loss": 1.4879,
+      "step": 1280
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.09486370384243133,
+      "learning_rate": 0.0009699322650437432,
+      "loss": 1.6916,
+      "step": 1281
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.09483802520074949,
+      "learning_rate": 0.00096987277063893,
+      "loss": 1.4527,
+      "step": 1282
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.10092644093150213,
+      "learning_rate": 0.0009698132192604238,
+      "loss": 1.4879,
+      "step": 1283
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.09844581657806153,
+      "learning_rate": 0.0009697536109154449,
+      "loss": 1.4846,
+      "step": 1284
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.09448314390473393,
+      "learning_rate": 0.0009696939456112213,
+      "loss": 1.318,
+      "step": 1285
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.11425995620933611,
+      "learning_rate": 0.0009696342233549873,
+      "loss": 1.3915,
+      "step": 1286
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.10553751796369051,
+      "learning_rate": 0.0009695744441539849,
+      "loss": 1.6516,
+      "step": 1287
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.09672534952850492,
+      "learning_rate": 0.0009695146080154621,
+      "loss": 1.4494,
+      "step": 1288
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.10035478781628406,
+      "learning_rate": 0.0009694547149466745,
+      "loss": 1.5295,
+      "step": 1289
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.08972138904342805,
+      "learning_rate": 0.0009693947649548842,
+      "loss": 1.474,
+      "step": 1290
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.08679630306391951,
+      "learning_rate": 0.0009693347580473604,
+      "loss": 1.4317,
+      "step": 1291
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.10871542085222537,
+      "learning_rate": 0.0009692746942313792,
+      "loss": 1.5484,
+      "step": 1292
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.08603429220157566,
+      "learning_rate": 0.0009692145735142235,
+      "loss": 1.4189,
+      "step": 1293
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.094369401781512,
+      "learning_rate": 0.000969154395903183,
+      "loss": 1.2444,
+      "step": 1294
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.08502135101989325,
+      "learning_rate": 0.0009690941614055546,
+      "loss": 1.4072,
+      "step": 1295
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.09178718049011819,
+      "learning_rate": 0.0009690338700286421,
+      "loss": 1.3757,
+      "step": 1296
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.08929961053282694,
+      "learning_rate": 0.0009689735217797557,
+      "loss": 1.423,
+      "step": 1297
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0829406241969028,
+      "learning_rate": 0.0009689131166662131,
+      "loss": 1.4598,
+      "step": 1298
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.09831046552987227,
+      "learning_rate": 0.0009688526546953384,
+      "loss": 1.5143,
+      "step": 1299
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.10233533693634345,
+      "learning_rate": 0.0009687921358744629,
+      "loss": 1.5625,
+      "step": 1300
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.12009618350664676,
+      "learning_rate": 0.0009687315602109248,
+      "loss": 1.5813,
+      "step": 1301
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.09579390968775611,
+      "learning_rate": 0.0009686709277120691,
+      "loss": 1.5,
+      "step": 1302
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.09347070874904188,
+      "learning_rate": 0.0009686102383852477,
+      "loss": 1.4762,
+      "step": 1303
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0974530705160833,
+      "learning_rate": 0.0009685494922378193,
+      "loss": 1.4729,
+      "step": 1304
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.09780466008302127,
+      "learning_rate": 0.0009684886892771497,
+      "loss": 1.5751,
+      "step": 1305
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.10446303843416181,
+      "learning_rate": 0.0009684278295106112,
+      "loss": 1.521,
+      "step": 1306
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.08583740614987581,
+      "learning_rate": 0.0009683669129455836,
+      "loss": 1.4352,
+      "step": 1307
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.09548449846071512,
+      "learning_rate": 0.0009683059395894532,
+      "loss": 1.6236,
+      "step": 1308
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.09317317974015707,
+      "learning_rate": 0.0009682449094496129,
+      "loss": 1.4718,
+      "step": 1309
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.09028344652051286,
+      "learning_rate": 0.0009681838225334632,
+      "loss": 1.5197,
+      "step": 1310
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1010454477834947,
+      "learning_rate": 0.0009681226788484107,
+      "loss": 1.4355,
+      "step": 1311
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.10063424616961393,
+      "learning_rate": 0.0009680614784018699,
+      "loss": 1.5216,
+      "step": 1312
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.08847368543043868,
+      "learning_rate": 0.000968000221201261,
+      "loss": 1.4489,
+      "step": 1313
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.08917200687658379,
+      "learning_rate": 0.0009679389072540118,
+      "loss": 1.5095,
+      "step": 1314
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.08780084899229014,
+      "learning_rate": 0.0009678775365675569,
+      "loss": 1.2328,
+      "step": 1315
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.09952472380916254,
+      "learning_rate": 0.0009678161091493377,
+      "loss": 1.5445,
+      "step": 1316
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0922044875770402,
+      "learning_rate": 0.0009677546250068024,
+      "loss": 1.4997,
+      "step": 1317
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.08679058782028819,
+      "learning_rate": 0.0009676930841474063,
+      "loss": 1.4818,
+      "step": 1318
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.09946020728542797,
+      "learning_rate": 0.0009676314865786113,
+      "loss": 1.541,
+      "step": 1319
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.09924599571927359,
+      "learning_rate": 0.0009675698323078864,
+      "loss": 1.4275,
+      "step": 1320
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.09736847437990914,
+      "learning_rate": 0.0009675081213427075,
+      "loss": 1.3534,
+      "step": 1321
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.08981117793178947,
+      "learning_rate": 0.0009674463536905569,
+      "loss": 1.5193,
+      "step": 1322
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0892544928899613,
+      "learning_rate": 0.0009673845293589245,
+      "loss": 1.4088,
+      "step": 1323
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.09692218740693645,
+      "learning_rate": 0.0009673226483553067,
+      "loss": 1.4853,
+      "step": 1324
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.090931858503463,
+      "learning_rate": 0.0009672607106872065,
+      "loss": 1.5478,
+      "step": 1325
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.08677645972181737,
+      "learning_rate": 0.0009671987163621343,
+      "loss": 1.4351,
+      "step": 1326
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0965811198069088,
+      "learning_rate": 0.0009671366653876072,
+      "loss": 1.44,
+      "step": 1327
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.10265572360163623,
+      "learning_rate": 0.0009670745577711489,
+      "loss": 1.4758,
+      "step": 1328
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.10813126200323721,
+      "learning_rate": 0.0009670123935202901,
+      "loss": 1.5154,
+      "step": 1329
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0818442543578839,
+      "learning_rate": 0.0009669501726425686,
+      "loss": 1.5325,
+      "step": 1330
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.08476502614974389,
+      "learning_rate": 0.0009668878951455289,
+      "loss": 1.4566,
+      "step": 1331
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0929172330057343,
+      "learning_rate": 0.0009668255610367222,
+      "loss": 1.4826,
+      "step": 1332
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.11067141747464249,
+      "learning_rate": 0.0009667631703237069,
+      "loss": 1.4778,
+      "step": 1333
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.09895618231426587,
+      "learning_rate": 0.000966700723014048,
+      "loss": 1.4972,
+      "step": 1334
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.09258309463650612,
+      "learning_rate": 0.0009666382191153175,
+      "loss": 1.5373,
+      "step": 1335
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.08478282889023635,
+      "learning_rate": 0.0009665756586350942,
+      "loss": 1.3624,
+      "step": 1336
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1029379375379849,
+      "learning_rate": 0.0009665130415809636,
+      "loss": 1.5687,
+      "step": 1337
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.10999399674084075,
+      "learning_rate": 0.0009664503679605186,
+      "loss": 1.4364,
+      "step": 1338
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1002629659584574,
+      "learning_rate": 0.0009663876377813583,
+      "loss": 1.4827,
+      "step": 1339
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.09189118197800461,
+      "learning_rate": 0.000966324851051089,
+      "loss": 1.5273,
+      "step": 1340
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.08846946549951583,
+      "learning_rate": 0.0009662620077773239,
+      "loss": 1.4333,
+      "step": 1341
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.09374605110911702,
+      "learning_rate": 0.0009661991079676828,
+      "loss": 1.547,
+      "step": 1342
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.08242169059262731,
+      "learning_rate": 0.0009661361516297928,
+      "loss": 1.4459,
+      "step": 1343
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.09832558372243909,
+      "learning_rate": 0.0009660731387712873,
+      "loss": 1.5649,
+      "step": 1344
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.10825358627457048,
+      "learning_rate": 0.0009660100693998071,
+      "loss": 1.4277,
+      "step": 1345
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.09313205123054559,
+      "learning_rate": 0.0009659469435229992,
+      "loss": 1.5624,
+      "step": 1346
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.07998615164356912,
+      "learning_rate": 0.0009658837611485181,
+      "loss": 1.4637,
+      "step": 1347
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.10552045428986864,
+      "learning_rate": 0.0009658205222840249,
+      "loss": 1.454,
+      "step": 1348
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.08907788707383209,
+      "learning_rate": 0.0009657572269371873,
+      "loss": 1.5136,
+      "step": 1349
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.08617641281827475,
+      "learning_rate": 0.0009656938751156802,
+      "loss": 1.4615,
+      "step": 1350
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.09921990951476314,
+      "learning_rate": 0.0009656304668271854,
+      "loss": 1.4423,
+      "step": 1351
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.10024579878753528,
+      "learning_rate": 0.000965567002079391,
+      "loss": 1.4746,
+      "step": 1352
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.09568464334273936,
+      "learning_rate": 0.0009655034808799928,
+      "loss": 1.5115,
+      "step": 1353
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0948152667533343,
+      "learning_rate": 0.0009654399032366925,
+      "loss": 1.482,
+      "step": 1354
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.09802476493056436,
+      "learning_rate": 0.0009653762691571995,
+      "loss": 1.427,
+      "step": 1355
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.08709415251464686,
+      "learning_rate": 0.0009653125786492294,
+      "loss": 1.4658,
+      "step": 1356
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.08686335295193764,
+      "learning_rate": 0.0009652488317205048,
+      "loss": 1.3828,
+      "step": 1357
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.10936667844927866,
+      "learning_rate": 0.0009651850283787555,
+      "loss": 1.6204,
+      "step": 1358
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.08318505997719156,
+      "learning_rate": 0.000965121168631718,
+      "loss": 1.3626,
+      "step": 1359
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0981715400028448,
+      "learning_rate": 0.000965057252487135,
+      "loss": 1.4695,
+      "step": 1360
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0855374725173818,
+      "learning_rate": 0.000964993279952757,
+      "loss": 1.4296,
+      "step": 1361
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.09090872788305977,
+      "learning_rate": 0.0009649292510363407,
+      "loss": 1.4942,
+      "step": 1362
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.08903024982584848,
+      "learning_rate": 0.0009648651657456497,
+      "loss": 1.53,
+      "step": 1363
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.09651472004451829,
+      "learning_rate": 0.0009648010240884549,
+      "loss": 1.4141,
+      "step": 1364
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0975116971187096,
+      "learning_rate": 0.0009647368260725335,
+      "loss": 1.6606,
+      "step": 1365
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.09109281987857183,
+      "learning_rate": 0.0009646725717056696,
+      "loss": 1.4704,
+      "step": 1366
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.08977013401911238,
+      "learning_rate": 0.0009646082609956546,
+      "loss": 1.4401,
+      "step": 1367
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.09619001203581924,
+      "learning_rate": 0.0009645438939502862,
+      "loss": 1.4719,
+      "step": 1368
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0920101285961825,
+      "learning_rate": 0.000964479470577369,
+      "loss": 1.4291,
+      "step": 1369
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.10189576673060297,
+      "learning_rate": 0.0009644149908847148,
+      "loss": 1.3505,
+      "step": 1370
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.10141333070010058,
+      "learning_rate": 0.0009643504548801418,
+      "loss": 1.5281,
+      "step": 1371
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.10183136911830679,
+      "learning_rate": 0.0009642858625714753,
+      "loss": 1.617,
+      "step": 1372
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.08938428018230365,
+      "learning_rate": 0.0009642212139665474,
+      "loss": 1.5241,
+      "step": 1373
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.09867932384400704,
+      "learning_rate": 0.0009641565090731968,
+      "loss": 1.4051,
+      "step": 1374
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.08685976647015854,
+      "learning_rate": 0.0009640917478992692,
+      "loss": 1.4531,
+      "step": 1375
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.08388522860676766,
+      "learning_rate": 0.0009640269304526175,
+      "loss": 1.4059,
+      "step": 1376
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.08755849762519224,
+      "learning_rate": 0.0009639620567411005,
+      "loss": 1.4757,
+      "step": 1377
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.08967987306203287,
+      "learning_rate": 0.0009638971267725846,
+      "loss": 1.4917,
+      "step": 1378
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.09337617930404216,
+      "learning_rate": 0.0009638321405549429,
+      "loss": 1.4998,
+      "step": 1379
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.11438394374416985,
+      "learning_rate": 0.0009637670980960549,
+      "loss": 1.3926,
+      "step": 1380
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.08596327108077496,
+      "learning_rate": 0.0009637019994038076,
+      "loss": 1.4768,
+      "step": 1381
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.10081100976304277,
+      "learning_rate": 0.0009636368444860941,
+      "loss": 1.4218,
+      "step": 1382
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.08628745782550797,
+      "learning_rate": 0.0009635716333508149,
+      "loss": 1.5145,
+      "step": 1383
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.09973150834603343,
+      "learning_rate": 0.000963506366005877,
+      "loss": 1.5565,
+      "step": 1384
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0915294973766297,
+      "learning_rate": 0.0009634410424591941,
+      "loss": 1.5018,
+      "step": 1385
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.09358282880478652,
+      "learning_rate": 0.0009633756627186874,
+      "loss": 1.4947,
+      "step": 1386
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.08854602719419898,
+      "learning_rate": 0.0009633102267922838,
+      "loss": 1.3823,
+      "step": 1387
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.09370252902472906,
+      "learning_rate": 0.0009632447346879181,
+      "loss": 1.372,
+      "step": 1388
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.07704849046364678,
+      "learning_rate": 0.0009631791864135313,
+      "loss": 1.3268,
+      "step": 1389
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.09768848364054729,
+      "learning_rate": 0.0009631135819770711,
+      "loss": 1.4281,
+      "step": 1390
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.10743879271717618,
+      "learning_rate": 0.0009630479213864927,
+      "loss": 1.6314,
+      "step": 1391
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.11841657156720103,
+      "learning_rate": 0.0009629822046497573,
+      "loss": 1.455,
+      "step": 1392
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0871394521914947,
+      "learning_rate": 0.0009629164317748335,
+      "loss": 1.4534,
+      "step": 1393
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.10329392146592385,
+      "learning_rate": 0.0009628506027696966,
+      "loss": 1.4848,
+      "step": 1394
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.08307046208995565,
+      "learning_rate": 0.0009627847176423282,
+      "loss": 1.4573,
+      "step": 1395
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.10441134569877149,
+      "learning_rate": 0.0009627187764007175,
+      "loss": 1.4169,
+      "step": 1396
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.11076009082556416,
+      "learning_rate": 0.0009626527790528599,
+      "loss": 1.4463,
+      "step": 1397
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.09972177969927494,
+      "learning_rate": 0.0009625867256067577,
+      "loss": 1.4483,
+      "step": 1398
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.10232915510331024,
+      "learning_rate": 0.0009625206160704203,
+      "loss": 1.4843,
+      "step": 1399
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.10033949503613594,
+      "learning_rate": 0.0009624544504518636,
+      "loss": 1.4372,
+      "step": 1400
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.10470825877339096,
+      "learning_rate": 0.0009623882287591106,
+      "loss": 1.4166,
+      "step": 1401
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.09177089511218502,
+      "learning_rate": 0.0009623219510001906,
+      "loss": 1.1729,
+      "step": 1402
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.09485966877207715,
+      "learning_rate": 0.0009622556171831403,
+      "loss": 1.5652,
+      "step": 1403
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0908345472809001,
+      "learning_rate": 0.0009621892273160027,
+      "loss": 1.4296,
+      "step": 1404
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.09258681147903781,
+      "learning_rate": 0.0009621227814068281,
+      "loss": 1.3104,
+      "step": 1405
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.09872348083382965,
+      "learning_rate": 0.0009620562794636728,
+      "loss": 1.4792,
+      "step": 1406
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.10549538108369423,
+      "learning_rate": 0.0009619897214946007,
+      "loss": 1.4798,
+      "step": 1407
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.10762569208689654,
+      "learning_rate": 0.0009619231075076823,
+      "loss": 1.5403,
+      "step": 1408
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.10595909019966,
+      "learning_rate": 0.0009618564375109945,
+      "loss": 1.4209,
+      "step": 1409
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.09530047576327778,
+      "learning_rate": 0.0009617897115126215,
+      "loss": 1.5453,
+      "step": 1410
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.09150305099250461,
+      "learning_rate": 0.0009617229295206537,
+      "loss": 1.4043,
+      "step": 1411
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.10734559081510649,
+      "learning_rate": 0.0009616560915431891,
+      "loss": 1.5386,
+      "step": 1412
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1166936881059637,
+      "learning_rate": 0.000961589197588332,
+      "loss": 1.4432,
+      "step": 1413
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.09399762085255639,
+      "learning_rate": 0.000961522247664193,
+      "loss": 1.3782,
+      "step": 1414
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.10108051188315073,
+      "learning_rate": 0.0009614552417788906,
+      "loss": 1.5388,
+      "step": 1415
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.08425065039992434,
+      "learning_rate": 0.0009613881799405491,
+      "loss": 1.3643,
+      "step": 1416
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.08774005124641272,
+      "learning_rate": 0.0009613210621573001,
+      "loss": 1.4577,
+      "step": 1417
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.11048355058020695,
+      "learning_rate": 0.0009612538884372821,
+      "loss": 1.4588,
+      "step": 1418
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.08641151132612267,
+      "learning_rate": 0.0009611866587886399,
+      "loss": 1.5007,
+      "step": 1419
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.09339672697113094,
+      "learning_rate": 0.0009611193732195254,
+      "loss": 1.4057,
+      "step": 1420
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.08382803898484295,
+      "learning_rate": 0.000961052031738097,
+      "loss": 1.4925,
+      "step": 1421
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.13267703370359948,
+      "learning_rate": 0.0009609846343525204,
+      "loss": 1.3566,
+      "step": 1422
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.09158613485789965,
+      "learning_rate": 0.0009609171810709676,
+      "loss": 1.4777,
+      "step": 1423
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.09234863197344007,
+      "learning_rate": 0.0009608496719016175,
+      "loss": 1.4714,
+      "step": 1424
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.08860250240147277,
+      "learning_rate": 0.0009607821068526559,
+      "loss": 1.5249,
+      "step": 1425
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.09605535819946916,
+      "learning_rate": 0.0009607144859322754,
+      "loss": 1.4138,
+      "step": 1426
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0910599326476764,
+      "learning_rate": 0.000960646809148675,
+      "loss": 1.393,
+      "step": 1427
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.08997170262147476,
+      "learning_rate": 0.000960579076510061,
+      "loss": 1.4954,
+      "step": 1428
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.09321262030864544,
+      "learning_rate": 0.0009605112880246462,
+      "loss": 1.4509,
+      "step": 1429
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.08691768022573092,
+      "learning_rate": 0.00096044344370065,
+      "loss": 1.6263,
+      "step": 1430
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.09869112519405952,
+      "learning_rate": 0.0009603755435462989,
+      "loss": 1.5487,
+      "step": 1431
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0929891550387196,
+      "learning_rate": 0.000960307587569826,
+      "loss": 1.5137,
+      "step": 1432
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.09569271308352895,
+      "learning_rate": 0.0009602395757794711,
+      "loss": 1.4777,
+      "step": 1433
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.09755014615911316,
+      "learning_rate": 0.000960171508183481,
+      "loss": 1.4682,
+      "step": 1434
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0972474147226531,
+      "learning_rate": 0.0009601033847901091,
+      "loss": 1.5142,
+      "step": 1435
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.08735796798182946,
+      "learning_rate": 0.0009600352056076154,
+      "loss": 1.4992,
+      "step": 1436
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.09119723664332861,
+      "learning_rate": 0.0009599669706442672,
+      "loss": 1.4638,
+      "step": 1437
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.12378601755710915,
+      "learning_rate": 0.000959898679908338,
+      "loss": 1.6005,
+      "step": 1438
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.10478109012333306,
+      "learning_rate": 0.0009598303334081085,
+      "loss": 1.4894,
+      "step": 1439
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.23570162511872436,
+      "learning_rate": 0.0009597619311518657,
+      "loss": 1.5644,
+      "step": 1440
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.10682124284391994,
+      "learning_rate": 0.0009596934731479036,
+      "loss": 1.5554,
+      "step": 1441
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.09295343956178324,
+      "learning_rate": 0.0009596249594045232,
+      "loss": 1.438,
+      "step": 1442
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.09713099725191914,
+      "learning_rate": 0.0009595563899300319,
+      "loss": 1.4028,
+      "step": 1443
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.08720151241947806,
+      "learning_rate": 0.000959487764732744,
+      "loss": 1.3532,
+      "step": 1444
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0886950843743924,
+      "learning_rate": 0.0009594190838209805,
+      "loss": 1.5788,
+      "step": 1445
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.08793516619885679,
+      "learning_rate": 0.0009593503472030692,
+      "loss": 1.3038,
+      "step": 1446
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.09144261031291229,
+      "learning_rate": 0.0009592815548873448,
+      "loss": 1.3458,
+      "step": 1447
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.09311955286422781,
+      "learning_rate": 0.0009592127068821484,
+      "loss": 1.5537,
+      "step": 1448
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.08723657754682848,
+      "learning_rate": 0.0009591438031958282,
+      "loss": 1.3372,
+      "step": 1449
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.09217854048138731,
+      "learning_rate": 0.0009590748438367388,
+      "loss": 1.5481,
+      "step": 1450
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.08500728494003494,
+      "learning_rate": 0.000959005828813242,
+      "loss": 1.4458,
+      "step": 1451
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0853018457118595,
+      "learning_rate": 0.0009589367581337061,
+      "loss": 1.424,
+      "step": 1452
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.09868598911419524,
+      "learning_rate": 0.000958867631806506,
+      "loss": 1.3293,
+      "step": 1453
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.08639367094360562,
+      "learning_rate": 0.0009587984498400235,
+      "loss": 1.3447,
+      "step": 1454
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.10359812768616228,
+      "learning_rate": 0.0009587292122426475,
+      "loss": 1.4611,
+      "step": 1455
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.11130702504760993,
+      "learning_rate": 0.0009586599190227728,
+      "loss": 1.4713,
+      "step": 1456
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.08535204266384723,
+      "learning_rate": 0.0009585905701888018,
+      "loss": 1.4174,
+      "step": 1457
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.09085953542296663,
+      "learning_rate": 0.0009585211657491431,
+      "loss": 1.4496,
+      "step": 1458
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.08848742414165559,
+      "learning_rate": 0.0009584517057122122,
+      "loss": 1.5021,
+      "step": 1459
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.08832297742063362,
+      "learning_rate": 0.0009583821900864315,
+      "loss": 1.494,
+      "step": 1460
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.08922415214466346,
+      "learning_rate": 0.0009583126188802302,
+      "loss": 1.5711,
+      "step": 1461
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.09999174702384374,
+      "learning_rate": 0.0009582429921020436,
+      "loss": 1.5518,
+      "step": 1462
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.09822083586247843,
+      "learning_rate": 0.0009581733097603145,
+      "loss": 1.3522,
+      "step": 1463
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.10384201828720639,
+      "learning_rate": 0.0009581035718634919,
+      "loss": 1.4244,
+      "step": 1464
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.09055085355297395,
+      "learning_rate": 0.0009580337784200319,
+      "loss": 1.5444,
+      "step": 1465
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.10518106325110177,
+      "learning_rate": 0.0009579639294383973,
+      "loss": 1.4364,
+      "step": 1466
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0967908735277497,
+      "learning_rate": 0.0009578940249270573,
+      "loss": 1.457,
+      "step": 1467
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.09207038006235847,
+      "learning_rate": 0.0009578240648944882,
+      "loss": 1.4476,
+      "step": 1468
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.10043286509181094,
+      "learning_rate": 0.000957754049349173,
+      "loss": 1.4658,
+      "step": 1469
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.09054729227475117,
+      "learning_rate": 0.0009576839782996012,
+      "loss": 1.4334,
+      "step": 1470
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.08377663291202349,
+      "learning_rate": 0.000957613851754269,
+      "loss": 1.4492,
+      "step": 1471
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.11060837805665544,
+      "learning_rate": 0.0009575436697216797,
+      "loss": 1.4966,
+      "step": 1472
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.09969183136235889,
+      "learning_rate": 0.0009574734322103431,
+      "loss": 1.5009,
+      "step": 1473
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.09967141930559241,
+      "learning_rate": 0.0009574031392287757,
+      "loss": 1.4341,
+      "step": 1474
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.09799525072093489,
+      "learning_rate": 0.0009573327907855006,
+      "loss": 1.4278,
+      "step": 1475
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.08604866662176865,
+      "learning_rate": 0.0009572623868890481,
+      "loss": 1.4587,
+      "step": 1476
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.09265415277930036,
+      "learning_rate": 0.0009571919275479548,
+      "loss": 1.4699,
+      "step": 1477
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.10062297583552167,
+      "learning_rate": 0.0009571214127707639,
+      "loss": 1.4039,
+      "step": 1478
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.09118049223741401,
+      "learning_rate": 0.000957050842566026,
+      "loss": 1.4081,
+      "step": 1479
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.08501681321065166,
+      "learning_rate": 0.0009569802169422976,
+      "loss": 1.4726,
+      "step": 1480
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.08422437325947779,
+      "learning_rate": 0.0009569095359081426,
+      "loss": 1.4094,
+      "step": 1481
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.09366136573080257,
+      "learning_rate": 0.000956838799472131,
+      "loss": 1.3966,
+      "step": 1482
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.08511044239192846,
+      "learning_rate": 0.00095676800764284,
+      "loss": 1.464,
+      "step": 1483
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.09087661694984668,
+      "learning_rate": 0.0009566971604288534,
+      "loss": 1.4871,
+      "step": 1484
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1192006425086519,
+      "learning_rate": 0.0009566262578387617,
+      "loss": 1.404,
+      "step": 1485
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.09748627055341871,
+      "learning_rate": 0.000956555299881162,
+      "loss": 1.3744,
+      "step": 1486
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.09993689583714448,
+      "learning_rate": 0.000956484286564658,
+      "loss": 1.5675,
+      "step": 1487
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.10586802308275939,
+      "learning_rate": 0.0009564132178978606,
+      "loss": 1.4087,
+      "step": 1488
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.09472377025775539,
+      "learning_rate": 0.0009563420938893871,
+      "loss": 1.4368,
+      "step": 1489
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.10012467647677623,
+      "learning_rate": 0.0009562709145478615,
+      "loss": 1.4937,
+      "step": 1490
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.10303077194874116,
+      "learning_rate": 0.0009561996798819145,
+      "loss": 1.4654,
+      "step": 1491
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.10113641879804244,
+      "learning_rate": 0.0009561283899001835,
+      "loss": 1.5285,
+      "step": 1492
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.09589102118147506,
+      "learning_rate": 0.0009560570446113128,
+      "loss": 1.4873,
+      "step": 1493
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.09126782272024617,
+      "learning_rate": 0.000955985644023953,
+      "loss": 1.5029,
+      "step": 1494
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.09001747858385097,
+      "learning_rate": 0.0009559141881467619,
+      "loss": 1.3982,
+      "step": 1495
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.10262554472649192,
+      "learning_rate": 0.0009558426769884039,
+      "loss": 1.5575,
+      "step": 1496
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.10556088502827525,
+      "learning_rate": 0.0009557711105575496,
+      "loss": 1.483,
+      "step": 1497
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.08980632344515092,
+      "learning_rate": 0.0009556994888628769,
+      "loss": 1.3076,
+      "step": 1498
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.090263236114018,
+      "learning_rate": 0.0009556278119130701,
+      "loss": 1.3918,
+      "step": 1499
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.09906411529902781,
+      "learning_rate": 0.0009555560797168204,
+      "loss": 1.5834,
+      "step": 1500
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.09241869135981694,
+      "learning_rate": 0.0009554842922828254,
+      "loss": 1.5699,
+      "step": 1501
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.09344833611604621,
+      "learning_rate": 0.0009554124496197897,
+      "loss": 1.3565,
+      "step": 1502
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.09286482726655389,
+      "learning_rate": 0.0009553405517364244,
+      "loss": 1.4,
+      "step": 1503
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.09532973376012242,
+      "learning_rate": 0.0009552685986414475,
+      "loss": 1.4654,
+      "step": 1504
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.09067680572772746,
+      "learning_rate": 0.0009551965903435835,
+      "loss": 1.4848,
+      "step": 1505
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1040664046288145,
+      "learning_rate": 0.0009551245268515636,
+      "loss": 1.4488,
+      "step": 1506
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.09429619546311979,
+      "learning_rate": 0.0009550524081741256,
+      "loss": 1.5326,
+      "step": 1507
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.10532072284542995,
+      "learning_rate": 0.0009549802343200145,
+      "loss": 1.3992,
+      "step": 1508
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.08912863106155763,
+      "learning_rate": 0.0009549080052979813,
+      "loss": 1.4448,
+      "step": 1509
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1003030251918555,
+      "learning_rate": 0.0009548357211167841,
+      "loss": 1.484,
+      "step": 1510
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.11243803152082767,
+      "learning_rate": 0.000954763381785188,
+      "loss": 1.4291,
+      "step": 1511
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1001886187096508,
+      "learning_rate": 0.0009546909873119636,
+      "loss": 1.4383,
+      "step": 1512
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.10231401178866115,
+      "learning_rate": 0.0009546185377058898,
+      "loss": 1.5126,
+      "step": 1513
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.104553472069901,
+      "learning_rate": 0.0009545460329757507,
+      "loss": 1.6627,
+      "step": 1514
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.08907543138688107,
+      "learning_rate": 0.0009544734731303382,
+      "loss": 1.5044,
+      "step": 1515
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.12886965839182934,
+      "learning_rate": 0.0009544008581784503,
+      "loss": 1.4528,
+      "step": 1516
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.09209918402576728,
+      "learning_rate": 0.0009543281881288918,
+      "loss": 1.468,
+      "step": 1517
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.09475658742189953,
+      "learning_rate": 0.0009542554629904741,
+      "loss": 1.3359,
+      "step": 1518
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.08776235492227999,
+      "learning_rate": 0.0009541826827720155,
+      "loss": 1.4278,
+      "step": 1519
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0973510550854703,
+      "learning_rate": 0.0009541098474823408,
+      "loss": 1.4128,
+      "step": 1520
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.08791160995249331,
+      "learning_rate": 0.0009540369571302815,
+      "loss": 1.4823,
+      "step": 1521
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1104543848288305,
+      "learning_rate": 0.0009539640117246759,
+      "loss": 1.474,
+      "step": 1522
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.09377728378127773,
+      "learning_rate": 0.0009538910112743687,
+      "loss": 1.5595,
+      "step": 1523
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.08352231010091461,
+      "learning_rate": 0.0009538179557882117,
+      "loss": 1.4802,
+      "step": 1524
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.10346239673786903,
+      "learning_rate": 0.000953744845275063,
+      "loss": 1.5015,
+      "step": 1525
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.10087785904923345,
+      "learning_rate": 0.0009536716797437875,
+      "loss": 1.5315,
+      "step": 1526
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.09040805736131872,
+      "learning_rate": 0.0009535984592032569,
+      "loss": 1.4685,
+      "step": 1527
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.12036277893925264,
+      "learning_rate": 0.0009535251836623491,
+      "loss": 1.39,
+      "step": 1528
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.09447315081864598,
+      "learning_rate": 0.0009534518531299494,
+      "loss": 1.5542,
+      "step": 1529
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.09873458361328201,
+      "learning_rate": 0.0009533784676149492,
+      "loss": 1.4234,
+      "step": 1530
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.09101268943669456,
+      "learning_rate": 0.0009533050271262467,
+      "loss": 1.4116,
+      "step": 1531
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.08897558088863952,
+      "learning_rate": 0.0009532315316727469,
+      "loss": 1.4038,
+      "step": 1532
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.11018902277583942,
+      "learning_rate": 0.0009531579812633615,
+      "loss": 1.4236,
+      "step": 1533
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.09125555762503207,
+      "learning_rate": 0.0009530843759070085,
+      "loss": 1.5578,
+      "step": 1534
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.09385367834635039,
+      "learning_rate": 0.0009530107156126129,
+      "loss": 1.5099,
+      "step": 1535
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.10683817721639628,
+      "learning_rate": 0.0009529370003891062,
+      "loss": 1.6342,
+      "step": 1536
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.08967972247987649,
+      "learning_rate": 0.0009528632302454268,
+      "loss": 1.4171,
+      "step": 1537
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.09753788211751449,
+      "learning_rate": 0.0009527894051905194,
+      "loss": 1.4745,
+      "step": 1538
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.09304012816220279,
+      "learning_rate": 0.0009527155252333357,
+      "loss": 1.5435,
+      "step": 1539
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.09474919727112473,
+      "learning_rate": 0.0009526415903828337,
+      "loss": 1.3861,
+      "step": 1540
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.07778975111369331,
+      "learning_rate": 0.0009525676006479784,
+      "loss": 1.4173,
+      "step": 1541
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.09020032560229345,
+      "learning_rate": 0.0009524935560377414,
+      "loss": 1.4258,
+      "step": 1542
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.09665307246990602,
+      "learning_rate": 0.0009524194565611006,
+      "loss": 1.4578,
+      "step": 1543
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.09800227849364905,
+      "learning_rate": 0.000952345302227041,
+      "loss": 1.367,
+      "step": 1544
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1039422730155937,
+      "learning_rate": 0.0009522710930445541,
+      "loss": 1.5948,
+      "step": 1545
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.08481742258534228,
+      "learning_rate": 0.000952196829022638,
+      "loss": 1.4945,
+      "step": 1546
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.08727988396404307,
+      "learning_rate": 0.0009521225101702973,
+      "loss": 1.4855,
+      "step": 1547
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.10107756346544976,
+      "learning_rate": 0.0009520481364965435,
+      "loss": 1.5414,
+      "step": 1548
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.08148558962045617,
+      "learning_rate": 0.0009519737080103948,
+      "loss": 1.4015,
+      "step": 1549
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.09117985257996772,
+      "learning_rate": 0.0009518992247208758,
+      "loss": 1.4934,
+      "step": 1550
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.10201481363275045,
+      "learning_rate": 0.000951824686637018,
+      "loss": 1.4628,
+      "step": 1551
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.09022581841205508,
+      "learning_rate": 0.000951750093767859,
+      "loss": 1.6013,
+      "step": 1552
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0953335804179374,
+      "learning_rate": 0.0009516754461224439,
+      "loss": 1.6041,
+      "step": 1553
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.09639577286814874,
+      "learning_rate": 0.0009516007437098238,
+      "loss": 1.554,
+      "step": 1554
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.09708674309060862,
+      "learning_rate": 0.0009515259865390564,
+      "loss": 1.3906,
+      "step": 1555
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.09782023489932025,
+      "learning_rate": 0.0009514511746192068,
+      "loss": 1.4169,
+      "step": 1556
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.08847376964278968,
+      "learning_rate": 0.0009513763079593456,
+      "loss": 1.489,
+      "step": 1557
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.09418702831874794,
+      "learning_rate": 0.0009513013865685511,
+      "loss": 1.4692,
+      "step": 1558
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0839674566660529,
+      "learning_rate": 0.0009512264104559077,
+      "loss": 1.4908,
+      "step": 1559
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.09666360133659271,
+      "learning_rate": 0.0009511513796305062,
+      "loss": 1.519,
+      "step": 1560
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.08317259532860914,
+      "learning_rate": 0.0009510762941014446,
+      "loss": 1.4297,
+      "step": 1561
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.09939011817545634,
+      "learning_rate": 0.0009510011538778274,
+      "loss": 1.5525,
+      "step": 1562
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.08154865062357083,
+      "learning_rate": 0.0009509259589687653,
+      "loss": 1.4989,
+      "step": 1563
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.09355040253719421,
+      "learning_rate": 0.000950850709383376,
+      "loss": 1.4582,
+      "step": 1564
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.09125277909364553,
+      "learning_rate": 0.0009507754051307841,
+      "loss": 1.5391,
+      "step": 1565
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.089502284335009,
+      "learning_rate": 0.0009507000462201201,
+      "loss": 1.543,
+      "step": 1566
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.08176070826278141,
+      "learning_rate": 0.0009506246326605219,
+      "loss": 1.5341,
+      "step": 1567
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.10665943107435298,
+      "learning_rate": 0.0009505491644611333,
+      "loss": 1.3918,
+      "step": 1568
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.09157527137567606,
+      "learning_rate": 0.0009504736416311053,
+      "loss": 1.5549,
+      "step": 1569
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.09586127992803163,
+      "learning_rate": 0.0009503980641795952,
+      "loss": 1.3776,
+      "step": 1570
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.09017513026010152,
+      "learning_rate": 0.0009503224321157671,
+      "loss": 1.4205,
+      "step": 1571
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.09741079955785971,
+      "learning_rate": 0.0009502467454487915,
+      "loss": 1.2827,
+      "step": 1572
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.10275736685666167,
+      "learning_rate": 0.0009501710041878458,
+      "loss": 1.4762,
+      "step": 1573
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.08569182696154201,
+      "learning_rate": 0.0009500952083421139,
+      "loss": 1.3994,
+      "step": 1574
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.09179001894163673,
+      "learning_rate": 0.0009500193579207863,
+      "loss": 1.5314,
+      "step": 1575
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.08676494976521502,
+      "learning_rate": 0.0009499434529330602,
+      "loss": 1.4571,
+      "step": 1576
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.09077194936057632,
+      "learning_rate": 0.000949867493388139,
+      "loss": 1.4957,
+      "step": 1577
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.10232739854820663,
+      "learning_rate": 0.0009497914792952333,
+      "loss": 1.5458,
+      "step": 1578
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.10029952697335129,
+      "learning_rate": 0.0009497154106635604,
+      "loss": 1.4446,
+      "step": 1579
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1156402004405788,
+      "learning_rate": 0.0009496392875023432,
+      "loss": 1.4567,
+      "step": 1580
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.09278124851672967,
+      "learning_rate": 0.0009495631098208124,
+      "loss": 1.4469,
+      "step": 1581
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.10394690523142269,
+      "learning_rate": 0.0009494868776282046,
+      "loss": 1.3859,
+      "step": 1582
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.09395591900009802,
+      "learning_rate": 0.0009494105909337633,
+      "loss": 1.5942,
+      "step": 1583
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.10303144085226874,
+      "learning_rate": 0.0009493342497467385,
+      "loss": 1.4295,
+      "step": 1584
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1037174632287724,
+      "learning_rate": 0.000949257854076387,
+      "loss": 1.4119,
+      "step": 1585
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.09553928851204532,
+      "learning_rate": 0.0009491814039319716,
+      "loss": 1.3801,
+      "step": 1586
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.09651654620253583,
+      "learning_rate": 0.0009491048993227625,
+      "loss": 1.4304,
+      "step": 1587
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.09455369631618026,
+      "learning_rate": 0.000949028340258036,
+      "loss": 1.5054,
+      "step": 1588
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.097039379644678,
+      "learning_rate": 0.0009489517267470753,
+      "loss": 1.5381,
+      "step": 1589
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.09872281104254445,
+      "learning_rate": 0.00094887505879917,
+      "loss": 1.4912,
+      "step": 1590
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.09155850201559067,
+      "learning_rate": 0.0009487983364236162,
+      "loss": 1.5075,
+      "step": 1591
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.08864721777779384,
+      "learning_rate": 0.0009487215596297169,
+      "loss": 1.5051,
+      "step": 1592
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.11145148010181871,
+      "learning_rate": 0.0009486447284267816,
+      "loss": 1.5501,
+      "step": 1593
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.09424231822009453,
+      "learning_rate": 0.0009485678428241262,
+      "loss": 1.4026,
+      "step": 1594
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0936796798826941,
+      "learning_rate": 0.0009484909028310734,
+      "loss": 1.4845,
+      "step": 1595
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.09572362072102075,
+      "learning_rate": 0.0009484139084569525,
+      "loss": 1.5573,
+      "step": 1596
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.10489996329436806,
+      "learning_rate": 0.0009483368597110992,
+      "loss": 1.3578,
+      "step": 1597
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.09287949322524128,
+      "learning_rate": 0.000948259756602856,
+      "loss": 1.3336,
+      "step": 1598
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.10417263018902385,
+      "learning_rate": 0.000948182599141572,
+      "loss": 1.5492,
+      "step": 1599
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.09698735516601625,
+      "learning_rate": 0.0009481053873366027,
+      "loss": 1.6122,
+      "step": 1600
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.10366781086993937,
+      "learning_rate": 0.0009480281211973103,
+      "loss": 1.4094,
+      "step": 1601
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.08318527269695515,
+      "learning_rate": 0.0009479508007330638,
+      "loss": 1.395,
+      "step": 1602
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.09637176359968291,
+      "learning_rate": 0.0009478734259532381,
+      "loss": 1.4232,
+      "step": 1603
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.10117308070962983,
+      "learning_rate": 0.0009477959968672156,
+      "loss": 1.4984,
+      "step": 1604
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.08695311552966221,
+      "learning_rate": 0.0009477185134843846,
+      "loss": 1.6225,
+      "step": 1605
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.08627423777317476,
+      "learning_rate": 0.0009476409758141405,
+      "loss": 1.5029,
+      "step": 1606
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.07656034118664419,
+      "learning_rate": 0.0009475633838658847,
+      "loss": 1.3344,
+      "step": 1607
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0928219643216207,
+      "learning_rate": 0.0009474857376490257,
+      "loss": 1.4961,
+      "step": 1608
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.10412888114786652,
+      "learning_rate": 0.0009474080371729782,
+      "loss": 1.3727,
+      "step": 1609
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.09068058658406214,
+      "learning_rate": 0.0009473302824471637,
+      "loss": 1.5552,
+      "step": 1610
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.08843922927390716,
+      "learning_rate": 0.0009472524734810104,
+      "loss": 1.5213,
+      "step": 1611
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0915814113209529,
+      "learning_rate": 0.0009471746102839527,
+      "loss": 1.5118,
+      "step": 1612
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.06836751650965489,
+      "learning_rate": 0.0009470966928654319,
+      "loss": 1.4489,
+      "step": 1613
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.07747282566984308,
+      "learning_rate": 0.0009470187212348957,
+      "loss": 1.4894,
+      "step": 1614
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.07132050314074537,
+      "learning_rate": 0.0009469406954017985,
+      "loss": 1.5092,
+      "step": 1615
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.08277949038549247,
+      "learning_rate": 0.000946862615375601,
+      "loss": 1.5643,
+      "step": 1616
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.09289209657839693,
+      "learning_rate": 0.000946784481165771,
+      "loss": 1.4531,
+      "step": 1617
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.10295338032257036,
+      "learning_rate": 0.0009467062927817822,
+      "loss": 1.4354,
+      "step": 1618
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.09135645998922941,
+      "learning_rate": 0.0009466280502331154,
+      "loss": 1.4541,
+      "step": 1619
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.11184134475928127,
+      "learning_rate": 0.0009465497535292579,
+      "loss": 1.5205,
+      "step": 1620
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.09717044875080767,
+      "learning_rate": 0.0009464714026797031,
+      "loss": 1.4884,
+      "step": 1621
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.09757801259769658,
+      "learning_rate": 0.0009463929976939515,
+      "loss": 1.425,
+      "step": 1622
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0922656442309496,
+      "learning_rate": 0.0009463145385815102,
+      "loss": 1.4102,
+      "step": 1623
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.10056802441985584,
+      "learning_rate": 0.0009462360253518923,
+      "loss": 1.5064,
+      "step": 1624
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.09757900282309952,
+      "learning_rate": 0.0009461574580146179,
+      "loss": 1.4064,
+      "step": 1625
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.09223895834044753,
+      "learning_rate": 0.0009460788365792135,
+      "loss": 1.4853,
+      "step": 1626
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.08517631548088779,
+      "learning_rate": 0.0009460001610552124,
+      "loss": 1.3385,
+      "step": 1627
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.09644877186160637,
+      "learning_rate": 0.000945921431452154,
+      "loss": 1.3694,
+      "step": 1628
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.10147088301807176,
+      "learning_rate": 0.000945842647779585,
+      "loss": 1.4762,
+      "step": 1629
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.10594360644469922,
+      "learning_rate": 0.0009457638100470577,
+      "loss": 1.5326,
+      "step": 1630
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0964403054717145,
+      "learning_rate": 0.0009456849182641317,
+      "loss": 1.4083,
+      "step": 1631
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.08982800262769298,
+      "learning_rate": 0.000945605972440373,
+      "loss": 1.3163,
+      "step": 1632
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.09752836436861666,
+      "learning_rate": 0.0009455269725853536,
+      "loss": 1.3745,
+      "step": 1633
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.09479524516483495,
+      "learning_rate": 0.000945447918708653,
+      "loss": 1.3966,
+      "step": 1634
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.08660628978231161,
+      "learning_rate": 0.0009453688108198566,
+      "loss": 1.3603,
+      "step": 1635
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.08391899236342358,
+      "learning_rate": 0.0009452896489285563,
+      "loss": 1.4881,
+      "step": 1636
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.10847952104104855,
+      "learning_rate": 0.0009452104330443511,
+      "loss": 1.5492,
+      "step": 1637
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.08318537065982123,
+      "learning_rate": 0.0009451311631768459,
+      "loss": 1.5448,
+      "step": 1638
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.09468223699769081,
+      "learning_rate": 0.0009450518393356527,
+      "loss": 1.4535,
+      "step": 1639
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.10397951977441247,
+      "learning_rate": 0.0009449724615303894,
+      "loss": 1.4966,
+      "step": 1640
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.10389824321917117,
+      "learning_rate": 0.0009448930297706813,
+      "loss": 1.484,
+      "step": 1641
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.08499587782031137,
+      "learning_rate": 0.0009448135440661595,
+      "loss": 1.4936,
+      "step": 1642
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.08757048555474968,
+      "learning_rate": 0.0009447340044264619,
+      "loss": 1.4992,
+      "step": 1643
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.09261434742761426,
+      "learning_rate": 0.0009446544108612331,
+      "loss": 1.4285,
+      "step": 1644
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.08674281372806956,
+      "learning_rate": 0.0009445747633801241,
+      "loss": 1.4837,
+      "step": 1645
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.08584946897397816,
+      "learning_rate": 0.0009444950619927924,
+      "loss": 1.5338,
+      "step": 1646
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.11737706230601937,
+      "learning_rate": 0.0009444153067089019,
+      "loss": 1.5565,
+      "step": 1647
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.09096614220056518,
+      "learning_rate": 0.0009443354975381234,
+      "loss": 1.382,
+      "step": 1648
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.09032154151221224,
+      "learning_rate": 0.0009442556344901339,
+      "loss": 1.3891,
+      "step": 1649
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1027879079616311,
+      "learning_rate": 0.000944175717574617,
+      "loss": 1.551,
+      "step": 1650
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.08719485311068445,
+      "learning_rate": 0.0009440957468012632,
+      "loss": 1.4096,
+      "step": 1651
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.09352779878047505,
+      "learning_rate": 0.0009440157221797692,
+      "loss": 1.4714,
+      "step": 1652
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.10486400709151468,
+      "learning_rate": 0.0009439356437198379,
+      "loss": 1.4014,
+      "step": 1653
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.08752718991568993,
+      "learning_rate": 0.0009438555114311795,
+      "loss": 1.3704,
+      "step": 1654
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.11686969419920402,
+      "learning_rate": 0.00094377532532351,
+      "loss": 1.3982,
+      "step": 1655
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0973101634579299,
+      "learning_rate": 0.0009436950854065524,
+      "loss": 1.5212,
+      "step": 1656
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.09899858128911071,
+      "learning_rate": 0.0009436147916900361,
+      "loss": 1.5066,
+      "step": 1657
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.11649096433856834,
+      "learning_rate": 0.0009435344441836969,
+      "loss": 1.4407,
+      "step": 1658
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0953453538897908,
+      "learning_rate": 0.0009434540428972772,
+      "loss": 1.4545,
+      "step": 1659
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.10133785841493663,
+      "learning_rate": 0.0009433735878405261,
+      "loss": 1.4837,
+      "step": 1660
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.10015520209464993,
+      "learning_rate": 0.0009432930790231988,
+      "loss": 1.5321,
+      "step": 1661
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.08948532164322566,
+      "learning_rate": 0.0009432125164550576,
+      "loss": 1.4165,
+      "step": 1662
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.09658583989362529,
+      "learning_rate": 0.0009431319001458704,
+      "loss": 1.4593,
+      "step": 1663
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.10145569880361127,
+      "learning_rate": 0.000943051230105413,
+      "loss": 1.426,
+      "step": 1664
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.09816258060491967,
+      "learning_rate": 0.0009429705063434664,
+      "loss": 1.4223,
+      "step": 1665
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.09250427904332187,
+      "learning_rate": 0.0009428897288698188,
+      "loss": 1.5038,
+      "step": 1666
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0853389109743307,
+      "learning_rate": 0.0009428088976942646,
+      "loss": 1.5678,
+      "step": 1667
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.07695692774289016,
+      "learning_rate": 0.0009427280128266049,
+      "loss": 1.4343,
+      "step": 1668
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.10008107955257982,
+      "learning_rate": 0.0009426470742766476,
+      "loss": 1.4508,
+      "step": 1669
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.08818370379739952,
+      "learning_rate": 0.0009425660820542062,
+      "loss": 1.4621,
+      "step": 1670
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0852469680143305,
+      "learning_rate": 0.0009424850361691017,
+      "loss": 1.5184,
+      "step": 1671
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.081114977796592,
+      "learning_rate": 0.0009424039366311612,
+      "loss": 1.4093,
+      "step": 1672
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.10402276311434469,
+      "learning_rate": 0.0009423227834502181,
+      "loss": 1.5449,
+      "step": 1673
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.08138128486557042,
+      "learning_rate": 0.0009422415766361126,
+      "loss": 1.4053,
+      "step": 1674
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.09250871509940908,
+      "learning_rate": 0.0009421603161986913,
+      "loss": 1.4838,
+      "step": 1675
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.09799638166091465,
+      "learning_rate": 0.0009420790021478072,
+      "loss": 1.3309,
+      "step": 1676
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.07799924026862029,
+      "learning_rate": 0.00094199763449332,
+      "loss": 1.3402,
+      "step": 1677
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.09607736488537287,
+      "learning_rate": 0.0009419162132450961,
+      "loss": 1.4766,
+      "step": 1678
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.08245856386036492,
+      "learning_rate": 0.0009418347384130076,
+      "loss": 1.3557,
+      "step": 1679
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.08653876091437132,
+      "learning_rate": 0.000941753210006934,
+      "loss": 1.4636,
+      "step": 1680
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.09659732334971687,
+      "learning_rate": 0.0009416716280367606,
+      "loss": 1.5785,
+      "step": 1681
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.09251463831839435,
+      "learning_rate": 0.0009415899925123795,
+      "loss": 1.4092,
+      "step": 1682
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0885401579842873,
+      "learning_rate": 0.0009415083034436895,
+      "loss": 1.4988,
+      "step": 1683
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.08458310827754215,
+      "learning_rate": 0.0009414265608405956,
+      "loss": 1.453,
+      "step": 1684
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.09539966740345308,
+      "learning_rate": 0.0009413447647130096,
+      "loss": 1.5588,
+      "step": 1685
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.087828154350039,
+      "learning_rate": 0.0009412629150708492,
+      "loss": 1.4878,
+      "step": 1686
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.09293970968778822,
+      "learning_rate": 0.0009411810119240389,
+      "loss": 1.4289,
+      "step": 1687
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.09332098625509935,
+      "learning_rate": 0.00094109905528251,
+      "loss": 1.3876,
+      "step": 1688
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.09483816513762469,
+      "learning_rate": 0.0009410170451562001,
+      "loss": 1.4632,
+      "step": 1689
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0878505960015256,
+      "learning_rate": 0.000940934981555053,
+      "loss": 1.4356,
+      "step": 1690
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.10200709452803336,
+      "learning_rate": 0.000940852864489019,
+      "loss": 1.3956,
+      "step": 1691
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.09522477733928604,
+      "learning_rate": 0.0009407706939680556,
+      "loss": 1.3232,
+      "step": 1692
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.08411835086532109,
+      "learning_rate": 0.0009406884700021259,
+      "loss": 1.5153,
+      "step": 1693
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.09014720566437971,
+      "learning_rate": 0.0009406061926012,
+      "loss": 1.546,
+      "step": 1694
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.07605071855611331,
+      "learning_rate": 0.0009405238617752543,
+      "loss": 1.4299,
+      "step": 1695
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.080623963055023,
+      "learning_rate": 0.0009404414775342715,
+      "loss": 1.4253,
+      "step": 1696
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.09427266407825832,
+      "learning_rate": 0.0009403590398882411,
+      "loss": 1.4959,
+      "step": 1697
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.07474659358381536,
+      "learning_rate": 0.0009402765488471591,
+      "loss": 1.3768,
+      "step": 1698
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.09176750550282939,
+      "learning_rate": 0.0009401940044210276,
+      "loss": 1.4923,
+      "step": 1699
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.10027187119937282,
+      "learning_rate": 0.0009401114066198555,
+      "loss": 1.4703,
+      "step": 1700
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.09729007735305473,
+      "learning_rate": 0.000940028755453658,
+      "loss": 1.4746,
+      "step": 1701
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.08153229489393464,
+      "learning_rate": 0.000939946050932457,
+      "loss": 1.4445,
+      "step": 1702
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.08757343940935985,
+      "learning_rate": 0.0009398632930662805,
+      "loss": 1.5132,
+      "step": 1703
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.08901895804515486,
+      "learning_rate": 0.0009397804818651634,
+      "loss": 1.2978,
+      "step": 1704
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.10619612933936191,
+      "learning_rate": 0.0009396976173391466,
+      "loss": 1.507,
+      "step": 1705
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0917857472797286,
+      "learning_rate": 0.000939614699498278,
+      "loss": 1.4033,
+      "step": 1706
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1134247857054161,
+      "learning_rate": 0.0009395317283526113,
+      "loss": 1.5523,
+      "step": 1707
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.09581787178579383,
+      "learning_rate": 0.0009394487039122072,
+      "loss": 1.4631,
+      "step": 1708
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.09259611299053887,
+      "learning_rate": 0.0009393656261871328,
+      "loss": 1.4405,
+      "step": 1709
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.09297130715435674,
+      "learning_rate": 0.0009392824951874617,
+      "loss": 1.6184,
+      "step": 1710
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.08872509563911614,
+      "learning_rate": 0.0009391993109232735,
+      "loss": 1.4219,
+      "step": 1711
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.08836517965035444,
+      "learning_rate": 0.0009391160734046547,
+      "loss": 1.5004,
+      "step": 1712
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.09390231142694527,
+      "learning_rate": 0.000939032782641698,
+      "loss": 1.4052,
+      "step": 1713
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.09998203205571347,
+      "learning_rate": 0.000938949438644503,
+      "loss": 1.4534,
+      "step": 1714
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.10024244141659862,
+      "learning_rate": 0.0009388660414231751,
+      "loss": 1.597,
+      "step": 1715
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.09070174363207298,
+      "learning_rate": 0.0009387825909878269,
+      "loss": 1.431,
+      "step": 1716
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.09080214985361508,
+      "learning_rate": 0.0009386990873485767,
+      "loss": 1.4065,
+      "step": 1717
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1004759223908107,
+      "learning_rate": 0.0009386155305155497,
+      "loss": 1.4715,
+      "step": 1718
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.09826529128818194,
+      "learning_rate": 0.0009385319204988776,
+      "loss": 1.436,
+      "step": 1719
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.09212018460786485,
+      "learning_rate": 0.0009384482573086983,
+      "loss": 1.4487,
+      "step": 1720
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.09053519573952286,
+      "learning_rate": 0.000938364540955156,
+      "loss": 1.6079,
+      "step": 1721
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.09161427273443738,
+      "learning_rate": 0.0009382807714484021,
+      "loss": 1.5146,
+      "step": 1722
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.09193193868164781,
+      "learning_rate": 0.0009381969487985935,
+      "loss": 1.4547,
+      "step": 1723
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.08852240861982075,
+      "learning_rate": 0.0009381130730158943,
+      "loss": 1.5276,
+      "step": 1724
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0914904802379876,
+      "learning_rate": 0.0009380291441104747,
+      "loss": 1.6143,
+      "step": 1725
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.08260988756714899,
+      "learning_rate": 0.000937945162092511,
+      "loss": 1.4669,
+      "step": 1726
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.10181657253275951,
+      "learning_rate": 0.0009378611269721866,
+      "loss": 1.4018,
+      "step": 1727
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.09565750562612672,
+      "learning_rate": 0.0009377770387596911,
+      "loss": 1.5622,
+      "step": 1728
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.08427791252695747,
+      "learning_rate": 0.0009376928974652205,
+      "loss": 1.494,
+      "step": 1729
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.08475458912711241,
+      "learning_rate": 0.0009376087030989771,
+      "loss": 1.4584,
+      "step": 1730
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.09136780435680898,
+      "learning_rate": 0.0009375244556711695,
+      "loss": 1.4727,
+      "step": 1731
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.08392899372822892,
+      "learning_rate": 0.0009374401551920135,
+      "loss": 1.4106,
+      "step": 1732
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.07494760250079167,
+      "learning_rate": 0.0009373558016717306,
+      "loss": 1.3448,
+      "step": 1733
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.10183183783607534,
+      "learning_rate": 0.000937271395120549,
+      "loss": 1.5347,
+      "step": 1734
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.09490611937533996,
+      "learning_rate": 0.0009371869355487031,
+      "loss": 1.5137,
+      "step": 1735
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.09095625419563282,
+      "learning_rate": 0.0009371024229664341,
+      "loss": 1.5858,
+      "step": 1736
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.09897520105592522,
+      "learning_rate": 0.0009370178573839894,
+      "loss": 1.502,
+      "step": 1737
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.08026803716653642,
+      "learning_rate": 0.000936933238811623,
+      "loss": 1.3948,
+      "step": 1738
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.08326463099403736,
+      "learning_rate": 0.0009368485672595948,
+      "loss": 1.3566,
+      "step": 1739
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.08141481156595314,
+      "learning_rate": 0.000936763842738172,
+      "loss": 1.4374,
+      "step": 1740
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0986349039516981,
+      "learning_rate": 0.0009366790652576274,
+      "loss": 1.5395,
+      "step": 1741
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.10291774398160058,
+      "learning_rate": 0.0009365942348282405,
+      "loss": 1.5838,
+      "step": 1742
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.10217497765301305,
+      "learning_rate": 0.0009365093514602978,
+      "loss": 1.5123,
+      "step": 1743
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.08112289391270511,
+      "learning_rate": 0.0009364244151640913,
+      "loss": 1.4214,
+      "step": 1744
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.08252851448595519,
+      "learning_rate": 0.0009363394259499197,
+      "loss": 1.5329,
+      "step": 1745
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.10139278527828859,
+      "learning_rate": 0.0009362543838280884,
+      "loss": 1.5226,
+      "step": 1746
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.09091939187498209,
+      "learning_rate": 0.0009361692888089092,
+      "loss": 1.5196,
+      "step": 1747
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.08879705035046179,
+      "learning_rate": 0.0009360841409027001,
+      "loss": 1.3142,
+      "step": 1748
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1033063834090903,
+      "learning_rate": 0.0009359989401197852,
+      "loss": 1.4918,
+      "step": 1749
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.08250221798770896,
+      "learning_rate": 0.000935913686470496,
+      "loss": 1.4754,
+      "step": 1750
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.09952553090757268,
+      "learning_rate": 0.0009358283799651694,
+      "loss": 1.5277,
+      "step": 1751
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.08770051236655776,
+      "learning_rate": 0.0009357430206141492,
+      "loss": 1.5348,
+      "step": 1752
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.08456069157036568,
+      "learning_rate": 0.0009356576084277855,
+      "loss": 1.594,
+      "step": 1753
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.08688679813984553,
+      "learning_rate": 0.0009355721434164348,
+      "loss": 1.4007,
+      "step": 1754
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.08459334076466057,
+      "learning_rate": 0.0009354866255904602,
+      "loss": 1.3808,
+      "step": 1755
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0750770133707275,
+      "learning_rate": 0.0009354010549602308,
+      "loss": 1.4116,
+      "step": 1756
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0810999858623209,
+      "learning_rate": 0.0009353154315361223,
+      "loss": 1.392,
+      "step": 1757
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.07561264712057969,
+      "learning_rate": 0.0009352297553285172,
+      "loss": 1.5725,
+      "step": 1758
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.07689189356736299,
+      "learning_rate": 0.0009351440263478036,
+      "loss": 1.3871,
+      "step": 1759
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.07656862328089326,
+      "learning_rate": 0.0009350582446043767,
+      "loss": 1.4561,
+      "step": 1760
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.07421858530277249,
+      "learning_rate": 0.0009349724101086379,
+      "loss": 1.3506,
+      "step": 1761
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.09125757959767364,
+      "learning_rate": 0.0009348865228709947,
+      "loss": 1.3664,
+      "step": 1762
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.08458607011462088,
+      "learning_rate": 0.0009348005829018612,
+      "loss": 1.3518,
+      "step": 1763
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.07846569043882619,
+      "learning_rate": 0.0009347145902116582,
+      "loss": 1.4436,
+      "step": 1764
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.07880792008907359,
+      "learning_rate": 0.0009346285448108124,
+      "loss": 1.3728,
+      "step": 1765
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.08379065827759669,
+      "learning_rate": 0.0009345424467097572,
+      "loss": 1.4667,
+      "step": 1766
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.07785121808069187,
+      "learning_rate": 0.0009344562959189321,
+      "loss": 1.4439,
+      "step": 1767
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.07867477807500672,
+      "learning_rate": 0.0009343700924487835,
+      "loss": 1.4389,
+      "step": 1768
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0823946804001334,
+      "learning_rate": 0.0009342838363097635,
+      "loss": 1.5018,
+      "step": 1769
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.08063914021381803,
+      "learning_rate": 0.0009341975275123313,
+      "loss": 1.406,
+      "step": 1770
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.08547202214736771,
+      "learning_rate": 0.0009341111660669519,
+      "loss": 1.328,
+      "step": 1771
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.10723639210392595,
+      "learning_rate": 0.0009340247519840969,
+      "loss": 1.5986,
+      "step": 1772
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.08744751839005566,
+      "learning_rate": 0.0009339382852742446,
+      "loss": 1.4943,
+      "step": 1773
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0947836846245832,
+      "learning_rate": 0.0009338517659478791,
+      "loss": 1.4841,
+      "step": 1774
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.07957752401569529,
+      "learning_rate": 0.0009337651940154914,
+      "loss": 1.415,
+      "step": 1775
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.09486280553870081,
+      "learning_rate": 0.0009336785694875785,
+      "loss": 1.4312,
+      "step": 1776
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0946131450018491,
+      "learning_rate": 0.0009335918923746438,
+      "loss": 1.4232,
+      "step": 1777
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0869396458702221,
+      "learning_rate": 0.0009335051626871973,
+      "loss": 1.416,
+      "step": 1778
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.09052244814090689,
+      "learning_rate": 0.0009334183804357555,
+      "loss": 1.4041,
+      "step": 1779
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.11469431378434027,
+      "learning_rate": 0.0009333315456308407,
+      "loss": 1.4579,
+      "step": 1780
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0906512085860017,
+      "learning_rate": 0.0009332446582829821,
+      "loss": 1.5208,
+      "step": 1781
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.09416622905854574,
+      "learning_rate": 0.0009331577184027149,
+      "loss": 1.3891,
+      "step": 1782
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.09651543652014465,
+      "learning_rate": 0.0009330707260005813,
+      "loss": 1.6078,
+      "step": 1783
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.09579404336719885,
+      "learning_rate": 0.000932983681087129,
+      "loss": 1.52,
+      "step": 1784
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.09205179631972425,
+      "learning_rate": 0.0009328965836729127,
+      "loss": 1.4295,
+      "step": 1785
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.10099213256917071,
+      "learning_rate": 0.0009328094337684932,
+      "loss": 1.5516,
+      "step": 1786
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.08243984686364722,
+      "learning_rate": 0.0009327222313844376,
+      "loss": 1.3988,
+      "step": 1787
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.10671922892355516,
+      "learning_rate": 0.0009326349765313199,
+      "loss": 1.3028,
+      "step": 1788
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.10306132239894458,
+      "learning_rate": 0.0009325476692197197,
+      "loss": 1.4138,
+      "step": 1789
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.09456390248444875,
+      "learning_rate": 0.0009324603094602232,
+      "loss": 1.5466,
+      "step": 1790
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.08567679617373725,
+      "learning_rate": 0.0009323728972634234,
+      "loss": 1.4125,
+      "step": 1791
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0866261102864471,
+      "learning_rate": 0.0009322854326399192,
+      "loss": 1.3749,
+      "step": 1792
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.10008267779048576,
+      "learning_rate": 0.000932197915600316,
+      "loss": 1.4912,
+      "step": 1793
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.09628776139314607,
+      "learning_rate": 0.0009321103461552254,
+      "loss": 1.5609,
+      "step": 1794
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.09887887458347153,
+      "learning_rate": 0.0009320227243152657,
+      "loss": 1.5033,
+      "step": 1795
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.08765493554091913,
+      "learning_rate": 0.0009319350500910612,
+      "loss": 1.4827,
+      "step": 1796
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.09325239873922055,
+      "learning_rate": 0.0009318473234932428,
+      "loss": 1.3376,
+      "step": 1797
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.08425381736541987,
+      "learning_rate": 0.0009317595445324476,
+      "loss": 1.4944,
+      "step": 1798
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0966079305940004,
+      "learning_rate": 0.0009316717132193192,
+      "loss": 1.5316,
+      "step": 1799
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.08757180961711196,
+      "learning_rate": 0.0009315838295645074,
+      "loss": 1.4882,
+      "step": 1800
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.08662799707706155,
+      "learning_rate": 0.0009314958935786683,
+      "loss": 1.46,
+      "step": 1801
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.08311178903942941,
+      "learning_rate": 0.0009314079052724644,
+      "loss": 1.574,
+      "step": 1802
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.09580809932720677,
+      "learning_rate": 0.0009313198646565648,
+      "loss": 1.4954,
+      "step": 1803
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.09515422211680134,
+      "learning_rate": 0.0009312317717416448,
+      "loss": 1.4624,
+      "step": 1804
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.09528198931281466,
+      "learning_rate": 0.0009311436265383856,
+      "loss": 1.5618,
+      "step": 1805
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1036689450583713,
+      "learning_rate": 0.0009310554290574753,
+      "loss": 1.4556,
+      "step": 1806
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.08480419301117718,
+      "learning_rate": 0.0009309671793096082,
+      "loss": 1.4717,
+      "step": 1807
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.08133107911744367,
+      "learning_rate": 0.0009308788773054848,
+      "loss": 1.3288,
+      "step": 1808
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.08016797148504663,
+      "learning_rate": 0.0009307905230558121,
+      "loss": 1.5397,
+      "step": 1809
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0846677701501156,
+      "learning_rate": 0.0009307021165713033,
+      "loss": 1.4324,
+      "step": 1810
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.07570819301145869,
+      "learning_rate": 0.000930613657862678,
+      "loss": 1.5013,
+      "step": 1811
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.09220020986485801,
+      "learning_rate": 0.0009305251469406622,
+      "loss": 1.3811,
+      "step": 1812
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.08711057852439665,
+      "learning_rate": 0.000930436583815988,
+      "loss": 1.5304,
+      "step": 1813
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0801584114606825,
+      "learning_rate": 0.0009303479684993942,
+      "loss": 1.4914,
+      "step": 1814
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.08411107652644935,
+      "learning_rate": 0.0009302593010016254,
+      "loss": 1.309,
+      "step": 1815
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.08740716260274044,
+      "learning_rate": 0.0009301705813334331,
+      "loss": 1.4615,
+      "step": 1816
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.08826543969180319,
+      "learning_rate": 0.000930081809505575,
+      "loss": 1.3672,
+      "step": 1817
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.08663114298057124,
+      "learning_rate": 0.0009299929855288145,
+      "loss": 1.3388,
+      "step": 1818
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.09701128888327752,
+      "learning_rate": 0.0009299041094139222,
+      "loss": 1.5865,
+      "step": 1819
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.07635149578067951,
+      "learning_rate": 0.0009298151811716744,
+      "loss": 1.5124,
+      "step": 1820
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.08728723564923248,
+      "learning_rate": 0.0009297262008128543,
+      "loss": 1.5027,
+      "step": 1821
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.11155604154795036,
+      "learning_rate": 0.0009296371683482508,
+      "loss": 1.3568,
+      "step": 1822
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.07719213570012251,
+      "learning_rate": 0.0009295480837886594,
+      "loss": 1.4425,
+      "step": 1823
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0911625047888657,
+      "learning_rate": 0.0009294589471448819,
+      "loss": 1.4705,
+      "step": 1824
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.09613111465408736,
+      "learning_rate": 0.0009293697584277265,
+      "loss": 1.4979,
+      "step": 1825
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.08512139925092915,
+      "learning_rate": 0.0009292805176480077,
+      "loss": 1.3785,
+      "step": 1826
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.09262471827262757,
+      "learning_rate": 0.0009291912248165461,
+      "loss": 1.4475,
+      "step": 1827
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.09367459711998498,
+      "learning_rate": 0.0009291018799441691,
+      "loss": 1.4571,
+      "step": 1828
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.09291744975591652,
+      "learning_rate": 0.0009290124830417096,
+      "loss": 1.6161,
+      "step": 1829
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.091514229624702,
+      "learning_rate": 0.0009289230341200075,
+      "loss": 1.4075,
+      "step": 1830
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.09264404283002275,
+      "learning_rate": 0.0009288335331899088,
+      "loss": 1.4796,
+      "step": 1831
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.09373072460820865,
+      "learning_rate": 0.0009287439802622659,
+      "loss": 1.4221,
+      "step": 1832
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.09399036796231679,
+      "learning_rate": 0.0009286543753479372,
+      "loss": 1.3825,
+      "step": 1833
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.08289694956094026,
+      "learning_rate": 0.0009285647184577877,
+      "loss": 1.477,
+      "step": 1834
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.08546486711268557,
+      "learning_rate": 0.0009284750096026887,
+      "loss": 1.3694,
+      "step": 1835
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.09047798016590947,
+      "learning_rate": 0.0009283852487935174,
+      "loss": 1.4847,
+      "step": 1836
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.10009191077886122,
+      "learning_rate": 0.0009282954360411577,
+      "loss": 1.4234,
+      "step": 1837
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0841123623143838,
+      "learning_rate": 0.0009282055713565001,
+      "loss": 1.5119,
+      "step": 1838
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0915824218936105,
+      "learning_rate": 0.0009281156547504408,
+      "loss": 1.4167,
+      "step": 1839
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.08648630759897831,
+      "learning_rate": 0.0009280256862338822,
+      "loss": 1.3795,
+      "step": 1840
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.08526753456498552,
+      "learning_rate": 0.0009279356658177336,
+      "loss": 1.3847,
+      "step": 1841
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.09626763852163273,
+      "learning_rate": 0.0009278455935129102,
+      "loss": 1.5616,
+      "step": 1842
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.08816486594899263,
+      "learning_rate": 0.0009277554693303337,
+      "loss": 1.53,
+      "step": 1843
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.09226706367835856,
+      "learning_rate": 0.0009276652932809315,
+      "loss": 1.4491,
+      "step": 1844
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.10737839815995892,
+      "learning_rate": 0.0009275750653756384,
+      "loss": 1.3005,
+      "step": 1845
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0827159786371698,
+      "learning_rate": 0.0009274847856253945,
+      "loss": 1.3935,
+      "step": 1846
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.08930362176980898,
+      "learning_rate": 0.0009273944540411465,
+      "loss": 1.4578,
+      "step": 1847
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.08293126307134922,
+      "learning_rate": 0.0009273040706338476,
+      "loss": 1.4298,
+      "step": 1848
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.09167982308662573,
+      "learning_rate": 0.0009272136354144569,
+      "loss": 1.4165,
+      "step": 1849
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.08142752813431287,
+      "learning_rate": 0.0009271231483939402,
+      "loss": 1.413,
+      "step": 1850
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.08088682067099412,
+      "learning_rate": 0.0009270326095832691,
+      "loss": 1.5194,
+      "step": 1851
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.080201882006416,
+      "learning_rate": 0.0009269420189934219,
+      "loss": 1.5184,
+      "step": 1852
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.08617551581904413,
+      "learning_rate": 0.000926851376635383,
+      "loss": 1.3672,
+      "step": 1853
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.08943700327583678,
+      "learning_rate": 0.0009267606825201433,
+      "loss": 1.4141,
+      "step": 1854
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0788005263395767,
+      "learning_rate": 0.0009266699366586993,
+      "loss": 1.4555,
+      "step": 1855
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.08619240998799432,
+      "learning_rate": 0.0009265791390620546,
+      "loss": 1.4135,
+      "step": 1856
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0801316495282694,
+      "learning_rate": 0.0009264882897412188,
+      "loss": 1.4703,
+      "step": 1857
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.08894081393129301,
+      "learning_rate": 0.0009263973887072074,
+      "loss": 1.5404,
+      "step": 1858
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.08415426907162747,
+      "learning_rate": 0.0009263064359710427,
+      "loss": 1.3856,
+      "step": 1859
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.08494026519042301,
+      "learning_rate": 0.0009262154315437528,
+      "loss": 1.3861,
+      "step": 1860
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.08459923609704402,
+      "learning_rate": 0.0009261243754363725,
+      "loss": 1.4656,
+      "step": 1861
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.08646099536983778,
+      "learning_rate": 0.0009260332676599425,
+      "loss": 1.4025,
+      "step": 1862
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0833393446485166,
+      "learning_rate": 0.0009259421082255103,
+      "loss": 1.313,
+      "step": 1863
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.10105293160333415,
+      "learning_rate": 0.0009258508971441288,
+      "loss": 1.4142,
+      "step": 1864
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.07753588468155745,
+      "learning_rate": 0.0009257596344268579,
+      "loss": 1.4151,
+      "step": 1865
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.07843557406004462,
+      "learning_rate": 0.0009256683200847637,
+      "loss": 1.5059,
+      "step": 1866
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.09283600002740316,
+      "learning_rate": 0.000925576954128918,
+      "loss": 1.4546,
+      "step": 1867
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.08221451122832747,
+      "learning_rate": 0.0009254855365703995,
+      "loss": 1.4822,
+      "step": 1868
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.09696442404894645,
+      "learning_rate": 0.0009253940674202929,
+      "loss": 1.5163,
+      "step": 1869
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.09377079226474856,
+      "learning_rate": 0.000925302546689689,
+      "loss": 1.5331,
+      "step": 1870
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.08495042450853789,
+      "learning_rate": 0.000925210974389685,
+      "loss": 1.4877,
+      "step": 1871
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.09270980057823298,
+      "learning_rate": 0.0009251193505313844,
+      "loss": 1.4065,
+      "step": 1872
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.10013493564774836,
+      "learning_rate": 0.0009250276751258972,
+      "loss": 1.3734,
+      "step": 1873
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.08111322828236464,
+      "learning_rate": 0.0009249359481843389,
+      "loss": 1.4621,
+      "step": 1874
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.095044532703512,
+      "learning_rate": 0.0009248441697178318,
+      "loss": 1.3761,
+      "step": 1875
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.09643327975927989,
+      "learning_rate": 0.0009247523397375047,
+      "loss": 1.5059,
+      "step": 1876
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.09213762752747705,
+      "learning_rate": 0.000924660458254492,
+      "loss": 1.4399,
+      "step": 1877
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.096760198515979,
+      "learning_rate": 0.0009245685252799345,
+      "loss": 1.4854,
+      "step": 1878
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.08833643778084667,
+      "learning_rate": 0.0009244765408249798,
+      "loss": 1.3744,
+      "step": 1879
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.10338132621574833,
+      "learning_rate": 0.0009243845049007811,
+      "loss": 1.6115,
+      "step": 1880
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.09608860137761568,
+      "learning_rate": 0.0009242924175184981,
+      "loss": 1.4643,
+      "step": 1881
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.09499464003051669,
+      "learning_rate": 0.0009242002786892967,
+      "loss": 1.4931,
+      "step": 1882
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.10553838276741273,
+      "learning_rate": 0.0009241080884243491,
+      "loss": 1.4613,
+      "step": 1883
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.08971818800379484,
+      "learning_rate": 0.0009240158467348337,
+      "loss": 1.3961,
+      "step": 1884
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.09005172515294688,
+      "learning_rate": 0.000923923553631935,
+      "loss": 1.3789,
+      "step": 1885
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.09165971227928925,
+      "learning_rate": 0.000923831209126844,
+      "loss": 1.5329,
+      "step": 1886
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.08771485926445989,
+      "learning_rate": 0.0009237388132307576,
+      "loss": 1.3577,
+      "step": 1887
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.10489309578393058,
+      "learning_rate": 0.0009236463659548793,
+      "loss": 1.4723,
+      "step": 1888
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.08272777189492383,
+      "learning_rate": 0.0009235538673104187,
+      "loss": 1.3486,
+      "step": 1889
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.08734761554544256,
+      "learning_rate": 0.0009234613173085913,
+      "loss": 1.3613,
+      "step": 1890
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0913984053355163,
+      "learning_rate": 0.0009233687159606194,
+      "loss": 1.3321,
+      "step": 1891
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.08114508390947732,
+      "learning_rate": 0.000923276063277731,
+      "loss": 1.5438,
+      "step": 1892
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.08283596903978362,
+      "learning_rate": 0.0009231833592711609,
+      "loss": 1.5123,
+      "step": 1893
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.08104943206708809,
+      "learning_rate": 0.0009230906039521494,
+      "loss": 1.4906,
+      "step": 1894
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0813732064731777,
+      "learning_rate": 0.0009229977973319436,
+      "loss": 1.5399,
+      "step": 1895
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.07189004509962865,
+      "learning_rate": 0.0009229049394217965,
+      "loss": 1.4665,
+      "step": 1896
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1086799985754889,
+      "learning_rate": 0.0009228120302329677,
+      "loss": 1.3363,
+      "step": 1897
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.08013314880463421,
+      "learning_rate": 0.0009227190697767224,
+      "loss": 1.5356,
+      "step": 1898
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.08018669322491434,
+      "learning_rate": 0.0009226260580643326,
+      "loss": 1.5013,
+      "step": 1899
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.09553634333447207,
+      "learning_rate": 0.0009225329951070763,
+      "loss": 1.4245,
+      "step": 1900
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.09467093564785904,
+      "learning_rate": 0.0009224398809162376,
+      "loss": 1.4783,
+      "step": 1901
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.08422828895427777,
+      "learning_rate": 0.0009223467155031068,
+      "loss": 1.4271,
+      "step": 1902
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0876539354690251,
+      "learning_rate": 0.000922253498878981,
+      "loss": 1.2882,
+      "step": 1903
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.09326919056385669,
+      "learning_rate": 0.0009221602310551625,
+      "loss": 1.4323,
+      "step": 1904
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.093287013902574,
+      "learning_rate": 0.0009220669120429608,
+      "loss": 1.476,
+      "step": 1905
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0839791046751122,
+      "learning_rate": 0.0009219735418536908,
+      "loss": 1.5382,
+      "step": 1906
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0969289999285875,
+      "learning_rate": 0.000921880120498674,
+      "loss": 1.4541,
+      "step": 1907
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.10043740648290224,
+      "learning_rate": 0.0009217866479892383,
+      "loss": 1.5245,
+      "step": 1908
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.09340595530802943,
+      "learning_rate": 0.0009216931243367173,
+      "loss": 1.4604,
+      "step": 1909
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.09369494142636196,
+      "learning_rate": 0.0009215995495524512,
+      "loss": 1.4864,
+      "step": 1910
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.09596598542059517,
+      "learning_rate": 0.0009215059236477863,
+      "loss": 1.3945,
+      "step": 1911
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.09961180356748962,
+      "learning_rate": 0.000921412246634075,
+      "loss": 1.4319,
+      "step": 1912
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.09037393165223366,
+      "learning_rate": 0.000921318518522676,
+      "loss": 1.5074,
+      "step": 1913
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.08625277482368682,
+      "learning_rate": 0.000921224739324954,
+      "loss": 1.3883,
+      "step": 1914
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.09714422223326019,
+      "learning_rate": 0.0009211309090522802,
+      "loss": 1.5628,
+      "step": 1915
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0909470499364441,
+      "learning_rate": 0.0009210370277160319,
+      "loss": 1.5049,
+      "step": 1916
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.09368988434281326,
+      "learning_rate": 0.0009209430953275923,
+      "loss": 1.5492,
+      "step": 1917
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.09939051407296741,
+      "learning_rate": 0.0009208491118983514,
+      "loss": 1.4222,
+      "step": 1918
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.11093468746461634,
+      "learning_rate": 0.0009207550774397047,
+      "loss": 1.5089,
+      "step": 1919
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.09192166444725798,
+      "learning_rate": 0.0009206609919630542,
+      "loss": 1.4475,
+      "step": 1920
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.08413223070653192,
+      "learning_rate": 0.0009205668554798084,
+      "loss": 1.3386,
+      "step": 1921
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.08125748886386505,
+      "learning_rate": 0.0009204726680013813,
+      "loss": 1.3634,
+      "step": 1922
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.08620903076980944,
+      "learning_rate": 0.0009203784295391937,
+      "loss": 1.401,
+      "step": 1923
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.10407022815151389,
+      "learning_rate": 0.0009202841401046722,
+      "loss": 1.4641,
+      "step": 1924
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.09106470624464104,
+      "learning_rate": 0.0009201897997092496,
+      "loss": 1.3364,
+      "step": 1925
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0793647200463425,
+      "learning_rate": 0.0009200954083643654,
+      "loss": 1.3409,
+      "step": 1926
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.11996672867870423,
+      "learning_rate": 0.0009200009660814645,
+      "loss": 1.4471,
+      "step": 1927
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.07598152126075328,
+      "learning_rate": 0.0009199064728719988,
+      "loss": 1.3633,
+      "step": 1928
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.09030240320487572,
+      "learning_rate": 0.0009198119287474254,
+      "loss": 1.4453,
+      "step": 1929
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.09808317088255411,
+      "learning_rate": 0.0009197173337192082,
+      "loss": 1.4955,
+      "step": 1930
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.09680838880881662,
+      "learning_rate": 0.0009196226877988174,
+      "loss": 1.5606,
+      "step": 1931
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.08010659787429597,
+      "learning_rate": 0.0009195279909977293,
+      "loss": 1.3131,
+      "step": 1932
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.09152502912300714,
+      "learning_rate": 0.0009194332433274256,
+      "loss": 1.5883,
+      "step": 1933
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.09644413439022238,
+      "learning_rate": 0.0009193384447993954,
+      "loss": 1.4096,
+      "step": 1934
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.07964388791093453,
+      "learning_rate": 0.0009192435954251328,
+      "loss": 1.3457,
+      "step": 1935
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.08985149308714338,
+      "learning_rate": 0.0009191486952161392,
+      "loss": 1.4786,
+      "step": 1936
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.08504235679832424,
+      "learning_rate": 0.0009190537441839211,
+      "loss": 1.5413,
+      "step": 1937
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.08693547735099826,
+      "learning_rate": 0.0009189587423399919,
+      "loss": 1.3525,
+      "step": 1938
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.08671648762518341,
+      "learning_rate": 0.0009188636896958707,
+      "loss": 1.5021,
+      "step": 1939
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0905281579702194,
+      "learning_rate": 0.0009187685862630833,
+      "loss": 1.4043,
+      "step": 1940
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.08064542998545053,
+      "learning_rate": 0.0009186734320531609,
+      "loss": 1.4366,
+      "step": 1941
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.08168312123459194,
+      "learning_rate": 0.0009185782270776416,
+      "loss": 1.433,
+      "step": 1942
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.07806262864437452,
+      "learning_rate": 0.0009184829713480691,
+      "loss": 1.4421,
+      "step": 1943
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.09621596143974015,
+      "learning_rate": 0.0009183876648759937,
+      "loss": 1.4011,
+      "step": 1944
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.08310495330480361,
+      "learning_rate": 0.0009182923076729714,
+      "loss": 1.4856,
+      "step": 1945
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.09214450573103178,
+      "learning_rate": 0.0009181968997505649,
+      "loss": 1.4795,
+      "step": 1946
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.09309818499806359,
+      "learning_rate": 0.0009181014411203425,
+      "loss": 1.4156,
+      "step": 1947
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.09404188405062014,
+      "learning_rate": 0.0009180059317938789,
+      "loss": 1.4638,
+      "step": 1948
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.08127932701261531,
+      "learning_rate": 0.000917910371782755,
+      "loss": 1.4105,
+      "step": 1949
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.06988438601558783,
+      "learning_rate": 0.0009178147610985577,
+      "loss": 1.4714,
+      "step": 1950
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.08197023756718551,
+      "learning_rate": 0.0009177190997528803,
+      "loss": 1.4166,
+      "step": 1951
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.08041723486575238,
+      "learning_rate": 0.0009176233877573219,
+      "loss": 1.4853,
+      "step": 1952
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.08324063551372075,
+      "learning_rate": 0.000917527625123488,
+      "loss": 1.3482,
+      "step": 1953
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.08663822722836335,
+      "learning_rate": 0.00091743181186299,
+      "loss": 1.4284,
+      "step": 1954
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.08043853843069798,
+      "learning_rate": 0.000917335947987446,
+      "loss": 1.5212,
+      "step": 1955
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.09342537247056798,
+      "learning_rate": 0.0009172400335084792,
+      "loss": 1.5348,
+      "step": 1956
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.08823417963705833,
+      "learning_rate": 0.0009171440684377202,
+      "loss": 1.4782,
+      "step": 1957
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.08640741433278323,
+      "learning_rate": 0.0009170480527868045,
+      "loss": 1.3428,
+      "step": 1958
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.08641452096333434,
+      "learning_rate": 0.0009169519865673747,
+      "loss": 1.4254,
+      "step": 1959
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.09594850199633233,
+      "learning_rate": 0.0009168558697910792,
+      "loss": 1.493,
+      "step": 1960
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0810561582133138,
+      "learning_rate": 0.0009167597024695722,
+      "loss": 1.3539,
+      "step": 1961
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.07901539053041998,
+      "learning_rate": 0.0009166634846145145,
+      "loss": 1.4775,
+      "step": 1962
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.09014507355204333,
+      "learning_rate": 0.000916567216237573,
+      "loss": 1.5478,
+      "step": 1963
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.08634547230819684,
+      "learning_rate": 0.0009164708973504204,
+      "loss": 1.3436,
+      "step": 1964
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.09372835977458653,
+      "learning_rate": 0.0009163745279647355,
+      "loss": 1.4296,
+      "step": 1965
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.08489228357702569,
+      "learning_rate": 0.0009162781080922038,
+      "loss": 1.5168,
+      "step": 1966
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.08701353461899991,
+      "learning_rate": 0.0009161816377445162,
+      "loss": 1.5228,
+      "step": 1967
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.08607045368609934,
+      "learning_rate": 0.0009160851169333704,
+      "loss": 1.5066,
+      "step": 1968
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.07958188338929062,
+      "learning_rate": 0.0009159885456704695,
+      "loss": 1.3882,
+      "step": 1969
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.09109091693797759,
+      "learning_rate": 0.0009158919239675235,
+      "loss": 1.4687,
+      "step": 1970
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.08095901944672143,
+      "learning_rate": 0.0009157952518362478,
+      "loss": 1.4857,
+      "step": 1971
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0809430968575337,
+      "learning_rate": 0.0009156985292883645,
+      "loss": 1.4591,
+      "step": 1972
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.09504100751905215,
+      "learning_rate": 0.0009156017563356013,
+      "loss": 1.4128,
+      "step": 1973
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.09268525098327988,
+      "learning_rate": 0.0009155049329896923,
+      "loss": 1.4546,
+      "step": 1974
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.09645365774327952,
+      "learning_rate": 0.0009154080592623777,
+      "loss": 1.3952,
+      "step": 1975
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.08312108781757994,
+      "learning_rate": 0.000915311135165404,
+      "loss": 1.3924,
+      "step": 1976
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.08589342731975778,
+      "learning_rate": 0.0009152141607105231,
+      "loss": 1.4111,
+      "step": 1977
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.08478583867251643,
+      "learning_rate": 0.0009151171359094939,
+      "loss": 1.4493,
+      "step": 1978
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.08228102758097187,
+      "learning_rate": 0.0009150200607740809,
+      "loss": 1.4154,
+      "step": 1979
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.08876973483667644,
+      "learning_rate": 0.0009149229353160545,
+      "loss": 1.4799,
+      "step": 1980
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.08620876609407999,
+      "learning_rate": 0.0009148257595471919,
+      "loss": 1.4748,
+      "step": 1981
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.08926651388447245,
+      "learning_rate": 0.0009147285334792759,
+      "loss": 1.5433,
+      "step": 1982
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.08686773533514985,
+      "learning_rate": 0.0009146312571240953,
+      "loss": 1.4099,
+      "step": 1983
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.08120790615360068,
+      "learning_rate": 0.0009145339304934453,
+      "loss": 1.5089,
+      "step": 1984
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.08751756915323748,
+      "learning_rate": 0.0009144365535991273,
+      "loss": 1.5103,
+      "step": 1985
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.08258101469539926,
+      "learning_rate": 0.0009143391264529482,
+      "loss": 1.4118,
+      "step": 1986
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.08068566310207773,
+      "learning_rate": 0.0009142416490667217,
+      "loss": 1.4599,
+      "step": 1987
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.09637149301141386,
+      "learning_rate": 0.000914144121452267,
+      "loss": 1.5734,
+      "step": 1988
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.08287725138258284,
+      "learning_rate": 0.0009140465436214099,
+      "loss": 1.5485,
+      "step": 1989
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.09651842539644462,
+      "learning_rate": 0.000913948915585982,
+      "loss": 1.4493,
+      "step": 1990
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.08806081162924627,
+      "learning_rate": 0.0009138512373578209,
+      "loss": 1.4323,
+      "step": 1991
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.07898507253280361,
+      "learning_rate": 0.0009137535089487705,
+      "loss": 1.4421,
+      "step": 1992
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.08991267965323817,
+      "learning_rate": 0.0009136557303706808,
+      "loss": 1.4571,
+      "step": 1993
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.08687067556252696,
+      "learning_rate": 0.0009135579016354077,
+      "loss": 1.5119,
+      "step": 1994
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.09059313463610325,
+      "learning_rate": 0.0009134600227548132,
+      "loss": 1.3702,
+      "step": 1995
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.08454581493932053,
+      "learning_rate": 0.0009133620937407656,
+      "loss": 1.3382,
+      "step": 1996
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0915977469098984,
+      "learning_rate": 0.0009132641146051391,
+      "loss": 1.4972,
+      "step": 1997
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.07616506753971135,
+      "learning_rate": 0.0009131660853598138,
+      "loss": 1.4425,
+      "step": 1998
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0853799975536159,
+      "learning_rate": 0.0009130680060166763,
+      "loss": 1.4682,
+      "step": 1999
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.08997371980517725,
+      "learning_rate": 0.0009129698765876191,
+      "loss": 1.4905,
+      "step": 2000
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0893355318664062,
+      "learning_rate": 0.0009128716970845406,
+      "loss": 1.4148,
+      "step": 2001
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0984766048951835,
+      "learning_rate": 0.0009127734675193454,
+      "loss": 1.4577,
+      "step": 2002
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.08685169232348386,
+      "learning_rate": 0.0009126751879039441,
+      "loss": 1.326,
+      "step": 2003
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.08337716785047515,
+      "learning_rate": 0.0009125768582502539,
+      "loss": 1.457,
+      "step": 2004
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.08013750991226204,
+      "learning_rate": 0.0009124784785701969,
+      "loss": 1.2967,
+      "step": 2005
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.08594967067757286,
+      "learning_rate": 0.0009123800488757026,
+      "loss": 1.5688,
+      "step": 2006
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0777155592804685,
+      "learning_rate": 0.0009122815691787055,
+      "loss": 1.4152,
+      "step": 2007
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.10406910057370224,
+      "learning_rate": 0.000912183039491147,
+      "loss": 1.614,
+      "step": 2008
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.08189508655131475,
+      "learning_rate": 0.0009120844598249737,
+      "loss": 1.447,
+      "step": 2009
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.07712372854960725,
+      "learning_rate": 0.0009119858301921391,
+      "loss": 1.3347,
+      "step": 2010
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.08994235194617171,
+      "learning_rate": 0.0009118871506046024,
+      "loss": 1.5152,
+      "step": 2011
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.07797437990098123,
+      "learning_rate": 0.0009117884210743286,
+      "loss": 1.4365,
+      "step": 2012
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.08655098692860116,
+      "learning_rate": 0.0009116896416132889,
+      "loss": 1.4759,
+      "step": 2013
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.07745160520338816,
+      "learning_rate": 0.000911590812233461,
+      "loss": 1.3671,
+      "step": 2014
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.10139052107795801,
+      "learning_rate": 0.0009114919329468282,
+      "loss": 1.4782,
+      "step": 2015
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.09445383935773374,
+      "learning_rate": 0.00091139300376538,
+      "loss": 1.5765,
+      "step": 2016
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0893852361489357,
+      "learning_rate": 0.0009112940247011116,
+      "loss": 1.4059,
+      "step": 2017
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.08429572837560623,
+      "learning_rate": 0.0009111949957660248,
+      "loss": 1.3523,
+      "step": 2018
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.08670828198710072,
+      "learning_rate": 0.0009110959169721271,
+      "loss": 1.3836,
+      "step": 2019
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.09593977547060985,
+      "learning_rate": 0.0009109967883314323,
+      "loss": 1.499,
+      "step": 2020
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0893023259917977,
+      "learning_rate": 0.00091089760985596,
+      "loss": 1.4865,
+      "step": 2021
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.08920793652767317,
+      "learning_rate": 0.0009107983815577359,
+      "loss": 1.5379,
+      "step": 2022
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.10048196567007192,
+      "learning_rate": 0.0009106991034487917,
+      "loss": 1.5585,
+      "step": 2023
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.09363391762575825,
+      "learning_rate": 0.0009105997755411655,
+      "loss": 1.5651,
+      "step": 2024
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.10401149059455012,
+      "learning_rate": 0.0009105003978469009,
+      "loss": 1.4563,
+      "step": 2025
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.08535302742501824,
+      "learning_rate": 0.0009104009703780478,
+      "loss": 1.4701,
+      "step": 2026
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.08868145003485764,
+      "learning_rate": 0.0009103014931466623,
+      "loss": 1.4583,
+      "step": 2027
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0879030351354358,
+      "learning_rate": 0.000910201966164806,
+      "loss": 1.3511,
+      "step": 2028
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.10253950527640521,
+      "learning_rate": 0.0009101023894445472,
+      "loss": 1.5729,
+      "step": 2029
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.08542785241162401,
+      "learning_rate": 0.0009100027629979599,
+      "loss": 1.4357,
+      "step": 2030
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.08579083004639929,
+      "learning_rate": 0.0009099030868371241,
+      "loss": 1.4863,
+      "step": 2031
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.09758998052839936,
+      "learning_rate": 0.0009098033609741259,
+      "loss": 1.5492,
+      "step": 2032
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.07844186539869633,
+      "learning_rate": 0.0009097035854210573,
+      "loss": 1.5178,
+      "step": 2033
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.08847073531592889,
+      "learning_rate": 0.0009096037601900166,
+      "loss": 1.474,
+      "step": 2034
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.08055118937273592,
+      "learning_rate": 0.0009095038852931077,
+      "loss": 1.3124,
+      "step": 2035
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.08954927397429793,
+      "learning_rate": 0.000909403960742441,
+      "loss": 1.3911,
+      "step": 2036
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.08871609293202781,
+      "learning_rate": 0.0009093039865501327,
+      "loss": 1.4989,
+      "step": 2037
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.08331819035983692,
+      "learning_rate": 0.0009092039627283049,
+      "loss": 1.5124,
+      "step": 2038
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.08617807710536765,
+      "learning_rate": 0.0009091038892890859,
+      "loss": 1.3432,
+      "step": 2039
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.08013091876615193,
+      "learning_rate": 0.0009090037662446099,
+      "loss": 1.4968,
+      "step": 2040
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.08088959403354601,
+      "learning_rate": 0.0009089035936070171,
+      "loss": 1.2544,
+      "step": 2041
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.07511260388633456,
+      "learning_rate": 0.0009088033713884541,
+      "loss": 1.5305,
+      "step": 2042
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.07857443163504174,
+      "learning_rate": 0.0009087030996010728,
+      "loss": 1.4741,
+      "step": 2043
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.08773673280492814,
+      "learning_rate": 0.0009086027782570316,
+      "loss": 1.4905,
+      "step": 2044
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.08617794607094241,
+      "learning_rate": 0.0009085024073684951,
+      "loss": 1.5055,
+      "step": 2045
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0949855047620933,
+      "learning_rate": 0.0009084019869476332,
+      "loss": 1.5818,
+      "step": 2046
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.09609447475098014,
+      "learning_rate": 0.0009083015170066224,
+      "loss": 1.4573,
+      "step": 2047
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.08109159950480951,
+      "learning_rate": 0.0009082009975576451,
+      "loss": 1.3369,
+      "step": 2048
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.08835922201145058,
+      "learning_rate": 0.0009081004286128895,
+      "loss": 1.3451,
+      "step": 2049
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.08859070142962146,
+      "learning_rate": 0.0009079998101845501,
+      "loss": 1.3428,
+      "step": 2050
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.088189800018709,
+      "learning_rate": 0.000907899142284827,
+      "loss": 1.5434,
+      "step": 2051
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.08644531320243522,
+      "learning_rate": 0.0009077984249259268,
+      "loss": 1.4702,
+      "step": 2052
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.07668741919840479,
+      "learning_rate": 0.0009076976581200615,
+      "loss": 1.3968,
+      "step": 2053
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.09581406401109188,
+      "learning_rate": 0.0009075968418794498,
+      "loss": 1.4724,
+      "step": 2054
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.09246875649499241,
+      "learning_rate": 0.0009074959762163157,
+      "loss": 1.4314,
+      "step": 2055
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.07492645924876302,
+      "learning_rate": 0.0009073950611428897,
+      "loss": 1.2898,
+      "step": 2056
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.06928869437837035,
+      "learning_rate": 0.000907294096671408,
+      "loss": 1.3936,
+      "step": 2057
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0832862559597914,
+      "learning_rate": 0.0009071930828141128,
+      "loss": 1.4466,
+      "step": 2058
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.08848451000796444,
+      "learning_rate": 0.0009070920195832527,
+      "loss": 1.4536,
+      "step": 2059
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.09506876450203977,
+      "learning_rate": 0.0009069909069910816,
+      "loss": 1.342,
+      "step": 2060
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.08569941811729877,
+      "learning_rate": 0.0009068897450498602,
+      "loss": 1.4359,
+      "step": 2061
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.08488207204085854,
+      "learning_rate": 0.0009067885337718543,
+      "loss": 1.5106,
+      "step": 2062
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.08690685472711807,
+      "learning_rate": 0.0009066872731693361,
+      "loss": 1.3516,
+      "step": 2063
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.08252793754556853,
+      "learning_rate": 0.0009065859632545841,
+      "loss": 1.496,
+      "step": 2064
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.09116501152269689,
+      "learning_rate": 0.0009064846040398822,
+      "loss": 1.3323,
+      "step": 2065
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.08508703996689969,
+      "learning_rate": 0.0009063831955375209,
+      "loss": 1.4011,
+      "step": 2066
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.08245437493180188,
+      "learning_rate": 0.0009062817377597961,
+      "loss": 1.5624,
+      "step": 2067
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.08352745314965869,
+      "learning_rate": 0.0009061802307190098,
+      "loss": 1.4388,
+      "step": 2068
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.08597478680811273,
+      "learning_rate": 0.0009060786744274703,
+      "loss": 1.4461,
+      "step": 2069
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.08312171106043391,
+      "learning_rate": 0.0009059770688974915,
+      "loss": 1.4689,
+      "step": 2070
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.09194463520309476,
+      "learning_rate": 0.0009058754141413935,
+      "loss": 1.4197,
+      "step": 2071
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.08150972185382112,
+      "learning_rate": 0.0009057737101715024,
+      "loss": 1.4696,
+      "step": 2072
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0853710443666307,
+      "learning_rate": 0.0009056719570001498,
+      "loss": 1.5233,
+      "step": 2073
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.07523008570520828,
+      "learning_rate": 0.000905570154639674,
+      "loss": 1.5505,
+      "step": 2074
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0847559640472571,
+      "learning_rate": 0.0009054683031024187,
+      "loss": 1.5916,
+      "step": 2075
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.08104363762554731,
+      "learning_rate": 0.0009053664024007337,
+      "loss": 1.4587,
+      "step": 2076
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.07962733245896239,
+      "learning_rate": 0.0009052644525469751,
+      "loss": 1.3431,
+      "step": 2077
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.08353363693288063,
+      "learning_rate": 0.0009051624535535044,
+      "loss": 1.4799,
+      "step": 2078
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.08825166774195803,
+      "learning_rate": 0.0009050604054326893,
+      "loss": 1.3508,
+      "step": 2079
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.07694662079941654,
+      "learning_rate": 0.0009049583081969037,
+      "loss": 1.3975,
+      "step": 2080
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.08547425993073962,
+      "learning_rate": 0.0009048561618585269,
+      "loss": 1.4588,
+      "step": 2081
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.07890146312971011,
+      "learning_rate": 0.000904753966429945,
+      "loss": 1.5806,
+      "step": 2082
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0895410367058908,
+      "learning_rate": 0.0009046517219235492,
+      "loss": 1.428,
+      "step": 2083
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.07457533928827222,
+      "learning_rate": 0.000904549428351737,
+      "loss": 1.4553,
+      "step": 2084
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.08100017328753376,
+      "learning_rate": 0.0009044470857269121,
+      "loss": 1.3618,
+      "step": 2085
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.07129459326988942,
+      "learning_rate": 0.0009043446940614835,
+      "loss": 1.4317,
+      "step": 2086
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.07500535174736596,
+      "learning_rate": 0.0009042422533678667,
+      "loss": 1.4146,
+      "step": 2087
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.08011378389844677,
+      "learning_rate": 0.0009041397636584831,
+      "loss": 1.3876,
+      "step": 2088
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.07943960209069717,
+      "learning_rate": 0.00090403722494576,
+      "loss": 1.428,
+      "step": 2089
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.07076112195478479,
+      "learning_rate": 0.0009039346372421304,
+      "loss": 1.3577,
+      "step": 2090
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.09256678327668848,
+      "learning_rate": 0.0009038320005600336,
+      "loss": 1.4306,
+      "step": 2091
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.07157860379515876,
+      "learning_rate": 0.0009037293149119144,
+      "loss": 1.4065,
+      "step": 2092
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.09035990774616685,
+      "learning_rate": 0.0009036265803102237,
+      "loss": 1.3061,
+      "step": 2093
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.09038176456247499,
+      "learning_rate": 0.0009035237967674188,
+      "loss": 1.4361,
+      "step": 2094
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.10021661807035723,
+      "learning_rate": 0.0009034209642959624,
+      "loss": 1.3928,
+      "step": 2095
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.08500099881896636,
+      "learning_rate": 0.0009033180829083232,
+      "loss": 1.2741,
+      "step": 2096
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.09099960063293876,
+      "learning_rate": 0.0009032151526169761,
+      "loss": 1.3831,
+      "step": 2097
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.09257103776160713,
+      "learning_rate": 0.0009031121734344016,
+      "loss": 1.3861,
+      "step": 2098
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.09067821153143035,
+      "learning_rate": 0.0009030091453730862,
+      "loss": 1.5433,
+      "step": 2099
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.10061981038222514,
+      "learning_rate": 0.0009029060684455228,
+      "loss": 1.6066,
+      "step": 2100
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.08322155025349774,
+      "learning_rate": 0.0009028029426642095,
+      "loss": 1.44,
+      "step": 2101
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.08632501692592569,
+      "learning_rate": 0.0009026997680416505,
+      "loss": 1.5127,
+      "step": 2102
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.10646380692549054,
+      "learning_rate": 0.0009025965445903565,
+      "loss": 1.5835,
+      "step": 2103
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0778664557671461,
+      "learning_rate": 0.0009024932723228436,
+      "loss": 1.4187,
+      "step": 2104
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.08218957602622107,
+      "learning_rate": 0.0009023899512516336,
+      "loss": 1.322,
+      "step": 2105
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.068440123958451,
+      "learning_rate": 0.0009022865813892549,
+      "loss": 1.4493,
+      "step": 2106
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.07736229830673369,
+      "learning_rate": 0.0009021831627482413,
+      "loss": 1.3777,
+      "step": 2107
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.07927886201598712,
+      "learning_rate": 0.0009020796953411327,
+      "loss": 1.4454,
+      "step": 2108
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.07994526360761076,
+      "learning_rate": 0.0009019761791804748,
+      "loss": 1.3731,
+      "step": 2109
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.07833679220856182,
+      "learning_rate": 0.0009018726142788194,
+      "loss": 1.3188,
+      "step": 2110
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.07289754206616753,
+      "learning_rate": 0.000901769000648724,
+      "loss": 1.4212,
+      "step": 2111
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.07449956522325085,
+      "learning_rate": 0.0009016653383027522,
+      "loss": 1.4958,
+      "step": 2112
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.07840040658702752,
+      "learning_rate": 0.0009015616272534734,
+      "loss": 1.3861,
+      "step": 2113
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.07392483160096817,
+      "learning_rate": 0.0009014578675134628,
+      "loss": 1.4716,
+      "step": 2114
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.06578630614556928,
+      "learning_rate": 0.0009013540590953017,
+      "loss": 1.5169,
+      "step": 2115
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.07803948013204608,
+      "learning_rate": 0.0009012502020115776,
+      "loss": 1.5624,
+      "step": 2116
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.076649100119888,
+      "learning_rate": 0.0009011462962748829,
+      "loss": 1.5829,
+      "step": 2117
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.07625429267408237,
+      "learning_rate": 0.0009010423418978168,
+      "loss": 1.4575,
+      "step": 2118
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.07959475385111861,
+      "learning_rate": 0.0009009383388929842,
+      "loss": 1.4098,
+      "step": 2119
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.07880013779769067,
+      "learning_rate": 0.0009008342872729957,
+      "loss": 1.4709,
+      "step": 2120
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.08272735817424194,
+      "learning_rate": 0.0009007301870504681,
+      "loss": 1.4538,
+      "step": 2121
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.07552018367416652,
+      "learning_rate": 0.0009006260382380238,
+      "loss": 1.3958,
+      "step": 2122
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0773802621109197,
+      "learning_rate": 0.0009005218408482911,
+      "loss": 1.6479,
+      "step": 2123
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0903702500400604,
+      "learning_rate": 0.0009004175948939044,
+      "loss": 1.4688,
+      "step": 2124
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.08143156259179056,
+      "learning_rate": 0.000900313300387504,
+      "loss": 1.3508,
+      "step": 2125
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.08345652362054003,
+      "learning_rate": 0.0009002089573417356,
+      "loss": 1.4798,
+      "step": 2126
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.08097286863224183,
+      "learning_rate": 0.0009001045657692517,
+      "loss": 1.4088,
+      "step": 2127
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.07816712018885351,
+      "learning_rate": 0.0009000001256827095,
+      "loss": 1.464,
+      "step": 2128
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.07995949220911897,
+      "learning_rate": 0.0008998956370947733,
+      "loss": 1.4584,
+      "step": 2129
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.08389083649029062,
+      "learning_rate": 0.0008997911000181122,
+      "loss": 1.4115,
+      "step": 2130
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.08706037657202785,
+      "learning_rate": 0.0008996865144654023,
+      "loss": 1.4264,
+      "step": 2131
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.08739980235747145,
+      "learning_rate": 0.0008995818804493243,
+      "loss": 1.4701,
+      "step": 2132
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.09140265561641635,
+      "learning_rate": 0.0008994771979825658,
+      "loss": 1.4218,
+      "step": 2133
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.07375929912993871,
+      "learning_rate": 0.0008993724670778198,
+      "loss": 1.4068,
+      "step": 2134
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.07953888777292227,
+      "learning_rate": 0.0008992676877477854,
+      "loss": 1.444,
+      "step": 2135
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.10877764802904497,
+      "learning_rate": 0.0008991628600051673,
+      "loss": 1.4785,
+      "step": 2136
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.07748952894124081,
+      "learning_rate": 0.0008990579838626764,
+      "loss": 1.4316,
+      "step": 2137
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.08527565495180588,
+      "learning_rate": 0.0008989530593330291,
+      "loss": 1.6002,
+      "step": 2138
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.08335708177946245,
+      "learning_rate": 0.0008988480864289481,
+      "loss": 1.3452,
+      "step": 2139
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.09759801081695489,
+      "learning_rate": 0.0008987430651631613,
+      "loss": 1.4563,
+      "step": 2140
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.07840457899332089,
+      "learning_rate": 0.0008986379955484036,
+      "loss": 1.4884,
+      "step": 2141
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0876443419021953,
+      "learning_rate": 0.0008985328775974142,
+      "loss": 1.5695,
+      "step": 2142
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.08807857999714565,
+      "learning_rate": 0.0008984277113229397,
+      "loss": 1.5175,
+      "step": 2143
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.10090005538461481,
+      "learning_rate": 0.0008983224967377315,
+      "loss": 1.435,
+      "step": 2144
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.09627909221834423,
+      "learning_rate": 0.0008982172338545474,
+      "loss": 1.5002,
+      "step": 2145
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0789701966323783,
+      "learning_rate": 0.0008981119226861508,
+      "loss": 1.6117,
+      "step": 2146
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0862375515699889,
+      "learning_rate": 0.0008980065632453111,
+      "loss": 1.5657,
+      "step": 2147
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.07179033638664258,
+      "learning_rate": 0.0008979011555448035,
+      "loss": 1.4099,
+      "step": 2148
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.07114245906288379,
+      "learning_rate": 0.0008977956995974089,
+      "loss": 1.4637,
+      "step": 2149
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.07414337930145709,
+      "learning_rate": 0.0008976901954159144,
+      "loss": 1.4752,
+      "step": 2150
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.07132830095164981,
+      "learning_rate": 0.0008975846430131127,
+      "loss": 1.4056,
+      "step": 2151
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0719025939859958,
+      "learning_rate": 0.0008974790424018022,
+      "loss": 1.4565,
+      "step": 2152
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.07354577248780032,
+      "learning_rate": 0.0008973733935947877,
+      "loss": 1.3983,
+      "step": 2153
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0787391972157109,
+      "learning_rate": 0.0008972676966048789,
+      "loss": 1.3842,
+      "step": 2154
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.07914358659288476,
+      "learning_rate": 0.0008971619514448927,
+      "loss": 1.4485,
+      "step": 2155
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.09372633769515623,
+      "learning_rate": 0.0008970561581276505,
+      "loss": 1.4679,
+      "step": 2156
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.07041013581411883,
+      "learning_rate": 0.0008969503166659803,
+      "loss": 1.3479,
+      "step": 2157
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.08613639135343044,
+      "learning_rate": 0.0008968444270727157,
+      "loss": 1.4191,
+      "step": 2158
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.08240147616175394,
+      "learning_rate": 0.0008967384893606962,
+      "loss": 1.4829,
+      "step": 2159
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.08323090670400539,
+      "learning_rate": 0.0008966325035427669,
+      "loss": 1.4426,
+      "step": 2160
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0830480211709058,
+      "learning_rate": 0.0008965264696317795,
+      "loss": 1.4769,
+      "step": 2161
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.08928546482396188,
+      "learning_rate": 0.0008964203876405903,
+      "loss": 1.3255,
+      "step": 2162
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0822658434232799,
+      "learning_rate": 0.0008963142575820626,
+      "loss": 1.4945,
+      "step": 2163
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.08491975646767387,
+      "learning_rate": 0.0008962080794690648,
+      "loss": 1.5602,
+      "step": 2164
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.08414408557119114,
+      "learning_rate": 0.0008961018533144716,
+      "loss": 1.4698,
+      "step": 2165
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0954940468086645,
+      "learning_rate": 0.000895995579131163,
+      "loss": 1.3421,
+      "step": 2166
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.08659721199046269,
+      "learning_rate": 0.0008958892569320251,
+      "loss": 1.446,
+      "step": 2167
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.08211464984956957,
+      "learning_rate": 0.00089578288672995,
+      "loss": 1.4428,
+      "step": 2168
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.08730211184796505,
+      "learning_rate": 0.0008956764685378356,
+      "loss": 1.5298,
+      "step": 2169
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.08028511736236937,
+      "learning_rate": 0.0008955700023685851,
+      "loss": 1.401,
+      "step": 2170
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.08607184275244723,
+      "learning_rate": 0.000895463488235108,
+      "loss": 1.2803,
+      "step": 2171
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.08878050364927839,
+      "learning_rate": 0.0008953569261503198,
+      "loss": 1.3938,
+      "step": 2172
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.09230301699295128,
+      "learning_rate": 0.0008952503161271413,
+      "loss": 1.5243,
+      "step": 2173
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.09530871727738145,
+      "learning_rate": 0.000895143658178499,
+      "loss": 1.4443,
+      "step": 2174
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.09034274785163224,
+      "learning_rate": 0.0008950369523173263,
+      "loss": 1.3993,
+      "step": 2175
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.08126098450813693,
+      "learning_rate": 0.000894930198556561,
+      "loss": 1.3772,
+      "step": 2176
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.09846129233960661,
+      "learning_rate": 0.0008948233969091477,
+      "loss": 1.4249,
+      "step": 2177
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.09346589113210531,
+      "learning_rate": 0.0008947165473880363,
+      "loss": 1.2923,
+      "step": 2178
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.07884971976680567,
+      "learning_rate": 0.0008946096500061828,
+      "loss": 1.5027,
+      "step": 2179
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0815049356861313,
+      "learning_rate": 0.0008945027047765488,
+      "loss": 1.5911,
+      "step": 2180
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.08053472358349446,
+      "learning_rate": 0.0008943957117121017,
+      "loss": 1.3357,
+      "step": 2181
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.07124111934527916,
+      "learning_rate": 0.0008942886708258148,
+      "loss": 1.485,
+      "step": 2182
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.08208081331893921,
+      "learning_rate": 0.0008941815821306674,
+      "loss": 1.3862,
+      "step": 2183
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.08992051131410057,
+      "learning_rate": 0.0008940744456396442,
+      "loss": 1.4123,
+      "step": 2184
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.08292039307031486,
+      "learning_rate": 0.0008939672613657359,
+      "loss": 1.4377,
+      "step": 2185
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.07751252629536556,
+      "learning_rate": 0.000893860029321939,
+      "loss": 1.3624,
+      "step": 2186
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.07378227106710288,
+      "learning_rate": 0.0008937527495212555,
+      "loss": 1.5427,
+      "step": 2187
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08362754852649955,
+      "learning_rate": 0.0008936454219766938,
+      "loss": 1.3911,
+      "step": 2188
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08561863317364699,
+      "learning_rate": 0.0008935380467012675,
+      "loss": 1.4485,
+      "step": 2189
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08022086430199703,
+      "learning_rate": 0.0008934306237079963,
+      "loss": 1.4691,
+      "step": 2190
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.10481201673736351,
+      "learning_rate": 0.0008933231530099058,
+      "loss": 1.4008,
+      "step": 2191
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.07735777171508976,
+      "learning_rate": 0.0008932156346200268,
+      "loss": 1.4511,
+      "step": 2192
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.07372539389115546,
+      "learning_rate": 0.0008931080685513966,
+      "loss": 1.5533,
+      "step": 2193
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08699225540912171,
+      "learning_rate": 0.0008930004548170577,
+      "loss": 1.5613,
+      "step": 2194
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08015320764450956,
+      "learning_rate": 0.0008928927934300588,
+      "loss": 1.4748,
+      "step": 2195
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08026380063427321,
+      "learning_rate": 0.0008927850844034544,
+      "loss": 1.5011,
+      "step": 2196
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08569620404860831,
+      "learning_rate": 0.0008926773277503041,
+      "loss": 1.4672,
+      "step": 2197
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08527228975419182,
+      "learning_rate": 0.0008925695234836742,
+      "loss": 1.4531,
+      "step": 2198
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08123604466644242,
+      "learning_rate": 0.0008924616716166363,
+      "loss": 1.4952,
+      "step": 2199
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08556822527789404,
+      "learning_rate": 0.0008923537721622674,
+      "loss": 1.5915,
+      "step": 2200
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.09081981002504544,
+      "learning_rate": 0.0008922458251336511,
+      "loss": 1.5223,
+      "step": 2201
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08036843164515005,
+      "learning_rate": 0.0008921378305438763,
+      "loss": 1.415,
+      "step": 2202
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08294936113634996,
+      "learning_rate": 0.0008920297884060376,
+      "loss": 1.3729,
+      "step": 2203
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.09402295203612535,
+      "learning_rate": 0.0008919216987332356,
+      "loss": 1.3907,
+      "step": 2204
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08251725989381521,
+      "learning_rate": 0.0008918135615385763,
+      "loss": 1.3759,
+      "step": 2205
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.07518200632745241,
+      "learning_rate": 0.0008917053768351719,
+      "loss": 1.3763,
+      "step": 2206
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.07912416047033925,
+      "learning_rate": 0.0008915971446361404,
+      "loss": 1.4037,
+      "step": 2207
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08064078885414537,
+      "learning_rate": 0.0008914888649546048,
+      "loss": 1.5326,
+      "step": 2208
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.07408548298786542,
+      "learning_rate": 0.0008913805378036948,
+      "loss": 1.5165,
+      "step": 2209
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08872131280264428,
+      "learning_rate": 0.0008912721631965453,
+      "loss": 1.4678,
+      "step": 2210
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08249217488465536,
+      "learning_rate": 0.0008911637411462969,
+      "loss": 1.3006,
+      "step": 2211
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08866808668942394,
+      "learning_rate": 0.0008910552716660965,
+      "loss": 1.4458,
+      "step": 2212
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.07822108916047592,
+      "learning_rate": 0.0008909467547690962,
+      "loss": 1.5045,
+      "step": 2213
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.07968046873758618,
+      "learning_rate": 0.0008908381904684542,
+      "loss": 1.3571,
+      "step": 2214
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08621063906251096,
+      "learning_rate": 0.0008907295787773339,
+      "loss": 1.4686,
+      "step": 2215
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.09541846290904246,
+      "learning_rate": 0.0008906209197089054,
+      "loss": 1.4994,
+      "step": 2216
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08307421917053688,
+      "learning_rate": 0.0008905122132763437,
+      "loss": 1.5249,
+      "step": 2217
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.07788724633936456,
+      "learning_rate": 0.0008904034594928296,
+      "loss": 1.4023,
+      "step": 2218
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0816366690950349,
+      "learning_rate": 0.0008902946583715503,
+      "loss": 1.43,
+      "step": 2219
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.09982940037534067,
+      "learning_rate": 0.0008901858099256981,
+      "loss": 1.4255,
+      "step": 2220
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08232288093656423,
+      "learning_rate": 0.0008900769141684712,
+      "loss": 1.3594,
+      "step": 2221
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08357733265593364,
+      "learning_rate": 0.0008899679711130737,
+      "loss": 1.4477,
+      "step": 2222
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08849502280367522,
+      "learning_rate": 0.0008898589807727153,
+      "loss": 1.3596,
+      "step": 2223
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.07174272902783962,
+      "learning_rate": 0.0008897499431606116,
+      "loss": 1.4807,
+      "step": 2224
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08742620366231663,
+      "learning_rate": 0.0008896408582899833,
+      "loss": 1.4376,
+      "step": 2225
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.07675971108920426,
+      "learning_rate": 0.0008895317261740579,
+      "loss": 1.4947,
+      "step": 2226
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.07584998544369173,
+      "learning_rate": 0.0008894225468260675,
+      "loss": 1.5446,
+      "step": 2227
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08675223323782724,
+      "learning_rate": 0.000889313320259251,
+      "loss": 1.5474,
+      "step": 2228
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.07847232947461194,
+      "learning_rate": 0.000889204046486852,
+      "loss": 1.3825,
+      "step": 2229
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08410254193952434,
+      "learning_rate": 0.0008890947255221209,
+      "loss": 1.4741,
+      "step": 2230
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08534168834441479,
+      "learning_rate": 0.0008889853573783127,
+      "loss": 1.359,
+      "step": 2231
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.07198918151582337,
+      "learning_rate": 0.0008888759420686889,
+      "loss": 1.3831,
+      "step": 2232
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.07677775674736524,
+      "learning_rate": 0.0008887664796065165,
+      "loss": 1.3148,
+      "step": 2233
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08207106759233733,
+      "learning_rate": 0.0008886569700050682,
+      "loss": 1.468,
+      "step": 2234
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.07396556465325442,
+      "learning_rate": 0.0008885474132776224,
+      "loss": 1.3863,
+      "step": 2235
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08104484567295037,
+      "learning_rate": 0.0008884378094374632,
+      "loss": 1.469,
+      "step": 2236
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.07786764881894506,
+      "learning_rate": 0.0008883281584978804,
+      "loss": 1.5114,
+      "step": 2237
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.07771523721695527,
+      "learning_rate": 0.0008882184604721697,
+      "loss": 1.5415,
+      "step": 2238
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08084233888893828,
+      "learning_rate": 0.000888108715373632,
+      "loss": 1.5652,
+      "step": 2239
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.07311662878971846,
+      "learning_rate": 0.0008879989232155748,
+      "loss": 1.4416,
+      "step": 2240
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.07765587834196362,
+      "learning_rate": 0.0008878890840113105,
+      "loss": 1.3599,
+      "step": 2241
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.07468966830819058,
+      "learning_rate": 0.0008877791977741575,
+      "loss": 1.4502,
+      "step": 2242
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0810774288676914,
+      "learning_rate": 0.0008876692645174399,
+      "loss": 1.5496,
+      "step": 2243
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.07578306736684104,
+      "learning_rate": 0.0008875592842544875,
+      "loss": 1.4876,
+      "step": 2244
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.07308565897661695,
+      "learning_rate": 0.0008874492569986357,
+      "loss": 1.4319,
+      "step": 2245
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08661688941721231,
+      "learning_rate": 0.0008873391827632258,
+      "loss": 1.3202,
+      "step": 2246
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08669507642175672,
+      "learning_rate": 0.0008872290615616046,
+      "loss": 1.5107,
+      "step": 2247
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08289264655816533,
+      "learning_rate": 0.0008871188934071246,
+      "loss": 1.3426,
+      "step": 2248
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08315127821419833,
+      "learning_rate": 0.0008870086783131444,
+      "loss": 1.3794,
+      "step": 2249
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08843733636930243,
+      "learning_rate": 0.0008868984162930275,
+      "loss": 1.4677,
+      "step": 2250
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.09172413612371813,
+      "learning_rate": 0.0008867881073601439,
+      "loss": 1.5372,
+      "step": 2251
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08078669573488897,
+      "learning_rate": 0.0008866777515278688,
+      "loss": 1.4151,
+      "step": 2252
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08090166791860437,
+      "learning_rate": 0.0008865673488095832,
+      "loss": 1.5728,
+      "step": 2253
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08000330075098684,
+      "learning_rate": 0.0008864568992186739,
+      "loss": 1.3872,
+      "step": 2254
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0756398228234597,
+      "learning_rate": 0.0008863464027685332,
+      "loss": 1.2759,
+      "step": 2255
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.09501349605436056,
+      "learning_rate": 0.0008862358594725595,
+      "loss": 1.6111,
+      "step": 2256
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08869060983281359,
+      "learning_rate": 0.0008861252693441559,
+      "loss": 1.5307,
+      "step": 2257
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08322655242755857,
+      "learning_rate": 0.0008860146323967324,
+      "loss": 1.3653,
+      "step": 2258
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.09070182705927224,
+      "learning_rate": 0.0008859039486437039,
+      "loss": 1.465,
+      "step": 2259
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08490542675613597,
+      "learning_rate": 0.0008857932180984914,
+      "loss": 1.3364,
+      "step": 2260
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08572743355151981,
+      "learning_rate": 0.000885682440774521,
+      "loss": 1.39,
+      "step": 2261
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08391700720736694,
+      "learning_rate": 0.000885571616685225,
+      "loss": 1.3427,
+      "step": 2262
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.07350389296018711,
+      "learning_rate": 0.0008854607458440412,
+      "loss": 1.3976,
+      "step": 2263
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0914364941499127,
+      "learning_rate": 0.000885349828264413,
+      "loss": 1.4906,
+      "step": 2264
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08023243376640941,
+      "learning_rate": 0.0008852388639597897,
+      "loss": 1.3761,
+      "step": 2265
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.07893137372515271,
+      "learning_rate": 0.0008851278529436261,
+      "loss": 1.5595,
+      "step": 2266
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.07561619587902403,
+      "learning_rate": 0.0008850167952293825,
+      "loss": 1.476,
+      "step": 2267
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08126009628490852,
+      "learning_rate": 0.0008849056908305252,
+      "loss": 1.5766,
+      "step": 2268
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0779068568614784,
+      "learning_rate": 0.0008847945397605258,
+      "loss": 1.4817,
+      "step": 2269
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.07932082446412245,
+      "learning_rate": 0.0008846833420328619,
+      "loss": 1.373,
+      "step": 2270
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08156960018218874,
+      "learning_rate": 0.0008845720976610168,
+      "loss": 1.4695,
+      "step": 2271
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.07735279185733433,
+      "learning_rate": 0.0008844608066584787,
+      "loss": 1.3753,
+      "step": 2272
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.07733770886342596,
+      "learning_rate": 0.0008843494690387426,
+      "loss": 1.4613,
+      "step": 2273
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.07890085671660646,
+      "learning_rate": 0.0008842380848153082,
+      "loss": 1.3932,
+      "step": 2274
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.07782994223247774,
+      "learning_rate": 0.0008841266540016813,
+      "loss": 1.4201,
+      "step": 2275
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0957307653994958,
+      "learning_rate": 0.0008840151766113735,
+      "loss": 1.3713,
+      "step": 2276
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08676129615166901,
+      "learning_rate": 0.0008839036526579014,
+      "loss": 1.4288,
+      "step": 2277
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08086343688253304,
+      "learning_rate": 0.000883792082154788,
+      "loss": 1.453,
+      "step": 2278
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.07603031386284381,
+      "learning_rate": 0.0008836804651155617,
+      "loss": 1.3063,
+      "step": 2279
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08419241615367228,
+      "learning_rate": 0.0008835688015537559,
+      "loss": 1.3828,
+      "step": 2280
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08829666758411504,
+      "learning_rate": 0.0008834570914829108,
+      "loss": 1.4098,
+      "step": 2281
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08693258350713497,
+      "learning_rate": 0.0008833453349165714,
+      "loss": 1.3494,
+      "step": 2282
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.09007341118626776,
+      "learning_rate": 0.0008832335318682883,
+      "loss": 1.4282,
+      "step": 2283
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.09274387264986217,
+      "learning_rate": 0.0008831216823516185,
+      "loss": 1.4668,
+      "step": 2284
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.09646572654831478,
+      "learning_rate": 0.0008830097863801238,
+      "loss": 1.5571,
+      "step": 2285
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08083030222951221,
+      "learning_rate": 0.000882897843967372,
+      "loss": 1.4943,
+      "step": 2286
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08450865254218719,
+      "learning_rate": 0.0008827858551269368,
+      "loss": 1.3344,
+      "step": 2287
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08067151928811613,
+      "learning_rate": 0.0008826738198723967,
+      "loss": 1.5423,
+      "step": 2288
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08614124183323224,
+      "learning_rate": 0.0008825617382173369,
+      "loss": 1.4565,
+      "step": 2289
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.07814908550447222,
+      "learning_rate": 0.0008824496101753473,
+      "loss": 1.459,
+      "step": 2290
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08058843744933337,
+      "learning_rate": 0.0008823374357600241,
+      "loss": 1.5628,
+      "step": 2291
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08686365181873504,
+      "learning_rate": 0.0008822252149849686,
+      "loss": 1.4197,
+      "step": 2292
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.07392644100827063,
+      "learning_rate": 0.000882112947863788,
+      "loss": 1.3753,
+      "step": 2293
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08686559375776513,
+      "learning_rate": 0.0008820006344100953,
+      "loss": 1.5226,
+      "step": 2294
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0891901138257802,
+      "learning_rate": 0.0008818882746375085,
+      "loss": 1.516,
+      "step": 2295
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.09554588650132052,
+      "learning_rate": 0.0008817758685596519,
+      "loss": 1.4844,
+      "step": 2296
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.07631449918448031,
+      "learning_rate": 0.0008816634161901552,
+      "loss": 1.3256,
+      "step": 2297
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08089428852812466,
+      "learning_rate": 0.0008815509175426534,
+      "loss": 1.3853,
+      "step": 2298
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.07814014705209726,
+      "learning_rate": 0.0008814383726307876,
+      "loss": 1.473,
+      "step": 2299
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.07405353182252997,
+      "learning_rate": 0.0008813257814682038,
+      "loss": 1.5659,
+      "step": 2300
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.07250750021173145,
+      "learning_rate": 0.0008812131440685544,
+      "loss": 1.4467,
+      "step": 2301
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0717452676620836,
+      "learning_rate": 0.0008811004604454973,
+      "loss": 1.4328,
+      "step": 2302
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.07509376431463823,
+      "learning_rate": 0.0008809877306126953,
+      "loss": 1.4588,
+      "step": 2303
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.09604822190961418,
+      "learning_rate": 0.0008808749545838176,
+      "loss": 1.4542,
+      "step": 2304
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08803782587082482,
+      "learning_rate": 0.0008807621323725386,
+      "loss": 1.4286,
+      "step": 2305
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.07727137861921227,
+      "learning_rate": 0.0008806492639925383,
+      "loss": 1.5215,
+      "step": 2306
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08376997789258082,
+      "learning_rate": 0.0008805363494575024,
+      "loss": 1.3851,
+      "step": 2307
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0819358725338881,
+      "learning_rate": 0.0008804233887811223,
+      "loss": 1.4812,
+      "step": 2308
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0792579194627983,
+      "learning_rate": 0.0008803103819770947,
+      "loss": 1.5521,
+      "step": 2309
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0915688602232919,
+      "learning_rate": 0.0008801973290591223,
+      "loss": 1.4422,
+      "step": 2310
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.09290993712517474,
+      "learning_rate": 0.0008800842300409129,
+      "loss": 1.4499,
+      "step": 2311
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.07260739372011728,
+      "learning_rate": 0.0008799710849361803,
+      "loss": 1.3542,
+      "step": 2312
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.07889876448184573,
+      "learning_rate": 0.0008798578937586436,
+      "loss": 1.5093,
+      "step": 2313
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.082364590057529,
+      "learning_rate": 0.0008797446565220278,
+      "loss": 1.3858,
+      "step": 2314
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.07912789012616007,
+      "learning_rate": 0.0008796313732400634,
+      "loss": 1.4733,
+      "step": 2315
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08590092205671555,
+      "learning_rate": 0.000879518043926486,
+      "loss": 1.5081,
+      "step": 2316
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.07329552774435327,
+      "learning_rate": 0.0008794046685950373,
+      "loss": 1.3782,
+      "step": 2317
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.07799343415492194,
+      "learning_rate": 0.0008792912472594647,
+      "loss": 1.3976,
+      "step": 2318
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08369021131091983,
+      "learning_rate": 0.0008791777799335205,
+      "loss": 1.5131,
+      "step": 2319
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08178300989609327,
+      "learning_rate": 0.0008790642666309637,
+      "loss": 1.5082,
+      "step": 2320
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08988780905708967,
+      "learning_rate": 0.0008789507073655574,
+      "loss": 1.3528,
+      "step": 2321
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08569905270305181,
+      "learning_rate": 0.0008788371021510713,
+      "loss": 1.5646,
+      "step": 2322
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08054189742708023,
+      "learning_rate": 0.0008787234510012807,
+      "loss": 1.3814,
+      "step": 2323
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08068380039302207,
+      "learning_rate": 0.000878609753929966,
+      "loss": 1.4276,
+      "step": 2324
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08722938708128018,
+      "learning_rate": 0.0008784960109509133,
+      "loss": 1.4702,
+      "step": 2325
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08364451612686977,
+      "learning_rate": 0.0008783822220779145,
+      "loss": 1.4256,
+      "step": 2326
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0754569829554669,
+      "learning_rate": 0.0008782683873247667,
+      "loss": 1.4047,
+      "step": 2327
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0885670001296458,
+      "learning_rate": 0.0008781545067052729,
+      "loss": 1.4172,
+      "step": 2328
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08249488016564695,
+      "learning_rate": 0.0008780405802332415,
+      "loss": 1.4541,
+      "step": 2329
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08681960813369974,
+      "learning_rate": 0.0008779266079224863,
+      "loss": 1.3966,
+      "step": 2330
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.09003709379265666,
+      "learning_rate": 0.0008778125897868272,
+      "loss": 1.5273,
+      "step": 2331
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08457069138096256,
+      "learning_rate": 0.0008776985258400889,
+      "loss": 1.5136,
+      "step": 2332
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08639359952616797,
+      "learning_rate": 0.0008775844160961023,
+      "loss": 1.3153,
+      "step": 2333
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08436050301536181,
+      "learning_rate": 0.0008774702605687035,
+      "loss": 1.4928,
+      "step": 2334
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.07796128523951937,
+      "learning_rate": 0.0008773560592717343,
+      "loss": 1.4667,
+      "step": 2335
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08239972146783438,
+      "learning_rate": 0.0008772418122190418,
+      "loss": 1.4009,
+      "step": 2336
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08335490168042457,
+      "learning_rate": 0.0008771275194244792,
+      "loss": 1.4612,
+      "step": 2337
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0857843475703198,
+      "learning_rate": 0.0008770131809019046,
+      "loss": 1.4773,
+      "step": 2338
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08868483682877011,
+      "learning_rate": 0.0008768987966651822,
+      "loss": 1.4839,
+      "step": 2339
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.07590656353342753,
+      "learning_rate": 0.0008767843667281812,
+      "loss": 1.3742,
+      "step": 2340
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08180431782754888,
+      "learning_rate": 0.0008766698911047768,
+      "loss": 1.4264,
+      "step": 2341
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0827417682256729,
+      "learning_rate": 0.0008765553698088496,
+      "loss": 1.3259,
+      "step": 2342
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08122944950217337,
+      "learning_rate": 0.0008764408028542854,
+      "loss": 1.4616,
+      "step": 2343
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.07970569960246655,
+      "learning_rate": 0.0008763261902549762,
+      "loss": 1.4272,
+      "step": 2344
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0831541958515407,
+      "learning_rate": 0.0008762115320248192,
+      "loss": 1.3979,
+      "step": 2345
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.06820614193604783,
+      "learning_rate": 0.0008760968281777167,
+      "loss": 1.4121,
+      "step": 2346
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08231977544674345,
+      "learning_rate": 0.0008759820787275773,
+      "loss": 1.3942,
+      "step": 2347
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.07261045445751502,
+      "learning_rate": 0.0008758672836883146,
+      "loss": 1.4361,
+      "step": 2348
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.07834877807278877,
+      "learning_rate": 0.0008757524430738479,
+      "loss": 1.5091,
+      "step": 2349
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08087675977722167,
+      "learning_rate": 0.0008756375568981023,
+      "loss": 1.5176,
+      "step": 2350
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.06467515099636481,
+      "learning_rate": 0.0008755226251750077,
+      "loss": 1.3714,
+      "step": 2351
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.07426388522565618,
+      "learning_rate": 0.0008754076479185001,
+      "loss": 1.518,
+      "step": 2352
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.09416405071741757,
+      "learning_rate": 0.000875292625142521,
+      "loss": 1.5035,
+      "step": 2353
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08717906392381507,
+      "learning_rate": 0.0008751775568610175,
+      "loss": 1.5043,
+      "step": 2354
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.07416649126779296,
+      "learning_rate": 0.0008750624430879416,
+      "loss": 1.5297,
+      "step": 2355
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.07998119133842944,
+      "learning_rate": 0.0008749472838372514,
+      "loss": 1.3105,
+      "step": 2356
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.081112375757246,
+      "learning_rate": 0.0008748320791229106,
+      "loss": 1.7473,
+      "step": 2357
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08904481652658924,
+      "learning_rate": 0.0008747168289588879,
+      "loss": 1.4059,
+      "step": 2358
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08234605951979214,
+      "learning_rate": 0.0008746015333591578,
+      "loss": 1.5228,
+      "step": 2359
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08955036857950854,
+      "learning_rate": 0.0008744861923377001,
+      "loss": 1.5783,
+      "step": 2360
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0831734267365419,
+      "learning_rate": 0.0008743708059085008,
+      "loss": 1.63,
+      "step": 2361
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08897953739111732,
+      "learning_rate": 0.0008742553740855505,
+      "loss": 1.3316,
+      "step": 2362
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0982441775876888,
+      "learning_rate": 0.000874139896882846,
+      "loss": 1.4574,
+      "step": 2363
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08749307153723353,
+      "learning_rate": 0.000874024374314389,
+      "loss": 1.4225,
+      "step": 2364
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08421424247218816,
+      "learning_rate": 0.0008739088063941874,
+      "loss": 1.4224,
+      "step": 2365
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.09632762229707822,
+      "learning_rate": 0.0008737931931362536,
+      "loss": 1.4607,
+      "step": 2366
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08000551688554444,
+      "learning_rate": 0.0008736775345546066,
+      "loss": 1.5745,
+      "step": 2367
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08320750171271744,
+      "learning_rate": 0.0008735618306632704,
+      "loss": 1.409,
+      "step": 2368
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.07967367321090332,
+      "learning_rate": 0.0008734460814762743,
+      "loss": 1.4925,
+      "step": 2369
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.07739999356529328,
+      "learning_rate": 0.0008733302870076534,
+      "loss": 1.5329,
+      "step": 2370
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0739284834248217,
+      "learning_rate": 0.000873214447271448,
+      "loss": 1.4682,
+      "step": 2371
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08430494075436981,
+      "learning_rate": 0.0008730985622817043,
+      "loss": 1.5422,
+      "step": 2372
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.07628909438924128,
+      "learning_rate": 0.0008729826320524736,
+      "loss": 1.3732,
+      "step": 2373
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.08018562323859652,
+      "learning_rate": 0.0008728666565978129,
+      "loss": 1.3432,
+      "step": 2374
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.07884604930066987,
+      "learning_rate": 0.0008727506359317847,
+      "loss": 1.4342,
+      "step": 2375
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.07219376890047094,
+      "learning_rate": 0.0008726345700684568,
+      "loss": 1.5017,
+      "step": 2376
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.07842816450456827,
+      "learning_rate": 0.0008725184590219026,
+      "loss": 1.3689,
+      "step": 2377
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.08355168533148163,
+      "learning_rate": 0.000872402302806201,
+      "loss": 1.5077,
+      "step": 2378
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.07483659212043353,
+      "learning_rate": 0.0008722861014354363,
+      "loss": 1.4426,
+      "step": 2379
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.08414078995361736,
+      "learning_rate": 0.0008721698549236982,
+      "loss": 1.5652,
+      "step": 2380
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.07285448936483976,
+      "learning_rate": 0.0008720535632850823,
+      "loss": 1.4901,
+      "step": 2381
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.07770740691795891,
+      "learning_rate": 0.0008719372265336892,
+      "loss": 1.4684,
+      "step": 2382
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.09889862035074047,
+      "learning_rate": 0.0008718208446836251,
+      "loss": 1.4999,
+      "step": 2383
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.07299850297868295,
+      "learning_rate": 0.0008717044177490017,
+      "loss": 1.3581,
+      "step": 2384
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.07882407017544525,
+      "learning_rate": 0.0008715879457439362,
+      "loss": 1.3601,
+      "step": 2385
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.08112163967024738,
+      "learning_rate": 0.0008714714286825511,
+      "loss": 1.3724,
+      "step": 2386
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.09481491559845347,
+      "learning_rate": 0.0008713548665789748,
+      "loss": 1.6153,
+      "step": 2387
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.07735816148190601,
+      "learning_rate": 0.0008712382594473404,
+      "loss": 1.3866,
+      "step": 2388
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.08755764137667417,
+      "learning_rate": 0.0008711216073017875,
+      "loss": 1.4494,
+      "step": 2389
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.08229766736555363,
+      "learning_rate": 0.0008710049101564601,
+      "loss": 1.4153,
+      "step": 2390
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0842704737549601,
+      "learning_rate": 0.0008708881680255083,
+      "loss": 1.4117,
+      "step": 2391
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.08135189581146728,
+      "learning_rate": 0.0008707713809230875,
+      "loss": 1.5633,
+      "step": 2392
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0820669553456813,
+      "learning_rate": 0.0008706545488633586,
+      "loss": 1.4688,
+      "step": 2393
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.07755666930453603,
+      "learning_rate": 0.0008705376718604877,
+      "loss": 1.4559,
+      "step": 2394
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0788686142056009,
+      "learning_rate": 0.0008704207499286467,
+      "loss": 1.4704,
+      "step": 2395
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.08433337387934393,
+      "learning_rate": 0.0008703037830820127,
+      "loss": 1.4576,
+      "step": 2396
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.07928307629309808,
+      "learning_rate": 0.0008701867713347684,
+      "loss": 1.4808,
+      "step": 2397
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.083526767428123,
+      "learning_rate": 0.0008700697147011018,
+      "loss": 1.4463,
+      "step": 2398
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0837424016622009,
+      "learning_rate": 0.0008699526131952064,
+      "loss": 1.4302,
+      "step": 2399
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.09432955877287996,
+      "learning_rate": 0.0008698354668312815,
+      "loss": 1.5715,
+      "step": 2400
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.08495084110646854,
+      "learning_rate": 0.0008697182756235311,
+      "loss": 1.45,
+      "step": 2401
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.07433196345469577,
+      "learning_rate": 0.0008696010395861651,
+      "loss": 1.4465,
+      "step": 2402
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.07958498308048166,
+      "learning_rate": 0.0008694837587333988,
+      "loss": 1.4197,
+      "step": 2403
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.07695594597259321,
+      "learning_rate": 0.000869366433079453,
+      "loss": 1.3394,
+      "step": 2404
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.08917274057254963,
+      "learning_rate": 0.0008692490626385538,
+      "loss": 1.5033,
+      "step": 2405
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.07292600891886494,
+      "learning_rate": 0.0008691316474249329,
+      "loss": 1.446,
+      "step": 2406
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.07778910311002209,
+      "learning_rate": 0.000869014187452827,
+      "loss": 1.5116,
+      "step": 2407
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0758528324629491,
+      "learning_rate": 0.0008688966827364788,
+      "loss": 1.4107,
+      "step": 2408
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.07499271508937456,
+      "learning_rate": 0.000868779133290136,
+      "loss": 1.417,
+      "step": 2409
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.09554549660511087,
+      "learning_rate": 0.0008686615391280518,
+      "loss": 1.4752,
+      "step": 2410
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.07605546223030824,
+      "learning_rate": 0.0008685439002644851,
+      "loss": 1.4271,
+      "step": 2411
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.087182582522994,
+      "learning_rate": 0.0008684262167136998,
+      "loss": 1.4356,
+      "step": 2412
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.07107863576670509,
+      "learning_rate": 0.0008683084884899656,
+      "loss": 1.3961,
+      "step": 2413
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.08099026302786344,
+      "learning_rate": 0.0008681907156075577,
+      "loss": 1.5513,
+      "step": 2414
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.09136275098009797,
+      "learning_rate": 0.0008680728980807559,
+      "loss": 1.5117,
+      "step": 2415
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.09153738020255894,
+      "learning_rate": 0.0008679550359238464,
+      "loss": 1.391,
+      "step": 2416
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.08700276520341825,
+      "learning_rate": 0.0008678371291511202,
+      "loss": 1.4262,
+      "step": 2417
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0845217493208463,
+      "learning_rate": 0.0008677191777768739,
+      "loss": 1.3337,
+      "step": 2418
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.07806179350815017,
+      "learning_rate": 0.0008676011818154097,
+      "loss": 1.4307,
+      "step": 2419
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.08075539501899112,
+      "learning_rate": 0.0008674831412810349,
+      "loss": 1.3418,
+      "step": 2420
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.08976538106363947,
+      "learning_rate": 0.0008673650561880622,
+      "loss": 1.433,
+      "step": 2421
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.08419710209291338,
+      "learning_rate": 0.0008672469265508099,
+      "loss": 1.3991,
+      "step": 2422
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.07548679020927829,
+      "learning_rate": 0.0008671287523836018,
+      "loss": 1.3889,
+      "step": 2423
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.07968383779478568,
+      "learning_rate": 0.0008670105337007667,
+      "loss": 1.4236,
+      "step": 2424
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.07402167066513765,
+      "learning_rate": 0.0008668922705166391,
+      "loss": 1.3511,
+      "step": 2425
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.07791837249202221,
+      "learning_rate": 0.0008667739628455591,
+      "loss": 1.6978,
+      "step": 2426
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.07326641874026783,
+      "learning_rate": 0.0008666556107018713,
+      "loss": 1.3809,
+      "step": 2427
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.08249221016536444,
+      "learning_rate": 0.0008665372140999268,
+      "loss": 1.4819,
+      "step": 2428
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.08468100454038449,
+      "learning_rate": 0.0008664187730540813,
+      "loss": 1.3896,
+      "step": 2429
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.07235421408348185,
+      "learning_rate": 0.0008663002875786965,
+      "loss": 1.3804,
+      "step": 2430
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.07356618584967646,
+      "learning_rate": 0.0008661817576881391,
+      "loss": 1.4013,
+      "step": 2431
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.08129543864427304,
+      "learning_rate": 0.0008660631833967809,
+      "loss": 1.4612,
+      "step": 2432
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.07326509117595839,
+      "learning_rate": 0.0008659445647189999,
+      "loss": 1.4255,
+      "step": 2433
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.07271571725202336,
+      "learning_rate": 0.0008658259016691786,
+      "loss": 1.3477,
+      "step": 2434
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.07563391254650749,
+      "learning_rate": 0.0008657071942617056,
+      "loss": 1.5405,
+      "step": 2435
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.07088489797268319,
+      "learning_rate": 0.0008655884425109747,
+      "loss": 1.4278,
+      "step": 2436
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.07552308628678947,
+      "learning_rate": 0.0008654696464313846,
+      "loss": 1.4229,
+      "step": 2437
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.08332445254415345,
+      "learning_rate": 0.0008653508060373399,
+      "loss": 1.4936,
+      "step": 2438
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.07147667803000843,
+      "learning_rate": 0.0008652319213432504,
+      "loss": 1.4365,
+      "step": 2439
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.07498735151078668,
+      "learning_rate": 0.0008651129923635314,
+      "loss": 1.4489,
+      "step": 2440
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.08279578112131421,
+      "learning_rate": 0.0008649940191126033,
+      "loss": 1.4801,
+      "step": 2441
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.07408268276283134,
+      "learning_rate": 0.0008648750016048921,
+      "loss": 1.3708,
+      "step": 2442
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.08279177095879077,
+      "learning_rate": 0.000864755939854829,
+      "loss": 1.4605,
+      "step": 2443
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.07600695074651367,
+      "learning_rate": 0.0008646368338768506,
+      "loss": 1.3836,
+      "step": 2444
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.08935962208751888,
+      "learning_rate": 0.0008645176836853992,
+      "loss": 1.434,
+      "step": 2445
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.07414169923498534,
+      "learning_rate": 0.0008643984892949217,
+      "loss": 1.4505,
+      "step": 2446
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.088469364176592,
+      "learning_rate": 0.0008642792507198713,
+      "loss": 1.3818,
+      "step": 2447
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0826174631745449,
+      "learning_rate": 0.0008641599679747059,
+      "loss": 1.5778,
+      "step": 2448
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.08789837591114688,
+      "learning_rate": 0.0008640406410738888,
+      "loss": 1.3421,
+      "step": 2449
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.07804401790078475,
+      "learning_rate": 0.000863921270031889,
+      "loss": 1.5028,
+      "step": 2450
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.07428320796104862,
+      "learning_rate": 0.0008638018548631805,
+      "loss": 1.4064,
+      "step": 2451
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.07560885338288993,
+      "learning_rate": 0.000863682395582243,
+      "loss": 1.3384,
+      "step": 2452
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.10187961286723403,
+      "learning_rate": 0.0008635628922035613,
+      "loss": 1.5161,
+      "step": 2453
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.09026702268572244,
+      "learning_rate": 0.0008634433447416253,
+      "loss": 1.4465,
+      "step": 2454
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.09035052032958152,
+      "learning_rate": 0.000863323753210931,
+      "loss": 1.441,
+      "step": 2455
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0813296004930123,
+      "learning_rate": 0.0008632041176259788,
+      "loss": 1.4335,
+      "step": 2456
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.08422711412673596,
+      "learning_rate": 0.0008630844380012754,
+      "loss": 1.4123,
+      "step": 2457
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.07926370932506693,
+      "learning_rate": 0.0008629647143513321,
+      "loss": 1.3936,
+      "step": 2458
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.07404514764363598,
+      "learning_rate": 0.0008628449466906658,
+      "loss": 1.3632,
+      "step": 2459
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.07673003267922554,
+      "learning_rate": 0.0008627251350337988,
+      "loss": 1.4799,
+      "step": 2460
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.07558809531135767,
+      "learning_rate": 0.0008626052793952588,
+      "loss": 1.3333,
+      "step": 2461
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.07736495424535647,
+      "learning_rate": 0.0008624853797895784,
+      "loss": 1.4643,
+      "step": 2462
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0781190229156017,
+      "learning_rate": 0.0008623654362312961,
+      "loss": 1.4538,
+      "step": 2463
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.08576676357567804,
+      "learning_rate": 0.0008622454487349554,
+      "loss": 1.4785,
+      "step": 2464
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.07382184959413687,
+      "learning_rate": 0.0008621254173151052,
+      "loss": 1.4344,
+      "step": 2465
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.08142733655694544,
+      "learning_rate": 0.0008620053419862997,
+      "loss": 1.4177,
+      "step": 2466
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0752615574922961,
+      "learning_rate": 0.0008618852227630985,
+      "loss": 1.4311,
+      "step": 2467
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.09972312920589078,
+      "learning_rate": 0.0008617650596600665,
+      "loss": 1.4834,
+      "step": 2468
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.08568112647402251,
+      "learning_rate": 0.0008616448526917736,
+      "loss": 1.4549,
+      "step": 2469
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0780979765408123,
+      "learning_rate": 0.0008615246018727956,
+      "loss": 1.4037,
+      "step": 2470
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.08294238148339915,
+      "learning_rate": 0.0008614043072177135,
+      "loss": 1.4814,
+      "step": 2471
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.09715355593692114,
+      "learning_rate": 0.000861283968741113,
+      "loss": 1.5874,
+      "step": 2472
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.07782506887676875,
+      "learning_rate": 0.0008611635864575857,
+      "loss": 1.381,
+      "step": 2473
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0906288348611541,
+      "learning_rate": 0.0008610431603817285,
+      "loss": 1.3718,
+      "step": 2474
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.07870778480964968,
+      "learning_rate": 0.0008609226905281433,
+      "loss": 1.4466,
+      "step": 2475
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.08602254355096633,
+      "learning_rate": 0.0008608021769114378,
+      "loss": 1.4097,
+      "step": 2476
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0854785814244712,
+      "learning_rate": 0.0008606816195462243,
+      "loss": 1.3345,
+      "step": 2477
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.08923277482885959,
+      "learning_rate": 0.000860561018447121,
+      "loss": 1.4728,
+      "step": 2478
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.09441553256745243,
+      "learning_rate": 0.0008604403736287512,
+      "loss": 1.4979,
+      "step": 2479
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.09169254563073491,
+      "learning_rate": 0.0008603196851057434,
+      "loss": 1.4615,
+      "step": 2480
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.09397528694611802,
+      "learning_rate": 0.0008601989528927317,
+      "loss": 1.3988,
+      "step": 2481
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.08594017810333246,
+      "learning_rate": 0.0008600781770043551,
+      "loss": 1.434,
+      "step": 2482
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0857484672914396,
+      "learning_rate": 0.0008599573574552582,
+      "loss": 1.4412,
+      "step": 2483
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.09556263713315377,
+      "learning_rate": 0.0008598364942600906,
+      "loss": 1.5596,
+      "step": 2484
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.096404618723962,
+      "learning_rate": 0.0008597155874335076,
+      "loss": 1.4049,
+      "step": 2485
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.08729667242752806,
+      "learning_rate": 0.0008595946369901696,
+      "loss": 1.4342,
+      "step": 2486
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.08010480744400537,
+      "learning_rate": 0.0008594736429447421,
+      "loss": 1.4179,
+      "step": 2487
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.07866884276153974,
+      "learning_rate": 0.0008593526053118961,
+      "loss": 1.5207,
+      "step": 2488
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.08333403772066363,
+      "learning_rate": 0.0008592315241063077,
+      "loss": 1.4315,
+      "step": 2489
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.08300905086491174,
+      "learning_rate": 0.0008591103993426588,
+      "loss": 1.3496,
+      "step": 2490
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0820231745923072,
+      "learning_rate": 0.0008589892310356357,
+      "loss": 1.4858,
+      "step": 2491
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.09302760868607428,
+      "learning_rate": 0.0008588680191999308,
+      "loss": 1.4802,
+      "step": 2492
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.08616521316743678,
+      "learning_rate": 0.0008587467638502413,
+      "loss": 1.4069,
+      "step": 2493
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.08199246375158696,
+      "learning_rate": 0.0008586254650012699,
+      "loss": 1.4902,
+      "step": 2494
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.08134043261911114,
+      "learning_rate": 0.0008585041226677247,
+      "loss": 1.4154,
+      "step": 2495
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.07958788079491519,
+      "learning_rate": 0.0008583827368643185,
+      "loss": 1.4978,
+      "step": 2496
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.07389846036126062,
+      "learning_rate": 0.0008582613076057699,
+      "loss": 1.4757,
+      "step": 2497
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.07637567948544043,
+      "learning_rate": 0.0008581398349068028,
+      "loss": 1.3928,
+      "step": 2498
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.07228880909881819,
+      "learning_rate": 0.0008580183187821459,
+      "loss": 1.496,
+      "step": 2499
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.07744674862621322,
+      "learning_rate": 0.0008578967592465335,
+      "loss": 1.409,
+      "step": 2500
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.07110313278941842,
+      "learning_rate": 0.0008577751563147054,
+      "loss": 1.4029,
+      "step": 2501
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0699785168285317,
+      "learning_rate": 0.000857653510001406,
+      "loss": 1.365,
+      "step": 2502
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.07645102695193777,
+      "learning_rate": 0.0008575318203213855,
+      "loss": 1.3709,
+      "step": 2503
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.09179887617243095,
+      "learning_rate": 0.0008574100872893992,
+      "loss": 1.6029,
+      "step": 2504
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.08142107716187588,
+      "learning_rate": 0.0008572883109202077,
+      "loss": 1.4783,
+      "step": 2505
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.06824772265618469,
+      "learning_rate": 0.0008571664912285766,
+      "loss": 1.3282,
+      "step": 2506
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.07416330481288567,
+      "learning_rate": 0.0008570446282292773,
+      "loss": 1.439,
+      "step": 2507
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.07513943521999795,
+      "learning_rate": 0.0008569227219370856,
+      "loss": 1.4978,
+      "step": 2508
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.07361208809988612,
+      "learning_rate": 0.0008568007723667837,
+      "loss": 1.6253,
+      "step": 2509
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.08213288675158097,
+      "learning_rate": 0.0008566787795331579,
+      "loss": 1.4,
+      "step": 2510
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.07926099951067463,
+      "learning_rate": 0.0008565567434510004,
+      "loss": 1.4767,
+      "step": 2511
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.07761340571003827,
+      "learning_rate": 0.0008564346641351087,
+      "loss": 1.3751,
+      "step": 2512
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.07060565203198378,
+      "learning_rate": 0.0008563125416002849,
+      "loss": 1.4009,
+      "step": 2513
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.07997901091917972,
+      "learning_rate": 0.0008561903758613372,
+      "loss": 1.5378,
+      "step": 2514
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.08769052562520391,
+      "learning_rate": 0.0008560681669330783,
+      "loss": 1.3733,
+      "step": 2515
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.07068911164600497,
+      "learning_rate": 0.0008559459148303268,
+      "loss": 1.4424,
+      "step": 2516
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.06882590599628988,
+      "learning_rate": 0.0008558236195679059,
+      "loss": 1.4711,
+      "step": 2517
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.07503865848801855,
+      "learning_rate": 0.0008557012811606444,
+      "loss": 1.5057,
+      "step": 2518
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.07369198331548805,
+      "learning_rate": 0.0008555788996233764,
+      "loss": 1.3597,
+      "step": 2519
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.06915275711769295,
+      "learning_rate": 0.0008554564749709408,
+      "loss": 1.5348,
+      "step": 2520
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.08524842375728962,
+      "learning_rate": 0.0008553340072181822,
+      "loss": 1.2874,
+      "step": 2521
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.08088460709388172,
+      "learning_rate": 0.0008552114963799502,
+      "loss": 1.4587,
+      "step": 2522
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.07020129575793528,
+      "learning_rate": 0.0008550889424710997,
+      "loss": 1.494,
+      "step": 2523
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.07286098460203513,
+      "learning_rate": 0.0008549663455064907,
+      "loss": 1.4065,
+      "step": 2524
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.07224682131553291,
+      "learning_rate": 0.0008548437055009886,
+      "loss": 1.3713,
+      "step": 2525
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.07626714388252791,
+      "learning_rate": 0.0008547210224694639,
+      "loss": 1.4525,
+      "step": 2526
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.08518727889962867,
+      "learning_rate": 0.0008545982964267922,
+      "loss": 1.3893,
+      "step": 2527
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.08097460248924065,
+      "learning_rate": 0.0008544755273878546,
+      "loss": 1.3658,
+      "step": 2528
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.07863853475459574,
+      "learning_rate": 0.0008543527153675374,
+      "loss": 1.5665,
+      "step": 2529
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.07424216152167827,
+      "learning_rate": 0.0008542298603807317,
+      "loss": 1.3939,
+      "step": 2530
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0711106261983796,
+      "learning_rate": 0.0008541069624423343,
+      "loss": 1.3313,
+      "step": 2531
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0786625769759333,
+      "learning_rate": 0.0008539840215672467,
+      "loss": 1.4449,
+      "step": 2532
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.07540816364950761,
+      "learning_rate": 0.0008538610377703764,
+      "loss": 1.5184,
+      "step": 2533
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.07406561827332242,
+      "learning_rate": 0.0008537380110666351,
+      "loss": 1.4328,
+      "step": 2534
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.07747251536991673,
+      "learning_rate": 0.0008536149414709404,
+      "loss": 1.4349,
+      "step": 2535
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.07339786673715235,
+      "learning_rate": 0.0008534918289982152,
+      "loss": 1.3211,
+      "step": 2536
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.07446062835008409,
+      "learning_rate": 0.0008533686736633868,
+      "loss": 1.396,
+      "step": 2537
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.08408711492254073,
+      "learning_rate": 0.0008532454754813886,
+      "loss": 1.4873,
+      "step": 2538
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.06694151801394763,
+      "learning_rate": 0.0008531222344671588,
+      "loss": 1.5546,
+      "step": 2539
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.10107174239142706,
+      "learning_rate": 0.0008529989506356406,
+      "loss": 1.4372,
+      "step": 2540
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.07716872380123085,
+      "learning_rate": 0.0008528756240017825,
+      "loss": 1.3469,
+      "step": 2541
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1048404291852175,
+      "learning_rate": 0.0008527522545805385,
+      "loss": 1.4357,
+      "step": 2542
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.06980478409748192,
+      "learning_rate": 0.0008526288423868675,
+      "loss": 1.3616,
+      "step": 2543
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.07265651507995816,
+      "learning_rate": 0.0008525053874357338,
+      "loss": 1.4286,
+      "step": 2544
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.07424307098433014,
+      "learning_rate": 0.0008523818897421065,
+      "loss": 1.4647,
+      "step": 2545
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.07441297327581105,
+      "learning_rate": 0.0008522583493209603,
+      "loss": 1.4718,
+      "step": 2546
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.07910137117621747,
+      "learning_rate": 0.0008521347661872748,
+      "loss": 1.5926,
+      "step": 2547
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.07998697396490068,
+      "learning_rate": 0.0008520111403560349,
+      "loss": 1.3263,
+      "step": 2548
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0724968400988003,
+      "learning_rate": 0.0008518874718422307,
+      "loss": 1.3337,
+      "step": 2549
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0798632264900906,
+      "learning_rate": 0.0008517637606608573,
+      "loss": 1.4341,
+      "step": 2550
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.07801198011377168,
+      "learning_rate": 0.0008516400068269152,
+      "loss": 1.5612,
+      "step": 2551
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.08604863144633705,
+      "learning_rate": 0.00085151621035541,
+      "loss": 1.4635,
+      "step": 2552
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.07636576667059033,
+      "learning_rate": 0.0008513923712613525,
+      "loss": 1.5159,
+      "step": 2553
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.10516077127557029,
+      "learning_rate": 0.0008512684895597586,
+      "loss": 1.4889,
+      "step": 2554
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.07562060657657407,
+      "learning_rate": 0.0008511445652656494,
+      "loss": 1.4407,
+      "step": 2555
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.07763797531070722,
+      "learning_rate": 0.0008510205983940507,
+      "loss": 1.5135,
+      "step": 2556
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.07375876092295945,
+      "learning_rate": 0.0008508965889599947,
+      "loss": 1.4025,
+      "step": 2557
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.07563084794765446,
+      "learning_rate": 0.0008507725369785173,
+      "loss": 1.3534,
+      "step": 2558
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.07692755695568271,
+      "learning_rate": 0.0008506484424646606,
+      "loss": 1.4164,
+      "step": 2559
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0808219714314826,
+      "learning_rate": 0.0008505243054334713,
+      "loss": 1.4273,
+      "step": 2560
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.08065608841873591,
+      "learning_rate": 0.0008504001259000016,
+      "loss": 1.4571,
+      "step": 2561
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.08063335526960716,
+      "learning_rate": 0.0008502759038793087,
+      "loss": 1.3944,
+      "step": 2562
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.09265606382023213,
+      "learning_rate": 0.0008501516393864548,
+      "loss": 1.5381,
+      "step": 2563
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.08276084737543599,
+      "learning_rate": 0.0008500273324365073,
+      "loss": 1.4669,
+      "step": 2564
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.08159564795043535,
+      "learning_rate": 0.0008499029830445393,
+      "loss": 1.4467,
+      "step": 2565
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.08220391890474721,
+      "learning_rate": 0.0008497785912256282,
+      "loss": 1.3983,
+      "step": 2566
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.07864749205521121,
+      "learning_rate": 0.000849654156994857,
+      "loss": 1.538,
+      "step": 2567
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.08184544063144142,
+      "learning_rate": 0.0008495296803673138,
+      "loss": 1.4729,
+      "step": 2568
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.08469200046218774,
+      "learning_rate": 0.0008494051613580918,
+      "loss": 1.402,
+      "step": 2569
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.07779454288092019,
+      "learning_rate": 0.0008492805999822897,
+      "loss": 1.4113,
+      "step": 2570
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.08809251512257028,
+      "learning_rate": 0.0008491559962550104,
+      "loss": 1.4379,
+      "step": 2571
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.08325662907420392,
+      "learning_rate": 0.0008490313501913629,
+      "loss": 1.3634,
+      "step": 2572
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.08503557494839924,
+      "learning_rate": 0.0008489066618064609,
+      "loss": 1.482,
+      "step": 2573
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.08179961293993213,
+      "learning_rate": 0.0008487819311154233,
+      "loss": 1.3169,
+      "step": 2574
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.07767794072518784,
+      "learning_rate": 0.0008486571581333742,
+      "loss": 1.4895,
+      "step": 2575
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.07652551678602128,
+      "learning_rate": 0.0008485323428754426,
+      "loss": 1.4348,
+      "step": 2576
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.07414488818852104,
+      "learning_rate": 0.0008484074853567629,
+      "loss": 1.395,
+      "step": 2577
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.08519549165590558,
+      "learning_rate": 0.0008482825855924743,
+      "loss": 1.4378,
+      "step": 2578
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.08324947940809344,
+      "learning_rate": 0.0008481576435977217,
+      "loss": 1.3522,
+      "step": 2579
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.06811742710699656,
+      "learning_rate": 0.0008480326593876544,
+      "loss": 1.3793,
+      "step": 2580
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0812754167751017,
+      "learning_rate": 0.0008479076329774274,
+      "loss": 1.4694,
+      "step": 2581
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.07372976296447055,
+      "learning_rate": 0.0008477825643822003,
+      "loss": 1.5546,
+      "step": 2582
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.08190879452155753,
+      "learning_rate": 0.0008476574536171385,
+      "loss": 1.3917,
+      "step": 2583
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.09192577655834207,
+      "learning_rate": 0.0008475323006974116,
+      "loss": 1.5549,
+      "step": 2584
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.07686509843963789,
+      "learning_rate": 0.0008474071056381953,
+      "loss": 1.4178,
+      "step": 2585
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0905695216161556,
+      "learning_rate": 0.0008472818684546697,
+      "loss": 1.4021,
+      "step": 2586
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.07791879082984891,
+      "learning_rate": 0.0008471565891620203,
+      "loss": 1.5132,
+      "step": 2587
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.07101522886996688,
+      "learning_rate": 0.0008470312677754377,
+      "loss": 1.4741,
+      "step": 2588
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.08034711963413725,
+      "learning_rate": 0.0008469059043101175,
+      "loss": 1.4286,
+      "step": 2589
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.07548169925316262,
+      "learning_rate": 0.0008467804987812603,
+      "loss": 1.3944,
+      "step": 2590
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.07519177174592058,
+      "learning_rate": 0.0008466550512040722,
+      "loss": 1.4444,
+      "step": 2591
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.09383438703269041,
+      "learning_rate": 0.0008465295615937641,
+      "loss": 1.4118,
+      "step": 2592
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.08051542055976478,
+      "learning_rate": 0.0008464040299655518,
+      "loss": 1.4393,
+      "step": 2593
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.08233997317534307,
+      "learning_rate": 0.0008462784563346567,
+      "loss": 1.401,
+      "step": 2594
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.08144426653897043,
+      "learning_rate": 0.000846152840716305,
+      "loss": 1.4978,
+      "step": 2595
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.09101065818594047,
+      "learning_rate": 0.000846027183125728,
+      "loss": 1.4853,
+      "step": 2596
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.08392381563918748,
+      "learning_rate": 0.0008459014835781621,
+      "loss": 1.3675,
+      "step": 2597
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0800643944017374,
+      "learning_rate": 0.0008457757420888488,
+      "loss": 1.4346,
+      "step": 2598
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.07982629773993684,
+      "learning_rate": 0.0008456499586730346,
+      "loss": 1.3716,
+      "step": 2599
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0933833278071163,
+      "learning_rate": 0.0008455241333459715,
+      "loss": 1.4097,
+      "step": 2600
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.09264156229360837,
+      "learning_rate": 0.0008453982661229158,
+      "loss": 1.4631,
+      "step": 2601
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.09061343077073224,
+      "learning_rate": 0.0008452723570191297,
+      "loss": 1.5063,
+      "step": 2602
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.07478352133945661,
+      "learning_rate": 0.0008451464060498799,
+      "loss": 1.2421,
+      "step": 2603
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.08105557941037968,
+      "learning_rate": 0.0008450204132304386,
+      "loss": 1.4022,
+      "step": 2604
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.08207495978087463,
+      "learning_rate": 0.0008448943785760826,
+      "loss": 1.3719,
+      "step": 2605
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.08368550950846189,
+      "learning_rate": 0.0008447683021020942,
+      "loss": 1.5827,
+      "step": 2606
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.07688854298372594,
+      "learning_rate": 0.0008446421838237605,
+      "loss": 1.4621,
+      "step": 2607
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.09196343428155447,
+      "learning_rate": 0.0008445160237563741,
+      "loss": 1.5451,
+      "step": 2608
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.07903413930876052,
+      "learning_rate": 0.0008443898219152319,
+      "loss": 1.4598,
+      "step": 2609
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.07076947983689727,
+      "learning_rate": 0.0008442635783156365,
+      "loss": 1.3268,
+      "step": 2610
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.07636266741339315,
+      "learning_rate": 0.0008441372929728954,
+      "loss": 1.5277,
+      "step": 2611
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.07600285349782061,
+      "learning_rate": 0.0008440109659023211,
+      "loss": 1.4652,
+      "step": 2612
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.07201022852211071,
+      "learning_rate": 0.0008438845971192313,
+      "loss": 1.4532,
+      "step": 2613
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.07285683346924836,
+      "learning_rate": 0.0008437581866389483,
+      "loss": 1.4461,
+      "step": 2614
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.07768183722711053,
+      "learning_rate": 0.0008436317344768003,
+      "loss": 1.585,
+      "step": 2615
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.08838640887202365,
+      "learning_rate": 0.0008435052406481196,
+      "loss": 1.4553,
+      "step": 2616
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.08074211986383892,
+      "learning_rate": 0.0008433787051682443,
+      "loss": 1.5233,
+      "step": 2617
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.07530166795895882,
+      "learning_rate": 0.0008432521280525174,
+      "loss": 1.543,
+      "step": 2618
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.07064169571659298,
+      "learning_rate": 0.0008431255093162864,
+      "loss": 1.5052,
+      "step": 2619
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.07175033898234041,
+      "learning_rate": 0.0008429988489749046,
+      "loss": 1.5151,
+      "step": 2620
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.07320433910283142,
+      "learning_rate": 0.0008428721470437296,
+      "loss": 1.3737,
+      "step": 2621
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.07414426014352578,
+      "learning_rate": 0.0008427454035381249,
+      "loss": 1.5339,
+      "step": 2622
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.09143642921759464,
+      "learning_rate": 0.0008426186184734585,
+      "loss": 1.4326,
+      "step": 2623
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0845165786223017,
+      "learning_rate": 0.0008424917918651031,
+      "loss": 1.4812,
+      "step": 2624
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0759915730522228,
+      "learning_rate": 0.0008423649237284376,
+      "loss": 1.4493,
+      "step": 2625
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0745678505883287,
+      "learning_rate": 0.0008422380140788445,
+      "loss": 1.36,
+      "step": 2626
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.08146179178868668,
+      "learning_rate": 0.0008421110629317123,
+      "loss": 1.4266,
+      "step": 2627
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.07628880445326346,
+      "learning_rate": 0.0008419840703024344,
+      "loss": 1.401,
+      "step": 2628
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.07686954348813545,
+      "learning_rate": 0.0008418570362064091,
+      "loss": 1.5006,
+      "step": 2629
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.07513837237629467,
+      "learning_rate": 0.0008417299606590393,
+      "loss": 1.4112,
+      "step": 2630
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.07439513015706296,
+      "learning_rate": 0.0008416028436757339,
+      "loss": 1.3948,
+      "step": 2631
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.07113589119232142,
+      "learning_rate": 0.0008414756852719059,
+      "loss": 1.3521,
+      "step": 2632
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0746111940741349,
+      "learning_rate": 0.0008413484854629739,
+      "loss": 1.5365,
+      "step": 2633
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.08053469900948855,
+      "learning_rate": 0.0008412212442643611,
+      "loss": 1.317,
+      "step": 2634
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.07311809517893839,
+      "learning_rate": 0.0008410939616914961,
+      "loss": 1.3688,
+      "step": 2635
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.08879393398483074,
+      "learning_rate": 0.0008409666377598124,
+      "loss": 1.4062,
+      "step": 2636
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.06956310232444521,
+      "learning_rate": 0.0008408392724847482,
+      "loss": 1.3274,
+      "step": 2637
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.07211476766646407,
+      "learning_rate": 0.0008407118658817474,
+      "loss": 1.4828,
+      "step": 2638
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.07084769306180853,
+      "learning_rate": 0.0008405844179662581,
+      "loss": 1.46,
+      "step": 2639
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.08089545995057705,
+      "learning_rate": 0.000840456928753734,
+      "loss": 1.4635,
+      "step": 2640
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.09611046115627091,
+      "learning_rate": 0.0008403293982596336,
+      "loss": 1.505,
+      "step": 2641
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.07674825520262066,
+      "learning_rate": 0.0008402018264994203,
+      "loss": 1.4323,
+      "step": 2642
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0753083169510489,
+      "learning_rate": 0.0008400742134885627,
+      "loss": 1.4169,
+      "step": 2643
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.07514595139372099,
+      "learning_rate": 0.0008399465592425342,
+      "loss": 1.3978,
+      "step": 2644
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.06846109808760739,
+      "learning_rate": 0.0008398188637768136,
+      "loss": 1.3855,
+      "step": 2645
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.08015966915623766,
+      "learning_rate": 0.0008396911271068841,
+      "loss": 1.4843,
+      "step": 2646
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.07651693334730639,
+      "learning_rate": 0.0008395633492482344,
+      "loss": 1.4069,
+      "step": 2647
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.08363488440862517,
+      "learning_rate": 0.0008394355302163578,
+      "loss": 1.5263,
+      "step": 2648
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0910487532185436,
+      "learning_rate": 0.0008393076700267532,
+      "loss": 1.5277,
+      "step": 2649
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.08272526983826728,
+      "learning_rate": 0.0008391797686949237,
+      "loss": 1.4217,
+      "step": 2650
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.08459734947024608,
+      "learning_rate": 0.000839051826236378,
+      "loss": 1.3437,
+      "step": 2651
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07350458824423961,
+      "learning_rate": 0.0008389238426666294,
+      "loss": 1.4984,
+      "step": 2652
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07125971137650254,
+      "learning_rate": 0.0008387958180011964,
+      "loss": 1.329,
+      "step": 2653
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0819044410255831,
+      "learning_rate": 0.0008386677522556025,
+      "loss": 1.372,
+      "step": 2654
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07611520811269741,
+      "learning_rate": 0.0008385396454453762,
+      "loss": 1.6093,
+      "step": 2655
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.06776836737705744,
+      "learning_rate": 0.0008384114975860507,
+      "loss": 1.4628,
+      "step": 2656
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07221637155692598,
+      "learning_rate": 0.0008382833086931642,
+      "loss": 1.2861,
+      "step": 2657
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07581032939812374,
+      "learning_rate": 0.0008381550787822605,
+      "loss": 1.588,
+      "step": 2658
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0767775626653725,
+      "learning_rate": 0.0008380268078688877,
+      "loss": 1.4529,
+      "step": 2659
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07200170746516522,
+      "learning_rate": 0.0008378984959685991,
+      "loss": 1.4092,
+      "step": 2660
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07172741580740591,
+      "learning_rate": 0.0008377701430969528,
+      "loss": 1.4825,
+      "step": 2661
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.06814074124177844,
+      "learning_rate": 0.0008376417492695123,
+      "loss": 1.4343,
+      "step": 2662
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.08848005244759329,
+      "learning_rate": 0.0008375133145018457,
+      "loss": 1.5059,
+      "step": 2663
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07264186405227538,
+      "learning_rate": 0.000837384838809526,
+      "loss": 1.279,
+      "step": 2664
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07900937324347093,
+      "learning_rate": 0.0008372563222081316,
+      "loss": 1.4473,
+      "step": 2665
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.08847583018877739,
+      "learning_rate": 0.0008371277647132453,
+      "loss": 1.5201,
+      "step": 2666
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07547179396139557,
+      "learning_rate": 0.0008369991663404555,
+      "loss": 1.332,
+      "step": 2667
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07603778101136066,
+      "learning_rate": 0.0008368705271053547,
+      "loss": 1.2928,
+      "step": 2668
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.06737944518535066,
+      "learning_rate": 0.0008367418470235413,
+      "loss": 1.3386,
+      "step": 2669
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.09022510747259038,
+      "learning_rate": 0.0008366131261106179,
+      "loss": 1.6164,
+      "step": 2670
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07686893921895585,
+      "learning_rate": 0.0008364843643821927,
+      "loss": 1.4379,
+      "step": 2671
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07283643227924404,
+      "learning_rate": 0.000836355561853878,
+      "loss": 1.4125,
+      "step": 2672
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07205487947966797,
+      "learning_rate": 0.0008362267185412919,
+      "loss": 1.4272,
+      "step": 2673
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07591904501548323,
+      "learning_rate": 0.0008360978344600572,
+      "loss": 1.3735,
+      "step": 2674
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0840464697876995,
+      "learning_rate": 0.0008359689096258011,
+      "loss": 1.4788,
+      "step": 2675
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07417724728902096,
+      "learning_rate": 0.0008358399440541567,
+      "loss": 1.4503,
+      "step": 2676
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07269091407325959,
+      "learning_rate": 0.0008357109377607611,
+      "loss": 1.4989,
+      "step": 2677
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.08418319284393984,
+      "learning_rate": 0.0008355818907612569,
+      "loss": 1.557,
+      "step": 2678
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07474993738977974,
+      "learning_rate": 0.0008354528030712915,
+      "loss": 1.4532,
+      "step": 2679
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07500813692751027,
+      "learning_rate": 0.0008353236747065174,
+      "loss": 1.634,
+      "step": 2680
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0829101197991892,
+      "learning_rate": 0.0008351945056825917,
+      "loss": 1.3884,
+      "step": 2681
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07835371691603794,
+      "learning_rate": 0.0008350652960151765,
+      "loss": 1.4707,
+      "step": 2682
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.08405599478984883,
+      "learning_rate": 0.0008349360457199391,
+      "loss": 1.4975,
+      "step": 2683
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07659122988719483,
+      "learning_rate": 0.0008348067548125514,
+      "loss": 1.2843,
+      "step": 2684
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07639967954015899,
+      "learning_rate": 0.0008346774233086904,
+      "loss": 1.4632,
+      "step": 2685
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.08634703761278015,
+      "learning_rate": 0.000834548051224038,
+      "loss": 1.4512,
+      "step": 2686
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07601354417275835,
+      "learning_rate": 0.0008344186385742811,
+      "loss": 1.455,
+      "step": 2687
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07770416940481581,
+      "learning_rate": 0.0008342891853751114,
+      "loss": 1.4682,
+      "step": 2688
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0738525816231597,
+      "learning_rate": 0.0008341596916422254,
+      "loss": 1.459,
+      "step": 2689
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.08505835511210237,
+      "learning_rate": 0.0008340301573913249,
+      "loss": 1.567,
+      "step": 2690
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07871515325468663,
+      "learning_rate": 0.0008339005826381161,
+      "loss": 1.3747,
+      "step": 2691
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07020707472104386,
+      "learning_rate": 0.0008337709673983106,
+      "loss": 1.4348,
+      "step": 2692
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07336686807263738,
+      "learning_rate": 0.0008336413116876245,
+      "loss": 1.3568,
+      "step": 2693
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07164094008232932,
+      "learning_rate": 0.0008335116155217793,
+      "loss": 1.4443,
+      "step": 2694
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.06686213937741443,
+      "learning_rate": 0.0008333818789165008,
+      "loss": 1.4464,
+      "step": 2695
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07568116953259008,
+      "learning_rate": 0.00083325210188752,
+      "loss": 1.3801,
+      "step": 2696
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07516356051721163,
+      "learning_rate": 0.0008331222844505733,
+      "loss": 1.2997,
+      "step": 2697
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.06746999225222641,
+      "learning_rate": 0.000832992426621401,
+      "loss": 1.4718,
+      "step": 2698
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07642008615200645,
+      "learning_rate": 0.000832862528415749,
+      "loss": 1.4935,
+      "step": 2699
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0784700225660804,
+      "learning_rate": 0.0008327325898493677,
+      "loss": 1.5099,
+      "step": 2700
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07330172661899201,
+      "learning_rate": 0.000832602610938013,
+      "loss": 1.5399,
+      "step": 2701
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0717009174794634,
+      "learning_rate": 0.000832472591697445,
+      "loss": 1.4603,
+      "step": 2702
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07255700608309591,
+      "learning_rate": 0.0008323425321434291,
+      "loss": 1.3573,
+      "step": 2703
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07581764931155754,
+      "learning_rate": 0.0008322124322917353,
+      "loss": 1.4361,
+      "step": 2704
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07880780191271873,
+      "learning_rate": 0.0008320822921581388,
+      "loss": 1.4721,
+      "step": 2705
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07603032304986217,
+      "learning_rate": 0.0008319521117584194,
+      "loss": 1.2977,
+      "step": 2706
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.06423068223877285,
+      "learning_rate": 0.0008318218911083623,
+      "loss": 1.4188,
+      "step": 2707
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.08181975268370313,
+      "learning_rate": 0.0008316916302237568,
+      "loss": 1.5171,
+      "step": 2708
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0788819316295538,
+      "learning_rate": 0.0008315613291203976,
+      "loss": 1.4912,
+      "step": 2709
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07534497094282523,
+      "learning_rate": 0.0008314309878140842,
+      "loss": 1.4276,
+      "step": 2710
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07599726403491437,
+      "learning_rate": 0.000831300606320621,
+      "loss": 1.3766,
+      "step": 2711
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.08399386546384041,
+      "learning_rate": 0.0008311701846558171,
+      "loss": 1.4363,
+      "step": 2712
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07194774780906135,
+      "learning_rate": 0.0008310397228354864,
+      "loss": 1.359,
+      "step": 2713
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07170651795131391,
+      "learning_rate": 0.0008309092208754483,
+      "loss": 1.4945,
+      "step": 2714
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.09471088333737213,
+      "learning_rate": 0.0008307786787915261,
+      "loss": 1.6064,
+      "step": 2715
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07001458324123662,
+      "learning_rate": 0.0008306480965995489,
+      "loss": 1.477,
+      "step": 2716
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07331420219072145,
+      "learning_rate": 0.0008305174743153499,
+      "loss": 1.4956,
+      "step": 2717
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.08307822550827987,
+      "learning_rate": 0.000830386811954768,
+      "loss": 1.4908,
+      "step": 2718
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.10553037644370117,
+      "learning_rate": 0.0008302561095336459,
+      "loss": 1.502,
+      "step": 2719
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07753527659679234,
+      "learning_rate": 0.0008301253670678319,
+      "loss": 1.3897,
+      "step": 2720
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07645511236420256,
+      "learning_rate": 0.0008299945845731792,
+      "loss": 1.4402,
+      "step": 2721
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07836695249123142,
+      "learning_rate": 0.0008298637620655453,
+      "loss": 1.3848,
+      "step": 2722
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.08857471945388738,
+      "learning_rate": 0.0008297328995607932,
+      "loss": 1.3904,
+      "step": 2723
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.08127929407000808,
+      "learning_rate": 0.0008296019970747902,
+      "loss": 1.2742,
+      "step": 2724
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07769020055838141,
+      "learning_rate": 0.0008294710546234086,
+      "loss": 1.2634,
+      "step": 2725
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07820145025470278,
+      "learning_rate": 0.0008293400722225259,
+      "loss": 1.4207,
+      "step": 2726
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.08332673123651665,
+      "learning_rate": 0.0008292090498880241,
+      "loss": 1.4028,
+      "step": 2727
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.08469613645080842,
+      "learning_rate": 0.0008290779876357899,
+      "loss": 1.507,
+      "step": 2728
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07603799229175152,
+      "learning_rate": 0.0008289468854817153,
+      "loss": 1.4011,
+      "step": 2729
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07966344010849837,
+      "learning_rate": 0.0008288157434416967,
+      "loss": 1.4921,
+      "step": 2730
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07132344853654753,
+      "learning_rate": 0.0008286845615316356,
+      "loss": 1.3162,
+      "step": 2731
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07104434671857382,
+      "learning_rate": 0.0008285533397674382,
+      "loss": 1.373,
+      "step": 2732
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.08032338161910683,
+      "learning_rate": 0.0008284220781650158,
+      "loss": 1.4925,
+      "step": 2733
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0877688716941064,
+      "learning_rate": 0.000828290776740284,
+      "loss": 1.4924,
+      "step": 2734
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07186097832409206,
+      "learning_rate": 0.0008281594355091641,
+      "loss": 1.3546,
+      "step": 2735
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07622048007118681,
+      "learning_rate": 0.0008280280544875811,
+      "loss": 1.4452,
+      "step": 2736
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07849727387690847,
+      "learning_rate": 0.0008278966336914655,
+      "loss": 1.4423,
+      "step": 2737
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07222521947527058,
+      "learning_rate": 0.0008277651731367528,
+      "loss": 1.4843,
+      "step": 2738
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07099999895203911,
+      "learning_rate": 0.0008276336728393828,
+      "loss": 1.4156,
+      "step": 2739
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.06822141452088458,
+      "learning_rate": 0.0008275021328153006,
+      "loss": 1.3963,
+      "step": 2740
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0740043671559136,
+      "learning_rate": 0.0008273705530804554,
+      "loss": 1.4871,
+      "step": 2741
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07094511989097507,
+      "learning_rate": 0.0008272389336508022,
+      "loss": 1.2901,
+      "step": 2742
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.07948023468384699,
+      "learning_rate": 0.0008271072745423,
+      "loss": 1.4011,
+      "step": 2743
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.08280207297295716,
+      "learning_rate": 0.0008269755757709132,
+      "loss": 1.5201,
+      "step": 2744
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.088306395939248,
+      "learning_rate": 0.0008268438373526106,
+      "loss": 1.517,
+      "step": 2745
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.07326544074639216,
+      "learning_rate": 0.0008267120593033659,
+      "loss": 1.4476,
+      "step": 2746
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.08336236489635436,
+      "learning_rate": 0.0008265802416391577,
+      "loss": 1.2915,
+      "step": 2747
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.08857755307752961,
+      "learning_rate": 0.0008264483843759691,
+      "loss": 1.4161,
+      "step": 2748
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.08720921111079,
+      "learning_rate": 0.0008263164875297887,
+      "loss": 1.3664,
+      "step": 2749
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0849530419155793,
+      "learning_rate": 0.0008261845511166092,
+      "loss": 1.4203,
+      "step": 2750
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.08352948658753696,
+      "learning_rate": 0.0008260525751524282,
+      "loss": 1.4433,
+      "step": 2751
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.07117468986953664,
+      "learning_rate": 0.0008259205596532484,
+      "loss": 1.4106,
+      "step": 2752
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.09710061791847775,
+      "learning_rate": 0.0008257885046350773,
+      "loss": 1.3963,
+      "step": 2753
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.08773828788389773,
+      "learning_rate": 0.0008256564101139266,
+      "loss": 1.4722,
+      "step": 2754
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0782548391737564,
+      "learning_rate": 0.0008255242761058135,
+      "loss": 1.3094,
+      "step": 2755
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.08430634648229657,
+      "learning_rate": 0.0008253921026267599,
+      "loss": 1.6252,
+      "step": 2756
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.08229486931029277,
+      "learning_rate": 0.0008252598896927918,
+      "loss": 1.4528,
+      "step": 2757
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.08570349190519552,
+      "learning_rate": 0.0008251276373199408,
+      "loss": 1.445,
+      "step": 2758
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0823944896586127,
+      "learning_rate": 0.0008249953455242429,
+      "loss": 1.4586,
+      "step": 2759
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.078848461846731,
+      "learning_rate": 0.000824863014321739,
+      "loss": 1.3897,
+      "step": 2760
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.07616890999329672,
+      "learning_rate": 0.0008247306437284747,
+      "loss": 1.3357,
+      "step": 2761
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.07291692672694469,
+      "learning_rate": 0.0008245982337605003,
+      "loss": 1.4341,
+      "step": 2762
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0799698621170182,
+      "learning_rate": 0.0008244657844338708,
+      "loss": 1.4776,
+      "step": 2763
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.07410913925115305,
+      "learning_rate": 0.0008243332957646464,
+      "loss": 1.4272,
+      "step": 2764
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0725370622239953,
+      "learning_rate": 0.0008242007677688918,
+      "loss": 1.3854,
+      "step": 2765
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.07919036331246422,
+      "learning_rate": 0.0008240682004626765,
+      "loss": 1.4331,
+      "step": 2766
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.07625402550594935,
+      "learning_rate": 0.0008239355938620745,
+      "loss": 1.4987,
+      "step": 2767
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.07087029054031294,
+      "learning_rate": 0.0008238029479831652,
+      "loss": 1.5275,
+      "step": 2768
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.09117193245742936,
+      "learning_rate": 0.0008236702628420319,
+      "loss": 1.5647,
+      "step": 2769
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.07650468483821572,
+      "learning_rate": 0.0008235375384547635,
+      "loss": 1.4506,
+      "step": 2770
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.09424438605028439,
+      "learning_rate": 0.000823404774837453,
+      "loss": 1.4702,
+      "step": 2771
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.07393962832466713,
+      "learning_rate": 0.0008232719720061987,
+      "loss": 1.3698,
+      "step": 2772
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.07453913303088434,
+      "learning_rate": 0.0008231391299771034,
+      "loss": 1.3223,
+      "step": 2773
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.08888193416857225,
+      "learning_rate": 0.0008230062487662745,
+      "loss": 1.5193,
+      "step": 2774
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.08072388459301796,
+      "learning_rate": 0.0008228733283898243,
+      "loss": 1.4469,
+      "step": 2775
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.08710297536435438,
+      "learning_rate": 0.00082274036886387,
+      "loss": 1.4754,
+      "step": 2776
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.08359284400448978,
+      "learning_rate": 0.0008226073702045333,
+      "loss": 1.5685,
+      "step": 2777
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.08749874887891788,
+      "learning_rate": 0.0008224743324279407,
+      "loss": 1.4436,
+      "step": 2778
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.07760028764078293,
+      "learning_rate": 0.0008223412555502236,
+      "loss": 1.5412,
+      "step": 2779
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.08406358133209071,
+      "learning_rate": 0.000822208139587518,
+      "loss": 1.5003,
+      "step": 2780
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.08354272711327344,
+      "learning_rate": 0.0008220749845559648,
+      "loss": 1.4887,
+      "step": 2781
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.07861669011751445,
+      "learning_rate": 0.0008219417904717091,
+      "loss": 1.3924,
+      "step": 2782
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.06729249498959294,
+      "learning_rate": 0.0008218085573509016,
+      "loss": 1.4451,
+      "step": 2783
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.07110046725712708,
+      "learning_rate": 0.0008216752852096969,
+      "loss": 1.5831,
+      "step": 2784
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.07728581028416445,
+      "learning_rate": 0.0008215419740642549,
+      "loss": 1.4519,
+      "step": 2785
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.07445156605390331,
+      "learning_rate": 0.0008214086239307401,
+      "loss": 1.4988,
+      "step": 2786
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0826710728733664,
+      "learning_rate": 0.0008212752348253216,
+      "loss": 1.3996,
+      "step": 2787
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.10887963866407403,
+      "learning_rate": 0.0008211418067641734,
+      "loss": 1.4847,
+      "step": 2788
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.08905891287841133,
+      "learning_rate": 0.0008210083397634738,
+      "loss": 1.4369,
+      "step": 2789
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.08396342066020057,
+      "learning_rate": 0.0008208748338394064,
+      "loss": 1.2997,
+      "step": 2790
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.08154082009916738,
+      "learning_rate": 0.000820741289008159,
+      "loss": 1.5037,
+      "step": 2791
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0800821125968288,
+      "learning_rate": 0.0008206077052859246,
+      "loss": 1.5425,
+      "step": 2792
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.07327743953485256,
+      "learning_rate": 0.0008204740826889008,
+      "loss": 1.4202,
+      "step": 2793
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.06753263368726979,
+      "learning_rate": 0.0008203404212332897,
+      "loss": 1.4971,
+      "step": 2794
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.06862134633875801,
+      "learning_rate": 0.0008202067209352979,
+      "loss": 1.3466,
+      "step": 2795
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.06855119036723545,
+      "learning_rate": 0.0008200729818111372,
+      "loss": 1.466,
+      "step": 2796
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.08170769665034927,
+      "learning_rate": 0.0008199392038770242,
+      "loss": 1.4341,
+      "step": 2797
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.06737320855709193,
+      "learning_rate": 0.0008198053871491797,
+      "loss": 1.6141,
+      "step": 2798
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0659153505087909,
+      "learning_rate": 0.0008196715316438294,
+      "loss": 1.4589,
+      "step": 2799
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.08287915047683615,
+      "learning_rate": 0.0008195376373772039,
+      "loss": 1.4108,
+      "step": 2800
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.08024241150674424,
+      "learning_rate": 0.0008194037043655382,
+      "loss": 1.2903,
+      "step": 2801
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.07159138451861921,
+      "learning_rate": 0.0008192697326250722,
+      "loss": 1.4705,
+      "step": 2802
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.07311652052094464,
+      "learning_rate": 0.0008191357221720506,
+      "loss": 1.4327,
+      "step": 2803
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0764745512460257,
+      "learning_rate": 0.0008190016730227224,
+      "loss": 1.5245,
+      "step": 2804
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.07772607729105947,
+      "learning_rate": 0.0008188675851933414,
+      "loss": 1.5384,
+      "step": 2805
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.06669285839551875,
+      "learning_rate": 0.0008187334587001664,
+      "loss": 1.3745,
+      "step": 2806
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.08076034380977436,
+      "learning_rate": 0.0008185992935594607,
+      "loss": 1.4328,
+      "step": 2807
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.08093218565950956,
+      "learning_rate": 0.0008184650897874923,
+      "loss": 1.4824,
+      "step": 2808
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.07044639355943785,
+      "learning_rate": 0.000818330847400534,
+      "loss": 1.4994,
+      "step": 2809
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.07088399140907917,
+      "learning_rate": 0.0008181965664148628,
+      "loss": 1.4263,
+      "step": 2810
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.07363030280128995,
+      "learning_rate": 0.0008180622468467611,
+      "loss": 1.2771,
+      "step": 2811
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.08144611471453442,
+      "learning_rate": 0.0008179278887125152,
+      "loss": 1.4055,
+      "step": 2812
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.07736857460180284,
+      "learning_rate": 0.000817793492028417,
+      "loss": 1.4156,
+      "step": 2813
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.06555042851554732,
+      "learning_rate": 0.0008176590568107622,
+      "loss": 1.4743,
+      "step": 2814
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.07479045299224463,
+      "learning_rate": 0.0008175245830758515,
+      "loss": 1.49,
+      "step": 2815
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0841811461554626,
+      "learning_rate": 0.0008173900708399906,
+      "loss": 1.4209,
+      "step": 2816
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.07004572589189338,
+      "learning_rate": 0.0008172555201194894,
+      "loss": 1.5112,
+      "step": 2817
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.07560307310975573,
+      "learning_rate": 0.0008171209309306625,
+      "loss": 1.4413,
+      "step": 2818
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.07308794549332547,
+      "learning_rate": 0.0008169863032898296,
+      "loss": 1.3943,
+      "step": 2819
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.07782205167088278,
+      "learning_rate": 0.0008168516372133145,
+      "loss": 1.4607,
+      "step": 2820
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.08067353235727805,
+      "learning_rate": 0.0008167169327174459,
+      "loss": 1.485,
+      "step": 2821
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.07605076294811133,
+      "learning_rate": 0.0008165821898185576,
+      "loss": 1.3621,
+      "step": 2822
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.07507805847597698,
+      "learning_rate": 0.0008164474085329872,
+      "loss": 1.4791,
+      "step": 2823
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.08074805704595682,
+      "learning_rate": 0.0008163125888770776,
+      "loss": 1.4916,
+      "step": 2824
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.08492739058974759,
+      "learning_rate": 0.0008161777308671762,
+      "loss": 1.529,
+      "step": 2825
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.06727666112619586,
+      "learning_rate": 0.0008160428345196347,
+      "loss": 1.3156,
+      "step": 2826
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0688311847976553,
+      "learning_rate": 0.00081590789985081,
+      "loss": 1.5281,
+      "step": 2827
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.09836022791746445,
+      "learning_rate": 0.0008157729268770635,
+      "loss": 1.4562,
+      "step": 2828
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.06422326070659888,
+      "learning_rate": 0.0008156379156147608,
+      "loss": 1.3688,
+      "step": 2829
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.07043174146263807,
+      "learning_rate": 0.0008155028660802728,
+      "loss": 1.5884,
+      "step": 2830
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.083964177686042,
+      "learning_rate": 0.0008153677782899745,
+      "loss": 1.5427,
+      "step": 2831
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.06899335085260903,
+      "learning_rate": 0.0008152326522602458,
+      "loss": 1.4434,
+      "step": 2832
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.08238764103901412,
+      "learning_rate": 0.0008150974880074713,
+      "loss": 1.5017,
+      "step": 2833
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.07391969606608612,
+      "learning_rate": 0.0008149622855480401,
+      "loss": 1.3963,
+      "step": 2834
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.08318721513910766,
+      "learning_rate": 0.000814827044898346,
+      "loss": 1.5438,
+      "step": 2835
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.08554886321365836,
+      "learning_rate": 0.0008146917660747872,
+      "loss": 1.4779,
+      "step": 2836
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.07698488554302305,
+      "learning_rate": 0.0008145564490937668,
+      "loss": 1.4578,
+      "step": 2837
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.07037277585204749,
+      "learning_rate": 0.0008144210939716927,
+      "loss": 1.4635,
+      "step": 2838
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.07461934381938339,
+      "learning_rate": 0.0008142857007249768,
+      "loss": 1.4102,
+      "step": 2839
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.08392078326405981,
+      "learning_rate": 0.0008141502693700363,
+      "loss": 1.4993,
+      "step": 2840
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.06748677350552709,
+      "learning_rate": 0.0008140147999232925,
+      "loss": 1.5196,
+      "step": 2841
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.07883616211063142,
+      "learning_rate": 0.0008138792924011717,
+      "loss": 1.2979,
+      "step": 2842
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.06773196082861774,
+      "learning_rate": 0.0008137437468201047,
+      "loss": 1.3348,
+      "step": 2843
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.06988800909616495,
+      "learning_rate": 0.0008136081631965267,
+      "loss": 1.4625,
+      "step": 2844
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.07477987235145722,
+      "learning_rate": 0.0008134725415468775,
+      "loss": 1.4027,
+      "step": 2845
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.07010726064819425,
+      "learning_rate": 0.0008133368818876022,
+      "loss": 1.4468,
+      "step": 2846
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.08261004610399639,
+      "learning_rate": 0.0008132011842351496,
+      "loss": 1.3365,
+      "step": 2847
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.08736930521087224,
+      "learning_rate": 0.0008130654486059737,
+      "loss": 1.566,
+      "step": 2848
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.06642310720948336,
+      "learning_rate": 0.0008129296750165329,
+      "loss": 1.4061,
+      "step": 2849
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.08933889777362511,
+      "learning_rate": 0.0008127938634832901,
+      "loss": 1.3926,
+      "step": 2850
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0724623698609669,
+      "learning_rate": 0.0008126580140227131,
+      "loss": 1.4211,
+      "step": 2851
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.07930876972688328,
+      "learning_rate": 0.0008125221266512739,
+      "loss": 1.4018,
+      "step": 2852
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.07774062520927012,
+      "learning_rate": 0.0008123862013854495,
+      "loss": 1.3741,
+      "step": 2853
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.07060780534531708,
+      "learning_rate": 0.0008122502382417211,
+      "loss": 1.507,
+      "step": 2854
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.07428692679133919,
+      "learning_rate": 0.0008121142372365749,
+      "loss": 1.4531,
+      "step": 2855
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.07165081198741169,
+      "learning_rate": 0.0008119781983865013,
+      "loss": 1.3349,
+      "step": 2856
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.07103201332670216,
+      "learning_rate": 0.0008118421217079958,
+      "loss": 1.4967,
+      "step": 2857
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.06745289271534266,
+      "learning_rate": 0.0008117060072175578,
+      "loss": 1.3217,
+      "step": 2858
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.07565056480064823,
+      "learning_rate": 0.0008115698549316919,
+      "loss": 1.3472,
+      "step": 2859
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.08917104470742244,
+      "learning_rate": 0.0008114336648669068,
+      "loss": 1.4251,
+      "step": 2860
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.07999502935604388,
+      "learning_rate": 0.0008112974370397163,
+      "loss": 1.4453,
+      "step": 2861
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.07927721882772928,
+      "learning_rate": 0.0008111611714666382,
+      "loss": 1.4793,
+      "step": 2862
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.07378233339106233,
+      "learning_rate": 0.0008110248681641956,
+      "loss": 1.3303,
+      "step": 2863
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0716702577172704,
+      "learning_rate": 0.0008108885271489153,
+      "loss": 1.39,
+      "step": 2864
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.07223233377589355,
+      "learning_rate": 0.0008107521484373292,
+      "loss": 1.433,
+      "step": 2865
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.07618471336989437,
+      "learning_rate": 0.000810615732045974,
+      "loss": 1.4637,
+      "step": 2866
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.07505002942237572,
+      "learning_rate": 0.0008104792779913903,
+      "loss": 1.5149,
+      "step": 2867
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.09599552930508799,
+      "learning_rate": 0.0008103427862901238,
+      "loss": 1.453,
+      "step": 2868
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.08413359365818811,
+      "learning_rate": 0.0008102062569587244,
+      "loss": 1.299,
+      "step": 2869
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.08334424197676094,
+      "learning_rate": 0.0008100696900137469,
+      "loss": 1.5173,
+      "step": 2870
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0681229530596204,
+      "learning_rate": 0.0008099330854717508,
+      "loss": 1.4892,
+      "step": 2871
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0826144462608642,
+      "learning_rate": 0.0008097964433492993,
+      "loss": 1.5718,
+      "step": 2872
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.07647809806526784,
+      "learning_rate": 0.0008096597636629612,
+      "loss": 1.3691,
+      "step": 2873
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.07280620512949906,
+      "learning_rate": 0.0008095230464293091,
+      "loss": 1.416,
+      "step": 2874
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0881966046359378,
+      "learning_rate": 0.0008093862916649207,
+      "loss": 1.4359,
+      "step": 2875
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.07054285637471477,
+      "learning_rate": 0.0008092494993863775,
+      "loss": 1.4334,
+      "step": 2876
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.07684187702267566,
+      "learning_rate": 0.0008091126696102665,
+      "loss": 1.3943,
+      "step": 2877
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.06921697707144545,
+      "learning_rate": 0.0008089758023531788,
+      "loss": 1.2889,
+      "step": 2878
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0741571622600297,
+      "learning_rate": 0.0008088388976317096,
+      "loss": 1.4221,
+      "step": 2879
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0770392792484649,
+      "learning_rate": 0.0008087019554624595,
+      "loss": 1.3828,
+      "step": 2880
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.07399574456270709,
+      "learning_rate": 0.000808564975862033,
+      "loss": 1.5514,
+      "step": 2881
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.08204482880103793,
+      "learning_rate": 0.0008084279588470393,
+      "loss": 1.4972,
+      "step": 2882
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.07952716032297752,
+      "learning_rate": 0.0008082909044340924,
+      "loss": 1.4671,
+      "step": 2883
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.08374489001130474,
+      "learning_rate": 0.0008081538126398105,
+      "loss": 1.4455,
+      "step": 2884
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.06973746802826421,
+      "learning_rate": 0.0008080166834808165,
+      "loss": 1.5941,
+      "step": 2885
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.07006537617171173,
+      "learning_rate": 0.0008078795169737376,
+      "loss": 1.4264,
+      "step": 2886
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.08481181694594125,
+      "learning_rate": 0.0008077423131352059,
+      "loss": 1.3789,
+      "step": 2887
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.09125579107905064,
+      "learning_rate": 0.0008076050719818577,
+      "loss": 1.4581,
+      "step": 2888
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.07959044783895391,
+      "learning_rate": 0.0008074677935303342,
+      "loss": 1.3779,
+      "step": 2889
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0794410279675721,
+      "learning_rate": 0.0008073304777972807,
+      "loss": 1.4972,
+      "step": 2890
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.07737642113539102,
+      "learning_rate": 0.0008071931247993472,
+      "loss": 1.378,
+      "step": 2891
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.07837475551562528,
+      "learning_rate": 0.0008070557345531881,
+      "loss": 1.4374,
+      "step": 2892
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0753630649514313,
+      "learning_rate": 0.0008069183070754628,
+      "loss": 1.4086,
+      "step": 2893
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0896110774192796,
+      "learning_rate": 0.0008067808423828347,
+      "loss": 1.4894,
+      "step": 2894
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0825189232098451,
+      "learning_rate": 0.0008066433404919718,
+      "loss": 1.5712,
+      "step": 2895
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.08237890387777524,
+      "learning_rate": 0.0008065058014195465,
+      "loss": 1.3692,
+      "step": 2896
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0742473531492679,
+      "learning_rate": 0.0008063682251822363,
+      "loss": 1.3707,
+      "step": 2897
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.07624379455220504,
+      "learning_rate": 0.0008062306117967225,
+      "loss": 1.3805,
+      "step": 2898
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0806083050065474,
+      "learning_rate": 0.0008060929612796914,
+      "loss": 1.3719,
+      "step": 2899
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.07716589591308537,
+      "learning_rate": 0.0008059552736478333,
+      "loss": 1.3963,
+      "step": 2900
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.09220293711869569,
+      "learning_rate": 0.0008058175489178436,
+      "loss": 1.4691,
+      "step": 2901
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.07788974238950698,
+      "learning_rate": 0.0008056797871064216,
+      "loss": 1.3894,
+      "step": 2902
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.10235235836865986,
+      "learning_rate": 0.0008055419882302719,
+      "loss": 1.5451,
+      "step": 2903
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.07739987925558103,
+      "learning_rate": 0.0008054041523061026,
+      "loss": 1.3031,
+      "step": 2904
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.09149966036697384,
+      "learning_rate": 0.000805266279350627,
+      "loss": 1.4051,
+      "step": 2905
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0873485165013668,
+      "learning_rate": 0.0008051283693805625,
+      "loss": 1.3646,
+      "step": 2906
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.08345322596031957,
+      "learning_rate": 0.0008049904224126312,
+      "loss": 1.4502,
+      "step": 2907
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.08088555125381701,
+      "learning_rate": 0.0008048524384635598,
+      "loss": 1.3885,
+      "step": 2908
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.08576366605222421,
+      "learning_rate": 0.0008047144175500794,
+      "loss": 1.452,
+      "step": 2909
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.07420507574850772,
+      "learning_rate": 0.0008045763596889253,
+      "loss": 1.4566,
+      "step": 2910
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0704001420236665,
+      "learning_rate": 0.0008044382648968374,
+      "loss": 1.3832,
+      "step": 2911
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.07461222698093337,
+      "learning_rate": 0.0008043001331905604,
+      "loss": 1.3567,
+      "step": 2912
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.08177185770423774,
+      "learning_rate": 0.0008041619645868433,
+      "loss": 1.4646,
+      "step": 2913
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.07573707444529373,
+      "learning_rate": 0.0008040237591024393,
+      "loss": 1.457,
+      "step": 2914
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.10223308961472137,
+      "learning_rate": 0.0008038855167541064,
+      "loss": 1.4352,
+      "step": 2915
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.08263870527187396,
+      "learning_rate": 0.0008037472375586067,
+      "loss": 1.2946,
+      "step": 2916
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.08385596883014022,
+      "learning_rate": 0.0008036089215327076,
+      "loss": 1.3963,
+      "step": 2917
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0765236562605998,
+      "learning_rate": 0.00080347056869318,
+      "loss": 1.4262,
+      "step": 2918
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.08764321552227011,
+      "learning_rate": 0.0008033321790567996,
+      "loss": 1.5388,
+      "step": 2919
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.07062198820275128,
+      "learning_rate": 0.0008031937526403469,
+      "loss": 1.3124,
+      "step": 2920
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.07193907290002885,
+      "learning_rate": 0.0008030552894606063,
+      "loss": 1.4052,
+      "step": 2921
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.08627791606011874,
+      "learning_rate": 0.0008029167895343671,
+      "loss": 1.494,
+      "step": 2922
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.07689630376058164,
+      "learning_rate": 0.0008027782528784228,
+      "loss": 1.5321,
+      "step": 2923
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.07838402069101297,
+      "learning_rate": 0.0008026396795095716,
+      "loss": 1.4008,
+      "step": 2924
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.07890700473282818,
+      "learning_rate": 0.000802501069444616,
+      "loss": 1.4139,
+      "step": 2925
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.07602723921168046,
+      "learning_rate": 0.0008023624227003626,
+      "loss": 1.5312,
+      "step": 2926
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.09021190956086227,
+      "learning_rate": 0.0008022237392936231,
+      "loss": 1.4854,
+      "step": 2927
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0772411567904348,
+      "learning_rate": 0.0008020850192412135,
+      "loss": 1.3465,
+      "step": 2928
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.07531806908392204,
+      "learning_rate": 0.0008019462625599536,
+      "loss": 1.493,
+      "step": 2929
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.07284679179896028,
+      "learning_rate": 0.0008018074692666686,
+      "loss": 1.4981,
+      "step": 2930
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.06456369250701985,
+      "learning_rate": 0.0008016686393781874,
+      "loss": 1.487,
+      "step": 2931
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.06991722508371778,
+      "learning_rate": 0.0008015297729113436,
+      "loss": 1.4999,
+      "step": 2932
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.08148887404697132,
+      "learning_rate": 0.0008013908698829752,
+      "loss": 1.3856,
+      "step": 2933
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.08756760737824867,
+      "learning_rate": 0.000801251930309925,
+      "loss": 1.5644,
+      "step": 2934
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.06420419497450133,
+      "learning_rate": 0.0008011129542090397,
+      "loss": 1.3941,
+      "step": 2935
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0796356391212337,
+      "learning_rate": 0.0008009739415971704,
+      "loss": 1.4468,
+      "step": 2936
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.07887732694508226,
+      "learning_rate": 0.0008008348924911732,
+      "loss": 1.5729,
+      "step": 2937
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.07020543228364666,
+      "learning_rate": 0.0008006958069079081,
+      "loss": 1.3993,
+      "step": 2938
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0671047003370231,
+      "learning_rate": 0.0008005566848642399,
+      "loss": 1.3402,
+      "step": 2939
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.07054047003240703,
+      "learning_rate": 0.0008004175263770372,
+      "loss": 1.4655,
+      "step": 2940
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.07740737220087766,
+      "learning_rate": 0.0008002783314631738,
+      "loss": 1.4764,
+      "step": 2941
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0728072252685772,
+      "learning_rate": 0.0008001391001395277,
+      "loss": 1.357,
+      "step": 2942
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.07554078318188316,
+      "learning_rate": 0.0007999998324229809,
+      "loss": 1.4548,
+      "step": 2943
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.08128226314789191,
+      "learning_rate": 0.0007998605283304201,
+      "loss": 1.5187,
+      "step": 2944
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.07245701659717234,
+      "learning_rate": 0.0007997211878787365,
+      "loss": 1.4871,
+      "step": 2945
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.08310930694834401,
+      "learning_rate": 0.0007995818110848256,
+      "loss": 1.437,
+      "step": 2946
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.08043969912327549,
+      "learning_rate": 0.0007994423979655872,
+      "loss": 1.4586,
+      "step": 2947
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.07315357331256875,
+      "learning_rate": 0.0007993029485379257,
+      "loss": 1.3868,
+      "step": 2948
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.08625764575142641,
+      "learning_rate": 0.0007991634628187499,
+      "loss": 1.5691,
+      "step": 2949
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.06859030969791598,
+      "learning_rate": 0.0007990239408249729,
+      "loss": 1.4585,
+      "step": 2950
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.08095069758093375,
+      "learning_rate": 0.0007988843825735121,
+      "loss": 1.4382,
+      "step": 2951
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.08455585505928823,
+      "learning_rate": 0.0007987447880812895,
+      "loss": 1.4736,
+      "step": 2952
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.07840102281047322,
+      "learning_rate": 0.0007986051573652315,
+      "loss": 1.3839,
+      "step": 2953
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0781115296791198,
+      "learning_rate": 0.0007984654904422685,
+      "loss": 1.5133,
+      "step": 2954
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0743966495861294,
+      "learning_rate": 0.0007983257873293362,
+      "loss": 1.2965,
+      "step": 2955
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0751376147423777,
+      "learning_rate": 0.0007981860480433733,
+      "loss": 1.4109,
+      "step": 2956
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.08181895305123754,
+      "learning_rate": 0.0007980462726013245,
+      "loss": 1.4691,
+      "step": 2957
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.08562272032257762,
+      "learning_rate": 0.0007979064610201372,
+      "loss": 1.3148,
+      "step": 2958
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.07781075518252807,
+      "learning_rate": 0.0007977666133167647,
+      "loss": 1.3035,
+      "step": 2959
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.09447045090218623,
+      "learning_rate": 0.0007976267295081636,
+      "loss": 1.3363,
+      "step": 2960
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.08987328702071136,
+      "learning_rate": 0.0007974868096112957,
+      "loss": 1.3668,
+      "step": 2961
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.07679392751255011,
+      "learning_rate": 0.0007973468536431266,
+      "loss": 1.2937,
+      "step": 2962
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.07944557506892269,
+      "learning_rate": 0.000797206861620626,
+      "loss": 1.376,
+      "step": 2963
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0879601918515955,
+      "learning_rate": 0.0007970668335607692,
+      "loss": 1.4672,
+      "step": 2964
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0875774791959148,
+      "learning_rate": 0.0007969267694805344,
+      "loss": 1.7008,
+      "step": 2965
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.07462856123424448,
+      "learning_rate": 0.0007967866693969053,
+      "loss": 1.4685,
+      "step": 2966
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.07553293553208955,
+      "learning_rate": 0.0007966465333268692,
+      "loss": 1.4977,
+      "step": 2967
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.07498275141260913,
+      "learning_rate": 0.0007965063612874184,
+      "loss": 1.4594,
+      "step": 2968
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.08857786850312585,
+      "learning_rate": 0.0007963661532955491,
+      "loss": 1.3205,
+      "step": 2969
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0781016657672057,
+      "learning_rate": 0.0007962259093682618,
+      "loss": 1.5386,
+      "step": 2970
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.06611466290116325,
+      "learning_rate": 0.0007960856295225618,
+      "loss": 1.412,
+      "step": 2971
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.05980479298650208,
+      "learning_rate": 0.0007959453137754586,
+      "loss": 1.2784,
+      "step": 2972
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.08170204436084891,
+      "learning_rate": 0.0007958049621439658,
+      "loss": 1.3268,
+      "step": 2973
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.07901322029342102,
+      "learning_rate": 0.0007956645746451014,
+      "loss": 1.4302,
+      "step": 2974
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.07307847800484599,
+      "learning_rate": 0.0007955241512958881,
+      "loss": 1.3754,
+      "step": 2975
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.06798109224771459,
+      "learning_rate": 0.0007953836921133526,
+      "loss": 1.4458,
+      "step": 2976
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.07147885396357032,
+      "learning_rate": 0.000795243197114526,
+      "loss": 1.4256,
+      "step": 2977
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.08764899736377428,
+      "learning_rate": 0.0007951026663164441,
+      "loss": 1.4224,
+      "step": 2978
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0789920953072474,
+      "learning_rate": 0.0007949620997361464,
+      "loss": 1.527,
+      "step": 2979
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.060860731754923306,
+      "learning_rate": 0.0007948214973906773,
+      "loss": 1.3997,
+      "step": 2980
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0708414939041876,
+      "learning_rate": 0.0007946808592970851,
+      "loss": 1.3102,
+      "step": 2981
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.07718632769563154,
+      "learning_rate": 0.0007945401854724231,
+      "loss": 1.2831,
+      "step": 2982
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.08056455667858022,
+      "learning_rate": 0.0007943994759337478,
+      "loss": 1.3903,
+      "step": 2983
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.07101624686847008,
+      "learning_rate": 0.0007942587306981213,
+      "loss": 1.4499,
+      "step": 2984
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.06746511771239394,
+      "learning_rate": 0.0007941179497826092,
+      "loss": 1.5146,
+      "step": 2985
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.07859460093405395,
+      "learning_rate": 0.0007939771332042817,
+      "loss": 1.5021,
+      "step": 2986
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.08026433583333177,
+      "learning_rate": 0.0007938362809802133,
+      "loss": 1.5269,
+      "step": 2987
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0804769915693462,
+      "learning_rate": 0.0007936953931274827,
+      "loss": 1.4494,
+      "step": 2988
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.07631714582546169,
+      "learning_rate": 0.0007935544696631734,
+      "loss": 1.4709,
+      "step": 2989
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.07816795866767308,
+      "learning_rate": 0.0007934135106043725,
+      "loss": 1.4538,
+      "step": 2990
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.07403291790022538,
+      "learning_rate": 0.000793272515968172,
+      "loss": 1.4281,
+      "step": 2991
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.07624985603510445,
+      "learning_rate": 0.0007931314857716676,
+      "loss": 1.5291,
+      "step": 2992
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0742004553934557,
+      "learning_rate": 0.0007929904200319602,
+      "loss": 1.4579,
+      "step": 2993
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.08297138817412927,
+      "learning_rate": 0.0007928493187661543,
+      "loss": 1.478,
+      "step": 2994
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.07212656405189363,
+      "learning_rate": 0.0007927081819913589,
+      "loss": 1.3083,
+      "step": 2995
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.07193865619974371,
+      "learning_rate": 0.0007925670097246871,
+      "loss": 1.4224,
+      "step": 2996
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.08509412142880574,
+      "learning_rate": 0.0007924258019832569,
+      "loss": 1.4168,
+      "step": 2997
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.07326614306013109,
+      "learning_rate": 0.00079228455878419,
+      "loss": 1.4902,
+      "step": 2998
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.08526895309448661,
+      "learning_rate": 0.0007921432801446127,
+      "loss": 1.3673,
+      "step": 2999
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.06863352180449873,
+      "learning_rate": 0.0007920019660816555,
+      "loss": 1.347,
+      "step": 3000
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.07099094175406699,
+      "learning_rate": 0.0007918606166124534,
+      "loss": 1.4413,
+      "step": 3001
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.07331999460617844,
+      "learning_rate": 0.000791719231754145,
+      "loss": 1.3459,
+      "step": 3002
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.06742208138232067,
+      "learning_rate": 0.0007915778115238743,
+      "loss": 1.3944,
+      "step": 3003
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.07564835035452432,
+      "learning_rate": 0.0007914363559387887,
+      "loss": 1.4459,
+      "step": 3004
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0775774753702783,
+      "learning_rate": 0.0007912948650160404,
+      "loss": 1.6455,
+      "step": 3005
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.06744949006518554,
+      "learning_rate": 0.0007911533387727852,
+      "loss": 1.377,
+      "step": 3006
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.07458417426910349,
+      "learning_rate": 0.0007910117772261839,
+      "loss": 1.3564,
+      "step": 3007
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.06923492807546411,
+      "learning_rate": 0.0007908701803934013,
+      "loss": 1.4969,
+      "step": 3008
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0741115394912985,
+      "learning_rate": 0.0007907285482916067,
+      "loss": 1.5018,
+      "step": 3009
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.07936484472761873,
+      "learning_rate": 0.0007905868809379734,
+      "loss": 1.5695,
+      "step": 3010
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.07847917265702553,
+      "learning_rate": 0.0007904451783496789,
+      "loss": 1.3488,
+      "step": 3011
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.07361634529449362,
+      "learning_rate": 0.0007903034405439051,
+      "loss": 1.3476,
+      "step": 3012
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.06785097870413606,
+      "learning_rate": 0.0007901616675378383,
+      "loss": 1.4224,
+      "step": 3013
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.08175453562213789,
+      "learning_rate": 0.0007900198593486691,
+      "loss": 1.4744,
+      "step": 3014
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.07547507633226375,
+      "learning_rate": 0.0007898780159935921,
+      "loss": 1.4698,
+      "step": 3015
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0743726151849371,
+      "learning_rate": 0.0007897361374898062,
+      "loss": 1.4479,
+      "step": 3016
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.08480416967435063,
+      "learning_rate": 0.0007895942238545148,
+      "loss": 1.385,
+      "step": 3017
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.07394595425252229,
+      "learning_rate": 0.0007894522751049253,
+      "loss": 1.4534,
+      "step": 3018
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0714067033877472,
+      "learning_rate": 0.0007893102912582497,
+      "loss": 1.4485,
+      "step": 3019
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.07582505910196573,
+      "learning_rate": 0.000789168272331704,
+      "loss": 1.4941,
+      "step": 3020
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0731420165751429,
+      "learning_rate": 0.0007890262183425081,
+      "loss": 1.4112,
+      "step": 3021
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.09305550460370213,
+      "learning_rate": 0.0007888841293078868,
+      "loss": 1.4552,
+      "step": 3022
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.06702964079143485,
+      "learning_rate": 0.0007887420052450689,
+      "loss": 1.4836,
+      "step": 3023
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07074665516379978,
+      "learning_rate": 0.0007885998461712875,
+      "loss": 1.4678,
+      "step": 3024
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07983639768866699,
+      "learning_rate": 0.0007884576521037798,
+      "loss": 1.4624,
+      "step": 3025
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07033851096059493,
+      "learning_rate": 0.0007883154230597873,
+      "loss": 1.3853,
+      "step": 3026
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.08476954997112443,
+      "learning_rate": 0.0007881731590565559,
+      "loss": 1.4094,
+      "step": 3027
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07706177346613348,
+      "learning_rate": 0.0007880308601113352,
+      "loss": 1.4241,
+      "step": 3028
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.08278153657107577,
+      "learning_rate": 0.0007878885262413799,
+      "loss": 1.3777,
+      "step": 3029
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.08990975215370425,
+      "learning_rate": 0.0007877461574639484,
+      "loss": 1.5666,
+      "step": 3030
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.08601960681426747,
+      "learning_rate": 0.0007876037537963033,
+      "loss": 1.3981,
+      "step": 3031
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07788237507460331,
+      "learning_rate": 0.0007874613152557113,
+      "loss": 1.3502,
+      "step": 3032
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.104418842475275,
+      "learning_rate": 0.0007873188418594438,
+      "loss": 1.5997,
+      "step": 3033
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.08013070374938921,
+      "learning_rate": 0.0007871763336247764,
+      "loss": 1.481,
+      "step": 3034
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07381930904128295,
+      "learning_rate": 0.0007870337905689881,
+      "loss": 1.3511,
+      "step": 3035
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.08219086071312165,
+      "learning_rate": 0.0007868912127093638,
+      "loss": 1.3737,
+      "step": 3036
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.08179829364428115,
+      "learning_rate": 0.0007867486000631903,
+      "loss": 1.595,
+      "step": 3037
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.09105328320067167,
+      "learning_rate": 0.0007866059526477607,
+      "loss": 1.5167,
+      "step": 3038
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07385748224030242,
+      "learning_rate": 0.0007864632704803711,
+      "loss": 1.3882,
+      "step": 3039
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.08146182608657129,
+      "learning_rate": 0.0007863205535783227,
+      "loss": 1.4446,
+      "step": 3040
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.08225645645470177,
+      "learning_rate": 0.0007861778019589198,
+      "loss": 1.3429,
+      "step": 3041
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07002468941388816,
+      "learning_rate": 0.000786035015639472,
+      "loss": 1.4645,
+      "step": 3042
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.08573645543248615,
+      "learning_rate": 0.0007858921946372922,
+      "loss": 1.352,
+      "step": 3043
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07785179379080004,
+      "learning_rate": 0.0007857493389696983,
+      "loss": 1.4217,
+      "step": 3044
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.06999357815245667,
+      "learning_rate": 0.0007856064486540119,
+      "loss": 1.5484,
+      "step": 3045
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07645852761959619,
+      "learning_rate": 0.0007854635237075591,
+      "loss": 1.382,
+      "step": 3046
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07006480211273053,
+      "learning_rate": 0.0007853205641476698,
+      "loss": 1.409,
+      "step": 3047
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07233610453690911,
+      "learning_rate": 0.0007851775699916786,
+      "loss": 1.5152,
+      "step": 3048
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.06938021129476968,
+      "learning_rate": 0.0007850345412569236,
+      "loss": 1.4692,
+      "step": 3049
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07696153387851845,
+      "learning_rate": 0.0007848914779607479,
+      "loss": 1.3746,
+      "step": 3050
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07222688770788331,
+      "learning_rate": 0.0007847483801204984,
+      "loss": 1.3984,
+      "step": 3051
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07217459083677298,
+      "learning_rate": 0.0007846052477535262,
+      "loss": 1.4673,
+      "step": 3052
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07665892010803753,
+      "learning_rate": 0.0007844620808771864,
+      "loss": 1.2931,
+      "step": 3053
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0753925094097213,
+      "learning_rate": 0.0007843188795088386,
+      "loss": 1.5428,
+      "step": 3054
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.10489383050028697,
+      "learning_rate": 0.0007841756436658464,
+      "loss": 1.3772,
+      "step": 3055
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07367833059294322,
+      "learning_rate": 0.0007840323733655779,
+      "loss": 1.4949,
+      "step": 3056
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0720997223768572,
+      "learning_rate": 0.0007838890686254048,
+      "loss": 1.3626,
+      "step": 3057
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07082156336721733,
+      "learning_rate": 0.0007837457294627035,
+      "loss": 1.4854,
+      "step": 3058
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.06935942782492194,
+      "learning_rate": 0.0007836023558948542,
+      "loss": 1.4135,
+      "step": 3059
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.08042147923845204,
+      "learning_rate": 0.0007834589479392415,
+      "loss": 1.2829,
+      "step": 3060
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07681957831308812,
+      "learning_rate": 0.0007833155056132543,
+      "loss": 1.5114,
+      "step": 3061
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.06971247023737136,
+      "learning_rate": 0.0007831720289342851,
+      "loss": 1.2874,
+      "step": 3062
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07685493671422122,
+      "learning_rate": 0.0007830285179197313,
+      "loss": 1.5248,
+      "step": 3063
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.08086964730881499,
+      "learning_rate": 0.0007828849725869939,
+      "loss": 1.575,
+      "step": 3064
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07938799977856249,
+      "learning_rate": 0.0007827413929534783,
+      "loss": 1.4236,
+      "step": 3065
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07732912826079967,
+      "learning_rate": 0.000782597779036594,
+      "loss": 1.5576,
+      "step": 3066
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07555603999638347,
+      "learning_rate": 0.0007824541308537547,
+      "loss": 1.437,
+      "step": 3067
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0723582025951719,
+      "learning_rate": 0.0007823104484223784,
+      "loss": 1.3081,
+      "step": 3068
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07712330288020602,
+      "learning_rate": 0.0007821667317598871,
+      "loss": 1.4435,
+      "step": 3069
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07022497073587027,
+      "learning_rate": 0.0007820229808837065,
+      "loss": 1.4977,
+      "step": 3070
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.08805211648596689,
+      "learning_rate": 0.0007818791958112672,
+      "loss": 1.4112,
+      "step": 3071
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.09148966969586016,
+      "learning_rate": 0.0007817353765600038,
+      "loss": 1.4515,
+      "step": 3072
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.06831467563512097,
+      "learning_rate": 0.0007815915231473547,
+      "loss": 1.4484,
+      "step": 3073
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07058078689286881,
+      "learning_rate": 0.0007814476355907625,
+      "loss": 1.5195,
+      "step": 3074
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07712659647369673,
+      "learning_rate": 0.0007813037139076742,
+      "loss": 1.5531,
+      "step": 3075
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07372955770119158,
+      "learning_rate": 0.0007811597581155407,
+      "loss": 1.2755,
+      "step": 3076
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.06877276645781746,
+      "learning_rate": 0.0007810157682318174,
+      "loss": 1.4825,
+      "step": 3077
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07132173224006651,
+      "learning_rate": 0.0007808717442739634,
+      "loss": 1.4652,
+      "step": 3078
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07358607939108996,
+      "learning_rate": 0.000780727686259442,
+      "loss": 1.4582,
+      "step": 3079
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07282302440663897,
+      "learning_rate": 0.0007805835942057209,
+      "loss": 1.4773,
+      "step": 3080
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.08644424392136274,
+      "learning_rate": 0.0007804394681302716,
+      "loss": 1.3965,
+      "step": 3081
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07859450169273975,
+      "learning_rate": 0.0007802953080505699,
+      "loss": 1.5057,
+      "step": 3082
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07110005546604672,
+      "learning_rate": 0.000780151113984096,
+      "loss": 1.4402,
+      "step": 3083
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.06966222966144436,
+      "learning_rate": 0.0007800068859483335,
+      "loss": 1.4002,
+      "step": 3084
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07523096465564282,
+      "learning_rate": 0.0007798626239607709,
+      "loss": 1.4039,
+      "step": 3085
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.08023929078514579,
+      "learning_rate": 0.0007797183280389002,
+      "loss": 1.3024,
+      "step": 3086
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.06607105968539327,
+      "learning_rate": 0.0007795739982002179,
+      "loss": 1.4149,
+      "step": 3087
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07990908371456909,
+      "learning_rate": 0.0007794296344622245,
+      "loss": 1.5067,
+      "step": 3088
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07774648613968008,
+      "learning_rate": 0.0007792852368424246,
+      "loss": 1.3315,
+      "step": 3089
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.06540989487465651,
+      "learning_rate": 0.0007791408053583269,
+      "loss": 1.3436,
+      "step": 3090
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07598791466657227,
+      "learning_rate": 0.0007789963400274443,
+      "loss": 1.4636,
+      "step": 3091
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07035326107353247,
+      "learning_rate": 0.0007788518408672934,
+      "loss": 1.4288,
+      "step": 3092
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07360792080801745,
+      "learning_rate": 0.0007787073078953955,
+      "loss": 1.4191,
+      "step": 3093
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.08166605972997906,
+      "learning_rate": 0.0007785627411292757,
+      "loss": 1.4531,
+      "step": 3094
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.08293183860962706,
+      "learning_rate": 0.0007784181405864633,
+      "loss": 1.4735,
+      "step": 3095
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07091869532476382,
+      "learning_rate": 0.0007782735062844914,
+      "loss": 1.5312,
+      "step": 3096
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07628603322010782,
+      "learning_rate": 0.0007781288382408975,
+      "loss": 1.4945,
+      "step": 3097
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07645251008994298,
+      "learning_rate": 0.0007779841364732231,
+      "loss": 1.5767,
+      "step": 3098
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07069182150804638,
+      "learning_rate": 0.0007778394009990138,
+      "loss": 1.3911,
+      "step": 3099
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.06384757627433746,
+      "learning_rate": 0.0007776946318358193,
+      "loss": 1.4573,
+      "step": 3100
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.06905081379306961,
+      "learning_rate": 0.0007775498290011934,
+      "loss": 1.4385,
+      "step": 3101
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.06830814690017548,
+      "learning_rate": 0.0007774049925126938,
+      "loss": 1.3747,
+      "step": 3102
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07537641175106087,
+      "learning_rate": 0.0007772601223878825,
+      "loss": 1.4257,
+      "step": 3103
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.06944422593512861,
+      "learning_rate": 0.0007771152186443254,
+      "loss": 1.3538,
+      "step": 3104
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.06475440686857296,
+      "learning_rate": 0.0007769702812995929,
+      "loss": 1.4781,
+      "step": 3105
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.06935888065280987,
+      "learning_rate": 0.0007768253103712588,
+      "loss": 1.5099,
+      "step": 3106
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0958797906744289,
+      "learning_rate": 0.0007766803058769015,
+      "loss": 1.423,
+      "step": 3107
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07001471875357002,
+      "learning_rate": 0.0007765352678341031,
+      "loss": 1.3799,
+      "step": 3108
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.06829461879326418,
+      "learning_rate": 0.0007763901962604502,
+      "loss": 1.4027,
+      "step": 3109
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07110317044671406,
+      "learning_rate": 0.0007762450911735333,
+      "loss": 1.3645,
+      "step": 3110
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.06932317253596268,
+      "learning_rate": 0.0007760999525909465,
+      "loss": 1.3177,
+      "step": 3111
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07623931912800835,
+      "learning_rate": 0.0007759547805302887,
+      "loss": 1.5691,
+      "step": 3112
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07738409838265192,
+      "learning_rate": 0.0007758095750091624,
+      "loss": 1.376,
+      "step": 3113
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07384334853814246,
+      "learning_rate": 0.0007756643360451742,
+      "loss": 1.4941,
+      "step": 3114
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.07783517274774741,
+      "learning_rate": 0.0007755190636559349,
+      "loss": 1.5442,
+      "step": 3115
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.08431183660573531,
+      "learning_rate": 0.0007753737578590594,
+      "loss": 1.3991,
+      "step": 3116
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07425496823855175,
+      "learning_rate": 0.0007752284186721664,
+      "loss": 1.4384,
+      "step": 3117
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07768753925552281,
+      "learning_rate": 0.0007750830461128786,
+      "loss": 1.5313,
+      "step": 3118
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0742130465574765,
+      "learning_rate": 0.0007749376401988232,
+      "loss": 1.4509,
+      "step": 3119
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07106119474212085,
+      "learning_rate": 0.0007747922009476311,
+      "loss": 1.4476,
+      "step": 3120
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.08026943237402184,
+      "learning_rate": 0.0007746467283769373,
+      "loss": 1.4237,
+      "step": 3121
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07251319491821662,
+      "learning_rate": 0.0007745012225043808,
+      "loss": 1.4031,
+      "step": 3122
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0838468737335798,
+      "learning_rate": 0.0007743556833476046,
+      "loss": 1.3595,
+      "step": 3123
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07248789364297477,
+      "learning_rate": 0.0007742101109242561,
+      "loss": 1.3831,
+      "step": 3124
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.06538596756504367,
+      "learning_rate": 0.0007740645052519863,
+      "loss": 1.4328,
+      "step": 3125
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07321164447995468,
+      "learning_rate": 0.0007739188663484504,
+      "loss": 1.4778,
+      "step": 3126
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.06658563944433109,
+      "learning_rate": 0.0007737731942313075,
+      "loss": 1.4636,
+      "step": 3127
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07882492086563908,
+      "learning_rate": 0.000773627488918221,
+      "loss": 1.4752,
+      "step": 3128
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07703285975834412,
+      "learning_rate": 0.0007734817504268581,
+      "loss": 1.4747,
+      "step": 3129
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07459758239690902,
+      "learning_rate": 0.0007733359787748903,
+      "loss": 1.3904,
+      "step": 3130
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07884448125379544,
+      "learning_rate": 0.0007731901739799926,
+      "loss": 1.4475,
+      "step": 3131
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07877035915224383,
+      "learning_rate": 0.0007730443360598446,
+      "loss": 1.2817,
+      "step": 3132
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07954104659452825,
+      "learning_rate": 0.0007728984650321294,
+      "loss": 1.3287,
+      "step": 3133
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07618254195620103,
+      "learning_rate": 0.0007727525609145345,
+      "loss": 1.3799,
+      "step": 3134
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07133598285920512,
+      "learning_rate": 0.0007726066237247513,
+      "loss": 1.4582,
+      "step": 3135
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.08030773599573748,
+      "learning_rate": 0.000772460653480475,
+      "loss": 1.3284,
+      "step": 3136
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07601094209678182,
+      "learning_rate": 0.0007723146501994053,
+      "loss": 1.4454,
+      "step": 3137
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07237544124895776,
+      "learning_rate": 0.0007721686138992456,
+      "loss": 1.4829,
+      "step": 3138
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07307223788740541,
+      "learning_rate": 0.0007720225445977029,
+      "loss": 1.5171,
+      "step": 3139
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07334412315927522,
+      "learning_rate": 0.000771876442312489,
+      "loss": 1.3808,
+      "step": 3140
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07045811403441511,
+      "learning_rate": 0.0007717303070613192,
+      "loss": 1.5331,
+      "step": 3141
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.06922844829063676,
+      "learning_rate": 0.0007715841388619129,
+      "loss": 1.3979,
+      "step": 3142
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07247091145004228,
+      "learning_rate": 0.0007714379377319933,
+      "loss": 1.3762,
+      "step": 3143
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.08291167413964351,
+      "learning_rate": 0.0007712917036892881,
+      "loss": 1.5631,
+      "step": 3144
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07776849587307075,
+      "learning_rate": 0.0007711454367515284,
+      "loss": 1.4031,
+      "step": 3145
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0775914855157828,
+      "learning_rate": 0.0007709991369364498,
+      "loss": 1.2737,
+      "step": 3146
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07038223844706518,
+      "learning_rate": 0.0007708528042617915,
+      "loss": 1.3721,
+      "step": 3147
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.06373412075036435,
+      "learning_rate": 0.0007707064387452972,
+      "loss": 1.4981,
+      "step": 3148
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07680543115273483,
+      "learning_rate": 0.0007705600404047135,
+      "loss": 1.5106,
+      "step": 3149
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07667822761501376,
+      "learning_rate": 0.0007704136092577925,
+      "loss": 1.4488,
+      "step": 3150
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0650475144264375,
+      "learning_rate": 0.000770267145322289,
+      "loss": 1.5037,
+      "step": 3151
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07429507685088853,
+      "learning_rate": 0.0007701206486159622,
+      "loss": 1.2532,
+      "step": 3152
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.06955586145840957,
+      "learning_rate": 0.0007699741191565758,
+      "loss": 1.4294,
+      "step": 3153
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.08122438166751517,
+      "learning_rate": 0.0007698275569618965,
+      "loss": 1.4821,
+      "step": 3154
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.08451174509670904,
+      "learning_rate": 0.0007696809620496958,
+      "loss": 1.4029,
+      "step": 3155
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.08009083722558016,
+      "learning_rate": 0.0007695343344377485,
+      "loss": 1.4404,
+      "step": 3156
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07499676196986182,
+      "learning_rate": 0.0007693876741438341,
+      "loss": 1.383,
+      "step": 3157
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0725984705093447,
+      "learning_rate": 0.0007692409811857356,
+      "loss": 1.5022,
+      "step": 3158
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07422957294380073,
+      "learning_rate": 0.0007690942555812397,
+      "loss": 1.3875,
+      "step": 3159
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.06613298070529067,
+      "learning_rate": 0.0007689474973481377,
+      "loss": 1.4326,
+      "step": 3160
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.08919731981533549,
+      "learning_rate": 0.0007688007065042245,
+      "loss": 1.4545,
+      "step": 3161
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07398471196751458,
+      "learning_rate": 0.0007686538830672988,
+      "loss": 1.4074,
+      "step": 3162
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.08089708897669948,
+      "learning_rate": 0.0007685070270551638,
+      "loss": 1.452,
+      "step": 3163
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0789578097029579,
+      "learning_rate": 0.000768360138485626,
+      "loss": 1.4309,
+      "step": 3164
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07689564974172196,
+      "learning_rate": 0.000768213217376496,
+      "loss": 1.4699,
+      "step": 3165
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07618594956580693,
+      "learning_rate": 0.0007680662637455889,
+      "loss": 1.3925,
+      "step": 3166
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07046309543070939,
+      "learning_rate": 0.0007679192776107232,
+      "loss": 1.4711,
+      "step": 3167
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.06065773332639102,
+      "learning_rate": 0.0007677722589897214,
+      "loss": 1.4663,
+      "step": 3168
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.06448888328620603,
+      "learning_rate": 0.0007676252079004101,
+      "loss": 1.2777,
+      "step": 3169
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.06919677462771456,
+      "learning_rate": 0.0007674781243606197,
+      "loss": 1.4472,
+      "step": 3170
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07706927662533691,
+      "learning_rate": 0.0007673310083881844,
+      "loss": 1.4423,
+      "step": 3171
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07029417746284516,
+      "learning_rate": 0.0007671838600009429,
+      "loss": 1.5247,
+      "step": 3172
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07006293784244212,
+      "learning_rate": 0.0007670366792167371,
+      "loss": 1.3477,
+      "step": 3173
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0665223617685133,
+      "learning_rate": 0.0007668894660534135,
+      "loss": 1.3575,
+      "step": 3174
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07626888473910773,
+      "learning_rate": 0.0007667422205288219,
+      "loss": 1.4666,
+      "step": 3175
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0680382115354333,
+      "learning_rate": 0.0007665949426608164,
+      "loss": 1.4727,
+      "step": 3176
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0718190962160836,
+      "learning_rate": 0.0007664476324672552,
+      "loss": 1.5404,
+      "step": 3177
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0793876214439945,
+      "learning_rate": 0.0007663002899659999,
+      "loss": 1.477,
+      "step": 3178
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07195150887665572,
+      "learning_rate": 0.0007661529151749163,
+      "loss": 1.4443,
+      "step": 3179
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07746803954967403,
+      "learning_rate": 0.0007660055081118742,
+      "loss": 1.5019,
+      "step": 3180
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07756374186181363,
+      "learning_rate": 0.0007658580687947473,
+      "loss": 1.5107,
+      "step": 3181
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.084088607358134,
+      "learning_rate": 0.0007657105972414128,
+      "loss": 1.35,
+      "step": 3182
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07775278682089769,
+      "learning_rate": 0.0007655630934697524,
+      "loss": 1.4803,
+      "step": 3183
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07602799563639731,
+      "learning_rate": 0.0007654155574976515,
+      "loss": 1.5302,
+      "step": 3184
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.09250492590276259,
+      "learning_rate": 0.0007652679893429993,
+      "loss": 1.5617,
+      "step": 3185
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07556500064817434,
+      "learning_rate": 0.0007651203890236886,
+      "loss": 1.3106,
+      "step": 3186
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0823390739699376,
+      "learning_rate": 0.0007649727565576168,
+      "loss": 1.399,
+      "step": 3187
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.08353330942823384,
+      "learning_rate": 0.0007648250919626849,
+      "loss": 1.5233,
+      "step": 3188
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07649777003122347,
+      "learning_rate": 0.0007646773952567975,
+      "loss": 1.4492,
+      "step": 3189
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07280004721530052,
+      "learning_rate": 0.0007645296664578635,
+      "loss": 1.3165,
+      "step": 3190
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.09933336670129464,
+      "learning_rate": 0.0007643819055837955,
+      "loss": 1.3888,
+      "step": 3191
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07779644484036152,
+      "learning_rate": 0.0007642341126525099,
+      "loss": 1.3763,
+      "step": 3192
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0683306880772141,
+      "learning_rate": 0.0007640862876819271,
+      "loss": 1.3403,
+      "step": 3193
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07300035128576285,
+      "learning_rate": 0.0007639384306899716,
+      "loss": 1.4203,
+      "step": 3194
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07191331871436503,
+      "learning_rate": 0.0007637905416945716,
+      "loss": 1.3708,
+      "step": 3195
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.06748665275475779,
+      "learning_rate": 0.0007636426207136587,
+      "loss": 1.3958,
+      "step": 3196
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.06979899017349758,
+      "learning_rate": 0.0007634946677651693,
+      "loss": 1.5443,
+      "step": 3197
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0665352354043053,
+      "learning_rate": 0.000763346682867043,
+      "loss": 1.4987,
+      "step": 3198
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.08238063923115757,
+      "learning_rate": 0.0007631986660372234,
+      "loss": 1.4835,
+      "step": 3199
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.06784751022670603,
+      "learning_rate": 0.0007630506172936583,
+      "loss": 1.5603,
+      "step": 3200
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.06156458689570208,
+      "learning_rate": 0.000762902536654299,
+      "loss": 1.4053,
+      "step": 3201
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07874911875458783,
+      "learning_rate": 0.0007627544241371005,
+      "loss": 1.4591,
+      "step": 3202
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.06958839969062831,
+      "learning_rate": 0.0007626062797600225,
+      "loss": 1.43,
+      "step": 3203
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07551242792555159,
+      "learning_rate": 0.0007624581035410277,
+      "loss": 1.539,
+      "step": 3204
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0802738339351694,
+      "learning_rate": 0.000762309895498083,
+      "loss": 1.411,
+      "step": 3205
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07610686443993148,
+      "learning_rate": 0.0007621616556491591,
+      "loss": 1.331,
+      "step": 3206
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07973442227199408,
+      "learning_rate": 0.0007620133840122305,
+      "loss": 1.3919,
+      "step": 3207
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.06935731716024028,
+      "learning_rate": 0.0007618650806052761,
+      "loss": 1.468,
+      "step": 3208
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.07140926619478406,
+      "learning_rate": 0.0007617167454462777,
+      "loss": 1.3073,
+      "step": 3209
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07628062933111013,
+      "learning_rate": 0.0007615683785532217,
+      "loss": 1.4563,
+      "step": 3210
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.06641544863858916,
+      "learning_rate": 0.000761419979944098,
+      "loss": 1.488,
+      "step": 3211
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07750941731981588,
+      "learning_rate": 0.0007612715496369006,
+      "loss": 1.4531,
+      "step": 3212
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.06603884034822871,
+      "learning_rate": 0.0007611230876496269,
+      "loss": 1.3834,
+      "step": 3213
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07814376352410252,
+      "learning_rate": 0.0007609745940002785,
+      "loss": 1.2924,
+      "step": 3214
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.06860503690532974,
+      "learning_rate": 0.0007608260687068611,
+      "loss": 1.3301,
+      "step": 3215
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07411296420440955,
+      "learning_rate": 0.0007606775117873836,
+      "loss": 1.4608,
+      "step": 3216
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0887118333530251,
+      "learning_rate": 0.0007605289232598591,
+      "loss": 1.4031,
+      "step": 3217
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0749536165963251,
+      "learning_rate": 0.0007603803031423045,
+      "loss": 1.3985,
+      "step": 3218
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.08599613367738648,
+      "learning_rate": 0.0007602316514527404,
+      "loss": 1.3853,
+      "step": 3219
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.06959589644535209,
+      "learning_rate": 0.0007600829682091915,
+      "loss": 1.3899,
+      "step": 3220
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.08227183337119814,
+      "learning_rate": 0.0007599342534296861,
+      "loss": 1.573,
+      "step": 3221
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07457434958035399,
+      "learning_rate": 0.0007597855071322561,
+      "loss": 1.4239,
+      "step": 3222
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07725146361782721,
+      "learning_rate": 0.000759636729334938,
+      "loss": 1.4454,
+      "step": 3223
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07127251988346796,
+      "learning_rate": 0.0007594879200557711,
+      "loss": 1.2921,
+      "step": 3224
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.08661988788655628,
+      "learning_rate": 0.0007593390793127996,
+      "loss": 1.5255,
+      "step": 3225
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.08081542170362471,
+      "learning_rate": 0.0007591902071240705,
+      "loss": 1.5518,
+      "step": 3226
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07345692626207229,
+      "learning_rate": 0.0007590413035076354,
+      "loss": 1.3901,
+      "step": 3227
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.08021573419868515,
+      "learning_rate": 0.0007588923684815489,
+      "loss": 1.4691,
+      "step": 3228
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07183007813806935,
+      "learning_rate": 0.0007587434020638704,
+      "loss": 1.474,
+      "step": 3229
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.09673214943620395,
+      "learning_rate": 0.0007585944042726622,
+      "loss": 1.4883,
+      "step": 3230
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07943320951000708,
+      "learning_rate": 0.0007584453751259911,
+      "loss": 1.4071,
+      "step": 3231
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07247765308432169,
+      "learning_rate": 0.0007582963146419273,
+      "loss": 1.4031,
+      "step": 3232
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0768200140003289,
+      "learning_rate": 0.0007581472228385447,
+      "loss": 1.4768,
+      "step": 3233
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.086314346381171,
+      "learning_rate": 0.0007579980997339215,
+      "loss": 1.5756,
+      "step": 3234
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.08993807019188155,
+      "learning_rate": 0.0007578489453461392,
+      "loss": 1.4346,
+      "step": 3235
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07458162456361211,
+      "learning_rate": 0.0007576997596932833,
+      "loss": 1.4711,
+      "step": 3236
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0715283955335303,
+      "learning_rate": 0.0007575505427934433,
+      "loss": 1.4222,
+      "step": 3237
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07396927269800081,
+      "learning_rate": 0.000757401294664712,
+      "loss": 1.3881,
+      "step": 3238
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07757426845121163,
+      "learning_rate": 0.0007572520153251861,
+      "loss": 1.5357,
+      "step": 3239
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07139276773731634,
+      "learning_rate": 0.0007571027047929668,
+      "loss": 1.3491,
+      "step": 3240
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.09502708923492989,
+      "learning_rate": 0.0007569533630861578,
+      "loss": 1.5269,
+      "step": 3241
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07195223372318488,
+      "learning_rate": 0.0007568039902228679,
+      "loss": 1.4009,
+      "step": 3242
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07778683580295905,
+      "learning_rate": 0.000756654586221209,
+      "loss": 1.3772,
+      "step": 3243
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.06633510179131384,
+      "learning_rate": 0.0007565051510992964,
+      "loss": 1.5195,
+      "step": 3244
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07960888862929236,
+      "learning_rate": 0.00075635568487525,
+      "loss": 1.4633,
+      "step": 3245
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07572686281344403,
+      "learning_rate": 0.0007562061875671932,
+      "loss": 1.4543,
+      "step": 3246
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07078481527058693,
+      "learning_rate": 0.0007560566591932526,
+      "loss": 1.4283,
+      "step": 3247
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07721076934087472,
+      "learning_rate": 0.0007559070997715596,
+      "loss": 1.4219,
+      "step": 3248
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07709818422571885,
+      "learning_rate": 0.0007557575093202483,
+      "loss": 1.4719,
+      "step": 3249
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.08369733445571949,
+      "learning_rate": 0.0007556078878574573,
+      "loss": 1.3737,
+      "step": 3250
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07127796374107943,
+      "learning_rate": 0.0007554582354013287,
+      "loss": 1.4437,
+      "step": 3251
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.06956043970136461,
+      "learning_rate": 0.0007553085519700084,
+      "loss": 1.4763,
+      "step": 3252
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07552454140482719,
+      "learning_rate": 0.0007551588375816461,
+      "loss": 1.4528,
+      "step": 3253
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07703097370477303,
+      "learning_rate": 0.0007550090922543948,
+      "loss": 1.432,
+      "step": 3254
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.08408237716390983,
+      "learning_rate": 0.000754859316006412,
+      "loss": 1.5888,
+      "step": 3255
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07580679587042956,
+      "learning_rate": 0.0007547095088558585,
+      "loss": 1.4903,
+      "step": 3256
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07470619755946706,
+      "learning_rate": 0.0007545596708208991,
+      "loss": 1.2541,
+      "step": 3257
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07443437030109887,
+      "learning_rate": 0.0007544098019197018,
+      "loss": 1.3443,
+      "step": 3258
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07786511617842874,
+      "learning_rate": 0.000754259902170439,
+      "loss": 1.4346,
+      "step": 3259
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07049486481909144,
+      "learning_rate": 0.0007541099715912867,
+      "loss": 1.4424,
+      "step": 3260
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.08884401595716221,
+      "learning_rate": 0.0007539600102004241,
+      "loss": 1.4576,
+      "step": 3261
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.06963460525741602,
+      "learning_rate": 0.0007538100180160347,
+      "loss": 1.5296,
+      "step": 3262
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07647861256873244,
+      "learning_rate": 0.0007536599950563056,
+      "loss": 1.4882,
+      "step": 3263
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07582018739217335,
+      "learning_rate": 0.0007535099413394276,
+      "loss": 1.3487,
+      "step": 3264
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0677356672438712,
+      "learning_rate": 0.0007533598568835953,
+      "loss": 1.3253,
+      "step": 3265
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.06655141971882164,
+      "learning_rate": 0.0007532097417070069,
+      "loss": 1.3951,
+      "step": 3266
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0672083493114287,
+      "learning_rate": 0.0007530595958278643,
+      "loss": 1.4441,
+      "step": 3267
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.06966617093307825,
+      "learning_rate": 0.0007529094192643733,
+      "loss": 1.4569,
+      "step": 3268
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07642161102040695,
+      "learning_rate": 0.0007527592120347433,
+      "loss": 1.4863,
+      "step": 3269
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07535741721980958,
+      "learning_rate": 0.0007526089741571875,
+      "loss": 1.5901,
+      "step": 3270
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.08060712796678715,
+      "learning_rate": 0.0007524587056499227,
+      "loss": 1.4188,
+      "step": 3271
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07216132537348179,
+      "learning_rate": 0.0007523084065311694,
+      "loss": 1.4918,
+      "step": 3272
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.06697670994819911,
+      "learning_rate": 0.000752158076819152,
+      "loss": 1.3135,
+      "step": 3273
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.09212419434556492,
+      "learning_rate": 0.0007520077165320986,
+      "loss": 1.5355,
+      "step": 3274
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07964494763502676,
+      "learning_rate": 0.0007518573256882406,
+      "loss": 1.4749,
+      "step": 3275
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.06205358368642162,
+      "learning_rate": 0.0007517069043058136,
+      "loss": 1.4527,
+      "step": 3276
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.06997752615570801,
+      "learning_rate": 0.0007515564524030567,
+      "loss": 1.4434,
+      "step": 3277
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.06796929318797444,
+      "learning_rate": 0.0007514059699982127,
+      "loss": 1.4697,
+      "step": 3278
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.06739848065311337,
+      "learning_rate": 0.0007512554571095281,
+      "loss": 1.5002,
+      "step": 3279
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07068450192460699,
+      "learning_rate": 0.0007511049137552533,
+      "loss": 1.387,
+      "step": 3280
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.06525573630565533,
+      "learning_rate": 0.0007509543399536418,
+      "loss": 1.347,
+      "step": 3281
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0718387775523518,
+      "learning_rate": 0.0007508037357229515,
+      "loss": 1.5561,
+      "step": 3282
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.06705072625049899,
+      "learning_rate": 0.0007506531010814435,
+      "loss": 1.4128,
+      "step": 3283
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.06879282519736174,
+      "learning_rate": 0.0007505024360473829,
+      "loss": 1.481,
+      "step": 3284
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07087927610595694,
+      "learning_rate": 0.0007503517406390384,
+      "loss": 1.5064,
+      "step": 3285
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.06462073478069816,
+      "learning_rate": 0.000750201014874682,
+      "loss": 1.3303,
+      "step": 3286
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07515022238790801,
+      "learning_rate": 0.0007500502587725903,
+      "loss": 1.4377,
+      "step": 3287
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.06472988386041287,
+      "learning_rate": 0.0007498994723510427,
+      "loss": 1.4442,
+      "step": 3288
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07501542260623986,
+      "learning_rate": 0.0007497486556283222,
+      "loss": 1.4461,
+      "step": 3289
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07074110541629863,
+      "learning_rate": 0.0007495978086227164,
+      "loss": 1.3875,
+      "step": 3290
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.08067924371663614,
+      "learning_rate": 0.0007494469313525158,
+      "loss": 1.3703,
+      "step": 3291
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07953533766463396,
+      "learning_rate": 0.0007492960238360147,
+      "loss": 1.3868,
+      "step": 3292
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.08099029033843955,
+      "learning_rate": 0.0007491450860915114,
+      "loss": 1.5645,
+      "step": 3293
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.06974776206856093,
+      "learning_rate": 0.0007489941181373075,
+      "loss": 1.39,
+      "step": 3294
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0663906501580651,
+      "learning_rate": 0.0007488431199917081,
+      "loss": 1.3797,
+      "step": 3295
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07187536093113295,
+      "learning_rate": 0.0007486920916730228,
+      "loss": 1.4383,
+      "step": 3296
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07034086526155568,
+      "learning_rate": 0.0007485410331995639,
+      "loss": 1.4984,
+      "step": 3297
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.09952125204166053,
+      "learning_rate": 0.0007483899445896478,
+      "loss": 1.387,
+      "step": 3298
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07855678239956934,
+      "learning_rate": 0.0007482388258615946,
+      "loss": 1.5485,
+      "step": 3299
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.06719366075858256,
+      "learning_rate": 0.000748087677033728,
+      "loss": 1.3693,
+      "step": 3300
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07047823572228767,
+      "learning_rate": 0.0007479364981243753,
+      "loss": 1.5318,
+      "step": 3301
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.07007505720471473,
+      "learning_rate": 0.0007477852891518675,
+      "loss": 1.5427,
+      "step": 3302
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07306598130413317,
+      "learning_rate": 0.000747634050134539,
+      "loss": 1.4338,
+      "step": 3303
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.073346910314022,
+      "learning_rate": 0.0007474827810907283,
+      "loss": 1.4186,
+      "step": 3304
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.08051731318932802,
+      "learning_rate": 0.0007473314820387771,
+      "loss": 1.3899,
+      "step": 3305
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07628081552060612,
+      "learning_rate": 0.0007471801529970311,
+      "loss": 1.4434,
+      "step": 3306
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07692904755262349,
+      "learning_rate": 0.0007470287939838393,
+      "loss": 1.4179,
+      "step": 3307
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07629253080964134,
+      "learning_rate": 0.0007468774050175547,
+      "loss": 1.495,
+      "step": 3308
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07601463459835986,
+      "learning_rate": 0.0007467259861165335,
+      "loss": 1.3929,
+      "step": 3309
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07143316010707389,
+      "learning_rate": 0.0007465745372991359,
+      "loss": 1.4335,
+      "step": 3310
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.06589330547380333,
+      "learning_rate": 0.0007464230585837257,
+      "loss": 1.4047,
+      "step": 3311
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07249848843228891,
+      "learning_rate": 0.0007462715499886701,
+      "loss": 1.4494,
+      "step": 3312
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07137267226329203,
+      "learning_rate": 0.00074612001153234,
+      "loss": 1.2914,
+      "step": 3313
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0871238453782044,
+      "learning_rate": 0.00074596844323311,
+      "loss": 1.3637,
+      "step": 3314
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.06963991804376052,
+      "learning_rate": 0.0007458168451093582,
+      "loss": 1.4503,
+      "step": 3315
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.08264139026068265,
+      "learning_rate": 0.0007456652171794665,
+      "loss": 1.4632,
+      "step": 3316
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07188881114245928,
+      "learning_rate": 0.0007455135594618204,
+      "loss": 1.5282,
+      "step": 3317
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.06770134356378205,
+      "learning_rate": 0.0007453618719748086,
+      "loss": 1.423,
+      "step": 3318
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07265567228128596,
+      "learning_rate": 0.0007452101547368241,
+      "loss": 1.2598,
+      "step": 3319
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.08608144387463132,
+      "learning_rate": 0.0007450584077662628,
+      "loss": 1.4486,
+      "step": 3320
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.08328290542930794,
+      "learning_rate": 0.0007449066310815249,
+      "loss": 1.3688,
+      "step": 3321
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07610336074240623,
+      "learning_rate": 0.0007447548247010137,
+      "loss": 1.3053,
+      "step": 3322
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07294867434292124,
+      "learning_rate": 0.000744602988643136,
+      "loss": 1.4548,
+      "step": 3323
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07649839853752352,
+      "learning_rate": 0.0007444511229263026,
+      "loss": 1.3756,
+      "step": 3324
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07738633235915615,
+      "learning_rate": 0.0007442992275689281,
+      "loss": 1.4607,
+      "step": 3325
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0820756094066135,
+      "learning_rate": 0.0007441473025894298,
+      "loss": 1.3747,
+      "step": 3326
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07086103194112217,
+      "learning_rate": 0.0007439953480062294,
+      "loss": 1.3998,
+      "step": 3327
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07017241619996507,
+      "learning_rate": 0.000743843363837752,
+      "loss": 1.4109,
+      "step": 3328
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07430812668013119,
+      "learning_rate": 0.0007436913501024259,
+      "loss": 1.4276,
+      "step": 3329
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07826138304497253,
+      "learning_rate": 0.0007435393068186835,
+      "loss": 1.3875,
+      "step": 3330
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07893852810631888,
+      "learning_rate": 0.0007433872340049607,
+      "loss": 1.417,
+      "step": 3331
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0685915410461669,
+      "learning_rate": 0.0007432351316796964,
+      "loss": 1.4295,
+      "step": 3332
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07565845102634473,
+      "learning_rate": 0.0007430829998613342,
+      "loss": 1.4313,
+      "step": 3333
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07870178690082014,
+      "learning_rate": 0.0007429308385683199,
+      "loss": 1.3445,
+      "step": 3334
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.06969600272935284,
+      "learning_rate": 0.0007427786478191041,
+      "loss": 1.3896,
+      "step": 3335
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07161250210757818,
+      "learning_rate": 0.0007426264276321401,
+      "loss": 1.5339,
+      "step": 3336
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.06567066213357739,
+      "learning_rate": 0.0007424741780258855,
+      "loss": 1.4913,
+      "step": 3337
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07739750277858809,
+      "learning_rate": 0.0007423218990188008,
+      "loss": 1.5223,
+      "step": 3338
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07139533267516067,
+      "learning_rate": 0.0007421695906293504,
+      "loss": 1.2608,
+      "step": 3339
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07084136654668771,
+      "learning_rate": 0.0007420172528760022,
+      "loss": 1.4166,
+      "step": 3340
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07461768305043502,
+      "learning_rate": 0.0007418648857772279,
+      "loss": 1.5376,
+      "step": 3341
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07018413349605858,
+      "learning_rate": 0.0007417124893515022,
+      "loss": 1.2933,
+      "step": 3342
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07748422851796735,
+      "learning_rate": 0.0007415600636173039,
+      "loss": 1.3764,
+      "step": 3343
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.06651878463581043,
+      "learning_rate": 0.0007414076085931152,
+      "loss": 1.274,
+      "step": 3344
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0720113569348327,
+      "learning_rate": 0.0007412551242974215,
+      "loss": 1.4276,
+      "step": 3345
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.06866646064645342,
+      "learning_rate": 0.0007411026107487123,
+      "loss": 1.5271,
+      "step": 3346
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07366727179489832,
+      "learning_rate": 0.0007409500679654805,
+      "loss": 1.4906,
+      "step": 3347
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.08059863399610857,
+      "learning_rate": 0.0007407974959662222,
+      "loss": 1.3035,
+      "step": 3348
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.08704144005558019,
+      "learning_rate": 0.0007406448947694373,
+      "loss": 1.4854,
+      "step": 3349
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.08134336654874601,
+      "learning_rate": 0.0007404922643936294,
+      "loss": 1.3461,
+      "step": 3350
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07515621562261053,
+      "learning_rate": 0.0007403396048573051,
+      "loss": 1.4072,
+      "step": 3351
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07475152903248071,
+      "learning_rate": 0.0007401869161789753,
+      "loss": 1.3376,
+      "step": 3352
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07219916506306422,
+      "learning_rate": 0.0007400341983771539,
+      "loss": 1.5695,
+      "step": 3353
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07830596989301176,
+      "learning_rate": 0.0007398814514703585,
+      "loss": 1.5121,
+      "step": 3354
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07554099739341379,
+      "learning_rate": 0.0007397286754771098,
+      "loss": 1.3901,
+      "step": 3355
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0806406228367244,
+      "learning_rate": 0.000739575870415933,
+      "loss": 1.4044,
+      "step": 3356
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07450558110927982,
+      "learning_rate": 0.0007394230363053558,
+      "loss": 1.4013,
+      "step": 3357
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07182394432047186,
+      "learning_rate": 0.0007392701731639102,
+      "loss": 1.5534,
+      "step": 3358
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0755364633765582,
+      "learning_rate": 0.0007391172810101311,
+      "loss": 1.4123,
+      "step": 3359
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07636094394024896,
+      "learning_rate": 0.0007389643598625573,
+      "loss": 1.4783,
+      "step": 3360
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.06831486564246876,
+      "learning_rate": 0.0007388114097397311,
+      "loss": 1.5466,
+      "step": 3361
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.08977748147634984,
+      "learning_rate": 0.0007386584306601983,
+      "loss": 1.4046,
+      "step": 3362
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0672322704542431,
+      "learning_rate": 0.0007385054226425077,
+      "loss": 1.3721,
+      "step": 3363
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.06880509467787907,
+      "learning_rate": 0.0007383523857052124,
+      "loss": 1.3872,
+      "step": 3364
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07513708330952967,
+      "learning_rate": 0.0007381993198668688,
+      "loss": 1.5438,
+      "step": 3365
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07848633488508752,
+      "learning_rate": 0.0007380462251460364,
+      "loss": 1.5115,
+      "step": 3366
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07274279733626927,
+      "learning_rate": 0.0007378931015612786,
+      "loss": 1.5028,
+      "step": 3367
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.06856498960103916,
+      "learning_rate": 0.0007377399491311619,
+      "loss": 1.4003,
+      "step": 3368
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.06815056499835738,
+      "learning_rate": 0.0007375867678742567,
+      "loss": 1.582,
+      "step": 3369
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07130247554924865,
+      "learning_rate": 0.0007374335578091371,
+      "loss": 1.3157,
+      "step": 3370
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0699724644388025,
+      "learning_rate": 0.0007372803189543798,
+      "loss": 1.4183,
+      "step": 3371
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07859318932566481,
+      "learning_rate": 0.0007371270513285659,
+      "loss": 1.5631,
+      "step": 3372
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07308852607081995,
+      "learning_rate": 0.0007369737549502796,
+      "loss": 1.576,
+      "step": 3373
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.06767306064322558,
+      "learning_rate": 0.0007368204298381085,
+      "loss": 1.4173,
+      "step": 3374
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07610629235626565,
+      "learning_rate": 0.0007366670760106438,
+      "loss": 1.3515,
+      "step": 3375
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07255829044756128,
+      "learning_rate": 0.0007365136934864803,
+      "loss": 1.3286,
+      "step": 3376
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.06113231650355408,
+      "learning_rate": 0.0007363602822842159,
+      "loss": 1.4304,
+      "step": 3377
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07358452028485662,
+      "learning_rate": 0.0007362068424224526,
+      "loss": 1.3216,
+      "step": 3378
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07125399818046416,
+      "learning_rate": 0.0007360533739197952,
+      "loss": 1.4562,
+      "step": 3379
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.09144438881226293,
+      "learning_rate": 0.0007358998767948525,
+      "loss": 1.478,
+      "step": 3380
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.09000168011921686,
+      "learning_rate": 0.0007357463510662364,
+      "loss": 1.5342,
+      "step": 3381
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0706911701954772,
+      "learning_rate": 0.0007355927967525626,
+      "loss": 1.3949,
+      "step": 3382
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07477870009184143,
+      "learning_rate": 0.0007354392138724499,
+      "loss": 1.4481,
+      "step": 3383
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.09106274263140833,
+      "learning_rate": 0.0007352856024445208,
+      "loss": 1.3723,
+      "step": 3384
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07556958378953439,
+      "learning_rate": 0.0007351319624874012,
+      "loss": 1.5089,
+      "step": 3385
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07314230545614694,
+      "learning_rate": 0.0007349782940197207,
+      "loss": 1.3267,
+      "step": 3386
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07184969570633701,
+      "learning_rate": 0.000734824597060112,
+      "loss": 1.5058,
+      "step": 3387
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0729652621412798,
+      "learning_rate": 0.000734670871627211,
+      "loss": 1.5736,
+      "step": 3388
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0669490998865407,
+      "learning_rate": 0.000734517117739658,
+      "loss": 1.4002,
+      "step": 3389
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.08135789295051453,
+      "learning_rate": 0.000734363335416096,
+      "loss": 1.4905,
+      "step": 3390
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07506628391924669,
+      "learning_rate": 0.0007342095246751717,
+      "loss": 1.4123,
+      "step": 3391
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07631817394827138,
+      "learning_rate": 0.000734055685535535,
+      "loss": 1.4101,
+      "step": 3392
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.08451380581854086,
+      "learning_rate": 0.0007339018180158394,
+      "loss": 1.4673,
+      "step": 3393
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.08132207812962348,
+      "learning_rate": 0.0007337479221347419,
+      "loss": 1.5261,
+      "step": 3394
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.07423312430722277,
+      "learning_rate": 0.0007335939979109032,
+      "loss": 1.4681,
+      "step": 3395
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.08032224968211471,
+      "learning_rate": 0.0007334400453629869,
+      "loss": 1.518,
+      "step": 3396
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.07632194005433686,
+      "learning_rate": 0.0007332860645096604,
+      "loss": 1.4886,
+      "step": 3397
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0864928518705132,
+      "learning_rate": 0.0007331320553695942,
+      "loss": 1.516,
+      "step": 3398
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.07020971282123498,
+      "learning_rate": 0.0007329780179614624,
+      "loss": 1.3371,
+      "step": 3399
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.06779508227886841,
+      "learning_rate": 0.000732823952303943,
+      "loss": 1.4788,
+      "step": 3400
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.08093791904028005,
+      "learning_rate": 0.0007326698584157167,
+      "loss": 1.5612,
+      "step": 3401
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.07297898365770618,
+      "learning_rate": 0.0007325157363154678,
+      "loss": 1.4382,
+      "step": 3402
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.07359410616916069,
+      "learning_rate": 0.0007323615860218843,
+      "loss": 1.4834,
+      "step": 3403
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.07731757154460116,
+      "learning_rate": 0.0007322074075536574,
+      "loss": 1.5044,
+      "step": 3404
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0669375876227775,
+      "learning_rate": 0.0007320532009294818,
+      "loss": 1.5491,
+      "step": 3405
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.08319722191595673,
+      "learning_rate": 0.0007318989661680556,
+      "loss": 1.5788,
+      "step": 3406
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.07349318095347086,
+      "learning_rate": 0.0007317447032880804,
+      "loss": 1.4061,
+      "step": 3407
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0815880060028677,
+      "learning_rate": 0.0007315904123082608,
+      "loss": 1.4521,
+      "step": 3408
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.07833449022531756,
+      "learning_rate": 0.0007314360932473054,
+      "loss": 1.301,
+      "step": 3409
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.06153440472323062,
+      "learning_rate": 0.0007312817461239258,
+      "loss": 1.4043,
+      "step": 3410
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.061429343476582014,
+      "learning_rate": 0.000731127370956837,
+      "loss": 1.3902,
+      "step": 3411
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.07111252029973611,
+      "learning_rate": 0.0007309729677647579,
+      "loss": 1.3566,
+      "step": 3412
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.07442921283389783,
+      "learning_rate": 0.00073081853656641,
+      "loss": 1.4205,
+      "step": 3413
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.06875680871458785,
+      "learning_rate": 0.0007306640773805188,
+      "loss": 1.5055,
+      "step": 3414
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.07011065801353429,
+      "learning_rate": 0.000730509590225813,
+      "loss": 1.4903,
+      "step": 3415
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.07315229939972583,
+      "learning_rate": 0.0007303550751210247,
+      "loss": 1.3406,
+      "step": 3416
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.06862713960228749,
+      "learning_rate": 0.0007302005320848894,
+      "loss": 1.487,
+      "step": 3417
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.06991335457962292,
+      "learning_rate": 0.0007300459611361461,
+      "loss": 1.3993,
+      "step": 3418
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.07522110372865594,
+      "learning_rate": 0.0007298913622935366,
+      "loss": 1.5515,
+      "step": 3419
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.08368471518379769,
+      "learning_rate": 0.0007297367355758071,
+      "loss": 1.4278,
+      "step": 3420
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.07070017111651242,
+      "learning_rate": 0.0007295820810017064,
+      "loss": 1.3956,
+      "step": 3421
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.07220151835057213,
+      "learning_rate": 0.0007294273985899867,
+      "loss": 1.4705,
+      "step": 3422
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.06758161171286328,
+      "learning_rate": 0.0007292726883594042,
+      "loss": 1.5548,
+      "step": 3423
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.08252864849353624,
+      "learning_rate": 0.0007291179503287178,
+      "loss": 1.4895,
+      "step": 3424
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.07274048956748501,
+      "learning_rate": 0.0007289631845166897,
+      "loss": 1.4906,
+      "step": 3425
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0645889905677694,
+      "learning_rate": 0.0007288083909420865,
+      "loss": 1.506,
+      "step": 3426
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.07054697947372947,
+      "learning_rate": 0.0007286535696236769,
+      "loss": 1.3929,
+      "step": 3427
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.06771033703226476,
+      "learning_rate": 0.0007284987205802338,
+      "loss": 1.3862,
+      "step": 3428
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.07230953142493653,
+      "learning_rate": 0.000728343843830533,
+      "loss": 1.5308,
+      "step": 3429
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.07990829409950602,
+      "learning_rate": 0.0007281889393933539,
+      "loss": 1.3315,
+      "step": 3430
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0744965070918347,
+      "learning_rate": 0.0007280340072874791,
+      "loss": 1.3883,
+      "step": 3431
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.06853594795545971,
+      "learning_rate": 0.000727879047531695,
+      "loss": 1.4513,
+      "step": 3432
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.07005750508920855,
+      "learning_rate": 0.0007277240601447907,
+      "loss": 1.429,
+      "step": 3433
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0746780149046262,
+      "learning_rate": 0.000727569045145559,
+      "loss": 1.4085,
+      "step": 3434
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.07182195957106077,
+      "learning_rate": 0.000727414002552796,
+      "loss": 1.4943,
+      "step": 3435
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0640184881855754,
+      "learning_rate": 0.0007272589323853012,
+      "loss": 1.4937,
+      "step": 3436
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.07500444842348676,
+      "learning_rate": 0.0007271038346618774,
+      "loss": 1.3062,
+      "step": 3437
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.07010662383143268,
+      "learning_rate": 0.0007269487094013306,
+      "loss": 1.5224,
+      "step": 3438
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.06754630061604039,
+      "learning_rate": 0.0007267935566224704,
+      "loss": 1.5214,
+      "step": 3439
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.07547928011898576,
+      "learning_rate": 0.0007266383763441097,
+      "loss": 1.5368,
+      "step": 3440
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.06936226935791721,
+      "learning_rate": 0.0007264831685850645,
+      "loss": 1.3726,
+      "step": 3441
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.07120368855493818,
+      "learning_rate": 0.0007263279333641541,
+      "loss": 1.5166,
+      "step": 3442
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.06496628037540313,
+      "learning_rate": 0.0007261726707002016,
+      "loss": 1.4256,
+      "step": 3443
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.07751887752981045,
+      "learning_rate": 0.0007260173806120331,
+      "loss": 1.4362,
+      "step": 3444
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.07593916993944279,
+      "learning_rate": 0.0007258620631184781,
+      "loss": 1.4158,
+      "step": 3445
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.06962702489132955,
+      "learning_rate": 0.0007257067182383692,
+      "loss": 1.4768,
+      "step": 3446
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.07130112366350376,
+      "learning_rate": 0.0007255513459905425,
+      "loss": 1.4573,
+      "step": 3447
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.07317457257010111,
+      "learning_rate": 0.0007253959463938375,
+      "loss": 1.5383,
+      "step": 3448
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.07639509895879555,
+      "learning_rate": 0.0007252405194670972,
+      "loss": 1.4555,
+      "step": 3449
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.08762753103718868,
+      "learning_rate": 0.0007250850652291671,
+      "loss": 1.4631,
+      "step": 3450
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.07417213080044456,
+      "learning_rate": 0.000724929583698897,
+      "loss": 1.4104,
+      "step": 3451
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.08182638136439328,
+      "learning_rate": 0.0007247740748951393,
+      "loss": 1.4581,
+      "step": 3452
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.07652625611553487,
+      "learning_rate": 0.0007246185388367502,
+      "loss": 1.495,
+      "step": 3453
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0718289444464132,
+      "learning_rate": 0.0007244629755425889,
+      "loss": 1.3931,
+      "step": 3454
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0776318362638091,
+      "learning_rate": 0.0007243073850315179,
+      "loss": 1.4296,
+      "step": 3455
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.07319195455280723,
+      "learning_rate": 0.0007241517673224032,
+      "loss": 1.4762,
+      "step": 3456
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0704550126688549,
+      "learning_rate": 0.000723996122434114,
+      "loss": 1.5216,
+      "step": 3457
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0773671225672999,
+      "learning_rate": 0.0007238404503855227,
+      "loss": 1.4112,
+      "step": 3458
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.09560707561440114,
+      "learning_rate": 0.000723684751195505,
+      "loss": 1.4316,
+      "step": 3459
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.08176097879461783,
+      "learning_rate": 0.0007235290248829402,
+      "loss": 1.4419,
+      "step": 3460
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.07671659208485172,
+      "learning_rate": 0.0007233732714667104,
+      "loss": 1.5268,
+      "step": 3461
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.07096830395444363,
+      "learning_rate": 0.0007232174909657014,
+      "loss": 1.4263,
+      "step": 3462
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.07937368779759209,
+      "learning_rate": 0.0007230616833988021,
+      "loss": 1.4219,
+      "step": 3463
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.07043297450143687,
+      "learning_rate": 0.0007229058487849045,
+      "loss": 1.5423,
+      "step": 3464
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.08169254273194243,
+      "learning_rate": 0.0007227499871429046,
+      "loss": 1.3532,
+      "step": 3465
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.06726101998135063,
+      "learning_rate": 0.0007225940984917007,
+      "loss": 1.5103,
+      "step": 3466
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.06720793619103706,
+      "learning_rate": 0.0007224381828501947,
+      "loss": 1.5341,
+      "step": 3467
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.08637904126374926,
+      "learning_rate": 0.0007222822402372924,
+      "loss": 1.4348,
+      "step": 3468
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.08033335965656004,
+      "learning_rate": 0.000722126270671902,
+      "loss": 1.3226,
+      "step": 3469
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.07863747479165964,
+      "learning_rate": 0.0007219702741729357,
+      "loss": 1.4313,
+      "step": 3470
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.08115522916601269,
+      "learning_rate": 0.0007218142507593084,
+      "loss": 1.5195,
+      "step": 3471
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.08244015470619837,
+      "learning_rate": 0.0007216582004499384,
+      "loss": 1.4175,
+      "step": 3472
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.08949002833198955,
+      "learning_rate": 0.0007215021232637474,
+      "loss": 1.4375,
+      "step": 3473
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.07748904819217635,
+      "learning_rate": 0.0007213460192196602,
+      "loss": 1.3619,
+      "step": 3474
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.09245287952959458,
+      "learning_rate": 0.0007211898883366052,
+      "loss": 1.5071,
+      "step": 3475
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.07437346670229164,
+      "learning_rate": 0.0007210337306335137,
+      "loss": 1.4178,
+      "step": 3476
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.07848104764122867,
+      "learning_rate": 0.0007208775461293205,
+      "loss": 1.4574,
+      "step": 3477
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.08088266548645225,
+      "learning_rate": 0.0007207213348429629,
+      "loss": 1.4889,
+      "step": 3478
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.08175567193562198,
+      "learning_rate": 0.0007205650967933829,
+      "loss": 1.5308,
+      "step": 3479
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.08083881794107464,
+      "learning_rate": 0.0007204088319995245,
+      "loss": 1.45,
+      "step": 3480
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.06983298049987251,
+      "learning_rate": 0.0007202525404803352,
+      "loss": 1.4103,
+      "step": 3481
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.08082054975206557,
+      "learning_rate": 0.0007200962222547662,
+      "loss": 1.3104,
+      "step": 3482
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.07562790021280924,
+      "learning_rate": 0.0007199398773417713,
+      "loss": 1.4955,
+      "step": 3483
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.06228637321505799,
+      "learning_rate": 0.000719783505760308,
+      "loss": 1.5076,
+      "step": 3484
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.07524344062108244,
+      "learning_rate": 0.000719627107529337,
+      "loss": 1.4938,
+      "step": 3485
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.06815009438550329,
+      "learning_rate": 0.0007194706826678222,
+      "loss": 1.5052,
+      "step": 3486
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0670706317496325,
+      "learning_rate": 0.0007193142311947302,
+      "loss": 1.291,
+      "step": 3487
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.07554509041215911,
+      "learning_rate": 0.0007191577531290318,
+      "loss": 1.5673,
+      "step": 3488
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.06758859386555005,
+      "learning_rate": 0.0007190012484897002,
+      "loss": 1.4189,
+      "step": 3489
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0725558291970073,
+      "learning_rate": 0.0007188447172957121,
+      "loss": 1.526,
+      "step": 3490
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.07233185809808401,
+      "learning_rate": 0.0007186881595660478,
+      "loss": 1.4618,
+      "step": 3491
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.06533375243840589,
+      "learning_rate": 0.0007185315753196899,
+      "loss": 1.4157,
+      "step": 3492
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.07160807457158654,
+      "learning_rate": 0.0007183749645756253,
+      "loss": 1.5502,
+      "step": 3493
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0695540672338539,
+      "learning_rate": 0.0007182183273528436,
+      "loss": 1.3664,
+      "step": 3494
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.07960679599217649,
+      "learning_rate": 0.000718061663670337,
+      "loss": 1.4714,
+      "step": 3495
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.06495251596402488,
+      "learning_rate": 0.0007179049735471021,
+      "loss": 1.3737,
+      "step": 3496
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.06138238935894003,
+      "learning_rate": 0.0007177482570021379,
+      "loss": 1.4541,
+      "step": 3497
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.06824731668030849,
+      "learning_rate": 0.0007175915140544469,
+      "loss": 1.429,
+      "step": 3498
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.07191700712025918,
+      "learning_rate": 0.0007174347447230346,
+      "loss": 1.4168,
+      "step": 3499
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.06375599026585484,
+      "learning_rate": 0.0007172779490269099,
+      "loss": 1.422,
+      "step": 3500
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0727035296847183,
+      "learning_rate": 0.0007171211269850847,
+      "loss": 1.4303,
+      "step": 3501
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.07744658478094242,
+      "learning_rate": 0.0007169642786165746,
+      "loss": 1.5019,
+      "step": 3502
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.07138038364485365,
+      "learning_rate": 0.0007168074039403975,
+      "loss": 1.3967,
+      "step": 3503
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.07553044790967374,
+      "learning_rate": 0.0007166505029755752,
+      "loss": 1.3925,
+      "step": 3504
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.08417883389077783,
+      "learning_rate": 0.0007164935757411327,
+      "loss": 1.4638,
+      "step": 3505
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0700501690822233,
+      "learning_rate": 0.0007163366222560976,
+      "loss": 1.4004,
+      "step": 3506
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.07009352089361452,
+      "learning_rate": 0.0007161796425395013,
+      "loss": 1.4417,
+      "step": 3507
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.08437481489426124,
+      "learning_rate": 0.0007160226366103781,
+      "loss": 1.3978,
+      "step": 3508
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.07281127210582462,
+      "learning_rate": 0.0007158656044877654,
+      "loss": 1.4749,
+      "step": 3509
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.08738001079910843,
+      "learning_rate": 0.000715708546190704,
+      "loss": 1.3974,
+      "step": 3510
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.07474336566461999,
+      "learning_rate": 0.0007155514617382377,
+      "loss": 1.4416,
+      "step": 3511
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.07599929215831347,
+      "learning_rate": 0.0007153943511494134,
+      "loss": 1.5088,
+      "step": 3512
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.07441053895292685,
+      "learning_rate": 0.0007152372144432817,
+      "loss": 1.4735,
+      "step": 3513
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.06776065211117123,
+      "learning_rate": 0.0007150800516388956,
+      "loss": 1.4267,
+      "step": 3514
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.07725836370655137,
+      "learning_rate": 0.0007149228627553117,
+      "loss": 1.4947,
+      "step": 3515
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.08250426681685036,
+      "learning_rate": 0.0007147656478115898,
+      "loss": 1.4598,
+      "step": 3516
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.079833072983792,
+      "learning_rate": 0.0007146084068267928,
+      "loss": 1.3901,
+      "step": 3517
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.07133873587178743,
+      "learning_rate": 0.0007144511398199865,
+      "loss": 1.3933,
+      "step": 3518
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.06894999027593159,
+      "learning_rate": 0.00071429384681024,
+      "loss": 1.4166,
+      "step": 3519
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.06676758690851145,
+      "learning_rate": 0.0007141365278166261,
+      "loss": 1.3517,
+      "step": 3520
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.06575167661410714,
+      "learning_rate": 0.0007139791828582196,
+      "loss": 1.4837,
+      "step": 3521
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.08017489354969888,
+      "learning_rate": 0.0007138218119540998,
+      "loss": 1.4512,
+      "step": 3522
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.07228178571734656,
+      "learning_rate": 0.000713664415123348,
+      "loss": 1.3205,
+      "step": 3523
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1026294128547175,
+      "learning_rate": 0.0007135069923850493,
+      "loss": 1.4643,
+      "step": 3524
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.08011786852797782,
+      "learning_rate": 0.0007133495437582916,
+      "loss": 1.442,
+      "step": 3525
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.07244582841262252,
+      "learning_rate": 0.0007131920692621663,
+      "loss": 1.2873,
+      "step": 3526
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.06822431369016423,
+      "learning_rate": 0.0007130345689157676,
+      "loss": 1.4352,
+      "step": 3527
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.07031653395975418,
+      "learning_rate": 0.000712877042738193,
+      "loss": 1.3986,
+      "step": 3528
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.07397680940283695,
+      "learning_rate": 0.000712719490748543,
+      "loss": 1.361,
+      "step": 3529
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.07489945680906114,
+      "learning_rate": 0.0007125619129659214,
+      "loss": 1.3775,
+      "step": 3530
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.06688488591400973,
+      "learning_rate": 0.0007124043094094352,
+      "loss": 1.4639,
+      "step": 3531
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0706049908559832,
+      "learning_rate": 0.0007122466800981939,
+      "loss": 1.5427,
+      "step": 3532
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0711990066874569,
+      "learning_rate": 0.0007120890250513111,
+      "loss": 1.515,
+      "step": 3533
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.07309355595657928,
+      "learning_rate": 0.0007119313442879028,
+      "loss": 1.5657,
+      "step": 3534
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.07203044835573408,
+      "learning_rate": 0.0007117736378270885,
+      "loss": 1.3643,
+      "step": 3535
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.07921231575470476,
+      "learning_rate": 0.0007116159056879904,
+      "loss": 1.3966,
+      "step": 3536
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.06516803459785417,
+      "learning_rate": 0.0007114581478897342,
+      "loss": 1.3938,
+      "step": 3537
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.07084548870515166,
+      "learning_rate": 0.0007113003644514485,
+      "loss": 1.4167,
+      "step": 3538
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0754906872344934,
+      "learning_rate": 0.0007111425553922653,
+      "loss": 1.4765,
+      "step": 3539
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.07277782098342274,
+      "learning_rate": 0.0007109847207313191,
+      "loss": 1.4083,
+      "step": 3540
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.06913253674908561,
+      "learning_rate": 0.0007108268604877483,
+      "loss": 1.4447,
+      "step": 3541
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.06659941330711411,
+      "learning_rate": 0.0007106689746806939,
+      "loss": 1.4691,
+      "step": 3542
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.07232512460358181,
+      "learning_rate": 0.0007105110633292999,
+      "loss": 1.5318,
+      "step": 3543
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.06723843964590365,
+      "learning_rate": 0.0007103531264527138,
+      "loss": 1.4276,
+      "step": 3544
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.06852588979292723,
+      "learning_rate": 0.000710195164070086,
+      "loss": 1.3008,
+      "step": 3545
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.06904335986327417,
+      "learning_rate": 0.0007100371762005697,
+      "loss": 1.2505,
+      "step": 3546
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.07517395552253944,
+      "learning_rate": 0.0007098791628633217,
+      "loss": 1.5369,
+      "step": 3547
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.07166106998321647,
+      "learning_rate": 0.0007097211240775018,
+      "loss": 1.2608,
+      "step": 3548
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.06562667167612078,
+      "learning_rate": 0.0007095630598622724,
+      "loss": 1.4766,
+      "step": 3549
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.07500007910028895,
+      "learning_rate": 0.0007094049702367997,
+      "loss": 1.1845,
+      "step": 3550
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.07829807898182478,
+      "learning_rate": 0.0007092468552202523,
+      "loss": 1.2858,
+      "step": 3551
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.06925524966397749,
+      "learning_rate": 0.0007090887148318023,
+      "loss": 1.2332,
+      "step": 3552
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.08475475670099092,
+      "learning_rate": 0.000708930549090625,
+      "loss": 1.5682,
+      "step": 3553
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.07622082830610633,
+      "learning_rate": 0.0007087723580158983,
+      "loss": 1.3916,
+      "step": 3554
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0766525857066531,
+      "learning_rate": 0.0007086141416268033,
+      "loss": 1.4084,
+      "step": 3555
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.08003974204009992,
+      "learning_rate": 0.0007084558999425245,
+      "loss": 1.4344,
+      "step": 3556
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.06920958308933454,
+      "learning_rate": 0.0007082976329822491,
+      "loss": 1.4378,
+      "step": 3557
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.07090901454865277,
+      "learning_rate": 0.0007081393407651675,
+      "loss": 1.4642,
+      "step": 3558
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.07078282523191834,
+      "learning_rate": 0.0007079810233104734,
+      "loss": 1.3724,
+      "step": 3559
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.06605378041470791,
+      "learning_rate": 0.0007078226806373631,
+      "loss": 1.4078,
+      "step": 3560
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.06343453006005209,
+      "learning_rate": 0.0007076643127650366,
+      "loss": 1.4076,
+      "step": 3561
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.07317786253087795,
+      "learning_rate": 0.0007075059197126961,
+      "loss": 1.5665,
+      "step": 3562
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0654489556452131,
+      "learning_rate": 0.0007073475014995472,
+      "loss": 1.4653,
+      "step": 3563
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.07119377427290143,
+      "learning_rate": 0.0007071890581447992,
+      "loss": 1.4013,
+      "step": 3564
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.06589837136177674,
+      "learning_rate": 0.0007070305896676634,
+      "loss": 1.3597,
+      "step": 3565
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.07099054244663582,
+      "learning_rate": 0.0007068720960873552,
+      "loss": 1.4718,
+      "step": 3566
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.06627867168212025,
+      "learning_rate": 0.0007067135774230919,
+      "loss": 1.5182,
+      "step": 3567
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.06392067010017136,
+      "learning_rate": 0.0007065550336940947,
+      "loss": 1.5218,
+      "step": 3568
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.06383872794322432,
+      "learning_rate": 0.0007063964649195874,
+      "loss": 1.3602,
+      "step": 3569
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.07302909210052719,
+      "learning_rate": 0.0007062378711187973,
+      "loss": 1.4906,
+      "step": 3570
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.07202766521247028,
+      "learning_rate": 0.0007060792523109544,
+      "loss": 1.5076,
+      "step": 3571
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.06485625650547473,
+      "learning_rate": 0.0007059206085152918,
+      "loss": 1.3075,
+      "step": 3572
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0715065717760381,
+      "learning_rate": 0.0007057619397510453,
+      "loss": 1.4296,
+      "step": 3573
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.08825733453760141,
+      "learning_rate": 0.0007056032460374541,
+      "loss": 1.4665,
+      "step": 3574
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0701463361815818,
+      "learning_rate": 0.0007054445273937609,
+      "loss": 1.3857,
+      "step": 3575
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.07630369200437422,
+      "learning_rate": 0.0007052857838392104,
+      "loss": 1.4023,
+      "step": 3576
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.06817368849988802,
+      "learning_rate": 0.0007051270153930506,
+      "loss": 1.5429,
+      "step": 3577
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.07273743830308618,
+      "learning_rate": 0.0007049682220745332,
+      "loss": 1.5339,
+      "step": 3578
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.06237971196922428,
+      "learning_rate": 0.0007048094039029122,
+      "loss": 1.4023,
+      "step": 3579
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.07174054000231482,
+      "learning_rate": 0.0007046505608974447,
+      "loss": 1.4179,
+      "step": 3580
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.07079421703558056,
+      "learning_rate": 0.0007044916930773915,
+      "loss": 1.4336,
+      "step": 3581
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.06324002319659304,
+      "learning_rate": 0.0007043328004620154,
+      "loss": 1.4002,
+      "step": 3582
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.07273252464213828,
+      "learning_rate": 0.0007041738830705827,
+      "loss": 1.3021,
+      "step": 3583
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0679012855779142,
+      "learning_rate": 0.0007040149409223628,
+      "loss": 1.5173,
+      "step": 3584
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.06706605104839712,
+      "learning_rate": 0.0007038559740366281,
+      "loss": 1.4245,
+      "step": 3585
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.07263886553942718,
+      "learning_rate": 0.0007036969824326535,
+      "loss": 1.5822,
+      "step": 3586
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.06272461509437303,
+      "learning_rate": 0.0007035379661297179,
+      "loss": 1.3431,
+      "step": 3587
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.07364340877937099,
+      "learning_rate": 0.0007033789251471019,
+      "loss": 1.3892,
+      "step": 3588
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.07060017034955898,
+      "learning_rate": 0.0007032198595040901,
+      "loss": 1.4748,
+      "step": 3589
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.06756244433398076,
+      "learning_rate": 0.00070306076921997,
+      "loss": 1.3551,
+      "step": 3590
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.07358338514614324,
+      "learning_rate": 0.0007029016543140311,
+      "loss": 1.407,
+      "step": 3591
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.09232691672724472,
+      "learning_rate": 0.0007027425148055677,
+      "loss": 1.4698,
+      "step": 3592
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.07306567061185729,
+      "learning_rate": 0.000702583350713875,
+      "loss": 1.3349,
+      "step": 3593
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.07152049501825936,
+      "learning_rate": 0.0007024241620582527,
+      "loss": 1.5199,
+      "step": 3594
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.07471021321320678,
+      "learning_rate": 0.0007022649488580029,
+      "loss": 1.4685,
+      "step": 3595
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.07674020487606688,
+      "learning_rate": 0.0007021057111324307,
+      "loss": 1.3409,
+      "step": 3596
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.08010512353343058,
+      "learning_rate": 0.0007019464489008443,
+      "loss": 1.4164,
+      "step": 3597
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.07294899590648436,
+      "learning_rate": 0.0007017871621825549,
+      "loss": 1.4573,
+      "step": 3598
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0714024285774514,
+      "learning_rate": 0.0007016278509968761,
+      "loss": 1.3979,
+      "step": 3599
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0715807774435257,
+      "learning_rate": 0.0007014685153631255,
+      "loss": 1.3887,
+      "step": 3600
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.07500788006512786,
+      "learning_rate": 0.0007013091553006227,
+      "loss": 1.4591,
+      "step": 3601
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.06521259328969803,
+      "learning_rate": 0.0007011497708286909,
+      "loss": 1.4541,
+      "step": 3602
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0741904167001245,
+      "learning_rate": 0.000700990361966656,
+      "loss": 1.4057,
+      "step": 3603
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.06183999889715281,
+      "learning_rate": 0.0007008309287338467,
+      "loss": 1.3977,
+      "step": 3604
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.06681395641341766,
+      "learning_rate": 0.0007006714711495949,
+      "loss": 1.5082,
+      "step": 3605
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.08653655603945522,
+      "learning_rate": 0.0007005119892332354,
+      "loss": 1.4142,
+      "step": 3606
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.07028847770806987,
+      "learning_rate": 0.0007003524830041059,
+      "loss": 1.4125,
+      "step": 3607
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.08717202095751826,
+      "learning_rate": 0.0007001929524815472,
+      "loss": 1.4099,
+      "step": 3608
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.07358796039969515,
+      "learning_rate": 0.0007000333976849028,
+      "loss": 1.4611,
+      "step": 3609
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.07321829406105443,
+      "learning_rate": 0.0006998738186335193,
+      "loss": 1.2502,
+      "step": 3610
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.07212265150735724,
+      "learning_rate": 0.0006997142153467461,
+      "loss": 1.5196,
+      "step": 3611
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.06711496628227426,
+      "learning_rate": 0.000699554587843936,
+      "loss": 1.4971,
+      "step": 3612
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0709609735167094,
+      "learning_rate": 0.0006993949361444441,
+      "loss": 1.3627,
+      "step": 3613
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.06737833681998814,
+      "learning_rate": 0.0006992352602676287,
+      "loss": 1.4435,
+      "step": 3614
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.08318606479339613,
+      "learning_rate": 0.0006990755602328512,
+      "loss": 1.3223,
+      "step": 3615
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.07678376698123786,
+      "learning_rate": 0.0006989158360594756,
+      "loss": 1.2798,
+      "step": 3616
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.07405997377374739,
+      "learning_rate": 0.0006987560877668692,
+      "loss": 1.4287,
+      "step": 3617
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0764351613615117,
+      "learning_rate": 0.0006985963153744019,
+      "loss": 1.5333,
+      "step": 3618
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.08385446808915825,
+      "learning_rate": 0.0006984365189014467,
+      "loss": 1.4834,
+      "step": 3619
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.06263563035341106,
+      "learning_rate": 0.0006982766983673795,
+      "loss": 1.3152,
+      "step": 3620
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.08852794236399607,
+      "learning_rate": 0.000698116853791579,
+      "loss": 1.4673,
+      "step": 3621
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.06941771637584818,
+      "learning_rate": 0.000697956985193427,
+      "loss": 1.4864,
+      "step": 3622
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.08562998996198683,
+      "learning_rate": 0.0006977970925923081,
+      "loss": 1.4882,
+      "step": 3623
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.07531704894846612,
+      "learning_rate": 0.0006976371760076099,
+      "loss": 1.4987,
+      "step": 3624
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0801714652638698,
+      "learning_rate": 0.0006974772354587226,
+      "loss": 1.4313,
+      "step": 3625
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.08020508619169321,
+      "learning_rate": 0.0006973172709650397,
+      "loss": 1.4755,
+      "step": 3626
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.07296454416739705,
+      "learning_rate": 0.0006971572825459576,
+      "loss": 1.4356,
+      "step": 3627
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.06951882359819238,
+      "learning_rate": 0.000696997270220875,
+      "loss": 1.4227,
+      "step": 3628
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.07644784282849344,
+      "learning_rate": 0.0006968372340091946,
+      "loss": 1.4035,
+      "step": 3629
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.07155938125174738,
+      "learning_rate": 0.0006966771739303206,
+      "loss": 1.4814,
+      "step": 3630
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.06280959644830418,
+      "learning_rate": 0.0006965170900036613,
+      "loss": 1.432,
+      "step": 3631
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.06420053763727787,
+      "learning_rate": 0.0006963569822486276,
+      "loss": 1.3966,
+      "step": 3632
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.06407629379194352,
+      "learning_rate": 0.0006961968506846327,
+      "loss": 1.4223,
+      "step": 3633
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.07341699164847276,
+      "learning_rate": 0.0006960366953310931,
+      "loss": 1.6418,
+      "step": 3634
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.07158103196668297,
+      "learning_rate": 0.0006958765162074287,
+      "loss": 1.3857,
+      "step": 3635
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.06656688307492928,
+      "learning_rate": 0.0006957163133330611,
+      "loss": 1.5035,
+      "step": 3636
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.07009898988268752,
+      "learning_rate": 0.0006955560867274159,
+      "loss": 1.4776,
+      "step": 3637
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.06893175987469181,
+      "learning_rate": 0.0006953958364099208,
+      "loss": 1.5117,
+      "step": 3638
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.07194928575269588,
+      "learning_rate": 0.0006952355624000072,
+      "loss": 1.4054,
+      "step": 3639
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.07005599291594371,
+      "learning_rate": 0.0006950752647171086,
+      "loss": 1.4454,
+      "step": 3640
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.06485794286202459,
+      "learning_rate": 0.0006949149433806614,
+      "loss": 1.2388,
+      "step": 3641
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.07321685193872544,
+      "learning_rate": 0.0006947545984101053,
+      "loss": 1.5212,
+      "step": 3642
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.06766799514535249,
+      "learning_rate": 0.000694594229824883,
+      "loss": 1.4703,
+      "step": 3643
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.07674140785209206,
+      "learning_rate": 0.0006944338376444393,
+      "loss": 1.4143,
+      "step": 3644
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.07113361066488795,
+      "learning_rate": 0.0006942734218882225,
+      "loss": 1.5061,
+      "step": 3645
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.07408116710866873,
+      "learning_rate": 0.0006941129825756836,
+      "loss": 1.3525,
+      "step": 3646
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.06834433253859368,
+      "learning_rate": 0.0006939525197262762,
+      "loss": 1.3372,
+      "step": 3647
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.07240980576151891,
+      "learning_rate": 0.0006937920333594572,
+      "loss": 1.3757,
+      "step": 3648
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.07630489881087277,
+      "learning_rate": 0.0006936315234946861,
+      "loss": 1.4788,
+      "step": 3649
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.07956854851302582,
+      "learning_rate": 0.0006934709901514251,
+      "loss": 1.4576,
+      "step": 3650
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0670356740719808,
+      "learning_rate": 0.0006933104333491398,
+      "loss": 1.5165,
+      "step": 3651
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.07588722767607452,
+      "learning_rate": 0.0006931498531072977,
+      "loss": 1.3835,
+      "step": 3652
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.09334811732319619,
+      "learning_rate": 0.0006929892494453703,
+      "loss": 1.4464,
+      "step": 3653
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.061988691124392446,
+      "learning_rate": 0.0006928286223828309,
+      "loss": 1.3886,
+      "step": 3654
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.07921738004769585,
+      "learning_rate": 0.0006926679719391562,
+      "loss": 1.4586,
+      "step": 3655
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.08326077624240993,
+      "learning_rate": 0.0006925072981338259,
+      "loss": 1.3614,
+      "step": 3656
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.08352543499375753,
+      "learning_rate": 0.0006923466009863218,
+      "loss": 1.4544,
+      "step": 3657
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.06887853383184417,
+      "learning_rate": 0.0006921858805161294,
+      "loss": 1.3777,
+      "step": 3658
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.06923628331383389,
+      "learning_rate": 0.0006920251367427361,
+      "loss": 1.5486,
+      "step": 3659
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.06519640964192866,
+      "learning_rate": 0.0006918643696856333,
+      "loss": 1.4187,
+      "step": 3660
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.07237657442273564,
+      "learning_rate": 0.0006917035793643141,
+      "loss": 1.5045,
+      "step": 3661
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0693432290313142,
+      "learning_rate": 0.0006915427657982751,
+      "loss": 1.3163,
+      "step": 3662
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.06416290954447852,
+      "learning_rate": 0.0006913819290070153,
+      "loss": 1.4348,
+      "step": 3663
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.07147056356269492,
+      "learning_rate": 0.0006912210690100369,
+      "loss": 1.4821,
+      "step": 3664
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.07969102834155034,
+      "learning_rate": 0.0006910601858268444,
+      "loss": 1.453,
+      "step": 3665
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.13718934471197883,
+      "learning_rate": 0.000690899279476946,
+      "loss": 1.4078,
+      "step": 3666
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.08114721434640836,
+      "learning_rate": 0.0006907383499798516,
+      "loss": 1.4907,
+      "step": 3667
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.07115178902771321,
+      "learning_rate": 0.0006905773973550748,
+      "loss": 1.4556,
+      "step": 3668
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.07942590621430361,
+      "learning_rate": 0.0006904164216221314,
+      "loss": 1.4596,
+      "step": 3669
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.07289576701035694,
+      "learning_rate": 0.0006902554228005405,
+      "loss": 1.4612,
+      "step": 3670
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.06528821995816773,
+      "learning_rate": 0.0006900944009098238,
+      "loss": 1.3585,
+      "step": 3671
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.07616160702579455,
+      "learning_rate": 0.0006899333559695056,
+      "loss": 1.3554,
+      "step": 3672
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.07769584578503212,
+      "learning_rate": 0.0006897722879991131,
+      "loss": 1.4175,
+      "step": 3673
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.07083227509463093,
+      "learning_rate": 0.0006896111970181764,
+      "loss": 1.4129,
+      "step": 3674
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06563564523573756,
+      "learning_rate": 0.0006894500830462285,
+      "loss": 1.3298,
+      "step": 3675
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0704221954251129,
+      "learning_rate": 0.0006892889461028047,
+      "loss": 1.4314,
+      "step": 3676
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.07517734381017364,
+      "learning_rate": 0.0006891277862074439,
+      "loss": 1.3131,
+      "step": 3677
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.07709956431855618,
+      "learning_rate": 0.0006889666033796869,
+      "loss": 1.4983,
+      "step": 3678
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06641564320298735,
+      "learning_rate": 0.0006888053976390776,
+      "loss": 1.4659,
+      "step": 3679
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.07326313513970828,
+      "learning_rate": 0.000688644169005163,
+      "loss": 1.4576,
+      "step": 3680
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06526229926022548,
+      "learning_rate": 0.0006884829174974927,
+      "loss": 1.5378,
+      "step": 3681
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06510469605310103,
+      "learning_rate": 0.0006883216431356187,
+      "loss": 1.2951,
+      "step": 3682
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06975087735626676,
+      "learning_rate": 0.0006881603459390964,
+      "loss": 1.535,
+      "step": 3683
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.07054312371923935,
+      "learning_rate": 0.0006879990259274832,
+      "loss": 1.6018,
+      "step": 3684
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06413316601711448,
+      "learning_rate": 0.0006878376831203401,
+      "loss": 1.4806,
+      "step": 3685
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06207448118614186,
+      "learning_rate": 0.0006876763175372305,
+      "loss": 1.4473,
+      "step": 3686
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.07947569981904305,
+      "learning_rate": 0.00068751492919772,
+      "loss": 1.2827,
+      "step": 3687
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0667703783383479,
+      "learning_rate": 0.0006873535181213784,
+      "loss": 1.4303,
+      "step": 3688
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06814443823594021,
+      "learning_rate": 0.0006871920843277764,
+      "loss": 1.3534,
+      "step": 3689
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06858806164536586,
+      "learning_rate": 0.000687030627836489,
+      "loss": 1.4704,
+      "step": 3690
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.07516155321113553,
+      "learning_rate": 0.0006868691486670932,
+      "loss": 1.4027,
+      "step": 3691
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06523038718461018,
+      "learning_rate": 0.0006867076468391688,
+      "loss": 1.4835,
+      "step": 3692
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0641694726958727,
+      "learning_rate": 0.0006865461223722986,
+      "loss": 1.4408,
+      "step": 3693
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.07188965078622174,
+      "learning_rate": 0.0006863845752860679,
+      "loss": 1.509,
+      "step": 3694
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06917751289317574,
+      "learning_rate": 0.0006862230056000648,
+      "loss": 1.4961,
+      "step": 3695
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06763332815454019,
+      "learning_rate": 0.0006860614133338804,
+      "loss": 1.3402,
+      "step": 3696
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06762636955642931,
+      "learning_rate": 0.0006858997985071081,
+      "loss": 1.5605,
+      "step": 3697
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06856232217515264,
+      "learning_rate": 0.0006857381611393445,
+      "loss": 1.4544,
+      "step": 3698
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06275521544299778,
+      "learning_rate": 0.0006855765012501883,
+      "loss": 1.4577,
+      "step": 3699
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06899956083233556,
+      "learning_rate": 0.0006854148188592418,
+      "loss": 1.4359,
+      "step": 3700
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06995023762873773,
+      "learning_rate": 0.000685253113986109,
+      "loss": 1.4413,
+      "step": 3701
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06988588357833313,
+      "learning_rate": 0.0006850913866503977,
+      "loss": 1.3826,
+      "step": 3702
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06786538775904234,
+      "learning_rate": 0.0006849296368717176,
+      "loss": 1.4923,
+      "step": 3703
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.07654915353921299,
+      "learning_rate": 0.0006847678646696813,
+      "loss": 1.3739,
+      "step": 3704
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0627728148075214,
+      "learning_rate": 0.0006846060700639046,
+      "loss": 1.3529,
+      "step": 3705
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06829333606625117,
+      "learning_rate": 0.0006844442530740055,
+      "loss": 1.5439,
+      "step": 3706
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06632683279100844,
+      "learning_rate": 0.0006842824137196046,
+      "loss": 1.3775,
+      "step": 3707
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0709408213687562,
+      "learning_rate": 0.000684120552020326,
+      "loss": 1.2745,
+      "step": 3708
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.07214752580069766,
+      "learning_rate": 0.0006839586679957956,
+      "loss": 1.5218,
+      "step": 3709
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.07159420982548895,
+      "learning_rate": 0.0006837967616656425,
+      "loss": 1.4378,
+      "step": 3710
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.07620806323467304,
+      "learning_rate": 0.0006836348330494984,
+      "loss": 1.4767,
+      "step": 3711
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0712872441268927,
+      "learning_rate": 0.0006834728821669977,
+      "loss": 1.3746,
+      "step": 3712
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06739384795782043,
+      "learning_rate": 0.0006833109090377775,
+      "loss": 1.3444,
+      "step": 3713
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.07259481271825156,
+      "learning_rate": 0.0006831489136814777,
+      "loss": 1.4627,
+      "step": 3714
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.08300629502199651,
+      "learning_rate": 0.0006829868961177406,
+      "loss": 1.3774,
+      "step": 3715
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06620793761290668,
+      "learning_rate": 0.0006828248563662116,
+      "loss": 1.441,
+      "step": 3716
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06753191484298786,
+      "learning_rate": 0.0006826627944465383,
+      "loss": 1.2897,
+      "step": 3717
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06968868045330964,
+      "learning_rate": 0.0006825007103783716,
+      "loss": 1.3708,
+      "step": 3718
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0664248081179177,
+      "learning_rate": 0.0006823386041813647,
+      "loss": 1.3146,
+      "step": 3719
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06767190375786564,
+      "learning_rate": 0.0006821764758751732,
+      "loss": 1.404,
+      "step": 3720
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.07605015217317455,
+      "learning_rate": 0.0006820143254794559,
+      "loss": 1.5113,
+      "step": 3721
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06485099522202713,
+      "learning_rate": 0.0006818521530138743,
+      "loss": 1.3003,
+      "step": 3722
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.07531942615179421,
+      "learning_rate": 0.0006816899584980922,
+      "loss": 1.4695,
+      "step": 3723
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.09489560161012517,
+      "learning_rate": 0.000681527741951776,
+      "loss": 1.3409,
+      "step": 3724
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.061931818640721746,
+      "learning_rate": 0.0006813655033945956,
+      "loss": 1.4266,
+      "step": 3725
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.08377945929406046,
+      "learning_rate": 0.0006812032428462225,
+      "loss": 1.4634,
+      "step": 3726
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.08635568739760313,
+      "learning_rate": 0.0006810409603263314,
+      "loss": 1.5708,
+      "step": 3727
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.07301256105687044,
+      "learning_rate": 0.0006808786558546,
+      "loss": 1.4444,
+      "step": 3728
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.07858240960791367,
+      "learning_rate": 0.0006807163294507078,
+      "loss": 1.4104,
+      "step": 3729
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.069090300634783,
+      "learning_rate": 0.0006805539811343376,
+      "loss": 1.3712,
+      "step": 3730
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.07759023420372646,
+      "learning_rate": 0.0006803916109251748,
+      "loss": 1.3338,
+      "step": 3731
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.07414115909337814,
+      "learning_rate": 0.0006802292188429072,
+      "loss": 1.5009,
+      "step": 3732
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0818109136073175,
+      "learning_rate": 0.0006800668049072256,
+      "loss": 1.3017,
+      "step": 3733
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.07277609899484332,
+      "learning_rate": 0.000679904369137823,
+      "loss": 1.4741,
+      "step": 3734
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06986648440783053,
+      "learning_rate": 0.0006797419115543954,
+      "loss": 1.3859,
+      "step": 3735
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.08023992915741134,
+      "learning_rate": 0.0006795794321766415,
+      "loss": 1.5204,
+      "step": 3736
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.07276280799914031,
+      "learning_rate": 0.0006794169310242624,
+      "loss": 1.4427,
+      "step": 3737
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.07170170736813007,
+      "learning_rate": 0.0006792544081169616,
+      "loss": 1.4303,
+      "step": 3738
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06361279357416362,
+      "learning_rate": 0.000679091863474446,
+      "loss": 1.5411,
+      "step": 3739
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.07108682039307701,
+      "learning_rate": 0.0006789292971164244,
+      "loss": 1.2778,
+      "step": 3740
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.07580525556322194,
+      "learning_rate": 0.0006787667090626089,
+      "loss": 1.3907,
+      "step": 3741
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06995561131507061,
+      "learning_rate": 0.0006786040993327135,
+      "loss": 1.4253,
+      "step": 3742
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.07492744663626666,
+      "learning_rate": 0.0006784414679464552,
+      "loss": 1.4251,
+      "step": 3743
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06882834924705727,
+      "learning_rate": 0.0006782788149235538,
+      "loss": 1.4079,
+      "step": 3744
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.07609185882270524,
+      "learning_rate": 0.0006781161402837316,
+      "loss": 1.5241,
+      "step": 3745
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0919088950874994,
+      "learning_rate": 0.0006779534440467133,
+      "loss": 1.4305,
+      "step": 3746
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.07213172396040231,
+      "learning_rate": 0.0006777907262322262,
+      "loss": 1.3415,
+      "step": 3747
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06770604290970164,
+      "learning_rate": 0.0006776279868600008,
+      "loss": 1.4458,
+      "step": 3748
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.07331521526131282,
+      "learning_rate": 0.0006774652259497696,
+      "loss": 1.4466,
+      "step": 3749
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06631965109609582,
+      "learning_rate": 0.0006773024435212678,
+      "loss": 1.45,
+      "step": 3750
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06771060610435131,
+      "learning_rate": 0.0006771396395942338,
+      "loss": 1.3031,
+      "step": 3751
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0725160661843529,
+      "learning_rate": 0.0006769768141884074,
+      "loss": 1.5257,
+      "step": 3752
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0793938270762842,
+      "learning_rate": 0.0006768139673235323,
+      "loss": 1.5339,
+      "step": 3753
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.07151774981622631,
+      "learning_rate": 0.0006766510990193541,
+      "loss": 1.3757,
+      "step": 3754
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06547987137883667,
+      "learning_rate": 0.0006764882092956211,
+      "loss": 1.3952,
+      "step": 3755
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.07314158842433856,
+      "learning_rate": 0.0006763252981720844,
+      "loss": 1.5099,
+      "step": 3756
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.07145510078013781,
+      "learning_rate": 0.0006761623656684973,
+      "loss": 1.3377,
+      "step": 3757
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.08644202316815978,
+      "learning_rate": 0.0006759994118046161,
+      "loss": 1.414,
+      "step": 3758
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.07282147693961499,
+      "learning_rate": 0.0006758364366001994,
+      "loss": 1.4244,
+      "step": 3759
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.07554452717292655,
+      "learning_rate": 0.0006756734400750087,
+      "loss": 1.3676,
+      "step": 3760
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0666807039375581,
+      "learning_rate": 0.0006755104222488076,
+      "loss": 1.4265,
+      "step": 3761
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.09012510595806483,
+      "learning_rate": 0.0006753473831413628,
+      "loss": 1.3391,
+      "step": 3762
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.07835217913269821,
+      "learning_rate": 0.0006751843227724432,
+      "loss": 1.5525,
+      "step": 3763
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06987950663692771,
+      "learning_rate": 0.0006750212411618206,
+      "loss": 1.3493,
+      "step": 3764
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06996075079577234,
+      "learning_rate": 0.000674858138329269,
+      "loss": 1.457,
+      "step": 3765
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.07173072264270693,
+      "learning_rate": 0.0006746950142945653,
+      "loss": 1.326,
+      "step": 3766
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.06810115938826602,
+      "learning_rate": 0.0006745318690774891,
+      "loss": 1.2359,
+      "step": 3767
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.07757166620232256,
+      "learning_rate": 0.0006743687026978219,
+      "loss": 1.4865,
+      "step": 3768
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.08521148854272749,
+      "learning_rate": 0.0006742055151753483,
+      "loss": 1.5518,
+      "step": 3769
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.07341229077115634,
+      "learning_rate": 0.0006740423065298556,
+      "loss": 1.5728,
+      "step": 3770
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.07251757762711362,
+      "learning_rate": 0.0006738790767811329,
+      "loss": 1.4438,
+      "step": 3771
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0754765370031017,
+      "learning_rate": 0.0006737158259489729,
+      "loss": 1.4546,
+      "step": 3772
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.06987041998936272,
+      "learning_rate": 0.0006735525540531702,
+      "loss": 1.4971,
+      "step": 3773
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.07846064142843433,
+      "learning_rate": 0.0006733892611135217,
+      "loss": 1.3602,
+      "step": 3774
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.08199997856494941,
+      "learning_rate": 0.0006732259471498278,
+      "loss": 1.5195,
+      "step": 3775
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.06705852406335323,
+      "learning_rate": 0.0006730626121818906,
+      "loss": 1.3127,
+      "step": 3776
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.07146928164080467,
+      "learning_rate": 0.0006728992562295148,
+      "loss": 1.4246,
+      "step": 3777
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.07467077035958926,
+      "learning_rate": 0.0006727358793125084,
+      "loss": 1.3603,
+      "step": 3778
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.06747803921749279,
+      "learning_rate": 0.0006725724814506809,
+      "loss": 1.401,
+      "step": 3779
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.07455585726671034,
+      "learning_rate": 0.0006724090626638451,
+      "loss": 1.3864,
+      "step": 3780
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.06935011523991552,
+      "learning_rate": 0.0006722456229718162,
+      "loss": 1.3373,
+      "step": 3781
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.07001330043357815,
+      "learning_rate": 0.0006720821623944117,
+      "loss": 1.3813,
+      "step": 3782
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.07006791244257965,
+      "learning_rate": 0.0006719186809514516,
+      "loss": 1.4948,
+      "step": 3783
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.08041409052706847,
+      "learning_rate": 0.000671755178662759,
+      "loss": 1.4941,
+      "step": 3784
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.07903064459799436,
+      "learning_rate": 0.0006715916555481585,
+      "loss": 1.3258,
+      "step": 3785
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.07790220948556671,
+      "learning_rate": 0.0006714281116274783,
+      "loss": 1.3577,
+      "step": 3786
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0683831054506245,
+      "learning_rate": 0.0006712645469205488,
+      "loss": 1.3955,
+      "step": 3787
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.07375602721280128,
+      "learning_rate": 0.0006711009614472022,
+      "loss": 1.344,
+      "step": 3788
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0753083239073712,
+      "learning_rate": 0.0006709373552272744,
+      "loss": 1.3137,
+      "step": 3789
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.08058980255270123,
+      "learning_rate": 0.0006707737282806029,
+      "loss": 1.4757,
+      "step": 3790
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.07188458605519603,
+      "learning_rate": 0.000670610080627028,
+      "loss": 1.4421,
+      "step": 3791
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0745797053981257,
+      "learning_rate": 0.0006704464122863928,
+      "loss": 1.385,
+      "step": 3792
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0809561549922701,
+      "learning_rate": 0.0006702827232785425,
+      "loss": 1.5045,
+      "step": 3793
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.07043268111949744,
+      "learning_rate": 0.0006701190136233249,
+      "loss": 1.4208,
+      "step": 3794
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.06464522265885646,
+      "learning_rate": 0.0006699552833405903,
+      "loss": 1.4705,
+      "step": 3795
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.06898157473720405,
+      "learning_rate": 0.0006697915324501918,
+      "loss": 1.3685,
+      "step": 3796
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.06720721473990293,
+      "learning_rate": 0.0006696277609719845,
+      "loss": 1.2577,
+      "step": 3797
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.069477039304262,
+      "learning_rate": 0.0006694639689258265,
+      "loss": 1.395,
+      "step": 3798
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.067690922335574,
+      "learning_rate": 0.0006693001563315782,
+      "loss": 1.4064,
+      "step": 3799
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.07169248794058057,
+      "learning_rate": 0.0006691363232091019,
+      "loss": 1.483,
+      "step": 3800
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.07878728528871834,
+      "learning_rate": 0.0006689724695782635,
+      "loss": 1.5265,
+      "step": 3801
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.07027898095463854,
+      "learning_rate": 0.0006688085954589307,
+      "loss": 1.4565,
+      "step": 3802
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0671644222122429,
+      "learning_rate": 0.0006686447008709735,
+      "loss": 1.3886,
+      "step": 3803
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.10803177504795355,
+      "learning_rate": 0.0006684807858342652,
+      "loss": 1.4621,
+      "step": 3804
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.06890699521351634,
+      "learning_rate": 0.0006683168503686806,
+      "loss": 1.3289,
+      "step": 3805
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.06272061534057344,
+      "learning_rate": 0.0006681528944940977,
+      "loss": 1.4191,
+      "step": 3806
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.06759035500761486,
+      "learning_rate": 0.0006679889182303966,
+      "loss": 1.4501,
+      "step": 3807
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.07934872841569579,
+      "learning_rate": 0.00066782492159746,
+      "loss": 1.3587,
+      "step": 3808
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.07583621080731881,
+      "learning_rate": 0.0006676609046151732,
+      "loss": 1.2725,
+      "step": 3809
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.08046256366810259,
+      "learning_rate": 0.0006674968673034235,
+      "loss": 1.3842,
+      "step": 3810
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.08949644141491607,
+      "learning_rate": 0.0006673328096821012,
+      "loss": 1.3413,
+      "step": 3811
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.07477384583849257,
+      "learning_rate": 0.0006671687317710989,
+      "loss": 1.4916,
+      "step": 3812
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.07549342598106959,
+      "learning_rate": 0.0006670046335903116,
+      "loss": 1.3989,
+      "step": 3813
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.07054448270490725,
+      "learning_rate": 0.0006668405151596367,
+      "loss": 1.3175,
+      "step": 3814
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.06726660936523654,
+      "learning_rate": 0.0006666763764989742,
+      "loss": 1.5434,
+      "step": 3815
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.07343583098757128,
+      "learning_rate": 0.0006665122176282264,
+      "loss": 1.4393,
+      "step": 3816
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.07954843590714228,
+      "learning_rate": 0.000666348038567298,
+      "loss": 1.4805,
+      "step": 3817
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0749554275646494,
+      "learning_rate": 0.0006661838393360966,
+      "loss": 1.3784,
+      "step": 3818
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.07420331079114773,
+      "learning_rate": 0.0006660196199545317,
+      "loss": 1.3676,
+      "step": 3819
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.06857050240511989,
+      "learning_rate": 0.0006658553804425156,
+      "loss": 1.4668,
+      "step": 3820
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.07062870193427842,
+      "learning_rate": 0.0006656911208199627,
+      "loss": 1.4805,
+      "step": 3821
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.07851558591594957,
+      "learning_rate": 0.0006655268411067903,
+      "loss": 1.4744,
+      "step": 3822
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.07572325313412566,
+      "learning_rate": 0.0006653625413229177,
+      "loss": 1.4193,
+      "step": 3823
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.07822957044039873,
+      "learning_rate": 0.000665198221488267,
+      "loss": 1.4386,
+      "step": 3824
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.07604473764552287,
+      "learning_rate": 0.0006650338816227622,
+      "loss": 1.3252,
+      "step": 3825
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.07524178279695189,
+      "learning_rate": 0.0006648695217463304,
+      "loss": 1.3825,
+      "step": 3826
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.08116651197896221,
+      "learning_rate": 0.0006647051418789007,
+      "loss": 1.3457,
+      "step": 3827
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.07209272417387406,
+      "learning_rate": 0.0006645407420404047,
+      "loss": 1.3494,
+      "step": 3828
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.07602965698815438,
+      "learning_rate": 0.0006643763222507765,
+      "loss": 1.4471,
+      "step": 3829
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.06920958645347326,
+      "learning_rate": 0.0006642118825299526,
+      "loss": 1.3819,
+      "step": 3830
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.06868419959482276,
+      "learning_rate": 0.0006640474228978716,
+      "loss": 1.3569,
+      "step": 3831
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.07442565049284244,
+      "learning_rate": 0.0006638829433744753,
+      "loss": 1.4242,
+      "step": 3832
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.07469968744098102,
+      "learning_rate": 0.0006637184439797069,
+      "loss": 1.3941,
+      "step": 3833
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.06954189557918407,
+      "learning_rate": 0.0006635539247335128,
+      "loss": 1.3249,
+      "step": 3834
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0788039300618776,
+      "learning_rate": 0.0006633893856558415,
+      "loss": 1.5244,
+      "step": 3835
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.07227186871954694,
+      "learning_rate": 0.000663224826766644,
+      "loss": 1.4414,
+      "step": 3836
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.07364965898774103,
+      "learning_rate": 0.0006630602480858734,
+      "loss": 1.4668,
+      "step": 3837
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0710617632092743,
+      "learning_rate": 0.0006628956496334856,
+      "loss": 1.4955,
+      "step": 3838
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.07483232033855866,
+      "learning_rate": 0.0006627310314294385,
+      "loss": 1.4276,
+      "step": 3839
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.07028946813885235,
+      "learning_rate": 0.0006625663934936932,
+      "loss": 1.4192,
+      "step": 3840
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.06907195642106281,
+      "learning_rate": 0.0006624017358462122,
+      "loss": 1.509,
+      "step": 3841
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.08082576245434056,
+      "learning_rate": 0.0006622370585069604,
+      "loss": 1.4527,
+      "step": 3842
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.07246371293472104,
+      "learning_rate": 0.0006620723614959063,
+      "loss": 1.4888,
+      "step": 3843
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.07731335402711258,
+      "learning_rate": 0.0006619076448330197,
+      "loss": 1.3546,
+      "step": 3844
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.09375234814524461,
+      "learning_rate": 0.0006617429085382727,
+      "loss": 1.5487,
+      "step": 3845
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.08419642597319515,
+      "learning_rate": 0.0006615781526316406,
+      "loss": 1.3322,
+      "step": 3846
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.07497704397986803,
+      "learning_rate": 0.0006614133771331006,
+      "loss": 1.3227,
+      "step": 3847
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.06673807622226641,
+      "learning_rate": 0.0006612485820626317,
+      "loss": 1.4786,
+      "step": 3848
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.06545079125780927,
+      "learning_rate": 0.0006610837674402167,
+      "loss": 1.4818,
+      "step": 3849
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.07666246902718891,
+      "learning_rate": 0.0006609189332858394,
+      "loss": 1.4476,
+      "step": 3850
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.07178457217246913,
+      "learning_rate": 0.0006607540796194866,
+      "loss": 1.4296,
+      "step": 3851
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.07540630488822511,
+      "learning_rate": 0.0006605892064611477,
+      "loss": 1.3576,
+      "step": 3852
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.06905586130396371,
+      "learning_rate": 0.0006604243138308137,
+      "loss": 1.4058,
+      "step": 3853
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0762169441747839,
+      "learning_rate": 0.0006602594017484785,
+      "loss": 1.4798,
+      "step": 3854
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.07945689113659538,
+      "learning_rate": 0.0006600944702341385,
+      "loss": 1.4323,
+      "step": 3855
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.06695641836147707,
+      "learning_rate": 0.0006599295193077922,
+      "loss": 1.514,
+      "step": 3856
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.06348489376560794,
+      "learning_rate": 0.0006597645489894399,
+      "loss": 1.3441,
+      "step": 3857
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.07358283615988383,
+      "learning_rate": 0.0006595995592990855,
+      "loss": 1.3343,
+      "step": 3858
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.06725547686839599,
+      "learning_rate": 0.0006594345502567342,
+      "loss": 1.4087,
+      "step": 3859
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.08059962050387262,
+      "learning_rate": 0.0006592695218823942,
+      "loss": 1.4242,
+      "step": 3860
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.07206294434314071,
+      "learning_rate": 0.0006591044741960754,
+      "loss": 1.3445,
+      "step": 3861
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.06549271529864518,
+      "learning_rate": 0.0006589394072177907,
+      "loss": 1.3822,
+      "step": 3862
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.07173780240938447,
+      "learning_rate": 0.0006587743209675549,
+      "loss": 1.5295,
+      "step": 3863
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.07004691337394642,
+      "learning_rate": 0.0006586092154653854,
+      "loss": 1.4876,
+      "step": 3864
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.07900190551261445,
+      "learning_rate": 0.0006584440907313016,
+      "loss": 1.4928,
+      "step": 3865
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.07316898820714804,
+      "learning_rate": 0.0006582789467853256,
+      "loss": 1.3457,
+      "step": 3866
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.06359569667254837,
+      "learning_rate": 0.0006581137836474816,
+      "loss": 1.3676,
+      "step": 3867
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.06782660467001123,
+      "learning_rate": 0.0006579486013377963,
+      "loss": 1.3953,
+      "step": 3868
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.06696194015947708,
+      "learning_rate": 0.0006577833998762985,
+      "loss": 1.4628,
+      "step": 3869
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0720934944292986,
+      "learning_rate": 0.0006576181792830193,
+      "loss": 1.395,
+      "step": 3870
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.05904143097036177,
+      "learning_rate": 0.0006574529395779928,
+      "loss": 1.2883,
+      "step": 3871
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.07144924842031333,
+      "learning_rate": 0.0006572876807812545,
+      "loss": 1.3083,
+      "step": 3872
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.07096535038028348,
+      "learning_rate": 0.0006571224029128425,
+      "loss": 1.4658,
+      "step": 3873
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.06858119737322739,
+      "learning_rate": 0.0006569571059927977,
+      "loss": 1.397,
+      "step": 3874
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.07465588152502295,
+      "learning_rate": 0.0006567917900411625,
+      "loss": 1.4628,
+      "step": 3875
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.07150645537132062,
+      "learning_rate": 0.0006566264550779821,
+      "loss": 1.3837,
+      "step": 3876
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.06856592343026248,
+      "learning_rate": 0.0006564611011233043,
+      "loss": 1.3578,
+      "step": 3877
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.06183930235911227,
+      "learning_rate": 0.0006562957281971785,
+      "loss": 1.3457,
+      "step": 3878
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.06662531178283607,
+      "learning_rate": 0.0006561303363196568,
+      "loss": 1.357,
+      "step": 3879
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.07518583399000245,
+      "learning_rate": 0.0006559649255107935,
+      "loss": 1.4672,
+      "step": 3880
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0693085282173378,
+      "learning_rate": 0.0006557994957906455,
+      "loss": 1.3528,
+      "step": 3881
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0728523880755816,
+      "learning_rate": 0.0006556340471792712,
+      "loss": 1.4816,
+      "step": 3882
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.06637118795337808,
+      "learning_rate": 0.0006554685796967323,
+      "loss": 1.4812,
+      "step": 3883
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.06861984917136406,
+      "learning_rate": 0.0006553030933630921,
+      "loss": 1.3983,
+      "step": 3884
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.06529551777213016,
+      "learning_rate": 0.0006551375881984165,
+      "loss": 1.4422,
+      "step": 3885
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.06858208005798797,
+      "learning_rate": 0.0006549720642227735,
+      "loss": 1.4688,
+      "step": 3886
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.07143616441393273,
+      "learning_rate": 0.0006548065214562332,
+      "loss": 1.3626,
+      "step": 3887
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.07404672141993965,
+      "learning_rate": 0.0006546409599188686,
+      "loss": 1.347,
+      "step": 3888
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.06567975995952656,
+      "learning_rate": 0.0006544753796307547,
+      "loss": 1.3886,
+      "step": 3889
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.06827963964218622,
+      "learning_rate": 0.0006543097806119682,
+      "loss": 1.3641,
+      "step": 3890
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.06502890786809211,
+      "learning_rate": 0.000654144162882589,
+      "loss": 1.5318,
+      "step": 3891
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.09549961420573892,
+      "learning_rate": 0.0006539785264626985,
+      "loss": 1.5141,
+      "step": 3892
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0696505156579795,
+      "learning_rate": 0.0006538128713723808,
+      "loss": 1.4501,
+      "step": 3893
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.07241761729384924,
+      "learning_rate": 0.0006536471976317223,
+      "loss": 1.4871,
+      "step": 3894
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.07391803138958859,
+      "learning_rate": 0.0006534815052608114,
+      "loss": 1.4029,
+      "step": 3895
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.06593299094089043,
+      "learning_rate": 0.0006533157942797387,
+      "loss": 1.5064,
+      "step": 3896
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.07575361796363428,
+      "learning_rate": 0.0006531500647085974,
+      "loss": 1.3223,
+      "step": 3897
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.07197208892800534,
+      "learning_rate": 0.0006529843165674828,
+      "loss": 1.3693,
+      "step": 3898
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0776799924076783,
+      "learning_rate": 0.0006528185498764924,
+      "loss": 1.4277,
+      "step": 3899
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.07989342835825704,
+      "learning_rate": 0.0006526527646557261,
+      "loss": 1.4741,
+      "step": 3900
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.07733258169149501,
+      "learning_rate": 0.0006524869609252856,
+      "loss": 1.4633,
+      "step": 3901
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.07556221365420827,
+      "learning_rate": 0.0006523211387052755,
+      "loss": 1.4267,
+      "step": 3902
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.07166604005031749,
+      "learning_rate": 0.0006521552980158023,
+      "loss": 1.4879,
+      "step": 3903
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.07005460949807596,
+      "learning_rate": 0.0006519894388769744,
+      "loss": 1.4087,
+      "step": 3904
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.08364807345793063,
+      "learning_rate": 0.0006518235613089034,
+      "loss": 1.461,
+      "step": 3905
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.07413803750026045,
+      "learning_rate": 0.0006516576653317019,
+      "loss": 1.3398,
+      "step": 3906
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.07896706543787335,
+      "learning_rate": 0.0006514917509654857,
+      "loss": 1.3738,
+      "step": 3907
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0638598561453725,
+      "learning_rate": 0.0006513258182303724,
+      "loss": 1.4045,
+      "step": 3908
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.06952205296431396,
+      "learning_rate": 0.000651159867146482,
+      "loss": 1.42,
+      "step": 3909
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.06892081357631127,
+      "learning_rate": 0.0006509938977339366,
+      "loss": 1.4329,
+      "step": 3910
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.06937126649817552,
+      "learning_rate": 0.0006508279100128605,
+      "loss": 1.3472,
+      "step": 3911
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.06116336612759358,
+      "learning_rate": 0.0006506619040033804,
+      "loss": 1.4974,
+      "step": 3912
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.06792960629230754,
+      "learning_rate": 0.0006504958797256249,
+      "loss": 1.5023,
+      "step": 3913
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.07305324582122641,
+      "learning_rate": 0.0006503298371997252,
+      "loss": 1.3766,
+      "step": 3914
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.06663241510812803,
+      "learning_rate": 0.0006501637764458145,
+      "loss": 1.3254,
+      "step": 3915
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0649843862468896,
+      "learning_rate": 0.0006499976974840281,
+      "loss": 1.4063,
+      "step": 3916
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.06951734575941149,
+      "learning_rate": 0.0006498316003345039,
+      "loss": 1.3195,
+      "step": 3917
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.062361862591796616,
+      "learning_rate": 0.0006496654850173815,
+      "loss": 1.3639,
+      "step": 3918
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.07616360631714768,
+      "learning_rate": 0.0006494993515528031,
+      "loss": 1.516,
+      "step": 3919
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.06141123114164976,
+      "learning_rate": 0.0006493331999609132,
+      "loss": 1.4371,
+      "step": 3920
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.06802623499257793,
+      "learning_rate": 0.0006491670302618576,
+      "loss": 1.4329,
+      "step": 3921
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0772677461754373,
+      "learning_rate": 0.0006490008424757855,
+      "loss": 1.5025,
+      "step": 3922
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0743215556309416,
+      "learning_rate": 0.0006488346366228475,
+      "loss": 1.3276,
+      "step": 3923
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.07205238649167922,
+      "learning_rate": 0.0006486684127231967,
+      "loss": 1.4187,
+      "step": 3924
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0643706136749596,
+      "learning_rate": 0.0006485021707969882,
+      "loss": 1.42,
+      "step": 3925
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.06063894800539872,
+      "learning_rate": 0.0006483359108643798,
+      "loss": 1.3124,
+      "step": 3926
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.06558860865385341,
+      "learning_rate": 0.0006481696329455307,
+      "loss": 1.5174,
+      "step": 3927
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.07291103363275614,
+      "learning_rate": 0.0006480033370606027,
+      "loss": 1.5637,
+      "step": 3928
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.06763422148856137,
+      "learning_rate": 0.0006478370232297599,
+      "loss": 1.5378,
+      "step": 3929
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.06303247852882345,
+      "learning_rate": 0.0006476706914731683,
+      "loss": 1.3136,
+      "step": 3930
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.08152561088657095,
+      "learning_rate": 0.0006475043418109965,
+      "loss": 1.4433,
+      "step": 3931
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.06088011791680538,
+      "learning_rate": 0.0006473379742634144,
+      "loss": 1.3877,
+      "step": 3932
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.07318502818351857,
+      "learning_rate": 0.0006471715888505951,
+      "loss": 1.4242,
+      "step": 3933
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.06649810034662097,
+      "learning_rate": 0.0006470051855927134,
+      "loss": 1.4482,
+      "step": 3934
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.07058262099194323,
+      "learning_rate": 0.000646838764509946,
+      "loss": 1.4439,
+      "step": 3935
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.06514689463950991,
+      "learning_rate": 0.0006466723256224723,
+      "loss": 1.4394,
+      "step": 3936
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.08531249688658517,
+      "learning_rate": 0.0006465058689504733,
+      "loss": 1.4133,
+      "step": 3937
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.09075806648354802,
+      "learning_rate": 0.0006463393945141327,
+      "loss": 1.4447,
+      "step": 3938
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.06925448902528702,
+      "learning_rate": 0.0006461729023336361,
+      "loss": 1.4942,
+      "step": 3939
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.08057013127356427,
+      "learning_rate": 0.0006460063924291709,
+      "loss": 1.4033,
+      "step": 3940
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0744006256177114,
+      "learning_rate": 0.0006458398648209274,
+      "loss": 1.4624,
+      "step": 3941
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.08683211100878094,
+      "learning_rate": 0.0006456733195290976,
+      "loss": 1.3895,
+      "step": 3942
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.07420107511822488,
+      "learning_rate": 0.0006455067565738755,
+      "loss": 1.4124,
+      "step": 3943
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.08515285186593229,
+      "learning_rate": 0.0006453401759754574,
+      "loss": 1.3782,
+      "step": 3944
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.07716068943132816,
+      "learning_rate": 0.0006451735777540421,
+      "loss": 1.4545,
+      "step": 3945
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.06785250313095319,
+      "learning_rate": 0.0006450069619298299,
+      "loss": 1.4342,
+      "step": 3946
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.08320823203786909,
+      "learning_rate": 0.0006448403285230238,
+      "loss": 1.3588,
+      "step": 3947
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.06851001898167915,
+      "learning_rate": 0.0006446736775538284,
+      "loss": 1.4696,
+      "step": 3948
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.07424712344584676,
+      "learning_rate": 0.0006445070090424507,
+      "loss": 1.368,
+      "step": 3949
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.08196924625145434,
+      "learning_rate": 0.0006443403230091001,
+      "loss": 1.5129,
+      "step": 3950
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.07066837376848595,
+      "learning_rate": 0.0006441736194739878,
+      "loss": 1.4107,
+      "step": 3951
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.07186081308313993,
+      "learning_rate": 0.0006440068984573271,
+      "loss": 1.3675,
+      "step": 3952
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.08857135342229146,
+      "learning_rate": 0.0006438401599793333,
+      "loss": 1.3819,
+      "step": 3953
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.07697014336064355,
+      "learning_rate": 0.0006436734040602244,
+      "loss": 1.4648,
+      "step": 3954
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0697777333276736,
+      "learning_rate": 0.0006435066307202197,
+      "loss": 1.5152,
+      "step": 3955
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.07935233015115574,
+      "learning_rate": 0.0006433398399795415,
+      "loss": 1.3593,
+      "step": 3956
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.07245156010257515,
+      "learning_rate": 0.0006431730318584135,
+      "loss": 1.5383,
+      "step": 3957
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.06912143950929635,
+      "learning_rate": 0.0006430062063770618,
+      "loss": 1.3255,
+      "step": 3958
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.06871716581567094,
+      "learning_rate": 0.0006428393635557146,
+      "loss": 1.377,
+      "step": 3959
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.07122351216156715,
+      "learning_rate": 0.000642672503414602,
+      "loss": 1.4079,
+      "step": 3960
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.07245998086323546,
+      "learning_rate": 0.0006425056259739566,
+      "loss": 1.5169,
+      "step": 3961
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.07059710490678917,
+      "learning_rate": 0.0006423387312540126,
+      "loss": 1.3649,
+      "step": 3962
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.07049724857677844,
+      "learning_rate": 0.0006421718192750069,
+      "loss": 1.451,
+      "step": 3963
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.06736834343072567,
+      "learning_rate": 0.000642004890057178,
+      "loss": 1.4212,
+      "step": 3964
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.07409561524105795,
+      "learning_rate": 0.0006418379436207664,
+      "loss": 1.5065,
+      "step": 3965
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.07636154384442553,
+      "learning_rate": 0.0006416709799860152,
+      "loss": 1.3027,
+      "step": 3966
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.07598801731212228,
+      "learning_rate": 0.0006415039991731694,
+      "loss": 1.4313,
+      "step": 3967
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.07390579715074876,
+      "learning_rate": 0.0006413370012024759,
+      "loss": 1.3851,
+      "step": 3968
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0717562024777435,
+      "learning_rate": 0.0006411699860941835,
+      "loss": 1.474,
+      "step": 3969
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.06909040510305756,
+      "learning_rate": 0.0006410029538685438,
+      "loss": 1.4227,
+      "step": 3970
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.06699697627532278,
+      "learning_rate": 0.0006408359045458099,
+      "loss": 1.5094,
+      "step": 3971
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.07224536624675192,
+      "learning_rate": 0.0006406688381462367,
+      "loss": 1.4341,
+      "step": 3972
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.07865011939575806,
+      "learning_rate": 0.0006405017546900822,
+      "loss": 1.4284,
+      "step": 3973
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.07174267385677884,
+      "learning_rate": 0.0006403346541976056,
+      "loss": 1.4537,
+      "step": 3974
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.06841466489928297,
+      "learning_rate": 0.0006401675366890682,
+      "loss": 1.4555,
+      "step": 3975
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.07168977043503581,
+      "learning_rate": 0.0006400004021847338,
+      "loss": 1.456,
+      "step": 3976
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.08316765009925761,
+      "learning_rate": 0.000639833250704868,
+      "loss": 1.5264,
+      "step": 3977
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.06598310254941155,
+      "learning_rate": 0.0006396660822697383,
+      "loss": 1.2632,
+      "step": 3978
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.06861447351985102,
+      "learning_rate": 0.000639498896899615,
+      "loss": 1.3635,
+      "step": 3979
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.06990865790262073,
+      "learning_rate": 0.0006393316946147692,
+      "loss": 1.3088,
+      "step": 3980
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.06726246864470845,
+      "learning_rate": 0.0006391644754354751,
+      "loss": 1.4502,
+      "step": 3981
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.07210289911724266,
+      "learning_rate": 0.0006389972393820087,
+      "loss": 1.4364,
+      "step": 3982
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0708583088350892,
+      "learning_rate": 0.0006388299864746477,
+      "loss": 1.4931,
+      "step": 3983
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.06838091953553109,
+      "learning_rate": 0.0006386627167336724,
+      "loss": 1.3996,
+      "step": 3984
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.07333635634297311,
+      "learning_rate": 0.0006384954301793647,
+      "loss": 1.424,
+      "step": 3985
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.06392704323596084,
+      "learning_rate": 0.0006383281268320083,
+      "loss": 1.4805,
+      "step": 3986
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.07239214592386142,
+      "learning_rate": 0.0006381608067118898,
+      "loss": 1.4503,
+      "step": 3987
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.08377596451570307,
+      "learning_rate": 0.0006379934698392972,
+      "loss": 1.4575,
+      "step": 3988
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.06512746205390593,
+      "learning_rate": 0.0006378261162345207,
+      "loss": 1.4052,
+      "step": 3989
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.06566868198185125,
+      "learning_rate": 0.0006376587459178525,
+      "loss": 1.4508,
+      "step": 3990
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.07464547519704835,
+      "learning_rate": 0.0006374913589095865,
+      "loss": 1.3952,
+      "step": 3991
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.07711648505702222,
+      "learning_rate": 0.0006373239552300194,
+      "loss": 1.3761,
+      "step": 3992
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.06850586335147511,
+      "learning_rate": 0.0006371565348994492,
+      "loss": 1.4297,
+      "step": 3993
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.07199508377082078,
+      "learning_rate": 0.0006369890979381765,
+      "loss": 1.5062,
+      "step": 3994
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.06794921663514014,
+      "learning_rate": 0.0006368216443665033,
+      "loss": 1.3346,
+      "step": 3995
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.06926739185641112,
+      "learning_rate": 0.000636654174204734,
+      "loss": 1.48,
+      "step": 3996
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.06888232643275677,
+      "learning_rate": 0.0006364866874731749,
+      "loss": 1.2917,
+      "step": 3997
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.07729538672205612,
+      "learning_rate": 0.0006363191841921344,
+      "loss": 1.4689,
+      "step": 3998
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0666543134995083,
+      "learning_rate": 0.0006361516643819229,
+      "loss": 1.3915,
+      "step": 3999
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.08126873599265884,
+      "learning_rate": 0.0006359841280628529,
+      "loss": 1.4609,
+      "step": 4000
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0778814602771884,
+      "learning_rate": 0.0006358165752552383,
+      "loss": 1.5199,
+      "step": 4001
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.06526420166226415,
+      "learning_rate": 0.0006356490059793959,
+      "loss": 1.4386,
+      "step": 4002
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.07223692530289383,
+      "learning_rate": 0.0006354814202556437,
+      "loss": 1.3302,
+      "step": 4003
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.06565906223942776,
+      "learning_rate": 0.0006353138181043024,
+      "loss": 1.4046,
+      "step": 4004
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.06969485520321976,
+      "learning_rate": 0.000635146199545694,
+      "loss": 1.4483,
+      "step": 4005
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0769829877285532,
+      "learning_rate": 0.000634978564600143,
+      "loss": 1.5446,
+      "step": 4006
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.06600204951040782,
+      "learning_rate": 0.0006348109132879758,
+      "loss": 1.4368,
+      "step": 4007
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.06277465263963822,
+      "learning_rate": 0.0006346432456295206,
+      "loss": 1.4953,
+      "step": 4008
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.070869092214546,
+      "learning_rate": 0.0006344755616451075,
+      "loss": 1.3439,
+      "step": 4009
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0851992383775605,
+      "learning_rate": 0.0006343078613550692,
+      "loss": 1.2783,
+      "step": 4010
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.06873785217748689,
+      "learning_rate": 0.0006341401447797395,
+      "loss": 1.4184,
+      "step": 4011
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.08105491442820943,
+      "learning_rate": 0.0006339724119394548,
+      "loss": 1.4541,
+      "step": 4012
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.09058309410357024,
+      "learning_rate": 0.0006338046628545533,
+      "loss": 1.3389,
+      "step": 4013
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.07765474200223683,
+      "learning_rate": 0.0006336368975453751,
+      "loss": 1.3905,
+      "step": 4014
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.08691514341346585,
+      "learning_rate": 0.0006334691160322625,
+      "loss": 1.3661,
+      "step": 4015
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.07644850905371951,
+      "learning_rate": 0.0006333013183355594,
+      "loss": 1.4375,
+      "step": 4016
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.06857380019815038,
+      "learning_rate": 0.0006331335044756118,
+      "loss": 1.3521,
+      "step": 4017
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.08305252497824059,
+      "learning_rate": 0.0006329656744727679,
+      "loss": 1.3916,
+      "step": 4018
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.08124314368644833,
+      "learning_rate": 0.0006327978283473775,
+      "loss": 1.3291,
+      "step": 4019
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0705597696963439,
+      "learning_rate": 0.0006326299661197925,
+      "loss": 1.3706,
+      "step": 4020
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.07269778983693065,
+      "learning_rate": 0.000632462087810367,
+      "loss": 1.3945,
+      "step": 4021
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.07146559700538933,
+      "learning_rate": 0.0006322941934394568,
+      "loss": 1.533,
+      "step": 4022
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.08857351618785715,
+      "learning_rate": 0.0006321262830274192,
+      "loss": 1.4558,
+      "step": 4023
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.07938377494020217,
+      "learning_rate": 0.0006319583565946147,
+      "loss": 1.3931,
+      "step": 4024
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.07280102266618659,
+      "learning_rate": 0.0006317904141614043,
+      "loss": 1.3522,
+      "step": 4025
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.07857482129301427,
+      "learning_rate": 0.0006316224557481518,
+      "loss": 1.4833,
+      "step": 4026
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.06592901984933465,
+      "learning_rate": 0.0006314544813752229,
+      "loss": 1.4067,
+      "step": 4027
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.07367040385418791,
+      "learning_rate": 0.0006312864910629848,
+      "loss": 1.4827,
+      "step": 4028
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0735499496321778,
+      "learning_rate": 0.0006311184848318071,
+      "loss": 1.4299,
+      "step": 4029
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.07313825624311018,
+      "learning_rate": 0.0006309504627020611,
+      "loss": 1.4739,
+      "step": 4030
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.06941465489720243,
+      "learning_rate": 0.0006307824246941199,
+      "loss": 1.4239,
+      "step": 4031
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.07456618633064974,
+      "learning_rate": 0.0006306143708283592,
+      "loss": 1.3066,
+      "step": 4032
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0651709204419574,
+      "learning_rate": 0.0006304463011251554,
+      "loss": 1.3473,
+      "step": 4033
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.06531237023428524,
+      "learning_rate": 0.000630278215604888,
+      "loss": 1.5217,
+      "step": 4034
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.08111952291273701,
+      "learning_rate": 0.0006301101142879378,
+      "loss": 1.302,
+      "step": 4035
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.07363733510463709,
+      "learning_rate": 0.0006299419971946876,
+      "loss": 1.446,
+      "step": 4036
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.07311494980959608,
+      "learning_rate": 0.0006297738643455224,
+      "loss": 1.4033,
+      "step": 4037
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.06715207919719457,
+      "learning_rate": 0.0006296057157608287,
+      "loss": 1.5365,
+      "step": 4038
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.06925573041694051,
+      "learning_rate": 0.0006294375514609951,
+      "loss": 1.3796,
+      "step": 4039
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0819579049167436,
+      "learning_rate": 0.0006292693714664122,
+      "loss": 1.416,
+      "step": 4040
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.06194321911570573,
+      "learning_rate": 0.0006291011757974722,
+      "loss": 1.4414,
+      "step": 4041
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.07743332118321378,
+      "learning_rate": 0.0006289329644745698,
+      "loss": 1.4206,
+      "step": 4042
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.07286834756737166,
+      "learning_rate": 0.0006287647375181009,
+      "loss": 1.4865,
+      "step": 4043
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.06525179714561152,
+      "learning_rate": 0.0006285964949484637,
+      "loss": 1.2784,
+      "step": 4044
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.07796288896846361,
+      "learning_rate": 0.0006284282367860579,
+      "loss": 1.4092,
+      "step": 4045
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.07380537860259485,
+      "learning_rate": 0.0006282599630512858,
+      "loss": 1.5595,
+      "step": 4046
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.07290283264188771,
+      "learning_rate": 0.000628091673764551,
+      "loss": 1.4883,
+      "step": 4047
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.07557073453800928,
+      "learning_rate": 0.0006279233689462591,
+      "loss": 1.3976,
+      "step": 4048
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.07190307484813235,
+      "learning_rate": 0.0006277550486168177,
+      "loss": 1.4121,
+      "step": 4049
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.07124234491193457,
+      "learning_rate": 0.0006275867127966363,
+      "loss": 1.3455,
+      "step": 4050
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.07187076898229122,
+      "learning_rate": 0.000627418361506126,
+      "loss": 1.3166,
+      "step": 4051
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.06811580777354202,
+      "learning_rate": 0.0006272499947657002,
+      "loss": 1.5047,
+      "step": 4052
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.07519583887683878,
+      "learning_rate": 0.000627081612595774,
+      "loss": 1.4374,
+      "step": 4053
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0658215338355946,
+      "learning_rate": 0.0006269132150167638,
+      "loss": 1.4198,
+      "step": 4054
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.06636046398423283,
+      "learning_rate": 0.0006267448020490889,
+      "loss": 1.5282,
+      "step": 4055
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0756700595936872,
+      "learning_rate": 0.0006265763737131698,
+      "loss": 1.4572,
+      "step": 4056
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.09031156746298717,
+      "learning_rate": 0.000626407930029429,
+      "loss": 1.4109,
+      "step": 4057
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.07865278513159098,
+      "learning_rate": 0.0006262394710182909,
+      "loss": 1.3605,
+      "step": 4058
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.06988033127417924,
+      "learning_rate": 0.0006260709967001816,
+      "loss": 1.4146,
+      "step": 4059
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.07532677831427176,
+      "learning_rate": 0.0006259025070955295,
+      "loss": 1.4004,
+      "step": 4060
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.07109168814063885,
+      "learning_rate": 0.0006257340022247643,
+      "loss": 1.5781,
+      "step": 4061
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.06307341769528307,
+      "learning_rate": 0.0006255654821083178,
+      "loss": 1.3837,
+      "step": 4062
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0657165372516695,
+      "learning_rate": 0.0006253969467666238,
+      "loss": 1.4479,
+      "step": 4063
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0767516467964421,
+      "learning_rate": 0.0006252283962201177,
+      "loss": 1.3942,
+      "step": 4064
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.07047450174961416,
+      "learning_rate": 0.0006250598304892368,
+      "loss": 1.4335,
+      "step": 4065
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0726381170938812,
+      "learning_rate": 0.0006248912495944203,
+      "loss": 1.4966,
+      "step": 4066
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.07488327336903305,
+      "learning_rate": 0.0006247226535561092,
+      "loss": 1.4666,
+      "step": 4067
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.06558892902830106,
+      "learning_rate": 0.0006245540423947462,
+      "loss": 1.4455,
+      "step": 4068
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.07251446239916863,
+      "learning_rate": 0.0006243854161307765,
+      "loss": 1.5224,
+      "step": 4069
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.08316498630023138,
+      "learning_rate": 0.0006242167747846461,
+      "loss": 1.5017,
+      "step": 4070
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.07975250717728095,
+      "learning_rate": 0.0006240481183768036,
+      "loss": 1.3195,
+      "step": 4071
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.07204884155011891,
+      "learning_rate": 0.000623879446927699,
+      "loss": 1.4155,
+      "step": 4072
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.07334104065864457,
+      "learning_rate": 0.0006237107604577843,
+      "loss": 1.3981,
+      "step": 4073
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.07842758263483142,
+      "learning_rate": 0.0006235420589875136,
+      "loss": 1.4745,
+      "step": 4074
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0705062668239952,
+      "learning_rate": 0.0006233733425373422,
+      "loss": 1.4089,
+      "step": 4075
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.07756434672164429,
+      "learning_rate": 0.0006232046111277277,
+      "loss": 1.5183,
+      "step": 4076
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.07135757491952958,
+      "learning_rate": 0.0006230358647791294,
+      "loss": 1.5568,
+      "step": 4077
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.08626731930825417,
+      "learning_rate": 0.0006228671035120082,
+      "loss": 1.3864,
+      "step": 4078
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.07264549223156583,
+      "learning_rate": 0.0006226983273468273,
+      "loss": 1.4313,
+      "step": 4079
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.06987639541786198,
+      "learning_rate": 0.0006225295363040511,
+      "loss": 1.361,
+      "step": 4080
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.06641572646235297,
+      "learning_rate": 0.0006223607304041462,
+      "loss": 1.5163,
+      "step": 4081
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.06436519422690486,
+      "learning_rate": 0.0006221919096675808,
+      "loss": 1.4704,
+      "step": 4082
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.07388014001773682,
+      "learning_rate": 0.0006220230741148253,
+      "loss": 1.5568,
+      "step": 4083
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.06842339021724836,
+      "learning_rate": 0.0006218542237663513,
+      "loss": 1.4529,
+      "step": 4084
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.07737167494281082,
+      "learning_rate": 0.0006216853586426325,
+      "loss": 1.4532,
+      "step": 4085
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.08744856475360263,
+      "learning_rate": 0.0006215164787641446,
+      "loss": 1.3995,
+      "step": 4086
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.08103564337406777,
+      "learning_rate": 0.0006213475841513646,
+      "loss": 1.2311,
+      "step": 4087
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.07470751509790251,
+      "learning_rate": 0.0006211786748247716,
+      "loss": 1.4645,
+      "step": 4088
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.06984095772926356,
+      "learning_rate": 0.0006210097508048466,
+      "loss": 1.3588,
+      "step": 4089
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.07684632265692232,
+      "learning_rate": 0.0006208408121120723,
+      "loss": 1.432,
+      "step": 4090
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0648317073790488,
+      "learning_rate": 0.0006206718587669326,
+      "loss": 1.3795,
+      "step": 4091
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.06610090338490285,
+      "learning_rate": 0.0006205028907899143,
+      "loss": 1.4399,
+      "step": 4092
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.06986369171000673,
+      "learning_rate": 0.0006203339082015048,
+      "loss": 1.3131,
+      "step": 4093
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.06672274588600151,
+      "learning_rate": 0.0006201649110221943,
+      "loss": 1.4506,
+      "step": 4094
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.07641605904306757,
+      "learning_rate": 0.000619995899272474,
+      "loss": 1.4011,
+      "step": 4095
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.07299058614515767,
+      "learning_rate": 0.0006198268729728371,
+      "loss": 1.3871,
+      "step": 4096
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.06910804786567874,
+      "learning_rate": 0.0006196578321437789,
+      "loss": 1.2873,
+      "step": 4097
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0663791199619923,
+      "learning_rate": 0.000619488776805796,
+      "loss": 1.4373,
+      "step": 4098
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.06725867211121205,
+      "learning_rate": 0.0006193197069793869,
+      "loss": 1.4147,
+      "step": 4099
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.06213689170540425,
+      "learning_rate": 0.000619150622685052,
+      "loss": 1.4147,
+      "step": 4100
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.058435811328203476,
+      "learning_rate": 0.0006189815239432935,
+      "loss": 1.4477,
+      "step": 4101
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.06735167267567684,
+      "learning_rate": 0.0006188124107746148,
+      "loss": 1.3274,
+      "step": 4102
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.07419176458112457,
+      "learning_rate": 0.0006186432831995218,
+      "loss": 1.3537,
+      "step": 4103
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.07341471855142019,
+      "learning_rate": 0.0006184741412385217,
+      "loss": 1.5122,
+      "step": 4104
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.06830118546681271,
+      "learning_rate": 0.0006183049849121233,
+      "loss": 1.3571,
+      "step": 4105
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0636175844946524,
+      "learning_rate": 0.000618135814240838,
+      "loss": 1.3388,
+      "step": 4106
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.06988407450517839,
+      "learning_rate": 0.0006179666292451775,
+      "loss": 1.4889,
+      "step": 4107
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.06755098834408893,
+      "learning_rate": 0.0006177974299456568,
+      "loss": 1.4158,
+      "step": 4108
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.07359010079927847,
+      "learning_rate": 0.0006176282163627917,
+      "loss": 1.4536,
+      "step": 4109
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.08025899963759743,
+      "learning_rate": 0.0006174589885170995,
+      "loss": 1.5464,
+      "step": 4110
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.08035685133786531,
+      "learning_rate": 0.0006172897464291004,
+      "loss": 1.3692,
+      "step": 4111
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.06498891842009204,
+      "learning_rate": 0.000617120490119315,
+      "loss": 1.5264,
+      "step": 4112
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.06996794393706789,
+      "learning_rate": 0.0006169512196082663,
+      "loss": 1.484,
+      "step": 4113
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.07656932570480635,
+      "learning_rate": 0.0006167819349164791,
+      "loss": 1.4086,
+      "step": 4114
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.07442145054353973,
+      "learning_rate": 0.0006166126360644797,
+      "loss": 1.3669,
+      "step": 4115
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.06996384412019334,
+      "learning_rate": 0.0006164433230727962,
+      "loss": 1.4832,
+      "step": 4116
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.07644256155086539,
+      "learning_rate": 0.0006162739959619583,
+      "loss": 1.4555,
+      "step": 4117
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.06486478925259731,
+      "learning_rate": 0.0006161046547524976,
+      "loss": 1.486,
+      "step": 4118
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.07282495629633286,
+      "learning_rate": 0.0006159352994649469,
+      "loss": 1.4068,
+      "step": 4119
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0756036520808425,
+      "learning_rate": 0.0006157659301198418,
+      "loss": 1.5446,
+      "step": 4120
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.07928837539192085,
+      "learning_rate": 0.0006155965467377186,
+      "loss": 1.3833,
+      "step": 4121
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.07648852141287911,
+      "learning_rate": 0.0006154271493391155,
+      "loss": 1.4019,
+      "step": 4122
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.08129914315194162,
+      "learning_rate": 0.0006152577379445725,
+      "loss": 1.4999,
+      "step": 4123
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.06435353991840793,
+      "learning_rate": 0.0006150883125746314,
+      "loss": 1.429,
+      "step": 4124
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.07344269169809117,
+      "learning_rate": 0.0006149188732498357,
+      "loss": 1.4553,
+      "step": 4125
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.06969141501711071,
+      "learning_rate": 0.0006147494199907305,
+      "loss": 1.3447,
+      "step": 4126
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.07913855361793129,
+      "learning_rate": 0.0006145799528178625,
+      "loss": 1.4756,
+      "step": 4127
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.07453073475167775,
+      "learning_rate": 0.0006144104717517802,
+      "loss": 1.5231,
+      "step": 4128
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.07847315579126765,
+      "learning_rate": 0.0006142409768130338,
+      "loss": 1.5112,
+      "step": 4129
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.06889484864529667,
+      "learning_rate": 0.0006140714680221749,
+      "loss": 1.3361,
+      "step": 4130
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.07564516476238581,
+      "learning_rate": 0.0006139019453997575,
+      "loss": 1.3954,
+      "step": 4131
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.06694427175894813,
+      "learning_rate": 0.0006137324089663365,
+      "loss": 1.4801,
+      "step": 4132
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.06993820649694431,
+      "learning_rate": 0.0006135628587424687,
+      "loss": 1.3924,
+      "step": 4133
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.06788981217297224,
+      "learning_rate": 0.0006133932947487129,
+      "loss": 1.3172,
+      "step": 4134
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.07762909048019521,
+      "learning_rate": 0.0006132237170056291,
+      "loss": 1.3408,
+      "step": 4135
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.07190854220257663,
+      "learning_rate": 0.0006130541255337791,
+      "loss": 1.3979,
+      "step": 4136
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.07586277240340727,
+      "learning_rate": 0.0006128845203537269,
+      "loss": 1.3706,
+      "step": 4137
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.08398736072378425,
+      "learning_rate": 0.0006127149014860374,
+      "loss": 1.4149,
+      "step": 4138
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.07868961218092786,
+      "learning_rate": 0.0006125452689512774,
+      "loss": 1.4638,
+      "step": 4139
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.06358548319471856,
+      "learning_rate": 0.0006123756227700155,
+      "loss": 1.3557,
+      "step": 4140
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.06691413699984927,
+      "learning_rate": 0.000612205962962822,
+      "loss": 1.4534,
+      "step": 4141
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.06411817333259076,
+      "learning_rate": 0.0006120362895502687,
+      "loss": 1.4033,
+      "step": 4142
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.061015608518283186,
+      "learning_rate": 0.0006118666025529292,
+      "loss": 1.4078,
+      "step": 4143
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.06242891829313593,
+      "learning_rate": 0.0006116969019913781,
+      "loss": 1.6057,
+      "step": 4144
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07128735272434375,
+      "learning_rate": 0.0006115271878861928,
+      "loss": 1.5334,
+      "step": 4145
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.06366021932361336,
+      "learning_rate": 0.0006113574602579515,
+      "loss": 1.376,
+      "step": 4146
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07725408870783299,
+      "learning_rate": 0.0006111877191272339,
+      "loss": 1.3857,
+      "step": 4147
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07624293185866599,
+      "learning_rate": 0.0006110179645146224,
+      "loss": 1.3786,
+      "step": 4148
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.06575853794546838,
+      "learning_rate": 0.0006108481964407,
+      "loss": 1.3347,
+      "step": 4149
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07375952009354378,
+      "learning_rate": 0.0006106784149260513,
+      "loss": 1.4017,
+      "step": 4150
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07059410446959481,
+      "learning_rate": 0.0006105086199912635,
+      "loss": 1.3164,
+      "step": 4151
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07501949153880226,
+      "learning_rate": 0.0006103388116569244,
+      "loss": 1.5203,
+      "step": 4152
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.06371702168963504,
+      "learning_rate": 0.000610168989943624,
+      "loss": 1.4431,
+      "step": 4153
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.06411394548266706,
+      "learning_rate": 0.0006099991548719539,
+      "loss": 1.5797,
+      "step": 4154
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07495407167224134,
+      "learning_rate": 0.000609829306462507,
+      "loss": 1.4326,
+      "step": 4155
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.06631718739285854,
+      "learning_rate": 0.0006096594447358778,
+      "loss": 1.4092,
+      "step": 4156
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.09938044865046552,
+      "learning_rate": 0.0006094895697126631,
+      "loss": 1.5385,
+      "step": 4157
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07461912150158716,
+      "learning_rate": 0.0006093196814134605,
+      "loss": 1.4419,
+      "step": 4158
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07312130807661461,
+      "learning_rate": 0.0006091497798588699,
+      "loss": 1.4032,
+      "step": 4159
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07582986448445779,
+      "learning_rate": 0.000608979865069492,
+      "loss": 1.4466,
+      "step": 4160
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07880974884557526,
+      "learning_rate": 0.0006088099370659297,
+      "loss": 1.4004,
+      "step": 4161
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07458495001160687,
+      "learning_rate": 0.0006086399958687875,
+      "loss": 1.4518,
+      "step": 4162
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.06578258899976172,
+      "learning_rate": 0.0006084700414986712,
+      "loss": 1.3905,
+      "step": 4163
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.06810011832709272,
+      "learning_rate": 0.0006083000739761884,
+      "loss": 1.5071,
+      "step": 4164
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07909807518347882,
+      "learning_rate": 0.0006081300933219485,
+      "loss": 1.5617,
+      "step": 4165
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.06909964402324427,
+      "learning_rate": 0.0006079600995565617,
+      "loss": 1.4311,
+      "step": 4166
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07644740121314376,
+      "learning_rate": 0.0006077900927006408,
+      "loss": 1.5397,
+      "step": 4167
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07791533648735414,
+      "learning_rate": 0.0006076200727747994,
+      "loss": 1.4011,
+      "step": 4168
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07737760669772996,
+      "learning_rate": 0.0006074500397996533,
+      "loss": 1.3309,
+      "step": 4169
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.06895771259558378,
+      "learning_rate": 0.0006072799937958195,
+      "loss": 1.3748,
+      "step": 4170
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07586645523731778,
+      "learning_rate": 0.0006071099347839164,
+      "loss": 1.5111,
+      "step": 4171
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07236277970051411,
+      "learning_rate": 0.0006069398627845645,
+      "loss": 1.3702,
+      "step": 4172
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07276198083117988,
+      "learning_rate": 0.0006067697778183856,
+      "loss": 1.3674,
+      "step": 4173
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.06649290069396598,
+      "learning_rate": 0.0006065996799060031,
+      "loss": 1.406,
+      "step": 4174
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.085528893791016,
+      "learning_rate": 0.0006064295690680418,
+      "loss": 1.4379,
+      "step": 4175
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07056553332509218,
+      "learning_rate": 0.0006062594453251285,
+      "loss": 1.5733,
+      "step": 4176
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.06910344638946603,
+      "learning_rate": 0.0006060893086978909,
+      "loss": 1.5036,
+      "step": 4177
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.06867054468060886,
+      "learning_rate": 0.0006059191592069589,
+      "loss": 1.367,
+      "step": 4178
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07692617680362644,
+      "learning_rate": 0.0006057489968729638,
+      "loss": 1.516,
+      "step": 4179
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.06656309146573545,
+      "learning_rate": 0.0006055788217165383,
+      "loss": 1.4344,
+      "step": 4180
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.06873108247286198,
+      "learning_rate": 0.0006054086337583165,
+      "loss": 1.4402,
+      "step": 4181
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07460649744446407,
+      "learning_rate": 0.0006052384330189346,
+      "loss": 1.4741,
+      "step": 4182
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07849888148581544,
+      "learning_rate": 0.0006050682195190299,
+      "loss": 1.3884,
+      "step": 4183
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0652682809816125,
+      "learning_rate": 0.0006048979932792413,
+      "loss": 1.3857,
+      "step": 4184
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07701868904645039,
+      "learning_rate": 0.0006047277543202095,
+      "loss": 1.4581,
+      "step": 4185
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.06849774253433671,
+      "learning_rate": 0.0006045575026625762,
+      "loss": 1.3599,
+      "step": 4186
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.09158825757368835,
+      "learning_rate": 0.0006043872383269854,
+      "loss": 1.3715,
+      "step": 4187
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07279400872351754,
+      "learning_rate": 0.000604216961334082,
+      "loss": 1.3506,
+      "step": 4188
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.06912852463603696,
+      "learning_rate": 0.0006040466717045127,
+      "loss": 1.4299,
+      "step": 4189
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.06763599779398562,
+      "learning_rate": 0.0006038763694589258,
+      "loss": 1.3763,
+      "step": 4190
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.08951066042952609,
+      "learning_rate": 0.0006037060546179712,
+      "loss": 1.4625,
+      "step": 4191
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.06326259809844488,
+      "learning_rate": 0.0006035357272022997,
+      "loss": 1.4229,
+      "step": 4192
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07451116529186678,
+      "learning_rate": 0.0006033653872325644,
+      "loss": 1.4435,
+      "step": 4193
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07508265893694721,
+      "learning_rate": 0.0006031950347294196,
+      "loss": 1.4549,
+      "step": 4194
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07826032094894797,
+      "learning_rate": 0.0006030246697135209,
+      "loss": 1.4891,
+      "step": 4195
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07645274230350402,
+      "learning_rate": 0.000602854292205526,
+      "loss": 1.3029,
+      "step": 4196
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.06920033529496816,
+      "learning_rate": 0.0006026839022260935,
+      "loss": 1.412,
+      "step": 4197
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.06704934333095056,
+      "learning_rate": 0.0006025134997958839,
+      "loss": 1.2994,
+      "step": 4198
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07423418599780926,
+      "learning_rate": 0.000602343084935559,
+      "loss": 1.3095,
+      "step": 4199
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07390960821320829,
+      "learning_rate": 0.0006021726576657822,
+      "loss": 1.3535,
+      "step": 4200
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07476816449840105,
+      "learning_rate": 0.0006020022180072184,
+      "loss": 1.4619,
+      "step": 4201
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07002103516627521,
+      "learning_rate": 0.0006018317659805341,
+      "loss": 1.3879,
+      "step": 4202
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0773920967736904,
+      "learning_rate": 0.000601661301606397,
+      "loss": 1.5078,
+      "step": 4203
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07217551586812339,
+      "learning_rate": 0.0006014908249054767,
+      "loss": 1.4657,
+      "step": 4204
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07543631118107905,
+      "learning_rate": 0.000601320335898444,
+      "loss": 1.4275,
+      "step": 4205
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07223024017477711,
+      "learning_rate": 0.0006011498346059712,
+      "loss": 1.3861,
+      "step": 4206
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07472220435048994,
+      "learning_rate": 0.0006009793210487322,
+      "loss": 1.4108,
+      "step": 4207
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07639838356520745,
+      "learning_rate": 0.0006008087952474024,
+      "loss": 1.4983,
+      "step": 4208
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0701511264724409,
+      "learning_rate": 0.0006006382572226587,
+      "loss": 1.3362,
+      "step": 4209
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.06770431572735086,
+      "learning_rate": 0.0006004677069951792,
+      "loss": 1.4401,
+      "step": 4210
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07739188467731939,
+      "learning_rate": 0.0006002971445856441,
+      "loss": 1.4421,
+      "step": 4211
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07539465339428007,
+      "learning_rate": 0.0006001265700147344,
+      "loss": 1.5035,
+      "step": 4212
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07078364592406552,
+      "learning_rate": 0.0005999559833031327,
+      "loss": 1.3512,
+      "step": 4213
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07865502303660855,
+      "learning_rate": 0.0005997853844715237,
+      "loss": 1.4592,
+      "step": 4214
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.06829634894502473,
+      "learning_rate": 0.0005996147735405925,
+      "loss": 1.3703,
+      "step": 4215
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0781852996100699,
+      "learning_rate": 0.0005994441505310269,
+      "loss": 1.3338,
+      "step": 4216
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07062610585794313,
+      "learning_rate": 0.0005992735154635151,
+      "loss": 1.4492,
+      "step": 4217
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07909815760336844,
+      "learning_rate": 0.0005991028683587471,
+      "loss": 1.3787,
+      "step": 4218
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07063792835909273,
+      "learning_rate": 0.000598932209237415,
+      "loss": 1.363,
+      "step": 4219
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.08213431406304056,
+      "learning_rate": 0.0005987615381202112,
+      "loss": 1.3396,
+      "step": 4220
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07766459822155773,
+      "learning_rate": 0.0005985908550278305,
+      "loss": 1.4481,
+      "step": 4221
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07358174371250564,
+      "learning_rate": 0.0005984201599809689,
+      "loss": 1.37,
+      "step": 4222
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.06590233723419557,
+      "learning_rate": 0.0005982494530003233,
+      "loss": 1.3558,
+      "step": 4223
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0682442044026668,
+      "learning_rate": 0.0005980787341065929,
+      "loss": 1.4652,
+      "step": 4224
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07891003356492753,
+      "learning_rate": 0.000597908003320478,
+      "loss": 1.4018,
+      "step": 4225
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.06639971946412793,
+      "learning_rate": 0.00059773726066268,
+      "loss": 1.3917,
+      "step": 4226
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07335602496727649,
+      "learning_rate": 0.0005975665061539022,
+      "loss": 1.3812,
+      "step": 4227
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07410909291067329,
+      "learning_rate": 0.0005973957398148493,
+      "loss": 1.6306,
+      "step": 4228
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.06263196785177541,
+      "learning_rate": 0.000597224961666227,
+      "loss": 1.4055,
+      "step": 4229
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.06865685568182275,
+      "learning_rate": 0.0005970541717287431,
+      "loss": 1.4365,
+      "step": 4230
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07007647676666574,
+      "learning_rate": 0.0005968833700231062,
+      "loss": 1.564,
+      "step": 4231
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0759327624640143,
+      "learning_rate": 0.0005967125565700265,
+      "loss": 1.3756,
+      "step": 4232
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0683659235338103,
+      "learning_rate": 0.0005965417313902162,
+      "loss": 1.3516,
+      "step": 4233
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.059114031755854154,
+      "learning_rate": 0.000596370894504388,
+      "loss": 1.3317,
+      "step": 4234
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.06596578435593728,
+      "learning_rate": 0.0005962000459332566,
+      "loss": 1.3701,
+      "step": 4235
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.056174330967718004,
+      "learning_rate": 0.000596029185697538,
+      "loss": 1.4079,
+      "step": 4236
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.06556568495064191,
+      "learning_rate": 0.0005958583138179494,
+      "loss": 1.3345,
+      "step": 4237
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.07692922309310339,
+      "learning_rate": 0.00059568743031521,
+      "loss": 1.4097,
+      "step": 4238
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.06730920807440371,
+      "learning_rate": 0.0005955165352100398,
+      "loss": 1.3266,
+      "step": 4239
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.06838790689808874,
+      "learning_rate": 0.0005953456285231602,
+      "loss": 1.433,
+      "step": 4240
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.06915277142703034,
+      "learning_rate": 0.0005951747102752946,
+      "loss": 1.429,
+      "step": 4241
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.07277351313611133,
+      "learning_rate": 0.0005950037804871673,
+      "loss": 1.5549,
+      "step": 4242
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.06650552884193628,
+      "learning_rate": 0.0005948328391795038,
+      "loss": 1.3798,
+      "step": 4243
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.07550184977976046,
+      "learning_rate": 0.000594661886373032,
+      "loss": 1.3489,
+      "step": 4244
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.06434529850760629,
+      "learning_rate": 0.00059449092208848,
+      "loss": 1.5048,
+      "step": 4245
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.06134605286144569,
+      "learning_rate": 0.0005943199463465779,
+      "loss": 1.4264,
+      "step": 4246
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.07135379478202215,
+      "learning_rate": 0.000594148959168057,
+      "loss": 1.4708,
+      "step": 4247
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0657081744632789,
+      "learning_rate": 0.0005939779605736504,
+      "loss": 1.4203,
+      "step": 4248
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.09721619161803492,
+      "learning_rate": 0.0005938069505840919,
+      "loss": 1.5514,
+      "step": 4249
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.06960076559949502,
+      "learning_rate": 0.0005936359292201174,
+      "loss": 1.4348,
+      "step": 4250
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0681511654307289,
+      "learning_rate": 0.0005934648965024636,
+      "loss": 1.2808,
+      "step": 4251
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0665974511558386,
+      "learning_rate": 0.0005932938524518689,
+      "loss": 1.548,
+      "step": 4252
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.06975298195353812,
+      "learning_rate": 0.0005931227970890731,
+      "loss": 1.5667,
+      "step": 4253
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0671332270408465,
+      "learning_rate": 0.000592951730434817,
+      "loss": 1.3778,
+      "step": 4254
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.062144346033571184,
+      "learning_rate": 0.0005927806525098431,
+      "loss": 1.4931,
+      "step": 4255
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.06924031944664585,
+      "learning_rate": 0.0005926095633348952,
+      "loss": 1.4989,
+      "step": 4256
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.06798196988097031,
+      "learning_rate": 0.0005924384629307184,
+      "loss": 1.4264,
+      "step": 4257
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.07981962984856707,
+      "learning_rate": 0.0005922673513180596,
+      "loss": 1.379,
+      "step": 4258
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.07231588001633618,
+      "learning_rate": 0.0005920962285176661,
+      "loss": 1.3043,
+      "step": 4259
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0742394001366356,
+      "learning_rate": 0.0005919250945502874,
+      "loss": 1.4649,
+      "step": 4260
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0699345660413472,
+      "learning_rate": 0.0005917539494366741,
+      "loss": 1.3223,
+      "step": 4261
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.07213363643829698,
+      "learning_rate": 0.0005915827931975782,
+      "loss": 1.3135,
+      "step": 4262
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.07930439093632695,
+      "learning_rate": 0.0005914116258537527,
+      "loss": 1.423,
+      "step": 4263
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.07094944522631236,
+      "learning_rate": 0.0005912404474259526,
+      "loss": 1.4871,
+      "step": 4264
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.07994927967283738,
+      "learning_rate": 0.0005910692579349336,
+      "loss": 1.4013,
+      "step": 4265
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0739574577729969,
+      "learning_rate": 0.0005908980574014532,
+      "loss": 1.3338,
+      "step": 4266
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.07033576100419545,
+      "learning_rate": 0.00059072684584627,
+      "loss": 1.3731,
+      "step": 4267
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0693006408871248,
+      "learning_rate": 0.0005905556232901437,
+      "loss": 1.4151,
+      "step": 4268
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0771275790905268,
+      "learning_rate": 0.000590384389753836,
+      "loss": 1.3806,
+      "step": 4269
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.08802543077110558,
+      "learning_rate": 0.0005902131452581096,
+      "loss": 1.4195,
+      "step": 4270
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.07815218988772281,
+      "learning_rate": 0.0005900418898237281,
+      "loss": 1.3446,
+      "step": 4271
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.07661324405037143,
+      "learning_rate": 0.0005898706234714572,
+      "loss": 1.5356,
+      "step": 4272
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.07920864004503088,
+      "learning_rate": 0.0005896993462220633,
+      "loss": 1.3312,
+      "step": 4273
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.07351431361264203,
+      "learning_rate": 0.0005895280580963144,
+      "loss": 1.3353,
+      "step": 4274
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0829204899421725,
+      "learning_rate": 0.0005893567591149799,
+      "loss": 1.4148,
+      "step": 4275
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0731711710914952,
+      "learning_rate": 0.0005891854492988302,
+      "loss": 1.4031,
+      "step": 4276
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.07697900216587132,
+      "learning_rate": 0.0005890141286686375,
+      "loss": 1.4344,
+      "step": 4277
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.08406706969300401,
+      "learning_rate": 0.0005888427972451746,
+      "loss": 1.4329,
+      "step": 4278
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.07000902617829333,
+      "learning_rate": 0.0005886714550492163,
+      "loss": 1.5284,
+      "step": 4279
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.07299706209596549,
+      "learning_rate": 0.0005885001021015383,
+      "loss": 1.5595,
+      "step": 4280
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.06511331156240723,
+      "learning_rate": 0.0005883287384229181,
+      "loss": 1.4026,
+      "step": 4281
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.08615882506688809,
+      "learning_rate": 0.0005881573640341335,
+      "loss": 1.3818,
+      "step": 4282
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0665931298491308,
+      "learning_rate": 0.0005879859789559649,
+      "loss": 1.489,
+      "step": 4283
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.06736534693938794,
+      "learning_rate": 0.0005878145832091929,
+      "loss": 1.4695,
+      "step": 4284
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.07637517648261326,
+      "learning_rate": 0.0005876431768145997,
+      "loss": 1.4606,
+      "step": 4285
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.07295821930998996,
+      "learning_rate": 0.0005874717597929696,
+      "loss": 1.2971,
+      "step": 4286
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.06826614255036653,
+      "learning_rate": 0.0005873003321650868,
+      "loss": 1.4135,
+      "step": 4287
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.07459899080496685,
+      "learning_rate": 0.0005871288939517377,
+      "loss": 1.4705,
+      "step": 4288
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.07846593078913912,
+      "learning_rate": 0.00058695744517371,
+      "loss": 1.4887,
+      "step": 4289
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.07214161019843333,
+      "learning_rate": 0.0005867859858517922,
+      "loss": 1.3555,
+      "step": 4290
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.06799052813777913,
+      "learning_rate": 0.0005866145160067745,
+      "loss": 1.5964,
+      "step": 4291
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.06470531737409344,
+      "learning_rate": 0.000586443035659448,
+      "loss": 1.4667,
+      "step": 4292
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.06948444252266078,
+      "learning_rate": 0.0005862715448306055,
+      "loss": 1.4447,
+      "step": 4293
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.06420616680388852,
+      "learning_rate": 0.0005861000435410407,
+      "loss": 1.4562,
+      "step": 4294
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.07606806237403266,
+      "learning_rate": 0.0005859285318115488,
+      "loss": 1.4291,
+      "step": 4295
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.06970676132945107,
+      "learning_rate": 0.0005857570096629262,
+      "loss": 1.5096,
+      "step": 4296
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.07218587347413118,
+      "learning_rate": 0.0005855854771159706,
+      "loss": 1.4344,
+      "step": 4297
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0869640303145645,
+      "learning_rate": 0.0005854139341914808,
+      "loss": 1.4323,
+      "step": 4298
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.06296273663509053,
+      "learning_rate": 0.0005852423809102568,
+      "loss": 1.3476,
+      "step": 4299
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.06801967942797818,
+      "learning_rate": 0.0005850708172931005,
+      "loss": 1.4381,
+      "step": 4300
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.07313413676647226,
+      "learning_rate": 0.0005848992433608142,
+      "loss": 1.4988,
+      "step": 4301
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.07855268328588744,
+      "learning_rate": 0.0005847276591342021,
+      "loss": 1.3393,
+      "step": 4302
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.07208810569830114,
+      "learning_rate": 0.0005845560646340691,
+      "loss": 1.4611,
+      "step": 4303
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.07485263168250995,
+      "learning_rate": 0.0005843844598812218,
+      "loss": 1.394,
+      "step": 4304
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.06973519268508456,
+      "learning_rate": 0.0005842128448964676,
+      "loss": 1.3173,
+      "step": 4305
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.06435508387831129,
+      "learning_rate": 0.000584041219700616,
+      "loss": 1.4103,
+      "step": 4306
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0719765280377028,
+      "learning_rate": 0.0005838695843144766,
+      "loss": 1.3714,
+      "step": 4307
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.08143997254921763,
+      "learning_rate": 0.000583697938758861,
+      "loss": 1.5099,
+      "step": 4308
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.06992837500062335,
+      "learning_rate": 0.0005835262830545816,
+      "loss": 1.3621,
+      "step": 4309
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0667326641216811,
+      "learning_rate": 0.0005833546172224526,
+      "loss": 1.3966,
+      "step": 4310
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.07025203654887667,
+      "learning_rate": 0.0005831829412832888,
+      "loss": 1.3464,
+      "step": 4311
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.07092819341479226,
+      "learning_rate": 0.0005830112552579067,
+      "loss": 1.4383,
+      "step": 4312
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.06969893660218596,
+      "learning_rate": 0.0005828395591671238,
+      "loss": 1.3822,
+      "step": 4313
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.07472487467182264,
+      "learning_rate": 0.0005826678530317585,
+      "loss": 1.3483,
+      "step": 4314
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.08045792356361967,
+      "learning_rate": 0.0005824961368726311,
+      "loss": 1.2365,
+      "step": 4315
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.07539767278084446,
+      "learning_rate": 0.0005823244107105627,
+      "loss": 1.4309,
+      "step": 4316
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.06989912203654702,
+      "learning_rate": 0.0005821526745663758,
+      "loss": 1.3415,
+      "step": 4317
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.07123258473151052,
+      "learning_rate": 0.000581980928460894,
+      "loss": 1.3155,
+      "step": 4318
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.07321426972149589,
+      "learning_rate": 0.0005818091724149417,
+      "loss": 1.5035,
+      "step": 4319
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.06861095456012398,
+      "learning_rate": 0.0005816374064493453,
+      "loss": 1.4533,
+      "step": 4320
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.062160527409850376,
+      "learning_rate": 0.000581465630584932,
+      "loss": 1.5065,
+      "step": 4321
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.08033920097956411,
+      "learning_rate": 0.0005812938448425299,
+      "loss": 1.4659,
+      "step": 4322
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.06743837135831846,
+      "learning_rate": 0.0005811220492429692,
+      "loss": 1.4259,
+      "step": 4323
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.07239672168722948,
+      "learning_rate": 0.0005809502438070801,
+      "loss": 1.3405,
+      "step": 4324
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.07073635590191434,
+      "learning_rate": 0.000580778428555695,
+      "loss": 1.3189,
+      "step": 4325
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.07614227248503913,
+      "learning_rate": 0.0005806066035096468,
+      "loss": 1.3165,
+      "step": 4326
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.06700790849180441,
+      "learning_rate": 0.0005804347686897702,
+      "loss": 1.3281,
+      "step": 4327
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.06791159995686306,
+      "learning_rate": 0.0005802629241169005,
+      "loss": 1.5144,
+      "step": 4328
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.06581161360888663,
+      "learning_rate": 0.0005800910698118745,
+      "loss": 1.6056,
+      "step": 4329
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.07091361131731538,
+      "learning_rate": 0.0005799192057955303,
+      "loss": 1.3676,
+      "step": 4330
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.06892622529489716,
+      "learning_rate": 0.0005797473320887068,
+      "loss": 1.3971,
+      "step": 4331
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.06963628707063342,
+      "learning_rate": 0.0005795754487122444,
+      "loss": 1.4121,
+      "step": 4332
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.07610530573128409,
+      "learning_rate": 0.0005794035556869843,
+      "loss": 1.4506,
+      "step": 4333
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.06066084045830262,
+      "learning_rate": 0.0005792316530337696,
+      "loss": 1.5658,
+      "step": 4334
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0625241482545445,
+      "learning_rate": 0.0005790597407734437,
+      "loss": 1.3852,
+      "step": 4335
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.07533639979979685,
+      "learning_rate": 0.0005788878189268516,
+      "loss": 1.4158,
+      "step": 4336
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.063951064948048,
+      "learning_rate": 0.0005787158875148397,
+      "loss": 1.6017,
+      "step": 4337
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.07185074481823407,
+      "learning_rate": 0.0005785439465582548,
+      "loss": 1.3789,
+      "step": 4338
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.067738052717109,
+      "learning_rate": 0.0005783719960779458,
+      "loss": 1.2404,
+      "step": 4339
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.06693252733909952,
+      "learning_rate": 0.000578200036094762,
+      "loss": 1.4805,
+      "step": 4340
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0707406767223583,
+      "learning_rate": 0.0005780280666295543,
+      "loss": 1.3044,
+      "step": 4341
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.07245024813674909,
+      "learning_rate": 0.0005778560877031743,
+      "loss": 1.4834,
+      "step": 4342
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.07076343834104014,
+      "learning_rate": 0.0005776840993364754,
+      "loss": 1.4862,
+      "step": 4343
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0771692706601467,
+      "learning_rate": 0.0005775121015503117,
+      "loss": 1.5121,
+      "step": 4344
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.07575869372023414,
+      "learning_rate": 0.0005773400943655385,
+      "loss": 1.4231,
+      "step": 4345
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.08331985864502259,
+      "learning_rate": 0.0005771680778030122,
+      "loss": 1.4944,
+      "step": 4346
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0978213404908466,
+      "learning_rate": 0.0005769960518835902,
+      "loss": 1.385,
+      "step": 4347
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.07030500514431032,
+      "learning_rate": 0.0005768240166281317,
+      "loss": 1.384,
+      "step": 4348
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0731870634540761,
+      "learning_rate": 0.0005766519720574963,
+      "loss": 1.5135,
+      "step": 4349
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.07700046097610479,
+      "learning_rate": 0.0005764799181925449,
+      "loss": 1.468,
+      "step": 4350
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0820989017585397,
+      "learning_rate": 0.0005763078550541399,
+      "loss": 1.4766,
+      "step": 4351
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.06727729877440962,
+      "learning_rate": 0.0005761357826631443,
+      "loss": 1.4807,
+      "step": 4352
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0710779421387419,
+      "learning_rate": 0.0005759637010404226,
+      "loss": 1.4129,
+      "step": 4353
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0792314675429011,
+      "learning_rate": 0.0005757916102068402,
+      "loss": 1.5648,
+      "step": 4354
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0694154213276982,
+      "learning_rate": 0.000575619510183264,
+      "loss": 1.3538,
+      "step": 4355
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0635064723999925,
+      "learning_rate": 0.0005754474009905613,
+      "loss": 1.3959,
+      "step": 4356
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.06970467537810597,
+      "learning_rate": 0.0005752752826496012,
+      "loss": 1.4013,
+      "step": 4357
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.07260049255850706,
+      "learning_rate": 0.0005751031551812536,
+      "loss": 1.3132,
+      "step": 4358
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.06955971587200312,
+      "learning_rate": 0.0005749310186063892,
+      "loss": 1.3798,
+      "step": 4359
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0679171340858405,
+      "learning_rate": 0.0005747588729458809,
+      "loss": 1.526,
+      "step": 4360
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0673009342150572,
+      "learning_rate": 0.0005745867182206012,
+      "loss": 1.4129,
+      "step": 4361
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.07326363249587493,
+      "learning_rate": 0.000574414554451425,
+      "loss": 1.4206,
+      "step": 4362
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.06410449649673688,
+      "learning_rate": 0.0005742423816592274,
+      "loss": 1.3288,
+      "step": 4363
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.06985286145175268,
+      "learning_rate": 0.0005740701998648849,
+      "loss": 1.3431,
+      "step": 4364
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.07041595467515156,
+      "learning_rate": 0.0005738980090892757,
+      "loss": 1.4546,
+      "step": 4365
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.07074849648123872,
+      "learning_rate": 0.0005737258093532781,
+      "loss": 1.4942,
+      "step": 4366
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.06451965099343698,
+      "learning_rate": 0.0005735536006777716,
+      "loss": 1.4386,
+      "step": 4367
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.07518999046838014,
+      "learning_rate": 0.0005733813830836376,
+      "loss": 1.5297,
+      "step": 4368
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.06507127189100723,
+      "learning_rate": 0.000573209156591758,
+      "loss": 1.4292,
+      "step": 4369
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.061908429397981614,
+      "learning_rate": 0.0005730369212230157,
+      "loss": 1.4905,
+      "step": 4370
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.07486397178116465,
+      "learning_rate": 0.0005728646769982951,
+      "loss": 1.3875,
+      "step": 4371
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.06995650744403155,
+      "learning_rate": 0.0005726924239384809,
+      "loss": 1.3832,
+      "step": 4372
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.06642041686352557,
+      "learning_rate": 0.0005725201620644598,
+      "loss": 1.3663,
+      "step": 4373
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.06838526746301918,
+      "learning_rate": 0.0005723478913971191,
+      "loss": 1.4206,
+      "step": 4374
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.06994726686310161,
+      "learning_rate": 0.0005721756119573471,
+      "loss": 1.3621,
+      "step": 4375
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.07335027824922698,
+      "learning_rate": 0.0005720033237660334,
+      "loss": 1.38,
+      "step": 4376
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.07132320370362646,
+      "learning_rate": 0.0005718310268440683,
+      "loss": 1.5475,
+      "step": 4377
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.07136276677595751,
+      "learning_rate": 0.0005716587212123435,
+      "loss": 1.4212,
+      "step": 4378
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.06487403418814532,
+      "learning_rate": 0.0005714864068917519,
+      "loss": 1.4789,
+      "step": 4379
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0772161895483952,
+      "learning_rate": 0.0005713140839031871,
+      "loss": 1.4534,
+      "step": 4380
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.06981128959729192,
+      "learning_rate": 0.0005711417522675437,
+      "loss": 1.4698,
+      "step": 4381
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.07765049782121027,
+      "learning_rate": 0.0005709694120057174,
+      "loss": 1.3064,
+      "step": 4382
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.07241987185501046,
+      "learning_rate": 0.0005707970631386054,
+      "loss": 1.4083,
+      "step": 4383
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.07081369496009181,
+      "learning_rate": 0.0005706247056871053,
+      "loss": 1.4168,
+      "step": 4384
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.06703084002299411,
+      "learning_rate": 0.0005704523396721162,
+      "loss": 1.4372,
+      "step": 4385
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.08033733458464856,
+      "learning_rate": 0.000570279965114538,
+      "loss": 1.2662,
+      "step": 4386
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0781908026380043,
+      "learning_rate": 0.0005701075820352718,
+      "loss": 1.4461,
+      "step": 4387
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.07027921368888441,
+      "learning_rate": 0.0005699351904552196,
+      "loss": 1.38,
+      "step": 4388
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.07984943706970755,
+      "learning_rate": 0.0005697627903952844,
+      "loss": 1.3059,
+      "step": 4389
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.07262529776501059,
+      "learning_rate": 0.0005695903818763703,
+      "loss": 1.439,
+      "step": 4390
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.06658673915041431,
+      "learning_rate": 0.0005694179649193826,
+      "loss": 1.2767,
+      "step": 4391
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.07230322011342759,
+      "learning_rate": 0.0005692455395452272,
+      "loss": 1.6068,
+      "step": 4392
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.07247693991750412,
+      "learning_rate": 0.0005690731057748116,
+      "loss": 1.4641,
+      "step": 4393
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.08824859279901634,
+      "learning_rate": 0.0005689006636290436,
+      "loss": 1.397,
+      "step": 4394
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.07435843692754146,
+      "learning_rate": 0.0005687282131288326,
+      "loss": 1.4981,
+      "step": 4395
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.06866404893479804,
+      "learning_rate": 0.000568555754295089,
+      "loss": 1.3269,
+      "step": 4396
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.06631490616956406,
+      "learning_rate": 0.0005683832871487237,
+      "loss": 1.3666,
+      "step": 4397
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.07461689282658565,
+      "learning_rate": 0.0005682108117106492,
+      "loss": 1.442,
+      "step": 4398
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.07629678720903356,
+      "learning_rate": 0.0005680383280017785,
+      "loss": 1.4704,
+      "step": 4399
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.10030040676241504,
+      "learning_rate": 0.000567865836043026,
+      "loss": 1.4009,
+      "step": 4400
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.07009548150800028,
+      "learning_rate": 0.0005676933358553068,
+      "loss": 1.3269,
+      "step": 4401
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.07800720543403114,
+      "learning_rate": 0.0005675208274595376,
+      "loss": 1.4252,
+      "step": 4402
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.06738892904108221,
+      "learning_rate": 0.0005673483108766348,
+      "loss": 1.3973,
+      "step": 4403
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.06857122139693983,
+      "learning_rate": 0.0005671757861275175,
+      "loss": 1.3929,
+      "step": 4404
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.07166471776508838,
+      "learning_rate": 0.0005670032532331045,
+      "loss": 1.4055,
+      "step": 4405
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.06517833141153938,
+      "learning_rate": 0.0005668307122143159,
+      "loss": 1.4336,
+      "step": 4406
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.07899339662584212,
+      "learning_rate": 0.0005666581630920731,
+      "loss": 1.428,
+      "step": 4407
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0858859644599195,
+      "learning_rate": 0.0005664856058872984,
+      "loss": 1.3167,
+      "step": 4408
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.06573428062527853,
+      "learning_rate": 0.0005663130406209144,
+      "loss": 1.3983,
+      "step": 4409
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.07532366412662411,
+      "learning_rate": 0.000566140467313846,
+      "loss": 1.5081,
+      "step": 4410
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.07402368057331848,
+      "learning_rate": 0.0005659678859870177,
+      "loss": 1.2766,
+      "step": 4411
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.07084658555164398,
+      "learning_rate": 0.0005657952966613558,
+      "loss": 1.4809,
+      "step": 4412
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.07293649298192152,
+      "learning_rate": 0.0005656226993577874,
+      "loss": 1.3181,
+      "step": 4413
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.07343140003558507,
+      "learning_rate": 0.0005654500940972404,
+      "loss": 1.3817,
+      "step": 4414
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.06442382898080595,
+      "learning_rate": 0.000565277480900644,
+      "loss": 1.565,
+      "step": 4415
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0758026294438705,
+      "learning_rate": 0.0005651048597889277,
+      "loss": 1.5896,
+      "step": 4416
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.07908556991375093,
+      "learning_rate": 0.000564932230783023,
+      "loss": 1.2737,
+      "step": 4417
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.06293706327093158,
+      "learning_rate": 0.0005647595939038615,
+      "loss": 1.4938,
+      "step": 4418
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.06813466656281142,
+      "learning_rate": 0.0005645869491723757,
+      "loss": 1.4309,
+      "step": 4419
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.07377659147668462,
+      "learning_rate": 0.0005644142966094997,
+      "loss": 1.4835,
+      "step": 4420
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.07262404263783771,
+      "learning_rate": 0.0005642416362361683,
+      "loss": 1.4194,
+      "step": 4421
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.06906843043045555,
+      "learning_rate": 0.0005640689680733171,
+      "loss": 1.4486,
+      "step": 4422
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.07374356166217189,
+      "learning_rate": 0.0005638962921418824,
+      "loss": 1.4561,
+      "step": 4423
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.07256179926993496,
+      "learning_rate": 0.0005637236084628022,
+      "loss": 1.4977,
+      "step": 4424
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.09589845080482406,
+      "learning_rate": 0.0005635509170570148,
+      "loss": 1.4129,
+      "step": 4425
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.074427280353466,
+      "learning_rate": 0.0005633782179454594,
+      "loss": 1.366,
+      "step": 4426
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.06692816156957947,
+      "learning_rate": 0.0005632055111490767,
+      "loss": 1.3508,
+      "step": 4427
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.07199780349642708,
+      "learning_rate": 0.0005630327966888079,
+      "loss": 1.4069,
+      "step": 4428
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0683222669972921,
+      "learning_rate": 0.0005628600745855952,
+      "loss": 1.5312,
+      "step": 4429
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.07072452845567084,
+      "learning_rate": 0.0005626873448603817,
+      "loss": 1.2589,
+      "step": 4430
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.07226687346488332,
+      "learning_rate": 0.0005625146075341115,
+      "loss": 1.5328,
+      "step": 4431
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.06829504835227422,
+      "learning_rate": 0.0005623418626277296,
+      "loss": 1.5685,
+      "step": 4432
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.06920654184104114,
+      "learning_rate": 0.000562169110162182,
+      "loss": 1.3811,
+      "step": 4433
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.07101655341266463,
+      "learning_rate": 0.0005619963501584154,
+      "loss": 1.6274,
+      "step": 4434
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0678822950684123,
+      "learning_rate": 0.0005618235826373776,
+      "loss": 1.4166,
+      "step": 4435
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.06572661972423154,
+      "learning_rate": 0.0005616508076200174,
+      "loss": 1.4391,
+      "step": 4436
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.06820610609290627,
+      "learning_rate": 0.0005614780251272842,
+      "loss": 1.5358,
+      "step": 4437
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.06550275046899529,
+      "learning_rate": 0.0005613052351801283,
+      "loss": 1.4511,
+      "step": 4438
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.08188025593494892,
+      "learning_rate": 0.0005611324377995016,
+      "loss": 1.4931,
+      "step": 4439
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.06581901611273966,
+      "learning_rate": 0.0005609596330063558,
+      "loss": 1.3614,
+      "step": 4440
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.06954853363417197,
+      "learning_rate": 0.0005607868208216445,
+      "loss": 1.3153,
+      "step": 4441
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.07447233452866256,
+      "learning_rate": 0.0005606140012663214,
+      "loss": 1.5368,
+      "step": 4442
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.06884509785274705,
+      "learning_rate": 0.0005604411743613418,
+      "loss": 1.51,
+      "step": 4443
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.07476522479204314,
+      "learning_rate": 0.0005602683401276614,
+      "loss": 1.3806,
+      "step": 4444
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.07323636031812013,
+      "learning_rate": 0.0005600954985862373,
+      "loss": 1.315,
+      "step": 4445
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.06912970850725218,
+      "learning_rate": 0.0005599226497580264,
+      "loss": 1.4559,
+      "step": 4446
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0730926337836783,
+      "learning_rate": 0.0005597497936639878,
+      "loss": 1.4216,
+      "step": 4447
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.07234954230583959,
+      "learning_rate": 0.0005595769303250809,
+      "loss": 1.4416,
+      "step": 4448
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0746131568968767,
+      "learning_rate": 0.0005594040597622653,
+      "loss": 1.3221,
+      "step": 4449
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.08148817831391064,
+      "learning_rate": 0.0005592311819965032,
+      "loss": 1.4065,
+      "step": 4450
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.07166491026144559,
+      "learning_rate": 0.0005590582970487558,
+      "loss": 1.345,
+      "step": 4451
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.07930429551915798,
+      "learning_rate": 0.0005588854049399864,
+      "loss": 1.4026,
+      "step": 4452
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.06973923917211951,
+      "learning_rate": 0.0005587125056911586,
+      "loss": 1.3695,
+      "step": 4453
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.07736874764036672,
+      "learning_rate": 0.000558539599323237,
+      "loss": 1.3889,
+      "step": 4454
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.07395876425197255,
+      "learning_rate": 0.0005583666858571873,
+      "loss": 1.4323,
+      "step": 4455
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.06956108570224902,
+      "learning_rate": 0.0005581937653139757,
+      "loss": 1.3301,
+      "step": 4456
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.070340521017195,
+      "learning_rate": 0.0005580208377145693,
+      "loss": 1.4881,
+      "step": 4457
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.07142190688929934,
+      "learning_rate": 0.0005578479030799362,
+      "loss": 1.3902,
+      "step": 4458
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.07158868756773161,
+      "learning_rate": 0.0005576749614310456,
+      "loss": 1.4741,
+      "step": 4459
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.06838868568645787,
+      "learning_rate": 0.0005575020127888672,
+      "loss": 1.4185,
+      "step": 4460
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.07183543922807431,
+      "learning_rate": 0.0005573290571743714,
+      "loss": 1.4307,
+      "step": 4461
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.06986236197336322,
+      "learning_rate": 0.0005571560946085298,
+      "loss": 1.4594,
+      "step": 4462
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.06696960142160002,
+      "learning_rate": 0.0005569831251123145,
+      "loss": 1.3465,
+      "step": 4463
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.07233568368627842,
+      "learning_rate": 0.0005568101487066991,
+      "loss": 1.307,
+      "step": 4464
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0775921186659931,
+      "learning_rate": 0.0005566371654126572,
+      "loss": 1.439,
+      "step": 4465
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.07908711625273704,
+      "learning_rate": 0.0005564641752511636,
+      "loss": 1.3838,
+      "step": 4466
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.07270285235982457,
+      "learning_rate": 0.0005562911782431943,
+      "loss": 1.4833,
+      "step": 4467
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.07069593509582667,
+      "learning_rate": 0.0005561181744097255,
+      "loss": 1.4951,
+      "step": 4468
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.07302606485108576,
+      "learning_rate": 0.0005559451637717345,
+      "loss": 1.376,
+      "step": 4469
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.06534902828368297,
+      "learning_rate": 0.0005557721463501997,
+      "loss": 1.395,
+      "step": 4470
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.06380982214192189,
+      "learning_rate": 0.0005555991221660998,
+      "loss": 1.3312,
+      "step": 4471
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.06465927979215724,
+      "learning_rate": 0.0005554260912404146,
+      "loss": 1.3872,
+      "step": 4472
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.07193715282214726,
+      "learning_rate": 0.0005552530535941248,
+      "loss": 1.3708,
+      "step": 4473
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.06858753038473571,
+      "learning_rate": 0.0005550800092482117,
+      "loss": 1.4944,
+      "step": 4474
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.06901279556324783,
+      "learning_rate": 0.0005549069582236576,
+      "loss": 1.3406,
+      "step": 4475
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.08512978246349323,
+      "learning_rate": 0.0005547339005414456,
+      "loss": 1.5978,
+      "step": 4476
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.07861647095932271,
+      "learning_rate": 0.0005545608362225594,
+      "loss": 1.4021,
+      "step": 4477
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.06934085764688186,
+      "learning_rate": 0.0005543877652879837,
+      "loss": 1.3235,
+      "step": 4478
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.07068395348115039,
+      "learning_rate": 0.0005542146877587041,
+      "loss": 1.4409,
+      "step": 4479
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.07537414137102522,
+      "learning_rate": 0.0005540416036557064,
+      "loss": 1.4742,
+      "step": 4480
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.06280318869662363,
+      "learning_rate": 0.0005538685129999782,
+      "loss": 1.4793,
+      "step": 4481
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.06802012013643947,
+      "learning_rate": 0.000553695415812507,
+      "loss": 1.4218,
+      "step": 4482
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.07184833062274965,
+      "learning_rate": 0.0005535223121142817,
+      "loss": 1.5405,
+      "step": 4483
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.07048066216163036,
+      "learning_rate": 0.0005533492019262913,
+      "loss": 1.5284,
+      "step": 4484
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.07208127756365079,
+      "learning_rate": 0.0005531760852695263,
+      "loss": 1.4083,
+      "step": 4485
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.06955811505268919,
+      "learning_rate": 0.0005530029621649777,
+      "loss": 1.3813,
+      "step": 4486
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.07251905452428771,
+      "learning_rate": 0.0005528298326336372,
+      "loss": 1.47,
+      "step": 4487
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0660393445159999,
+      "learning_rate": 0.0005526566966964972,
+      "loss": 1.4468,
+      "step": 4488
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.07447153416480336,
+      "learning_rate": 0.0005524835543745515,
+      "loss": 1.2929,
+      "step": 4489
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.07285995323219417,
+      "learning_rate": 0.0005523104056887936,
+      "loss": 1.3571,
+      "step": 4490
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.07066125440052619,
+      "learning_rate": 0.0005521372506602187,
+      "loss": 1.3609,
+      "step": 4491
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.06989235987621864,
+      "learning_rate": 0.0005519640893098227,
+      "loss": 1.4226,
+      "step": 4492
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.06912670079475867,
+      "learning_rate": 0.0005517909216586015,
+      "loss": 1.4072,
+      "step": 4493
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.08376149611071462,
+      "learning_rate": 0.0005516177477275523,
+      "loss": 1.4395,
+      "step": 4494
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.07540252701541866,
+      "learning_rate": 0.0005514445675376735,
+      "loss": 1.3781,
+      "step": 4495
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.07338801331012032,
+      "learning_rate": 0.0005512713811099636,
+      "loss": 1.4511,
+      "step": 4496
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.08017833620848996,
+      "learning_rate": 0.0005510981884654217,
+      "loss": 1.3373,
+      "step": 4497
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.07695481291241366,
+      "learning_rate": 0.0005509249896250485,
+      "loss": 1.3973,
+      "step": 4498
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.07799573575606975,
+      "learning_rate": 0.0005507517846098447,
+      "loss": 1.3026,
+      "step": 4499
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.06890484703699848,
+      "learning_rate": 0.0005505785734408121,
+      "loss": 1.4129,
+      "step": 4500
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.07170266528926741,
+      "learning_rate": 0.0005504053561389531,
+      "loss": 1.4825,
+      "step": 4501
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.08265522375488303,
+      "learning_rate": 0.0005502321327252708,
+      "loss": 1.413,
+      "step": 4502
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.06619748791951723,
+      "learning_rate": 0.0005500589032207692,
+      "loss": 1.375,
+      "step": 4503
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.07303920274619637,
+      "learning_rate": 0.000549885667646453,
+      "loss": 1.4716,
+      "step": 4504
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.08679445748778027,
+      "learning_rate": 0.0005497124260233277,
+      "loss": 1.5006,
+      "step": 4505
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.06787411422320562,
+      "learning_rate": 0.0005495391783723994,
+      "loss": 1.4461,
+      "step": 4506
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0813816825415281,
+      "learning_rate": 0.0005493659247146749,
+      "loss": 1.6357,
+      "step": 4507
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.07138574670659505,
+      "learning_rate": 0.0005491926650711619,
+      "loss": 1.4653,
+      "step": 4508
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.06750733162437504,
+      "learning_rate": 0.0005490193994628685,
+      "loss": 1.4337,
+      "step": 4509
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0987904441000795,
+      "learning_rate": 0.000548846127910804,
+      "loss": 1.2916,
+      "step": 4510
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.07504186207018469,
+      "learning_rate": 0.000548672850435978,
+      "loss": 1.3991,
+      "step": 4511
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.08576301042708652,
+      "learning_rate": 0.0005484995670594012,
+      "loss": 1.5069,
+      "step": 4512
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.07276167204073009,
+      "learning_rate": 0.0005483262778020846,
+      "loss": 1.4622,
+      "step": 4513
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.08992303325112609,
+      "learning_rate": 0.0005481529826850403,
+      "loss": 1.3602,
+      "step": 4514
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.07875033982616096,
+      "learning_rate": 0.000547979681729281,
+      "loss": 1.585,
+      "step": 4515
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.07207386003591792,
+      "learning_rate": 0.0005478063749558195,
+      "loss": 1.4695,
+      "step": 4516
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0719171163854882,
+      "learning_rate": 0.0005476330623856705,
+      "loss": 1.2742,
+      "step": 4517
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.07184078857477177,
+      "learning_rate": 0.0005474597440398483,
+      "loss": 1.4122,
+      "step": 4518
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.06633376011230552,
+      "learning_rate": 0.0005472864199393687,
+      "loss": 1.308,
+      "step": 4519
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.07621419629579644,
+      "learning_rate": 0.0005471130901052476,
+      "loss": 1.5846,
+      "step": 4520
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.06731631619116417,
+      "learning_rate": 0.0005469397545585019,
+      "loss": 1.3949,
+      "step": 4521
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.07302884182715914,
+      "learning_rate": 0.000546766413320149,
+      "loss": 1.4557,
+      "step": 4522
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.06445838444376298,
+      "learning_rate": 0.0005465930664112073,
+      "loss": 1.4074,
+      "step": 4523
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.06288037927464722,
+      "learning_rate": 0.000546419713852696,
+      "loss": 1.3382,
+      "step": 4524
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.07343137466979509,
+      "learning_rate": 0.000546246355665634,
+      "loss": 1.4763,
+      "step": 4525
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.09605837351043238,
+      "learning_rate": 0.000546072991871042,
+      "loss": 1.4698,
+      "step": 4526
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.07278503081830355,
+      "learning_rate": 0.0005458996224899409,
+      "loss": 1.4603,
+      "step": 4527
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.06792443909696488,
+      "learning_rate": 0.0005457262475433523,
+      "loss": 1.3286,
+      "step": 4528
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.07308926990410777,
+      "learning_rate": 0.0005455528670522987,
+      "loss": 1.3996,
+      "step": 4529
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.07682204161835285,
+      "learning_rate": 0.0005453794810378028,
+      "loss": 1.367,
+      "step": 4530
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.06996761782406571,
+      "learning_rate": 0.0005452060895208883,
+      "loss": 1.3614,
+      "step": 4531
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0725382634877934,
+      "learning_rate": 0.0005450326925225798,
+      "loss": 1.3161,
+      "step": 4532
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.06988327489990985,
+      "learning_rate": 0.000544859290063902,
+      "loss": 1.4512,
+      "step": 4533
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.08344302247555796,
+      "learning_rate": 0.0005446858821658805,
+      "loss": 1.4152,
+      "step": 4534
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.07425148455363119,
+      "learning_rate": 0.000544512468849542,
+      "loss": 1.3646,
+      "step": 4535
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.06737391959124295,
+      "learning_rate": 0.0005443390501359129,
+      "loss": 1.3049,
+      "step": 4536
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.06954957142687951,
+      "learning_rate": 0.0005441656260460212,
+      "loss": 1.366,
+      "step": 4537
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.07154336570704602,
+      "learning_rate": 0.0005439921966008953,
+      "loss": 1.5324,
+      "step": 4538
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.08572842758012887,
+      "learning_rate": 0.0005438187618215636,
+      "loss": 1.4391,
+      "step": 4539
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.06743260458188859,
+      "learning_rate": 0.0005436453217290562,
+      "loss": 1.3816,
+      "step": 4540
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.07155234240297914,
+      "learning_rate": 0.000543471876344403,
+      "loss": 1.4001,
+      "step": 4541
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.07525337751704271,
+      "learning_rate": 0.0005432984256886347,
+      "loss": 1.5117,
+      "step": 4542
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.07358363432443406,
+      "learning_rate": 0.0005431249697827832,
+      "loss": 1.335,
+      "step": 4543
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.07454610224380064,
+      "learning_rate": 0.0005429515086478805,
+      "loss": 1.3017,
+      "step": 4544
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.07046647316226497,
+      "learning_rate": 0.0005427780423049593,
+      "loss": 1.3929,
+      "step": 4545
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0766116705717234,
+      "learning_rate": 0.0005426045707750529,
+      "loss": 1.4339,
+      "step": 4546
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.09166538682860578,
+      "learning_rate": 0.0005424310940791953,
+      "loss": 1.4755,
+      "step": 4547
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.07278076098680313,
+      "learning_rate": 0.0005422576122384215,
+      "loss": 1.5623,
+      "step": 4548
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.06308207348750028,
+      "learning_rate": 0.0005420841252737664,
+      "loss": 1.3334,
+      "step": 4549
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.07808545423036595,
+      "learning_rate": 0.0005419106332062661,
+      "loss": 1.5023,
+      "step": 4550
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.07914617981438024,
+      "learning_rate": 0.000541737136056957,
+      "loss": 1.4335,
+      "step": 4551
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.07101119935426858,
+      "learning_rate": 0.0005415636338468762,
+      "loss": 1.3554,
+      "step": 4552
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.07221781995588977,
+      "learning_rate": 0.0005413901265970616,
+      "loss": 1.3392,
+      "step": 4553
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.06859955402895859,
+      "learning_rate": 0.0005412166143285514,
+      "loss": 1.3689,
+      "step": 4554
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.06889021096031589,
+      "learning_rate": 0.0005410430970623847,
+      "loss": 1.4025,
+      "step": 4555
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.07536620068094364,
+      "learning_rate": 0.0005408695748196009,
+      "loss": 1.3711,
+      "step": 4556
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.07286775325340998,
+      "learning_rate": 0.0005406960476212403,
+      "loss": 1.4687,
+      "step": 4557
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.07270129112592764,
+      "learning_rate": 0.0005405225154883435,
+      "loss": 1.5237,
+      "step": 4558
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.07454308740712048,
+      "learning_rate": 0.0005403489784419521,
+      "loss": 1.5203,
+      "step": 4559
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.07490105304575796,
+      "learning_rate": 0.000540175436503108,
+      "loss": 1.3618,
+      "step": 4560
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.07055456412536586,
+      "learning_rate": 0.0005400018896928537,
+      "loss": 1.3952,
+      "step": 4561
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.07493601484219946,
+      "learning_rate": 0.0005398283380322323,
+      "loss": 1.4886,
+      "step": 4562
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.07994133908501737,
+      "learning_rate": 0.0005396547815422877,
+      "loss": 1.3913,
+      "step": 4563
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.07205633839504125,
+      "learning_rate": 0.000539481220244064,
+      "loss": 1.4247,
+      "step": 4564
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.08296967780452945,
+      "learning_rate": 0.0005393076541586062,
+      "loss": 1.4473,
+      "step": 4565
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.06787442123709032,
+      "learning_rate": 0.0005391340833069601,
+      "loss": 1.3086,
+      "step": 4566
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.08012266054574767,
+      "learning_rate": 0.0005389605077101712,
+      "loss": 1.4119,
+      "step": 4567
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.08161258295766774,
+      "learning_rate": 0.0005387869273892865,
+      "loss": 1.5046,
+      "step": 4568
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.07775070711351714,
+      "learning_rate": 0.0005386133423653532,
+      "loss": 1.3888,
+      "step": 4569
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.06586472603954369,
+      "learning_rate": 0.0005384397526594189,
+      "loss": 1.3614,
+      "step": 4570
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.08135731486330744,
+      "learning_rate": 0.0005382661582925322,
+      "loss": 1.3754,
+      "step": 4571
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.06833444123583289,
+      "learning_rate": 0.000538092559285742,
+      "loss": 1.5292,
+      "step": 4572
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.07780963232580691,
+      "learning_rate": 0.0005379189556600974,
+      "loss": 1.3701,
+      "step": 4573
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.06877199878994798,
+      "learning_rate": 0.0005377453474366489,
+      "loss": 1.3532,
+      "step": 4574
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.06300244430878904,
+      "learning_rate": 0.0005375717346364468,
+      "loss": 1.3087,
+      "step": 4575
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.08631255658908871,
+      "learning_rate": 0.0005373981172805421,
+      "loss": 1.5267,
+      "step": 4576
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0741591794846563,
+      "learning_rate": 0.0005372244953899872,
+      "loss": 1.4152,
+      "step": 4577
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.07917543188100162,
+      "learning_rate": 0.0005370508689858336,
+      "loss": 1.5721,
+      "step": 4578
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.07534740443199708,
+      "learning_rate": 0.0005368772380891345,
+      "loss": 1.3377,
+      "step": 4579
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.06688936328416421,
+      "learning_rate": 0.0005367036027209431,
+      "loss": 1.4143,
+      "step": 4580
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.06971105161281534,
+      "learning_rate": 0.0005365299629023133,
+      "loss": 1.4915,
+      "step": 4581
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.07018375199107363,
+      "learning_rate": 0.0005363563186542997,
+      "loss": 1.3651,
+      "step": 4582
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0783407287552422,
+      "learning_rate": 0.000536182669997957,
+      "loss": 1.4741,
+      "step": 4583
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.06865204082546987,
+      "learning_rate": 0.0005360090169543409,
+      "loss": 1.5494,
+      "step": 4584
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.07492596982291116,
+      "learning_rate": 0.0005358353595445074,
+      "loss": 1.4175,
+      "step": 4585
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.06726307354329226,
+      "learning_rate": 0.0005356616977895129,
+      "loss": 1.4331,
+      "step": 4586
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.09002694081447887,
+      "learning_rate": 0.0005354880317104144,
+      "loss": 1.4617,
+      "step": 4587
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.069512338457349,
+      "learning_rate": 0.0005353143613282702,
+      "loss": 1.4518,
+      "step": 4588
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.07455424036661215,
+      "learning_rate": 0.0005351406866641377,
+      "loss": 1.5088,
+      "step": 4589
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.07415562850466256,
+      "learning_rate": 0.0005349670077390757,
+      "loss": 1.3662,
+      "step": 4590
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0715278258922769,
+      "learning_rate": 0.0005347933245741435,
+      "loss": 1.3138,
+      "step": 4591
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.07669012583344414,
+      "learning_rate": 0.0005346196371904009,
+      "loss": 1.488,
+      "step": 4592
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.06974715535155264,
+      "learning_rate": 0.0005344459456089078,
+      "loss": 1.5462,
+      "step": 4593
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.07338763101109938,
+      "learning_rate": 0.0005342722498507251,
+      "loss": 1.485,
+      "step": 4594
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0694636156603943,
+      "learning_rate": 0.0005340985499369137,
+      "loss": 1.3388,
+      "step": 4595
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0705713385482358,
+      "learning_rate": 0.0005339248458885359,
+      "loss": 1.4198,
+      "step": 4596
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.08029313414307719,
+      "learning_rate": 0.0005337511377266535,
+      "loss": 1.6143,
+      "step": 4597
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.06277611025631606,
+      "learning_rate": 0.0005335774254723293,
+      "loss": 1.3245,
+      "step": 4598
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.07068327245009445,
+      "learning_rate": 0.0005334037091466264,
+      "loss": 1.4243,
+      "step": 4599
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.07234677948360466,
+      "learning_rate": 0.0005332299887706087,
+      "loss": 1.4454,
+      "step": 4600
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.07374854687243138,
+      "learning_rate": 0.0005330562643653401,
+      "loss": 1.484,
+      "step": 4601
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.08347732349695075,
+      "learning_rate": 0.0005328825359518858,
+      "loss": 1.4723,
+      "step": 4602
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.06287568215439522,
+      "learning_rate": 0.0005327088035513105,
+      "loss": 1.4508,
+      "step": 4603
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.07172576192681832,
+      "learning_rate": 0.00053253506718468,
+      "loss": 1.3933,
+      "step": 4604
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.07005732728591357,
+      "learning_rate": 0.0005323613268730605,
+      "loss": 1.4256,
+      "step": 4605
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0654226164057184,
+      "learning_rate": 0.0005321875826375185,
+      "loss": 1.464,
+      "step": 4606
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.06238379236280632,
+      "learning_rate": 0.000532013834499121,
+      "loss": 1.4955,
+      "step": 4607
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.06840929269433478,
+      "learning_rate": 0.0005318400824789359,
+      "loss": 1.4424,
+      "step": 4608
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.07194276160370738,
+      "learning_rate": 0.0005316663265980309,
+      "loss": 1.2864,
+      "step": 4609
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.05734857788180746,
+      "learning_rate": 0.0005314925668774747,
+      "loss": 1.3987,
+      "step": 4610
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.06876114472500308,
+      "learning_rate": 0.0005313188033383363,
+      "loss": 1.4162,
+      "step": 4611
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.06978076302819285,
+      "learning_rate": 0.0005311450360016847,
+      "loss": 1.4863,
+      "step": 4612
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.061183894124280415,
+      "learning_rate": 0.0005309712648885904,
+      "loss": 1.3364,
+      "step": 4613
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.07043102493486708,
+      "learning_rate": 0.0005307974900201235,
+      "loss": 1.5243,
+      "step": 4614
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.06643923473201709,
+      "learning_rate": 0.0005306237114173545,
+      "loss": 1.5084,
+      "step": 4615
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.07447149586592648,
+      "learning_rate": 0.0005304499291013551,
+      "loss": 1.3401,
+      "step": 4616
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.06198491339411245,
+      "learning_rate": 0.0005302761430931968,
+      "loss": 1.3923,
+      "step": 4617
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.07738412696592357,
+      "learning_rate": 0.0005301023534139516,
+      "loss": 1.3069,
+      "step": 4618
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.08466989307130728,
+      "learning_rate": 0.0005299285600846926,
+      "loss": 1.4284,
+      "step": 4619
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.06723599229505743,
+      "learning_rate": 0.0005297547631264922,
+      "loss": 1.347,
+      "step": 4620
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.07153350664015816,
+      "learning_rate": 0.0005295809625604244,
+      "loss": 1.4633,
+      "step": 4621
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.07407785833339331,
+      "learning_rate": 0.0005294071584075629,
+      "loss": 1.3834,
+      "step": 4622
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0764897207373361,
+      "learning_rate": 0.0005292333506889819,
+      "loss": 1.3923,
+      "step": 4623
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0747599974968833,
+      "learning_rate": 0.0005290595394257564,
+      "loss": 1.3799,
+      "step": 4624
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.07394829288250857,
+      "learning_rate": 0.0005288857246389617,
+      "loss": 1.357,
+      "step": 4625
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.07255830568972636,
+      "learning_rate": 0.000528711906349673,
+      "loss": 1.3896,
+      "step": 4626
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.08099083798541638,
+      "learning_rate": 0.0005285380845789669,
+      "loss": 1.4059,
+      "step": 4627
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.07223762794677366,
+      "learning_rate": 0.0005283642593479197,
+      "loss": 1.4438,
+      "step": 4628
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.07054943655122418,
+      "learning_rate": 0.0005281904306776082,
+      "loss": 1.3298,
+      "step": 4629
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.07290095631200455,
+      "learning_rate": 0.0005280165985891098,
+      "loss": 1.4432,
+      "step": 4630
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.06882239361849153,
+      "learning_rate": 0.0005278427631035022,
+      "loss": 1.3861,
+      "step": 4631
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0774646267226411,
+      "learning_rate": 0.0005276689242418635,
+      "loss": 1.3935,
+      "step": 4632
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.07159465617312386,
+      "learning_rate": 0.0005274950820252725,
+      "loss": 1.4455,
+      "step": 4633
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.07941785759449224,
+      "learning_rate": 0.000527321236474808,
+      "loss": 1.457,
+      "step": 4634
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.08133325764192359,
+      "learning_rate": 0.0005271473876115494,
+      "loss": 1.4672,
+      "step": 4635
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.07566371625930349,
+      "learning_rate": 0.0005269735354565763,
+      "loss": 1.4521,
+      "step": 4636
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.07032893323259092,
+      "learning_rate": 0.0005267996800309692,
+      "loss": 1.5184,
+      "step": 4637
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.07951209008943733,
+      "learning_rate": 0.0005266258213558084,
+      "loss": 1.5179,
+      "step": 4638
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0868690888705973,
+      "learning_rate": 0.0005264519594521751,
+      "loss": 1.4653,
+      "step": 4639
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.07875135983511622,
+      "learning_rate": 0.0005262780943411504,
+      "loss": 1.4295,
+      "step": 4640
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.07812734317245881,
+      "learning_rate": 0.0005261042260438163,
+      "loss": 1.3016,
+      "step": 4641
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.07367453442907974,
+      "learning_rate": 0.0005259303545812546,
+      "loss": 1.4366,
+      "step": 4642
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.07696299129620354,
+      "learning_rate": 0.000525756479974548,
+      "loss": 1.4112,
+      "step": 4643
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0726616742047467,
+      "learning_rate": 0.0005255826022447796,
+      "loss": 1.5034,
+      "step": 4644
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.07007504723694576,
+      "learning_rate": 0.0005254087214130324,
+      "loss": 1.5622,
+      "step": 4645
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.09152270269567035,
+      "learning_rate": 0.0005252348375003902,
+      "loss": 1.4707,
+      "step": 4646
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.07062202199553434,
+      "learning_rate": 0.0005250609505279369,
+      "loss": 1.4475,
+      "step": 4647
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.07991628604235958,
+      "learning_rate": 0.0005248870605167569,
+      "loss": 1.5251,
+      "step": 4648
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.07321059260782213,
+      "learning_rate": 0.000524713167487935,
+      "loss": 1.3062,
+      "step": 4649
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.06430883856415529,
+      "learning_rate": 0.0005245392714625564,
+      "loss": 1.4268,
+      "step": 4650
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.07432342350797429,
+      "learning_rate": 0.0005243653724617067,
+      "loss": 1.3743,
+      "step": 4651
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.07084662663170704,
+      "learning_rate": 0.0005241914705064713,
+      "loss": 1.3945,
+      "step": 4652
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.07007986233491623,
+      "learning_rate": 0.0005240175656179368,
+      "loss": 1.4225,
+      "step": 4653
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.09277118002504049,
+      "learning_rate": 0.0005238436578171898,
+      "loss": 1.4031,
+      "step": 4654
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.07412570293831583,
+      "learning_rate": 0.0005236697471253167,
+      "loss": 1.4623,
+      "step": 4655
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0703204816023259,
+      "learning_rate": 0.0005234958335634057,
+      "loss": 1.2797,
+      "step": 4656
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.06764260000870075,
+      "learning_rate": 0.0005233219171525436,
+      "loss": 1.5262,
+      "step": 4657
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.06908150282828376,
+      "learning_rate": 0.0005231479979138186,
+      "loss": 1.5008,
+      "step": 4658
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.06594842101397937,
+      "learning_rate": 0.0005229740758683192,
+      "loss": 1.5086,
+      "step": 4659
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.07234559595933199,
+      "learning_rate": 0.0005228001510371337,
+      "loss": 1.3922,
+      "step": 4660
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.07615374538885955,
+      "learning_rate": 0.0005226262234413514,
+      "loss": 1.3848,
+      "step": 4661
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.07029248323561711,
+      "learning_rate": 0.0005224522931020616,
+      "loss": 1.445,
+      "step": 4662
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0714406791924088,
+      "learning_rate": 0.0005222783600403536,
+      "loss": 1.5777,
+      "step": 4663
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0748912481646773,
+      "learning_rate": 0.0005221044242773177,
+      "loss": 1.2854,
+      "step": 4664
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.06734632544256124,
+      "learning_rate": 0.0005219304858340443,
+      "loss": 1.5012,
+      "step": 4665
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.06856504138644735,
+      "learning_rate": 0.0005217565447316238,
+      "loss": 1.4233,
+      "step": 4666
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.07015175624730764,
+      "learning_rate": 0.0005215826009911474,
+      "loss": 1.5598,
+      "step": 4667
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.07662387766996255,
+      "learning_rate": 0.0005214086546337061,
+      "loss": 1.4279,
+      "step": 4668
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.07579601306263993,
+      "learning_rate": 0.0005212347056803916,
+      "loss": 1.4488,
+      "step": 4669
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.08407512048240309,
+      "learning_rate": 0.0005210607541522958,
+      "loss": 1.4203,
+      "step": 4670
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.06957229687866279,
+      "learning_rate": 0.000520886800070511,
+      "loss": 1.3616,
+      "step": 4671
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.06557597435831303,
+      "learning_rate": 0.0005207128434561297,
+      "loss": 1.2343,
+      "step": 4672
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.10138332283543867,
+      "learning_rate": 0.0005205388843302446,
+      "loss": 1.3596,
+      "step": 4673
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.059370486462788453,
+      "learning_rate": 0.0005203649227139491,
+      "loss": 1.4954,
+      "step": 4674
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.06702999161881795,
+      "learning_rate": 0.0005201909586283365,
+      "loss": 1.419,
+      "step": 4675
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.07669954138606098,
+      "learning_rate": 0.0005200169920945005,
+      "loss": 1.344,
+      "step": 4676
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.07665471343138372,
+      "learning_rate": 0.0005198430231335352,
+      "loss": 1.3518,
+      "step": 4677
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.07485280276094071,
+      "learning_rate": 0.000519669051766535,
+      "loss": 1.3616,
+      "step": 4678
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.08512190117286093,
+      "learning_rate": 0.0005194950780145945,
+      "loss": 1.4598,
+      "step": 4679
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.07047653270509795,
+      "learning_rate": 0.0005193211018988084,
+      "loss": 1.3529,
+      "step": 4680
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.06881206519124212,
+      "learning_rate": 0.0005191471234402723,
+      "loss": 1.4288,
+      "step": 4681
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0804348080365886,
+      "learning_rate": 0.0005189731426600813,
+      "loss": 1.3915,
+      "step": 4682
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.06573587273377841,
+      "learning_rate": 0.0005187991595793314,
+      "loss": 1.4739,
+      "step": 4683
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.06985488464953007,
+      "learning_rate": 0.0005186251742191187,
+      "loss": 1.4218,
+      "step": 4684
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.08365828393922413,
+      "learning_rate": 0.0005184511866005392,
+      "loss": 1.5171,
+      "step": 4685
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.06799405095670837,
+      "learning_rate": 0.0005182771967446899,
+      "loss": 1.4046,
+      "step": 4686
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.07745680866308548,
+      "learning_rate": 0.0005181032046726674,
+      "loss": 1.3601,
+      "step": 4687
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.07401868558341498,
+      "learning_rate": 0.0005179292104055689,
+      "loss": 1.408,
+      "step": 4688
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.08185325765888811,
+      "learning_rate": 0.0005177552139644919,
+      "loss": 1.3529,
+      "step": 4689
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.06555219557774031,
+      "learning_rate": 0.0005175812153705339,
+      "loss": 1.3481,
+      "step": 4690
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.07034421079868959,
+      "learning_rate": 0.000517407214644793,
+      "loss": 1.4121,
+      "step": 4691
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.06276669263952558,
+      "learning_rate": 0.0005172332118083673,
+      "loss": 1.3234,
+      "step": 4692
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.07723946949855989,
+      "learning_rate": 0.0005170592068823553,
+      "loss": 1.5377,
+      "step": 4693
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.08084112493527873,
+      "learning_rate": 0.0005168851998878555,
+      "loss": 1.3488,
+      "step": 4694
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0727018555758983,
+      "learning_rate": 0.0005167111908459672,
+      "loss": 1.3023,
+      "step": 4695
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.07233243830653298,
+      "learning_rate": 0.0005165371797777894,
+      "loss": 1.4445,
+      "step": 4696
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.07291339507881155,
+      "learning_rate": 0.0005163631667044213,
+      "loss": 1.4742,
+      "step": 4697
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.07670065393263528,
+      "learning_rate": 0.000516189151646963,
+      "loss": 1.4396,
+      "step": 4698
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.07358001838784738,
+      "learning_rate": 0.0005160151346265142,
+      "loss": 1.409,
+      "step": 4699
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.08157498764671632,
+      "learning_rate": 0.0005158411156641751,
+      "loss": 1.3561,
+      "step": 4700
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.07501124171108461,
+      "learning_rate": 0.0005156670947810462,
+      "loss": 1.4614,
+      "step": 4701
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.06975465507140635,
+      "learning_rate": 0.000515493071998228,
+      "loss": 1.3463,
+      "step": 4702
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.07219506375827546,
+      "learning_rate": 0.0005153190473368213,
+      "loss": 1.4587,
+      "step": 4703
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.07223521378851384,
+      "learning_rate": 0.0005151450208179276,
+      "loss": 1.3898,
+      "step": 4704
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.08370766965759803,
+      "learning_rate": 0.0005149709924626476,
+      "loss": 1.3624,
+      "step": 4705
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.08079188900750653,
+      "learning_rate": 0.0005147969622920832,
+      "loss": 1.3803,
+      "step": 4706
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.07792965487429286,
+      "learning_rate": 0.0005146229303273363,
+      "loss": 1.4281,
+      "step": 4707
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.08238838852280057,
+      "learning_rate": 0.0005144488965895084,
+      "loss": 1.4642,
+      "step": 4708
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0669417006150072,
+      "learning_rate": 0.0005142748610997023,
+      "loss": 1.484,
+      "step": 4709
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.07877637474591272,
+      "learning_rate": 0.0005141008238790199,
+      "loss": 1.3851,
+      "step": 4710
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.07109987453919168,
+      "learning_rate": 0.0005139267849485639,
+      "loss": 1.3506,
+      "step": 4711
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.07645449263811037,
+      "learning_rate": 0.0005137527443294374,
+      "loss": 1.4377,
+      "step": 4712
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.06840270491375017,
+      "learning_rate": 0.0005135787020427432,
+      "loss": 1.3547,
+      "step": 4713
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.07120138812630702,
+      "learning_rate": 0.0005134046581095844,
+      "loss": 1.4399,
+      "step": 4714
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.07151681302142116,
+      "learning_rate": 0.0005132306125510648,
+      "loss": 1.2813,
+      "step": 4715
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.06240065323117308,
+      "learning_rate": 0.0005130565653882877,
+      "loss": 1.3654,
+      "step": 4716
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.06194617156336227,
+      "learning_rate": 0.0005128825166423569,
+      "loss": 1.5099,
+      "step": 4717
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.059304667134885136,
+      "learning_rate": 0.0005127084663343769,
+      "loss": 1.4544,
+      "step": 4718
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.06483159155925523,
+      "learning_rate": 0.0005125344144854513,
+      "loss": 1.5357,
+      "step": 4719
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.06246551571893033,
+      "learning_rate": 0.0005123603611166848,
+      "loss": 1.5015,
+      "step": 4720
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.061595172916569754,
+      "learning_rate": 0.0005121863062491818,
+      "loss": 1.3058,
+      "step": 4721
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.06544023845066463,
+      "learning_rate": 0.0005120122499040473,
+      "loss": 1.392,
+      "step": 4722
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0759664433784072,
+      "learning_rate": 0.0005118381921023859,
+      "loss": 1.4097,
+      "step": 4723
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.07274017908537568,
+      "learning_rate": 0.0005116641328653031,
+      "loss": 1.5345,
+      "step": 4724
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.06619000110896354,
+      "learning_rate": 0.0005114900722139039,
+      "loss": 1.4603,
+      "step": 4725
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.07000980439320045,
+      "learning_rate": 0.0005113160101692938,
+      "loss": 1.4563,
+      "step": 4726
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0685154283152642,
+      "learning_rate": 0.0005111419467525786,
+      "loss": 1.4495,
+      "step": 4727
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.07039172540427986,
+      "learning_rate": 0.0005109678819848637,
+      "loss": 1.403,
+      "step": 4728
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.06839571480775747,
+      "learning_rate": 0.0005107938158872554,
+      "loss": 1.4476,
+      "step": 4729
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.06564357301738481,
+      "learning_rate": 0.0005106197484808598,
+      "loss": 1.368,
+      "step": 4730
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.06865827489358711,
+      "learning_rate": 0.0005104456797867831,
+      "loss": 1.3423,
+      "step": 4731
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.07352389360835015,
+      "learning_rate": 0.0005102716098261315,
+      "loss": 1.4653,
+      "step": 4732
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.07081825517552962,
+      "learning_rate": 0.0005100975386200119,
+      "loss": 1.3664,
+      "step": 4733
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.07545040261023422,
+      "learning_rate": 0.000509923466189531,
+      "loss": 1.5089,
+      "step": 4734
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.07863333546994716,
+      "learning_rate": 0.0005097493925557956,
+      "loss": 1.5012,
+      "step": 4735
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.06897139505039224,
+      "learning_rate": 0.0005095753177399127,
+      "loss": 1.4103,
+      "step": 4736
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0709893551702056,
+      "learning_rate": 0.0005094012417629895,
+      "loss": 1.5152,
+      "step": 4737
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.07522261575946469,
+      "learning_rate": 0.0005092271646461334,
+      "loss": 1.3445,
+      "step": 4738
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.07113884668637967,
+      "learning_rate": 0.0005090530864104517,
+      "loss": 1.3651,
+      "step": 4739
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.07430397727410497,
+      "learning_rate": 0.000508879007077052,
+      "loss": 1.5024,
+      "step": 4740
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.06937378871027804,
+      "learning_rate": 0.0005087049266670424,
+      "loss": 1.4918,
+      "step": 4741
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.07196532177141281,
+      "learning_rate": 0.0005085308452015301,
+      "loss": 1.5521,
+      "step": 4742
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.07511206297719938,
+      "learning_rate": 0.0005083567627016235,
+      "loss": 1.5095,
+      "step": 4743
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.06208055254053003,
+      "learning_rate": 0.0005081826791884307,
+      "loss": 1.4788,
+      "step": 4744
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.06275328428534259,
+      "learning_rate": 0.0005080085946830596,
+      "loss": 1.2569,
+      "step": 4745
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.07479166475912431,
+      "learning_rate": 0.0005078345092066191,
+      "loss": 1.3658,
+      "step": 4746
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.06983578026027626,
+      "learning_rate": 0.0005076604227802171,
+      "loss": 1.506,
+      "step": 4747
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.059381228868428844,
+      "learning_rate": 0.0005074863354249625,
+      "loss": 1.5236,
+      "step": 4748
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.07064049104101132,
+      "learning_rate": 0.000507312247161964,
+      "loss": 1.2982,
+      "step": 4749
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.06537241706728257,
+      "learning_rate": 0.0005071381580123302,
+      "loss": 1.3971,
+      "step": 4750
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0705342768829638,
+      "learning_rate": 0.0005069640679971702,
+      "loss": 1.556,
+      "step": 4751
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.07234045693835783,
+      "learning_rate": 0.000506789977137593,
+      "loss": 1.4512,
+      "step": 4752
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.07044135313910474,
+      "learning_rate": 0.0005066158854547075,
+      "loss": 1.4174,
+      "step": 4753
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.07391834794874395,
+      "learning_rate": 0.0005064417929696232,
+      "loss": 1.378,
+      "step": 4754
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0717476739675131,
+      "learning_rate": 0.0005062676997034493,
+      "loss": 1.3867,
+      "step": 4755
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.067547995646433,
+      "learning_rate": 0.0005060936056772951,
+      "loss": 1.5415,
+      "step": 4756
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0709070460555177,
+      "learning_rate": 0.0005059195109122705,
+      "loss": 1.4631,
+      "step": 4757
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0732869766299112,
+      "learning_rate": 0.0005057454154294846,
+      "loss": 1.4731,
+      "step": 4758
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.07631628853739582,
+      "learning_rate": 0.0005055713192500472,
+      "loss": 1.3617,
+      "step": 4759
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.07557827019800746,
+      "learning_rate": 0.0005053972223950682,
+      "loss": 1.3886,
+      "step": 4760
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.07541011499079218,
+      "learning_rate": 0.0005052231248856573,
+      "loss": 1.4401,
+      "step": 4761
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.06878411120055702,
+      "learning_rate": 0.0005050490267429246,
+      "loss": 1.4652,
+      "step": 4762
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.06739762564942098,
+      "learning_rate": 0.00050487492798798,
+      "loss": 1.5507,
+      "step": 4763
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.08015595700952176,
+      "learning_rate": 0.0005047008286419336,
+      "loss": 1.3467,
+      "step": 4764
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.07001966186468692,
+      "learning_rate": 0.0005045267287258953,
+      "loss": 1.4121,
+      "step": 4765
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.07030643248207329,
+      "learning_rate": 0.0005043526282609757,
+      "loss": 1.4879,
+      "step": 4766
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.08827989390479461,
+      "learning_rate": 0.0005041785272682849,
+      "loss": 1.3802,
+      "step": 4767
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.06827504201270179,
+      "learning_rate": 0.0005040044257689333,
+      "loss": 1.2448,
+      "step": 4768
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.07915638860324421,
+      "learning_rate": 0.0005038303237840314,
+      "loss": 1.4055,
+      "step": 4769
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.06918765384054784,
+      "learning_rate": 0.0005036562213346893,
+      "loss": 1.538,
+      "step": 4770
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.08004483251964818,
+      "learning_rate": 0.0005034821184420179,
+      "loss": 1.279,
+      "step": 4771
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.06912086940038036,
+      "learning_rate": 0.0005033080151271276,
+      "loss": 1.3917,
+      "step": 4772
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0714885845925155,
+      "learning_rate": 0.000503133911411129,
+      "loss": 1.394,
+      "step": 4773
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.06714730640721422,
+      "learning_rate": 0.0005029598073151329,
+      "loss": 1.4416,
+      "step": 4774
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.07827813518251049,
+      "learning_rate": 0.00050278570286025,
+      "loss": 1.3962,
+      "step": 4775
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.08034014575300731,
+      "learning_rate": 0.0005026115980675908,
+      "loss": 1.3307,
+      "step": 4776
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0765192032903928,
+      "learning_rate": 0.0005024374929582664,
+      "loss": 1.4161,
+      "step": 4777
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.06608261104152828,
+      "learning_rate": 0.0005022633875533879,
+      "loss": 1.2704,
+      "step": 4778
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.07850238429821357,
+      "learning_rate": 0.0005020892818740656,
+      "loss": 1.4342,
+      "step": 4779
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.06403022438627028,
+      "learning_rate": 0.0005019151759414107,
+      "loss": 1.4186,
+      "step": 4780
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.06799933698291283,
+      "learning_rate": 0.0005017410697765342,
+      "loss": 1.362,
+      "step": 4781
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0665150830459829,
+      "learning_rate": 0.0005015669634005467,
+      "loss": 1.4916,
+      "step": 4782
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.07236361062725231,
+      "learning_rate": 0.0005013928568345597,
+      "loss": 1.2738,
+      "step": 4783
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.06861181480465103,
+      "learning_rate": 0.000501218750099684,
+      "loss": 1.3728,
+      "step": 4784
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.06525807506942764,
+      "learning_rate": 0.0005010446432170306,
+      "loss": 1.5133,
+      "step": 4785
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.06543208409660094,
+      "learning_rate": 0.0005008705362077108,
+      "loss": 1.3821,
+      "step": 4786
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.07321867883420735,
+      "learning_rate": 0.0005006964290928351,
+      "loss": 1.3791,
+      "step": 4787
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.07359931820206309,
+      "learning_rate": 0.0005005223218935152,
+      "loss": 1.3444,
+      "step": 4788
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.07192584249908926,
+      "learning_rate": 0.0005003482146308621,
+      "loss": 1.3597,
+      "step": 4789
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.06602677971933513,
+      "learning_rate": 0.0005001741073259866,
+      "loss": 1.5547,
+      "step": 4790
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.07961813739728019,
+      "learning_rate": 0.0005,
+      "loss": 1.4441,
+      "step": 4791
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.07216239496007772,
+      "learning_rate": 0.0004998258926740136,
+      "loss": 1.3991,
+      "step": 4792
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.06996627259186475,
+      "learning_rate": 0.0004996517853691379,
+      "loss": 1.498,
+      "step": 4793
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.06657604028460089,
+      "learning_rate": 0.0004994776781064847,
+      "loss": 1.3081,
+      "step": 4794
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.06303048383434355,
+      "learning_rate": 0.0004993035709071648,
+      "loss": 1.2912,
+      "step": 4795
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.06521095894949946,
+      "learning_rate": 0.0004991294637922893,
+      "loss": 1.3,
+      "step": 4796
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.07399255311468003,
+      "learning_rate": 0.0004989553567829695,
+      "loss": 1.4204,
+      "step": 4797
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.08417173627975091,
+      "learning_rate": 0.000498781249900316,
+      "loss": 1.514,
+      "step": 4798
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.07708032383741255,
+      "learning_rate": 0.0004986071431654404,
+      "loss": 1.4721,
+      "step": 4799
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.06307820528877853,
+      "learning_rate": 0.0004984330365994535,
+      "loss": 1.373,
+      "step": 4800
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.06111579146187217,
+      "learning_rate": 0.000498258930223466,
+      "loss": 1.3943,
+      "step": 4801
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.06554523449766973,
+      "learning_rate": 0.0004980848240585895,
+      "loss": 1.3626,
+      "step": 4802
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.06604543741412207,
+      "learning_rate": 0.0004979107181259345,
+      "loss": 1.4796,
+      "step": 4803
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.06646732193706745,
+      "learning_rate": 0.0004977366124466121,
+      "loss": 1.337,
+      "step": 4804
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.07501224583003234,
+      "learning_rate": 0.0004975625070417335,
+      "loss": 1.4135,
+      "step": 4805
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0737723782461909,
+      "learning_rate": 0.0004973884019324092,
+      "loss": 1.4717,
+      "step": 4806
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.07017623757888976,
+      "learning_rate": 0.0004972142971397503,
+      "loss": 1.2908,
+      "step": 4807
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.08090698115220855,
+      "learning_rate": 0.0004970401926848673,
+      "loss": 1.3435,
+      "step": 4808
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.07428142762072636,
+      "learning_rate": 0.0004968660885888712,
+      "loss": 1.3949,
+      "step": 4809
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.06726515819501672,
+      "learning_rate": 0.0004966919848728726,
+      "loss": 1.5103,
+      "step": 4810
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.07380195335443142,
+      "learning_rate": 0.0004965178815579822,
+      "loss": 1.3751,
+      "step": 4811
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.07761558508982735,
+      "learning_rate": 0.0004963437786653108,
+      "loss": 1.3809,
+      "step": 4812
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.07763719108725761,
+      "learning_rate": 0.0004961696762159687,
+      "loss": 1.4223,
+      "step": 4813
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.07189756416959095,
+      "learning_rate": 0.0004959955742310667,
+      "loss": 1.3861,
+      "step": 4814
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.08204004830906156,
+      "learning_rate": 0.0004958214727317151,
+      "loss": 1.3729,
+      "step": 4815
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.07748801457158774,
+      "learning_rate": 0.0004956473717390242,
+      "loss": 1.4368,
+      "step": 4816
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.07114876371367315,
+      "learning_rate": 0.0004954732712741046,
+      "loss": 1.3806,
+      "step": 4817
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.07188339735039269,
+      "learning_rate": 0.0004952991713580667,
+      "loss": 1.2483,
+      "step": 4818
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.07930257872447899,
+      "learning_rate": 0.0004951250720120203,
+      "loss": 1.3644,
+      "step": 4819
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.08348873172896026,
+      "learning_rate": 0.0004949509732570756,
+      "loss": 1.3957,
+      "step": 4820
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.08389458458452004,
+      "learning_rate": 0.0004947768751143428,
+      "loss": 1.4093,
+      "step": 4821
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.06612652000902616,
+      "learning_rate": 0.000494602777604932,
+      "loss": 1.3784,
+      "step": 4822
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0741760050819813,
+      "learning_rate": 0.0004944286807499529,
+      "loss": 1.4181,
+      "step": 4823
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.07490795828156847,
+      "learning_rate": 0.0004942545845705155,
+      "loss": 1.3859,
+      "step": 4824
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.07099564390158039,
+      "learning_rate": 0.0004940804890877297,
+      "loss": 1.437,
+      "step": 4825
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.07139635992893732,
+      "learning_rate": 0.0004939063943227048,
+      "loss": 1.2719,
+      "step": 4826
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.07376969080532272,
+      "learning_rate": 0.0004937323002965506,
+      "loss": 1.4981,
+      "step": 4827
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.07045853122954072,
+      "learning_rate": 0.0004935582070303767,
+      "loss": 1.3566,
+      "step": 4828
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0717218632271678,
+      "learning_rate": 0.0004933841145452926,
+      "loss": 1.2817,
+      "step": 4829
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.06912439103196728,
+      "learning_rate": 0.0004932100228624072,
+      "loss": 1.3627,
+      "step": 4830
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.06397379656010241,
+      "learning_rate": 0.0004930359320028299,
+      "loss": 1.3331,
+      "step": 4831
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.072065307732877,
+      "learning_rate": 0.0004928618419876698,
+      "loss": 1.4236,
+      "step": 4832
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.06265824186043788,
+      "learning_rate": 0.0004926877528380362,
+      "loss": 1.4034,
+      "step": 4833
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.07470555319601731,
+      "learning_rate": 0.0004925136645750376,
+      "loss": 1.3959,
+      "step": 4834
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.06716092261155211,
+      "learning_rate": 0.0004923395772197829,
+      "loss": 1.3977,
+      "step": 4835
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.07475858200018855,
+      "learning_rate": 0.000492165490793381,
+      "loss": 1.3418,
+      "step": 4836
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.07327276783550837,
+      "learning_rate": 0.0004919914053169404,
+      "loss": 1.5526,
+      "step": 4837
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.06446215898606858,
+      "learning_rate": 0.0004918173208115694,
+      "loss": 1.4026,
+      "step": 4838
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.06850668413178283,
+      "learning_rate": 0.0004916432372983767,
+      "loss": 1.4541,
+      "step": 4839
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.07883512917030402,
+      "learning_rate": 0.0004914691547984701,
+      "loss": 1.4152,
+      "step": 4840
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.06626183570546103,
+      "learning_rate": 0.0004912950733329579,
+      "loss": 1.5067,
+      "step": 4841
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.06982230220511904,
+      "learning_rate": 0.000491120992922948,
+      "loss": 1.4825,
+      "step": 4842
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.06640439609436137,
+      "learning_rate": 0.0004909469135895484,
+      "loss": 1.5727,
+      "step": 4843
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.07230302368111746,
+      "learning_rate": 0.0004907728353538667,
+      "loss": 1.3462,
+      "step": 4844
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.07980176876442475,
+      "learning_rate": 0.0004905987582370106,
+      "loss": 1.2809,
+      "step": 4845
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.07376211478806756,
+      "learning_rate": 0.0004904246822600874,
+      "loss": 1.3073,
+      "step": 4846
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.08760449272236642,
+      "learning_rate": 0.0004902506074442044,
+      "loss": 1.4225,
+      "step": 4847
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.06840610524758446,
+      "learning_rate": 0.000490076533810469,
+      "loss": 1.2817,
+      "step": 4848
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.07559520805663945,
+      "learning_rate": 0.0004899024613799881,
+      "loss": 1.5285,
+      "step": 4849
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.06808732086358632,
+      "learning_rate": 0.0004897283901738686,
+      "loss": 1.3411,
+      "step": 4850
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.06805750834540934,
+      "learning_rate": 0.0004895543202132172,
+      "loss": 1.431,
+      "step": 4851
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.08061022722716565,
+      "learning_rate": 0.0004893802515191403,
+      "loss": 1.4008,
+      "step": 4852
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.07093877908970345,
+      "learning_rate": 0.0004892061841127446,
+      "loss": 1.3412,
+      "step": 4853
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.07059607839069051,
+      "learning_rate": 0.0004890321180151364,
+      "loss": 1.3494,
+      "step": 4854
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.07822255808458586,
+      "learning_rate": 0.0004888580532474216,
+      "loss": 1.5409,
+      "step": 4855
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.07227079108820397,
+      "learning_rate": 0.0004886839898307062,
+      "loss": 1.473,
+      "step": 4856
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.07038880965726567,
+      "learning_rate": 0.0004885099277860961,
+      "loss": 1.3489,
+      "step": 4857
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.06991878687842776,
+      "learning_rate": 0.0004883358671346968,
+      "loss": 1.4336,
+      "step": 4858
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.06717727331715036,
+      "learning_rate": 0.000488161807897614,
+      "loss": 1.4411,
+      "step": 4859
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0774671198594342,
+      "learning_rate": 0.00048798775009595285,
+      "loss": 1.4623,
+      "step": 4860
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.06980051958932858,
+      "learning_rate": 0.0004878136937508183,
+      "loss": 1.3775,
+      "step": 4861
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.06641487458265112,
+      "learning_rate": 0.00048763963888331544,
+      "loss": 1.4088,
+      "step": 4862
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.076824429835218,
+      "learning_rate": 0.00048746558551454876,
+      "loss": 1.3507,
+      "step": 4863
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.06712314946460401,
+      "learning_rate": 0.00048729153366562324,
+      "loss": 1.364,
+      "step": 4864
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.06919590987860398,
+      "learning_rate": 0.00048711748335764305,
+      "loss": 1.4172,
+      "step": 4865
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.07129027417864583,
+      "learning_rate": 0.00048694343461171233,
+      "loss": 1.5116,
+      "step": 4866
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0732730577898096,
+      "learning_rate": 0.0004867693874489353,
+      "loss": 1.504,
+      "step": 4867
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.06935474063557696,
+      "learning_rate": 0.0004865953418904156,
+      "loss": 1.5817,
+      "step": 4868
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.06946055369077489,
+      "learning_rate": 0.0004864212979572569,
+      "loss": 1.5422,
+      "step": 4869
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.08287847765182309,
+      "learning_rate": 0.0004862472556705626,
+      "loss": 1.4387,
+      "step": 4870
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.06424473267777256,
+      "learning_rate": 0.00048607321505143614,
+      "loss": 1.3416,
+      "step": 4871
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.07389225955640676,
+      "learning_rate": 0.0004858991761209803,
+      "loss": 1.3793,
+      "step": 4872
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.07360082163492014,
+      "learning_rate": 0.0004857251389002979,
+      "loss": 1.344,
+      "step": 4873
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.07449052954350506,
+      "learning_rate": 0.0004855511034104916,
+      "loss": 1.3882,
+      "step": 4874
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.08073253698261196,
+      "learning_rate": 0.0004853770696726638,
+      "loss": 1.3821,
+      "step": 4875
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.07201609348875974,
+      "learning_rate": 0.0004852030377079168,
+      "loss": 1.3864,
+      "step": 4876
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.07729040284480578,
+      "learning_rate": 0.00048502900753735246,
+      "loss": 1.3482,
+      "step": 4877
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.08204100045657077,
+      "learning_rate": 0.0004848549791820725,
+      "loss": 1.3878,
+      "step": 4878
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.07915089902414485,
+      "learning_rate": 0.0004846809526631786,
+      "loss": 1.4156,
+      "step": 4879
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.06668874663977116,
+      "learning_rate": 0.00048450692800177205,
+      "loss": 1.4023,
+      "step": 4880
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.07506899782913949,
+      "learning_rate": 0.00048433290521895375,
+      "loss": 1.5963,
+      "step": 4881
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.08009982251249653,
+      "learning_rate": 0.000484158884335825,
+      "loss": 1.5212,
+      "step": 4882
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.07433691781861437,
+      "learning_rate": 0.00048398486537348583,
+      "loss": 1.4625,
+      "step": 4883
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.07517421414037016,
+      "learning_rate": 0.0004838108483530371,
+      "loss": 1.3501,
+      "step": 4884
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.07718074248971674,
+      "learning_rate": 0.00048363683329557877,
+      "loss": 1.449,
+      "step": 4885
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.06772550140183944,
+      "learning_rate": 0.0004834628202222107,
+      "loss": 1.4537,
+      "step": 4886
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.08030132699081302,
+      "learning_rate": 0.00048328880915403285,
+      "loss": 1.3553,
+      "step": 4887
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.08022333563213116,
+      "learning_rate": 0.0004831148001121445,
+      "loss": 1.3818,
+      "step": 4888
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.08425144353421923,
+      "learning_rate": 0.0004829407931176447,
+      "loss": 1.4234,
+      "step": 4889
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.07772699541497993,
+      "learning_rate": 0.00048276678819163265,
+      "loss": 1.275,
+      "step": 4890
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.07129243130758527,
+      "learning_rate": 0.00048259278535520703,
+      "loss": 1.3798,
+      "step": 4891
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.08053342693305993,
+      "learning_rate": 0.0004824187846294662,
+      "loss": 1.5514,
+      "step": 4892
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.08309886930824148,
+      "learning_rate": 0.00048224478603550833,
+      "loss": 1.538,
+      "step": 4893
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.06876084759596014,
+      "learning_rate": 0.0004820707895944312,
+      "loss": 1.4856,
+      "step": 4894
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.07354818557342893,
+      "learning_rate": 0.00048189679532733274,
+      "loss": 1.3469,
+      "step": 4895
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.07530483097981298,
+      "learning_rate": 0.00048172280325531027,
+      "loss": 1.3601,
+      "step": 4896
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.077507644997754,
+      "learning_rate": 0.0004815488133994608,
+      "loss": 1.4887,
+      "step": 4897
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.07757860383810837,
+      "learning_rate": 0.0004813748257808814,
+      "loss": 1.4837,
+      "step": 4898
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0844479797681231,
+      "learning_rate": 0.00048120084042066865,
+      "loss": 1.4884,
+      "step": 4899
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.09641157890597221,
+      "learning_rate": 0.0004810268573399187,
+      "loss": 1.4827,
+      "step": 4900
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.07859998643895252,
+      "learning_rate": 0.0004808528765597278,
+      "loss": 1.4992,
+      "step": 4901
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.07276275280157116,
+      "learning_rate": 0.00048067889810119157,
+      "loss": 1.3915,
+      "step": 4902
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.08248731446994635,
+      "learning_rate": 0.00048050492198540575,
+      "loss": 1.5134,
+      "step": 4903
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.07469307310941618,
+      "learning_rate": 0.00048033094823346517,
+      "loss": 1.429,
+      "step": 4904
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.06984611564695324,
+      "learning_rate": 0.00048015697686646486,
+      "loss": 1.3318,
+      "step": 4905
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.07699548220008161,
+      "learning_rate": 0.00047998300790549957,
+      "loss": 1.4534,
+      "step": 4906
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.07689484583591935,
+      "learning_rate": 0.0004798090413716636,
+      "loss": 1.458,
+      "step": 4907
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0694871817732402,
+      "learning_rate": 0.00047963507728605105,
+      "loss": 1.4913,
+      "step": 4908
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.07682175823997765,
+      "learning_rate": 0.00047946111566975544,
+      "loss": 1.4395,
+      "step": 4909
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.07852039121044253,
+      "learning_rate": 0.00047928715654387043,
+      "loss": 1.3262,
+      "step": 4910
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.07051721959880726,
+      "learning_rate": 0.0004791131999294891,
+      "loss": 1.5198,
+      "step": 4911
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0830403180149037,
+      "learning_rate": 0.00047893924584770423,
+      "loss": 1.4268,
+      "step": 4912
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.07212490399033414,
+      "learning_rate": 0.0004787652943196087,
+      "loss": 1.3805,
+      "step": 4913
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.07701948410209183,
+      "learning_rate": 0.0004785913453662941,
+      "loss": 1.4523,
+      "step": 4914
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.08254302412735906,
+      "learning_rate": 0.00047841739900885284,
+      "loss": 1.3042,
+      "step": 4915
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.06816403715521224,
+      "learning_rate": 0.0004782434552683763,
+      "loss": 1.4307,
+      "step": 4916
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.07860013899326017,
+      "learning_rate": 0.0004780695141659557,
+      "loss": 1.6049,
+      "step": 4917
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.06599668818079388,
+      "learning_rate": 0.0004778955757226823,
+      "loss": 1.4359,
+      "step": 4918
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0784233713723611,
+      "learning_rate": 0.0004777216399596465,
+      "loss": 1.2344,
+      "step": 4919
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0788183292534905,
+      "learning_rate": 0.0004775477068979385,
+      "loss": 1.416,
+      "step": 4920
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.07366391835272944,
+      "learning_rate": 0.00047737377655864867,
+      "loss": 1.3873,
+      "step": 4921
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0644438910372059,
+      "learning_rate": 0.00047719984896286635,
+      "loss": 1.2997,
+      "step": 4922
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0819510510639513,
+      "learning_rate": 0.0004770259241316809,
+      "loss": 1.3432,
+      "step": 4923
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.07460646311584986,
+      "learning_rate": 0.00047685200208618164,
+      "loss": 1.4545,
+      "step": 4924
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.07142513170538935,
+      "learning_rate": 0.00047667808284745656,
+      "loss": 1.456,
+      "step": 4925
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0638211680555277,
+      "learning_rate": 0.0004765041664365945,
+      "loss": 1.384,
+      "step": 4926
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.06456251777135609,
+      "learning_rate": 0.00047633025287468323,
+      "loss": 1.3933,
+      "step": 4927
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.07219441230096289,
+      "learning_rate": 0.00047615634218281034,
+      "loss": 1.4249,
+      "step": 4928
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.06770038061326396,
+      "learning_rate": 0.0004759824343820632,
+      "loss": 1.3451,
+      "step": 4929
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.06401901008264176,
+      "learning_rate": 0.00047580852949352876,
+      "loss": 1.3593,
+      "step": 4930
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.06572917995129061,
+      "learning_rate": 0.0004756346275382934,
+      "loss": 1.4683,
+      "step": 4931
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.06690009643906077,
+      "learning_rate": 0.00047546072853744357,
+      "loss": 1.4288,
+      "step": 4932
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.06554492012605238,
+      "learning_rate": 0.00047528683251206493,
+      "loss": 1.5643,
+      "step": 4933
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0632941403255531,
+      "learning_rate": 0.00047511293948324324,
+      "loss": 1.2482,
+      "step": 4934
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.06731580775226245,
+      "learning_rate": 0.0004749390494720633,
+      "loss": 1.4638,
+      "step": 4935
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.07153188163916746,
+      "learning_rate": 0.00047476516249960994,
+      "loss": 1.4326,
+      "step": 4936
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.05895315648337665,
+      "learning_rate": 0.00047459127858696763,
+      "loss": 1.519,
+      "step": 4937
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.07150295093078468,
+      "learning_rate": 0.00047441739775522045,
+      "loss": 1.3454,
+      "step": 4938
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.06787792092572041,
+      "learning_rate": 0.000474243520025452,
+      "loss": 1.3111,
+      "step": 4939
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.06984567395943,
+      "learning_rate": 0.00047406964541874544,
+      "loss": 1.4355,
+      "step": 4940
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.07572599576818485,
+      "learning_rate": 0.00047389577395618387,
+      "loss": 1.5107,
+      "step": 4941
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.08604265411994148,
+      "learning_rate": 0.0004737219056588497,
+      "loss": 1.4844,
+      "step": 4942
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.06685170487349368,
+      "learning_rate": 0.00047354804054782493,
+      "loss": 1.3661,
+      "step": 4943
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.07421917272121109,
+      "learning_rate": 0.0004733741786441916,
+      "loss": 1.3045,
+      "step": 4944
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.08034262065898659,
+      "learning_rate": 0.00047320031996903094,
+      "loss": 1.3913,
+      "step": 4945
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.07111550377539654,
+      "learning_rate": 0.0004730264645434238,
+      "loss": 1.4248,
+      "step": 4946
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.07320459036647926,
+      "learning_rate": 0.0004728526123884508,
+      "loss": 1.378,
+      "step": 4947
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.08675173001306088,
+      "learning_rate": 0.0004726787635251921,
+      "loss": 1.3621,
+      "step": 4948
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.07183641238182674,
+      "learning_rate": 0.00047250491797472754,
+      "loss": 1.438,
+      "step": 4949
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0757586168872166,
+      "learning_rate": 0.00047233107575813657,
+      "loss": 1.5881,
+      "step": 4950
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.07513930507667875,
+      "learning_rate": 0.0004721572368964979,
+      "loss": 1.5212,
+      "step": 4951
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.08132533337284426,
+      "learning_rate": 0.0004719834014108903,
+      "loss": 1.4006,
+      "step": 4952
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.07089111386192162,
+      "learning_rate": 0.00047180956932239186,
+      "loss": 1.3653,
+      "step": 4953
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.07935963141630312,
+      "learning_rate": 0.00047163574065208034,
+      "loss": 1.4387,
+      "step": 4954
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.07808586869134204,
+      "learning_rate": 0.000471461915421033,
+      "loss": 1.4203,
+      "step": 4955
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.07325513089570637,
+      "learning_rate": 0.00047128809365032707,
+      "loss": 1.4985,
+      "step": 4956
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.08026244304016605,
+      "learning_rate": 0.0004711142753610385,
+      "loss": 1.3556,
+      "step": 4957
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.07566667413783172,
+      "learning_rate": 0.0004709404605742437,
+      "loss": 1.3653,
+      "step": 4958
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.07429766224344789,
+      "learning_rate": 0.0004707666493110182,
+      "loss": 1.509,
+      "step": 4959
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0794655393764005,
+      "learning_rate": 0.00047059284159243727,
+      "loss": 1.4117,
+      "step": 4960
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.07529539563266788,
+      "learning_rate": 0.0004704190374395757,
+      "loss": 1.3844,
+      "step": 4961
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.07016820804619064,
+      "learning_rate": 0.00047024523687350773,
+      "loss": 1.4303,
+      "step": 4962
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.07816173017864488,
+      "learning_rate": 0.0004700714399153075,
+      "loss": 1.3392,
+      "step": 4963
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.06755985966842121,
+      "learning_rate": 0.0004698976465860483,
+      "loss": 1.4734,
+      "step": 4964
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.07271241598715444,
+      "learning_rate": 0.0004697238569068033,
+      "loss": 1.3556,
+      "step": 4965
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.06993775956506902,
+      "learning_rate": 0.0004695500708986451,
+      "loss": 1.5511,
+      "step": 4966
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.07252382373357674,
+      "learning_rate": 0.00046937628858264555,
+      "loss": 1.4274,
+      "step": 4967
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.08483848022811791,
+      "learning_rate": 0.0004692025099798767,
+      "loss": 1.4444,
+      "step": 4968
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0715220291276646,
+      "learning_rate": 0.0004690287351114097,
+      "loss": 1.3493,
+      "step": 4969
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.07550992035600171,
+      "learning_rate": 0.00046885496399831536,
+      "loss": 1.3109,
+      "step": 4970
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.06715757190667862,
+      "learning_rate": 0.0004686811966616639,
+      "loss": 1.4589,
+      "step": 4971
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.08206633402247311,
+      "learning_rate": 0.00046850743312252537,
+      "loss": 1.332,
+      "step": 4972
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0812947283676224,
+      "learning_rate": 0.00046833367340196915,
+      "loss": 1.3324,
+      "step": 4973
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.08377563101034731,
+      "learning_rate": 0.0004681599175210641,
+      "loss": 1.4677,
+      "step": 4974
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.07932030491381303,
+      "learning_rate": 0.000467986165500879,
+      "loss": 1.3165,
+      "step": 4975
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.07764813651091716,
+      "learning_rate": 0.0004678124173624816,
+      "loss": 1.4508,
+      "step": 4976
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.06502997929922044,
+      "learning_rate": 0.00046763867312693975,
+      "loss": 1.4183,
+      "step": 4977
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.08892043190200546,
+      "learning_rate": 0.0004674649328153202,
+      "loss": 1.3739,
+      "step": 4978
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.07284899072991152,
+      "learning_rate": 0.0004672911964486896,
+      "loss": 1.2622,
+      "step": 4979
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.07680367290052509,
+      "learning_rate": 0.00046711746404811435,
+      "loss": 1.5053,
+      "step": 4980
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.07623392032659518,
+      "learning_rate": 0.0004669437356346599,
+      "loss": 1.3632,
+      "step": 4981
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.07287035493877488,
+      "learning_rate": 0.0004667700112293913,
+      "loss": 1.4439,
+      "step": 4982
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.07694402909619838,
+      "learning_rate": 0.0004665962908533736,
+      "loss": 1.2747,
+      "step": 4983
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.06519847555927114,
+      "learning_rate": 0.00046642257452767085,
+      "loss": 1.3187,
+      "step": 4984
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0753700653266978,
+      "learning_rate": 0.00046624886227334653,
+      "loss": 1.4432,
+      "step": 4985
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.06877063017007487,
+      "learning_rate": 0.0004660751541114641,
+      "loss": 1.4588,
+      "step": 4986
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.06914165470466452,
+      "learning_rate": 0.00046590145006308626,
+      "loss": 1.3748,
+      "step": 4987
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.07154339906462596,
+      "learning_rate": 0.0004657277501492751,
+      "loss": 1.3536,
+      "step": 4988
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.061994530518439867,
+      "learning_rate": 0.0004655540543910924,
+      "loss": 1.3233,
+      "step": 4989
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.07749942945527018,
+      "learning_rate": 0.0004653803628095992,
+      "loss": 1.4189,
+      "step": 4990
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.06924609370873705,
+      "learning_rate": 0.00046520667542585654,
+      "loss": 1.363,
+      "step": 4991
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0765738814212289,
+      "learning_rate": 0.0004650329922609244,
+      "loss": 1.3572,
+      "step": 4992
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.07121048885445819,
+      "learning_rate": 0.0004648593133358624,
+      "loss": 1.4481,
+      "step": 4993
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.06260534252699797,
+      "learning_rate": 0.0004646856386717299,
+      "loss": 1.4498,
+      "step": 4994
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.06744698219437496,
+      "learning_rate": 0.0004645119682895855,
+      "loss": 1.4486,
+      "step": 4995
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.07100221736713873,
+      "learning_rate": 0.0004643383022104871,
+      "loss": 1.4091,
+      "step": 4996
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.06257697855926746,
+      "learning_rate": 0.00046416464045549266,
+      "loss": 1.3761,
+      "step": 4997
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0644591791026162,
+      "learning_rate": 0.0004639909830456592,
+      "loss": 1.3716,
+      "step": 4998
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.06888699844748095,
+      "learning_rate": 0.0004638173300020431,
+      "loss": 1.4956,
+      "step": 4999
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.06980960282827073,
+      "learning_rate": 0.0004636436813457005,
+      "loss": 1.3363,
+      "step": 5000
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.06117394784086422,
+      "learning_rate": 0.0004634700370976867,
+      "loss": 1.5449,
+      "step": 5001
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.07078852629655827,
+      "learning_rate": 0.00046329639727905696,
+      "loss": 1.3434,
+      "step": 5002
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.06698154268968336,
+      "learning_rate": 0.00046312276191086567,
+      "loss": 1.4607,
+      "step": 5003
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.06296613954642143,
+      "learning_rate": 0.0004629491310141665,
+      "loss": 1.391,
+      "step": 5004
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.07509008854360427,
+      "learning_rate": 0.00046277550461001297,
+      "loss": 1.4132,
+      "step": 5005
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.06856227972641615,
+      "learning_rate": 0.00046260188271945784,
+      "loss": 1.3535,
+      "step": 5006
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.07113060912731874,
+      "learning_rate": 0.0004624282653635534,
+      "loss": 1.4244,
+      "step": 5007
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.08050206389441429,
+      "learning_rate": 0.00046225465256335117,
+      "loss": 1.4556,
+      "step": 5008
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.06897136804433396,
+      "learning_rate": 0.0004620810443399028,
+      "loss": 1.3154,
+      "step": 5009
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0757574536258104,
+      "learning_rate": 0.0004619074407142582,
+      "loss": 1.5634,
+      "step": 5010
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.07923391536996084,
+      "learning_rate": 0.0004617338417074679,
+      "loss": 1.4741,
+      "step": 5011
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.08502196581005937,
+      "learning_rate": 0.00046156024734058114,
+      "loss": 1.4186,
+      "step": 5012
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.07460264161817834,
+      "learning_rate": 0.0004613866576346468,
+      "loss": 1.3466,
+      "step": 5013
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.07965479727859294,
+      "learning_rate": 0.0004612130726107135,
+      "loss": 1.5325,
+      "step": 5014
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.07123399184880777,
+      "learning_rate": 0.0004610394922898289,
+      "loss": 1.36,
+      "step": 5015
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.07470424341472832,
+      "learning_rate": 0.00046086591669303997,
+      "loss": 1.2811,
+      "step": 5016
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.06729237043022075,
+      "learning_rate": 0.0004606923458413937,
+      "loss": 1.5231,
+      "step": 5017
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.07790469758464379,
+      "learning_rate": 0.000460518779755936,
+      "loss": 1.317,
+      "step": 5018
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0712544564620146,
+      "learning_rate": 0.00046034521845771256,
+      "loss": 1.4068,
+      "step": 5019
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.08929737083626135,
+      "learning_rate": 0.00046017166196776787,
+      "loss": 1.3364,
+      "step": 5020
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.06544812497801969,
+      "learning_rate": 0.00045999811030714643,
+      "loss": 1.4243,
+      "step": 5021
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.060540401781816446,
+      "learning_rate": 0.0004598245634968921,
+      "loss": 1.3251,
+      "step": 5022
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0753615240710179,
+      "learning_rate": 0.000459651021558048,
+      "loss": 1.3198,
+      "step": 5023
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.08191977033070631,
+      "learning_rate": 0.0004594774845116565,
+      "loss": 1.3843,
+      "step": 5024
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.07716411871829355,
+      "learning_rate": 0.00045930395237875983,
+      "loss": 1.2769,
+      "step": 5025
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.07458412119990497,
+      "learning_rate": 0.0004591304251803992,
+      "loss": 1.6653,
+      "step": 5026
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.06024873879839457,
+      "learning_rate": 0.0004589569029376153,
+      "loss": 1.2628,
+      "step": 5027
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.07220559994143401,
+      "learning_rate": 0.00045878338567144854,
+      "loss": 1.4473,
+      "step": 5028
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.08935952893918574,
+      "learning_rate": 0.0004586098734029384,
+      "loss": 1.4616,
+      "step": 5029
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.07031227068731213,
+      "learning_rate": 0.0004584363661531239,
+      "loss": 1.3846,
+      "step": 5030
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.07215927891814075,
+      "learning_rate": 0.00045826286394304316,
+      "loss": 1.4024,
+      "step": 5031
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.06502965527596266,
+      "learning_rate": 0.00045808936679373396,
+      "loss": 1.4927,
+      "step": 5032
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.07239411343302822,
+      "learning_rate": 0.00045791587472623365,
+      "loss": 1.3573,
+      "step": 5033
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.06741578912286643,
+      "learning_rate": 0.0004577423877615786,
+      "loss": 1.4525,
+      "step": 5034
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.08016804263999956,
+      "learning_rate": 0.00045756890592080473,
+      "loss": 1.6092,
+      "step": 5035
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.08362373973885047,
+      "learning_rate": 0.0004573954292249471,
+      "loss": 1.3283,
+      "step": 5036
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.07866004625026597,
+      "learning_rate": 0.00045722195769504084,
+      "loss": 1.4516,
+      "step": 5037
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.08537438521785196,
+      "learning_rate": 0.0004570484913521196,
+      "loss": 1.4975,
+      "step": 5038
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.07595030335723768,
+      "learning_rate": 0.00045687503021721673,
+      "loss": 1.4266,
+      "step": 5039
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.07736407471090755,
+      "learning_rate": 0.00045670157431136545,
+      "loss": 1.4982,
+      "step": 5040
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0772172433915742,
+      "learning_rate": 0.00045652812365559725,
+      "loss": 1.4539,
+      "step": 5041
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.06913636508981376,
+      "learning_rate": 0.00045635467827094404,
+      "loss": 1.3728,
+      "step": 5042
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.07308174202558122,
+      "learning_rate": 0.00045618123817843656,
+      "loss": 1.2952,
+      "step": 5043
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.07119469079079288,
+      "learning_rate": 0.00045600780339910487,
+      "loss": 1.4836,
+      "step": 5044
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.07229097508085704,
+      "learning_rate": 0.0004558343739539788,
+      "loss": 1.3501,
+      "step": 5045
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.07533651801310617,
+      "learning_rate": 0.00045566094986408716,
+      "loss": 1.506,
+      "step": 5046
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.07069151684931094,
+      "learning_rate": 0.0004554875311504581,
+      "loss": 1.421,
+      "step": 5047
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.07104602711266829,
+      "learning_rate": 0.0004553141178341195,
+      "loss": 1.302,
+      "step": 5048
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0678684179203935,
+      "learning_rate": 0.00045514070993609806,
+      "loss": 1.2245,
+      "step": 5049
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.06362739487382921,
+      "learning_rate": 0.00045496730747742023,
+      "loss": 1.4206,
+      "step": 5050
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.07856895251337907,
+      "learning_rate": 0.00045479391047911186,
+      "loss": 1.4577,
+      "step": 5051
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.07390180709879325,
+      "learning_rate": 0.00045462051896219736,
+      "loss": 1.2967,
+      "step": 5052
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.06240181363019398,
+      "learning_rate": 0.0004544471329477015,
+      "loss": 1.3616,
+      "step": 5053
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0769563634912301,
+      "learning_rate": 0.0004542737524566478,
+      "loss": 1.3491,
+      "step": 5054
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0772523406670056,
+      "learning_rate": 0.00045410037751005916,
+      "loss": 1.5153,
+      "step": 5055
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.08634109712464239,
+      "learning_rate": 0.0004539270081289581,
+      "loss": 1.4924,
+      "step": 5056
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.07431762471246574,
+      "learning_rate": 0.0004537536443343662,
+      "loss": 1.3931,
+      "step": 5057
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0733649512462882,
+      "learning_rate": 0.0004535802861473042,
+      "loss": 1.3972,
+      "step": 5058
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.07901391203531107,
+      "learning_rate": 0.0004534069335887926,
+      "loss": 1.4817,
+      "step": 5059
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.07381995246703327,
+      "learning_rate": 0.000453233586679851,
+      "loss": 1.3819,
+      "step": 5060
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.07247418114870308,
+      "learning_rate": 0.0004530602454414982,
+      "loss": 1.3994,
+      "step": 5061
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.08511342964700407,
+      "learning_rate": 0.00045288690989475264,
+      "loss": 1.5179,
+      "step": 5062
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.07305640500166688,
+      "learning_rate": 0.0004527135800606314,
+      "loss": 1.352,
+      "step": 5063
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.07150911228214982,
+      "learning_rate": 0.0004525402559601517,
+      "loss": 1.4765,
+      "step": 5064
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.06880615891115242,
+      "learning_rate": 0.0004523669376143296,
+      "loss": 1.383,
+      "step": 5065
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.07549278756591758,
+      "learning_rate": 0.00045219362504418057,
+      "loss": 1.4114,
+      "step": 5066
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.08469046092663751,
+      "learning_rate": 0.00045202031827071916,
+      "loss": 1.458,
+      "step": 5067
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.06922994658984552,
+      "learning_rate": 0.0004518470173149597,
+      "loss": 1.4044,
+      "step": 5068
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.07163348102127615,
+      "learning_rate": 0.00045167372219791544,
+      "loss": 1.3636,
+      "step": 5069
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.07201028025488586,
+      "learning_rate": 0.00045150043294059876,
+      "loss": 1.4034,
+      "step": 5070
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.07383927628148199,
+      "learning_rate": 0.000451327149564022,
+      "loss": 1.4078,
+      "step": 5071
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.07201378668042176,
+      "learning_rate": 0.00045115387208919625,
+      "loss": 1.4128,
+      "step": 5072
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.07129619592093052,
+      "learning_rate": 0.0004509806005371317,
+      "loss": 1.458,
+      "step": 5073
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.07247629019072475,
+      "learning_rate": 0.0004508073349288384,
+      "loss": 1.3639,
+      "step": 5074
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1810663332686453,
+      "learning_rate": 0.0004506340752853252,
+      "loss": 1.4445,
+      "step": 5075
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.07918478567312066,
+      "learning_rate": 0.0004504608216276007,
+      "loss": 1.3372,
+      "step": 5076
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.07803524195785769,
+      "learning_rate": 0.0004502875739766724,
+      "loss": 1.4658,
+      "step": 5077
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.07553645875301666,
+      "learning_rate": 0.00045011433235354697,
+      "loss": 1.372,
+      "step": 5078
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.06978572187693603,
+      "learning_rate": 0.0004499410967792308,
+      "loss": 1.4196,
+      "step": 5079
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.06894296278745332,
+      "learning_rate": 0.00044976786727472936,
+      "loss": 1.3758,
+      "step": 5080
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.07302702532691431,
+      "learning_rate": 0.000449594643861047,
+      "loss": 1.3609,
+      "step": 5081
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0809007009229191,
+      "learning_rate": 0.00044942142655918796,
+      "loss": 1.3801,
+      "step": 5082
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.07892839586023646,
+      "learning_rate": 0.0004492482153901554,
+      "loss": 1.4028,
+      "step": 5083
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.06926018562785509,
+      "learning_rate": 0.00044907501037495155,
+      "loss": 1.4091,
+      "step": 5084
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.07193791827021002,
+      "learning_rate": 0.0004489018115345784,
+      "loss": 1.2789,
+      "step": 5085
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.07040347068886654,
+      "learning_rate": 0.0004487286188900365,
+      "loss": 1.3098,
+      "step": 5086
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.07605857094439195,
+      "learning_rate": 0.00044855543246232653,
+      "loss": 1.3199,
+      "step": 5087
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.07217955296294648,
+      "learning_rate": 0.00044838225227244766,
+      "loss": 1.4811,
+      "step": 5088
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.07890617853340841,
+      "learning_rate": 0.0004482090783413986,
+      "loss": 1.3277,
+      "step": 5089
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0720778390267614,
+      "learning_rate": 0.00044803591069017746,
+      "loss": 1.5122,
+      "step": 5090
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.07516660854615728,
+      "learning_rate": 0.0004478627493397813,
+      "loss": 1.3591,
+      "step": 5091
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.07776960659986273,
+      "learning_rate": 0.0004476895943112064,
+      "loss": 1.504,
+      "step": 5092
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.06708427653195059,
+      "learning_rate": 0.0004475164456254488,
+      "loss": 1.3771,
+      "step": 5093
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0806630681787127,
+      "learning_rate": 0.0004473433033035028,
+      "loss": 1.4429,
+      "step": 5094
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.06961576483572854,
+      "learning_rate": 0.00044717016736636295,
+      "loss": 1.3978,
+      "step": 5095
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.07131127279311827,
+      "learning_rate": 0.0004469970378350224,
+      "loss": 1.3899,
+      "step": 5096
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.06955600958029026,
+      "learning_rate": 0.00044682391473047366,
+      "loss": 1.4512,
+      "step": 5097
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.06367222085226946,
+      "learning_rate": 0.0004466507980737087,
+      "loss": 1.3487,
+      "step": 5098
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.08530236840783734,
+      "learning_rate": 0.0004464776878857184,
+      "loss": 1.3804,
+      "step": 5099
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.07515745169686079,
+      "learning_rate": 0.000446304584187493,
+      "loss": 1.409,
+      "step": 5100
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.059871998178720805,
+      "learning_rate": 0.0004461314870000217,
+      "loss": 1.3639,
+      "step": 5101
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.06668678886268802,
+      "learning_rate": 0.0004459583963442935,
+      "loss": 1.4565,
+      "step": 5102
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.06941757215145417,
+      "learning_rate": 0.000445785312241296,
+      "loss": 1.4375,
+      "step": 5103
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0787820736556593,
+      "learning_rate": 0.0004456122347120164,
+      "loss": 1.4982,
+      "step": 5104
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.06498468989445924,
+      "learning_rate": 0.0004454391637774408,
+      "loss": 1.4231,
+      "step": 5105
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.07333559120436196,
+      "learning_rate": 0.0004452660994585545,
+      "loss": 1.4282,
+      "step": 5106
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.06590759319312026,
+      "learning_rate": 0.00044509304177634245,
+      "loss": 1.4366,
+      "step": 5107
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.07684068243714061,
+      "learning_rate": 0.00044491999075178844,
+      "loss": 1.4612,
+      "step": 5108
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.06923333908354005,
+      "learning_rate": 0.0004447469464058753,
+      "loss": 1.2806,
+      "step": 5109
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.07182215488774789,
+      "learning_rate": 0.00044457390875958546,
+      "loss": 1.6128,
+      "step": 5110
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.06531612126963693,
+      "learning_rate": 0.0004444008778339003,
+      "loss": 1.5235,
+      "step": 5111
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.06562690937863525,
+      "learning_rate": 0.0004442278536498003,
+      "loss": 1.4009,
+      "step": 5112
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.06630177290727778,
+      "learning_rate": 0.00044405483622826544,
+      "loss": 1.3705,
+      "step": 5113
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.07884292129674725,
+      "learning_rate": 0.0004438818255902746,
+      "loss": 1.4984,
+      "step": 5114
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.07161075616132401,
+      "learning_rate": 0.00044370882175680585,
+      "loss": 1.4684,
+      "step": 5115
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.07517122372057895,
+      "learning_rate": 0.00044353582474883645,
+      "loss": 1.3819,
+      "step": 5116
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.07602287354464461,
+      "learning_rate": 0.0004433628345873429,
+      "loss": 1.3158,
+      "step": 5117
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0617182529028056,
+      "learning_rate": 0.000443189851293301,
+      "loss": 1.4243,
+      "step": 5118
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.07441545210439454,
+      "learning_rate": 0.0004430168748876855,
+      "loss": 1.3834,
+      "step": 5119
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.06804643338631677,
+      "learning_rate": 0.00044284390539147024,
+      "loss": 1.4772,
+      "step": 5120
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.06762244884350038,
+      "learning_rate": 0.00044267094282562865,
+      "loss": 1.3884,
+      "step": 5121
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.07151377791730641,
+      "learning_rate": 0.00044249798721113286,
+      "loss": 1.3676,
+      "step": 5122
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.06859534272623603,
+      "learning_rate": 0.0004423250385689542,
+      "loss": 1.4373,
+      "step": 5123
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0777754907921359,
+      "learning_rate": 0.0004421520969200636,
+      "loss": 1.41,
+      "step": 5124
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.06955694265720186,
+      "learning_rate": 0.0004419791622854308,
+      "loss": 1.5175,
+      "step": 5125
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.06921231387512972,
+      "learning_rate": 0.00044180623468602457,
+      "loss": 1.3898,
+      "step": 5126
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0718189778864757,
+      "learning_rate": 0.00044163331414281287,
+      "loss": 1.3719,
+      "step": 5127
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.07074598625511523,
+      "learning_rate": 0.000441460400676763,
+      "loss": 1.5005,
+      "step": 5128
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.06483544761867031,
+      "learning_rate": 0.00044128749430884153,
+      "loss": 1.5393,
+      "step": 5129
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.06688587191843826,
+      "learning_rate": 0.00044111459506001373,
+      "loss": 1.3699,
+      "step": 5130
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.07058261521922415,
+      "learning_rate": 0.00044094170295124423,
+      "loss": 1.3599,
+      "step": 5131
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.08377792244328604,
+      "learning_rate": 0.0004407688180034968,
+      "loss": 1.5726,
+      "step": 5132
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.08170757981847657,
+      "learning_rate": 0.0004405959402377345,
+      "loss": 1.4028,
+      "step": 5133
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.07094933529123704,
+      "learning_rate": 0.00044042306967491927,
+      "loss": 1.4782,
+      "step": 5134
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.07833739407165695,
+      "learning_rate": 0.0004402502063360121,
+      "loss": 1.439,
+      "step": 5135
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.07612811567792596,
+      "learning_rate": 0.00044007735024197375,
+      "loss": 1.4726,
+      "step": 5136
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.07233376447979718,
+      "learning_rate": 0.0004399045014137629,
+      "loss": 1.484,
+      "step": 5137
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.07426330129609558,
+      "learning_rate": 0.00043973165987233853,
+      "loss": 1.4103,
+      "step": 5138
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0703597646707655,
+      "learning_rate": 0.00043955882563865824,
+      "loss": 1.406,
+      "step": 5139
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0747228033894531,
+      "learning_rate": 0.0004393859987336786,
+      "loss": 1.3999,
+      "step": 5140
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.06645446667750779,
+      "learning_rate": 0.0004392131791783556,
+      "loss": 1.4284,
+      "step": 5141
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.06742755209718568,
+      "learning_rate": 0.0004390403669936443,
+      "loss": 1.5311,
+      "step": 5142
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.06452897694766252,
+      "learning_rate": 0.0004388675622004985,
+      "loss": 1.5793,
+      "step": 5143
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.07123286746462744,
+      "learning_rate": 0.00043869476481987166,
+      "loss": 1.3647,
+      "step": 5144
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.080404488096446,
+      "learning_rate": 0.00043852197487271596,
+      "loss": 1.3003,
+      "step": 5145
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.06632696267396859,
+      "learning_rate": 0.00043834919237998275,
+      "loss": 1.3213,
+      "step": 5146
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.072257797841883,
+      "learning_rate": 0.0004381764173626225,
+      "loss": 1.4167,
+      "step": 5147
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.07509748873673347,
+      "learning_rate": 0.0004380036498415847,
+      "loss": 1.3869,
+      "step": 5148
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0757740088264892,
+      "learning_rate": 0.0004378308898378181,
+      "loss": 1.5651,
+      "step": 5149
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.08298039825124719,
+      "learning_rate": 0.0004376581373722705,
+      "loss": 1.5168,
+      "step": 5150
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.06581738179640959,
+      "learning_rate": 0.0004374853924658886,
+      "loss": 1.4759,
+      "step": 5151
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.07360615892941527,
+      "learning_rate": 0.00043731265513961837,
+      "loss": 1.4205,
+      "step": 5152
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0761582288989777,
+      "learning_rate": 0.00043713992541440495,
+      "loss": 1.4479,
+      "step": 5153
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.07168122590048633,
+      "learning_rate": 0.0004369672033111921,
+      "loss": 1.2982,
+      "step": 5154
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.06899343353227715,
+      "learning_rate": 0.0004367944888509233,
+      "loss": 1.4615,
+      "step": 5155
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0851060919458039,
+      "learning_rate": 0.00043662178205454064,
+      "loss": 1.416,
+      "step": 5156
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.07129039991770322,
+      "learning_rate": 0.0004364490829429855,
+      "loss": 1.4104,
+      "step": 5157
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.07358697165011492,
+      "learning_rate": 0.000436276391537198,
+      "loss": 1.4107,
+      "step": 5158
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.06873354857206569,
+      "learning_rate": 0.0004361037078581176,
+      "loss": 1.4866,
+      "step": 5159
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.06921469555044571,
+      "learning_rate": 0.00043593103192668306,
+      "loss": 1.2982,
+      "step": 5160
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.08636862001391342,
+      "learning_rate": 0.00043575836376383173,
+      "loss": 1.3917,
+      "step": 5161
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0737680093296688,
+      "learning_rate": 0.0004355857033905003,
+      "loss": 1.3104,
+      "step": 5162
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.07065380230712737,
+      "learning_rate": 0.0004354130508276243,
+      "loss": 1.412,
+      "step": 5163
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0777158993181712,
+      "learning_rate": 0.0004352404060961387,
+      "loss": 1.368,
+      "step": 5164
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.07163302597716577,
+      "learning_rate": 0.00043506776921697703,
+      "loss": 1.3967,
+      "step": 5165
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.07395361060929702,
+      "learning_rate": 0.0004348951402110721,
+      "loss": 1.4889,
+      "step": 5166
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.07668834064910103,
+      "learning_rate": 0.0004347225190993563,
+      "loss": 1.3945,
+      "step": 5167
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.07379835994602744,
+      "learning_rate": 0.00043454990590275966,
+      "loss": 1.2941,
+      "step": 5168
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.07353437475217274,
+      "learning_rate": 0.00043437730064221274,
+      "loss": 1.4592,
+      "step": 5169
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.07180642897850834,
+      "learning_rate": 0.00043420470333864437,
+      "loss": 1.5081,
+      "step": 5170
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.07480628068562017,
+      "learning_rate": 0.0004340321140129824,
+      "loss": 1.4688,
+      "step": 5171
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.08326786901526848,
+      "learning_rate": 0.0004338595326861542,
+      "loss": 1.3402,
+      "step": 5172
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0776630677788923,
+      "learning_rate": 0.00043368695937908564,
+      "loss": 1.396,
+      "step": 5173
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0858497428936398,
+      "learning_rate": 0.00043351439411270175,
+      "loss": 1.6149,
+      "step": 5174
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.07842818825426176,
+      "learning_rate": 0.00043334183690792687,
+      "loss": 1.3343,
+      "step": 5175
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.07189766698764516,
+      "learning_rate": 0.00043316928778568413,
+      "loss": 1.4638,
+      "step": 5176
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.08069092707873271,
+      "learning_rate": 0.0004329967467668955,
+      "loss": 1.6077,
+      "step": 5177
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.06730626948761761,
+      "learning_rate": 0.00043282421387248266,
+      "loss": 1.4001,
+      "step": 5178
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.09400270596405078,
+      "learning_rate": 0.0004326516891233652,
+      "loss": 1.4182,
+      "step": 5179
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.07152190135429132,
+      "learning_rate": 0.00043247917254046265,
+      "loss": 1.4702,
+      "step": 5180
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.07839367829975089,
+      "learning_rate": 0.0004323066641446932,
+      "loss": 1.3874,
+      "step": 5181
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.08025327287685355,
+      "learning_rate": 0.00043213416395697406,
+      "loss": 1.4811,
+      "step": 5182
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0675459205855955,
+      "learning_rate": 0.0004319616719982216,
+      "loss": 1.5319,
+      "step": 5183
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.08674747999573565,
+      "learning_rate": 0.00043178918828935093,
+      "loss": 1.3552,
+      "step": 5184
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.06783046466875754,
+      "learning_rate": 0.0004316167128512763,
+      "loss": 1.4993,
+      "step": 5185
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0766712737981568,
+      "learning_rate": 0.000431444245704911,
+      "loss": 1.3712,
+      "step": 5186
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.07620524476335197,
+      "learning_rate": 0.00043127178687116734,
+      "loss": 1.2,
+      "step": 5187
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0698091025429192,
+      "learning_rate": 0.0004310993363709563,
+      "loss": 1.451,
+      "step": 5188
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.07871163624673415,
+      "learning_rate": 0.0004309268942251887,
+      "loss": 1.4297,
+      "step": 5189
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.07820467834592452,
+      "learning_rate": 0.0004307544604547728,
+      "loss": 1.3821,
+      "step": 5190
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.07426071313960025,
+      "learning_rate": 0.00043058203508061755,
+      "loss": 1.5387,
+      "step": 5191
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.07152198924937515,
+      "learning_rate": 0.00043040961812362984,
+      "loss": 1.4379,
+      "step": 5192
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.07643513013877934,
+      "learning_rate": 0.00043023720960471567,
+      "loss": 1.3828,
+      "step": 5193
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.07377944276619808,
+      "learning_rate": 0.0004300648095447805,
+      "loss": 1.3577,
+      "step": 5194
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.07677807667832921,
+      "learning_rate": 0.0004298924179647283,
+      "loss": 1.4854,
+      "step": 5195
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.07254987146980628,
+      "learning_rate": 0.00042972003488546206,
+      "loss": 1.3919,
+      "step": 5196
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.07606417273762252,
+      "learning_rate": 0.00042954766032788386,
+      "loss": 1.4481,
+      "step": 5197
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.06574538650509663,
+      "learning_rate": 0.00042937529431289476,
+      "loss": 1.4083,
+      "step": 5198
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.08056330530383395,
+      "learning_rate": 0.00042920293686139483,
+      "loss": 1.3452,
+      "step": 5199
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.07254927775620691,
+      "learning_rate": 0.00042903058799428277,
+      "loss": 1.354,
+      "step": 5200
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.08042786998418212,
+      "learning_rate": 0.0004288582477324566,
+      "loss": 1.4644,
+      "step": 5201
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.07551326119414072,
+      "learning_rate": 0.0004286859160968131,
+      "loss": 1.4498,
+      "step": 5202
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0716710195185741,
+      "learning_rate": 0.0004285135931082481,
+      "loss": 1.3936,
+      "step": 5203
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.07607372161962338,
+      "learning_rate": 0.0004283412787876565,
+      "loss": 1.4397,
+      "step": 5204
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0767839216278687,
+      "learning_rate": 0.0004281689731559318,
+      "loss": 1.5554,
+      "step": 5205
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.08170307596491168,
+      "learning_rate": 0.00042799667623396676,
+      "loss": 1.4555,
+      "step": 5206
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.07293635890780277,
+      "learning_rate": 0.000427824388042653,
+      "loss": 1.5309,
+      "step": 5207
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.06361479379989111,
+      "learning_rate": 0.00042765210860288097,
+      "loss": 1.3286,
+      "step": 5208
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.06701607782127969,
+      "learning_rate": 0.0004274798379355402,
+      "loss": 1.3905,
+      "step": 5209
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.08632887914379973,
+      "learning_rate": 0.00042730757606151927,
+      "loss": 1.4785,
+      "step": 5210
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.06647493250126521,
+      "learning_rate": 0.0004271353230017052,
+      "loss": 1.2555,
+      "step": 5211
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.06693961197696197,
+      "learning_rate": 0.0004269630787769845,
+      "loss": 1.4008,
+      "step": 5212
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0692409003412069,
+      "learning_rate": 0.0004267908434082421,
+      "loss": 1.4041,
+      "step": 5213
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.06737205948239505,
+      "learning_rate": 0.0004266186169163624,
+      "loss": 1.3695,
+      "step": 5214
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.06829870460828749,
+      "learning_rate": 0.0004264463993222285,
+      "loss": 1.3776,
+      "step": 5215
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.06568111411168952,
+      "learning_rate": 0.0004262741906467221,
+      "loss": 1.5051,
+      "step": 5216
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.07099711966396452,
+      "learning_rate": 0.0004261019909107243,
+      "loss": 1.4215,
+      "step": 5217
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.06926480274268498,
+      "learning_rate": 0.000425929800135115,
+      "loss": 1.4874,
+      "step": 5218
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.07346867907033706,
+      "learning_rate": 0.0004257576183407726,
+      "loss": 1.4229,
+      "step": 5219
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.06602439589330132,
+      "learning_rate": 0.0004255854455485753,
+      "loss": 1.5083,
+      "step": 5220
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.08114794312950216,
+      "learning_rate": 0.0004254132817793989,
+      "loss": 1.3613,
+      "step": 5221
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.07671248645303533,
+      "learning_rate": 0.0004252411270541193,
+      "loss": 1.5036,
+      "step": 5222
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.07326874251575916,
+      "learning_rate": 0.0004250689813936108,
+      "loss": 1.4248,
+      "step": 5223
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.06407015535943295,
+      "learning_rate": 0.00042489684481874655,
+      "loss": 1.3891,
+      "step": 5224
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.06590779720525618,
+      "learning_rate": 0.00042472471735039894,
+      "loss": 1.2699,
+      "step": 5225
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.07001647041392903,
+      "learning_rate": 0.0004245525990094388,
+      "loss": 1.4988,
+      "step": 5226
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.06950346761352938,
+      "learning_rate": 0.00042438048981673613,
+      "loss": 1.5613,
+      "step": 5227
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0681780359829946,
+      "learning_rate": 0.00042420838979315975,
+      "loss": 1.3695,
+      "step": 5228
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.07072671929945357,
+      "learning_rate": 0.0004240362989595774,
+      "loss": 1.3276,
+      "step": 5229
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.07027533413197294,
+      "learning_rate": 0.00042386421733685574,
+      "loss": 1.414,
+      "step": 5230
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.07273910038980894,
+      "learning_rate": 0.00042369214494586025,
+      "loss": 1.3791,
+      "step": 5231
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.07673259671575565,
+      "learning_rate": 0.00042352008180745527,
+      "loss": 1.4342,
+      "step": 5232
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.08629006705556348,
+      "learning_rate": 0.0004233480279425039,
+      "loss": 1.4249,
+      "step": 5233
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.07021380822979276,
+      "learning_rate": 0.00042317598337186845,
+      "loss": 1.3425,
+      "step": 5234
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0696246874174527,
+      "learning_rate": 0.0004230039481164099,
+      "loss": 1.396,
+      "step": 5235
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0771553120990635,
+      "learning_rate": 0.00042283192219698797,
+      "loss": 1.3683,
+      "step": 5236
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.06880356851494933,
+      "learning_rate": 0.00042265990563446166,
+      "loss": 1.3027,
+      "step": 5237
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.07126039016589009,
+      "learning_rate": 0.0004224878984496884,
+      "loss": 1.4568,
+      "step": 5238
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.07841400172673146,
+      "learning_rate": 0.00042231590066352454,
+      "loss": 1.353,
+      "step": 5239
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0691154552365852,
+      "learning_rate": 0.00042214391229682564,
+      "loss": 1.3609,
+      "step": 5240
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.08822877788777062,
+      "learning_rate": 0.0004219719333704458,
+      "loss": 1.5357,
+      "step": 5241
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.07321921465130389,
+      "learning_rate": 0.00042179996390523817,
+      "loss": 1.4246,
+      "step": 5242
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.07663401686520896,
+      "learning_rate": 0.0004216280039220544,
+      "loss": 1.4507,
+      "step": 5243
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.06902686680633362,
+      "learning_rate": 0.00042145605344174524,
+      "loss": 1.3971,
+      "step": 5244
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.06877425014001001,
+      "learning_rate": 0.0004212841124851605,
+      "loss": 1.4818,
+      "step": 5245
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0796041285678449,
+      "learning_rate": 0.00042111218107314846,
+      "loss": 1.4862,
+      "step": 5246
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.07232671407705622,
+      "learning_rate": 0.00042094025922655636,
+      "loss": 1.493,
+      "step": 5247
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.06841978145994511,
+      "learning_rate": 0.0004207683469662305,
+      "loss": 1.61,
+      "step": 5248
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.08341542608065328,
+      "learning_rate": 0.0004205964443130157,
+      "loss": 1.4736,
+      "step": 5249
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.08253205503423554,
+      "learning_rate": 0.0004204245512877557,
+      "loss": 1.4058,
+      "step": 5250
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.07622013336203287,
+      "learning_rate": 0.00042025266791129325,
+      "loss": 1.3469,
+      "step": 5251
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.07590685646308382,
+      "learning_rate": 0.00042008079420446985,
+      "loss": 1.4038,
+      "step": 5252
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.06613348327522989,
+      "learning_rate": 0.0004199089301881256,
+      "loss": 1.2766,
+      "step": 5253
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.06636202703584451,
+      "learning_rate": 0.0004197370758830997,
+      "loss": 1.4253,
+      "step": 5254
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.08224651639354623,
+      "learning_rate": 0.0004195652313102299,
+      "loss": 1.3336,
+      "step": 5255
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07657486580957353,
+      "learning_rate": 0.00041939339649035325,
+      "loss": 1.4397,
+      "step": 5256
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07238429465194317,
+      "learning_rate": 0.0004192215714443052,
+      "loss": 1.3522,
+      "step": 5257
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07108040706690585,
+      "learning_rate": 0.00041904975619292003,
+      "loss": 1.4461,
+      "step": 5258
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07753429464243641,
+      "learning_rate": 0.00041887795075703095,
+      "loss": 1.4958,
+      "step": 5259
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.0755302483106443,
+      "learning_rate": 0.00041870615515747007,
+      "loss": 1.4597,
+      "step": 5260
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.06787790634763187,
+      "learning_rate": 0.00041853436941506817,
+      "loss": 1.3877,
+      "step": 5261
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.0661392999583233,
+      "learning_rate": 0.00041836259355065473,
+      "loss": 1.3623,
+      "step": 5262
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07417584421347115,
+      "learning_rate": 0.00041819082758505857,
+      "loss": 1.4985,
+      "step": 5263
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.0684843530224598,
+      "learning_rate": 0.00041801907153910623,
+      "loss": 1.3694,
+      "step": 5264
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.0726643145095135,
+      "learning_rate": 0.0004178473254336242,
+      "loss": 1.4311,
+      "step": 5265
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07855844880744575,
+      "learning_rate": 0.00041767558928943734,
+      "loss": 1.5054,
+      "step": 5266
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.06463843675977816,
+      "learning_rate": 0.0004175038631273689,
+      "loss": 1.4252,
+      "step": 5267
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07200443421609361,
+      "learning_rate": 0.0004173321469682415,
+      "loss": 1.5673,
+      "step": 5268
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07157206049559349,
+      "learning_rate": 0.0004171604408328764,
+      "loss": 1.4307,
+      "step": 5269
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07888986817590185,
+      "learning_rate": 0.00041698874474209327,
+      "loss": 1.355,
+      "step": 5270
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.08408446086913257,
+      "learning_rate": 0.0004168170587167111,
+      "loss": 1.5721,
+      "step": 5271
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07150037519993824,
+      "learning_rate": 0.0004166453827775474,
+      "loss": 1.4548,
+      "step": 5272
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.0723367901238978,
+      "learning_rate": 0.00041647371694541845,
+      "loss": 1.3577,
+      "step": 5273
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.0668665999900835,
+      "learning_rate": 0.00041630206124113923,
+      "loss": 1.3156,
+      "step": 5274
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07303698085136547,
+      "learning_rate": 0.0004161304156855235,
+      "loss": 1.3796,
+      "step": 5275
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07114217269771465,
+      "learning_rate": 0.00041595878029938415,
+      "loss": 1.4391,
+      "step": 5276
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07038735670023649,
+      "learning_rate": 0.0004157871551035324,
+      "loss": 1.3433,
+      "step": 5277
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.0765105156773969,
+      "learning_rate": 0.0004156155401187783,
+      "loss": 1.3778,
+      "step": 5278
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.06795495327597158,
+      "learning_rate": 0.00041544393536593096,
+      "loss": 1.3867,
+      "step": 5279
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07040727125995247,
+      "learning_rate": 0.000415272340865798,
+      "loss": 1.419,
+      "step": 5280
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07715273193285668,
+      "learning_rate": 0.0004151007566391857,
+      "loss": 1.3902,
+      "step": 5281
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07115238949389142,
+      "learning_rate": 0.0004149291827068995,
+      "loss": 1.3634,
+      "step": 5282
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.06990145449700264,
+      "learning_rate": 0.00041475761908974315,
+      "loss": 1.3727,
+      "step": 5283
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07417513717261448,
+      "learning_rate": 0.00041458606580851943,
+      "loss": 1.3928,
+      "step": 5284
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07342873182353658,
+      "learning_rate": 0.0004144145228840296,
+      "loss": 1.3938,
+      "step": 5285
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07729370889778275,
+      "learning_rate": 0.00041424299033707384,
+      "loss": 1.3694,
+      "step": 5286
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07807411980809978,
+      "learning_rate": 0.00041407146818845124,
+      "loss": 1.4838,
+      "step": 5287
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07769125995931969,
+      "learning_rate": 0.0004138999564589594,
+      "loss": 1.3846,
+      "step": 5288
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07203153523933757,
+      "learning_rate": 0.00041372845516939456,
+      "loss": 1.4529,
+      "step": 5289
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.0852466033918914,
+      "learning_rate": 0.000413556964340552,
+      "loss": 1.435,
+      "step": 5290
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07876953151957321,
+      "learning_rate": 0.0004133854839932256,
+      "loss": 1.3949,
+      "step": 5291
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07922352701188885,
+      "learning_rate": 0.00041321401414820785,
+      "loss": 1.3702,
+      "step": 5292
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07789017134336326,
+      "learning_rate": 0.00041304255482628997,
+      "loss": 1.399,
+      "step": 5293
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07120922513834405,
+      "learning_rate": 0.00041287110604826233,
+      "loss": 1.2416,
+      "step": 5294
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.09864215287276547,
+      "learning_rate": 0.0004126996678349133,
+      "loss": 1.4479,
+      "step": 5295
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07540022268253116,
+      "learning_rate": 0.0004125282402070306,
+      "loss": 1.5072,
+      "step": 5296
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07117404624697435,
+      "learning_rate": 0.0004123568231854003,
+      "loss": 1.4258,
+      "step": 5297
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07632072817859777,
+      "learning_rate": 0.00041218541679080724,
+      "loss": 1.3239,
+      "step": 5298
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07545200184891443,
+      "learning_rate": 0.00041201402104403516,
+      "loss": 1.4645,
+      "step": 5299
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.0755586931749317,
+      "learning_rate": 0.00041184263596586644,
+      "loss": 1.3638,
+      "step": 5300
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07534667284843548,
+      "learning_rate": 0.00041167126157708194,
+      "loss": 1.2639,
+      "step": 5301
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07371800282548627,
+      "learning_rate": 0.0004114998978984616,
+      "loss": 1.3628,
+      "step": 5302
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.06979442007264192,
+      "learning_rate": 0.0004113285449507837,
+      "loss": 1.5124,
+      "step": 5303
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07518472070533627,
+      "learning_rate": 0.00041115720275482535,
+      "loss": 1.3375,
+      "step": 5304
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07145575246092586,
+      "learning_rate": 0.0004109858713313628,
+      "loss": 1.268,
+      "step": 5305
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07227095780800502,
+      "learning_rate": 0.0004108145507011698,
+      "loss": 1.4299,
+      "step": 5306
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.06962460632342404,
+      "learning_rate": 0.0004106432408850202,
+      "loss": 1.3067,
+      "step": 5307
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.06967858569287429,
+      "learning_rate": 0.0004104719419036856,
+      "loss": 1.4634,
+      "step": 5308
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07452803560588936,
+      "learning_rate": 0.00041030065377793673,
+      "loss": 1.2936,
+      "step": 5309
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07198736708156497,
+      "learning_rate": 0.0004101293765285429,
+      "loss": 1.399,
+      "step": 5310
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07088761102492468,
+      "learning_rate": 0.00040995811017627195,
+      "loss": 1.5214,
+      "step": 5311
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.08037863075518545,
+      "learning_rate": 0.0004097868547418905,
+      "loss": 1.538,
+      "step": 5312
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.06702118564346261,
+      "learning_rate": 0.00040961561024616393,
+      "loss": 1.4452,
+      "step": 5313
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07138907507596554,
+      "learning_rate": 0.00040944437670985636,
+      "loss": 1.4375,
+      "step": 5314
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.0713287767088554,
+      "learning_rate": 0.00040927315415373013,
+      "loss": 1.3658,
+      "step": 5315
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07246596365203058,
+      "learning_rate": 0.00040910194259854705,
+      "loss": 1.3858,
+      "step": 5316
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.0638642019848533,
+      "learning_rate": 0.00040893074206506646,
+      "loss": 1.449,
+      "step": 5317
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07467248996983068,
+      "learning_rate": 0.0004087595525740475,
+      "loss": 1.45,
+      "step": 5318
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.06856809835411172,
+      "learning_rate": 0.0004085883741462474,
+      "loss": 1.4768,
+      "step": 5319
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07499617753743705,
+      "learning_rate": 0.00040841720680242187,
+      "loss": 1.4109,
+      "step": 5320
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07075707467749452,
+      "learning_rate": 0.0004082460505633259,
+      "loss": 1.5193,
+      "step": 5321
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.08615689422162827,
+      "learning_rate": 0.0004080749054497126,
+      "loss": 1.3668,
+      "step": 5322
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.08010696554699227,
+      "learning_rate": 0.00040790377148233406,
+      "loss": 1.5512,
+      "step": 5323
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.06867799500138569,
+      "learning_rate": 0.00040773264868194047,
+      "loss": 1.5605,
+      "step": 5324
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07358887873840544,
+      "learning_rate": 0.0004075615370692815,
+      "loss": 1.3062,
+      "step": 5325
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07190535444709516,
+      "learning_rate": 0.0004073904366651049,
+      "loss": 1.5042,
+      "step": 5326
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07395900114719453,
+      "learning_rate": 0.0004072193474901571,
+      "loss": 1.4081,
+      "step": 5327
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07036662502059775,
+      "learning_rate": 0.0004070482695651832,
+      "loss": 1.4774,
+      "step": 5328
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.0837011412853149,
+      "learning_rate": 0.000406877202910927,
+      "loss": 1.4898,
+      "step": 5329
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.06325539977822217,
+      "learning_rate": 0.00040670614754813105,
+      "loss": 1.3729,
+      "step": 5330
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.06890633970155305,
+      "learning_rate": 0.0004065351034975364,
+      "loss": 1.3829,
+      "step": 5331
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.06839743814713234,
+      "learning_rate": 0.0004063640707798826,
+      "loss": 1.5331,
+      "step": 5332
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07358862957787274,
+      "learning_rate": 0.00040619304941590806,
+      "loss": 1.3844,
+      "step": 5333
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.08022228572519535,
+      "learning_rate": 0.00040602203942634974,
+      "loss": 1.4749,
+      "step": 5334
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07053418115513409,
+      "learning_rate": 0.00040585104083194296,
+      "loss": 1.5135,
+      "step": 5335
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07580591745380308,
+      "learning_rate": 0.0004056800536534222,
+      "loss": 1.476,
+      "step": 5336
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.0740733400567423,
+      "learning_rate": 0.0004055090779115202,
+      "loss": 1.4435,
+      "step": 5337
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07623581006510774,
+      "learning_rate": 0.0004053381136269682,
+      "loss": 1.3887,
+      "step": 5338
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07726090719594393,
+      "learning_rate": 0.0004051671608204962,
+      "loss": 1.3204,
+      "step": 5339
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07643636380776819,
+      "learning_rate": 0.00040499621951283285,
+      "loss": 1.5598,
+      "step": 5340
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.06916237569138638,
+      "learning_rate": 0.0004048252897247054,
+      "loss": 1.2734,
+      "step": 5341
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07440005449676713,
+      "learning_rate": 0.00040465437147683985,
+      "loss": 1.477,
+      "step": 5342
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.0768591188002704,
+      "learning_rate": 0.0004044834647899603,
+      "loss": 1.4269,
+      "step": 5343
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07706826003310004,
+      "learning_rate": 0.00040431256968479,
+      "loss": 1.3852,
+      "step": 5344
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07365496429342674,
+      "learning_rate": 0.00040414168618205053,
+      "loss": 1.5324,
+      "step": 5345
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.06961222207303024,
+      "learning_rate": 0.000403970814302462,
+      "loss": 1.4392,
+      "step": 5346
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07663822923497113,
+      "learning_rate": 0.0004037999540667436,
+      "loss": 1.3213,
+      "step": 5347
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.08041052042443397,
+      "learning_rate": 0.00040362910549561216,
+      "loss": 1.4161,
+      "step": 5348
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.07464715250262671,
+      "learning_rate": 0.000403458268609784,
+      "loss": 1.5022,
+      "step": 5349
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.06798489895346704,
+      "learning_rate": 0.0004032874434299735,
+      "loss": 1.329,
+      "step": 5350
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.07330991474639785,
+      "learning_rate": 0.0004031166299768939,
+      "loss": 1.4582,
+      "step": 5351
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.07128785978243318,
+      "learning_rate": 0.000402945828271257,
+      "loss": 1.3826,
+      "step": 5352
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.07672704144785629,
+      "learning_rate": 0.000402775038333773,
+      "loss": 1.3935,
+      "step": 5353
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.06971968315018112,
+      "learning_rate": 0.00040260426018515086,
+      "loss": 1.3436,
+      "step": 5354
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.08551247843667656,
+      "learning_rate": 0.0004024334938460978,
+      "loss": 1.5431,
+      "step": 5355
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.06496244821036254,
+      "learning_rate": 0.00040226273933732006,
+      "loss": 1.4034,
+      "step": 5356
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.06880854364756596,
+      "learning_rate": 0.0004020919966795221,
+      "loss": 1.3686,
+      "step": 5357
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.061631643039155914,
+      "learning_rate": 0.0004019212658934073,
+      "loss": 1.4865,
+      "step": 5358
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.0876656022645856,
+      "learning_rate": 0.0004017505469996769,
+      "loss": 1.4607,
+      "step": 5359
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.07698134695084392,
+      "learning_rate": 0.0004015798400190314,
+      "loss": 1.3628,
+      "step": 5360
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.07046999280271365,
+      "learning_rate": 0.0004014091449721696,
+      "loss": 1.3769,
+      "step": 5361
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.07410553795969785,
+      "learning_rate": 0.00040123846187978897,
+      "loss": 1.337,
+      "step": 5362
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.07474813390239679,
+      "learning_rate": 0.00040106779076258515,
+      "loss": 1.3158,
+      "step": 5363
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.07356400677418831,
+      "learning_rate": 0.00040089713164125285,
+      "loss": 1.4531,
+      "step": 5364
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.08282254384195048,
+      "learning_rate": 0.00040072648453648506,
+      "loss": 1.4346,
+      "step": 5365
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.08105180775455986,
+      "learning_rate": 0.00040055584946897317,
+      "loss": 1.5177,
+      "step": 5366
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.06655886062838566,
+      "learning_rate": 0.00040038522645940744,
+      "loss": 1.3714,
+      "step": 5367
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.0762914733486254,
+      "learning_rate": 0.0004002146155284764,
+      "loss": 1.355,
+      "step": 5368
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.07308623399730424,
+      "learning_rate": 0.0004000440166968674,
+      "loss": 1.3991,
+      "step": 5369
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.07344200850911864,
+      "learning_rate": 0.0003998734299852659,
+      "loss": 1.4345,
+      "step": 5370
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.07950989021718434,
+      "learning_rate": 0.00039970285541435604,
+      "loss": 1.2364,
+      "step": 5371
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.06865838341036416,
+      "learning_rate": 0.0003995322930048208,
+      "loss": 1.3799,
+      "step": 5372
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.08467264426498419,
+      "learning_rate": 0.0003993617427773415,
+      "loss": 1.4482,
+      "step": 5373
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.07733714287405516,
+      "learning_rate": 0.0003991912047525976,
+      "loss": 1.4416,
+      "step": 5374
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.11476822294757576,
+      "learning_rate": 0.0003990206789512679,
+      "loss": 1.4573,
+      "step": 5375
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.07610655086519266,
+      "learning_rate": 0.00039885016539402895,
+      "loss": 1.3066,
+      "step": 5376
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.07716286349323345,
+      "learning_rate": 0.0003986796641015561,
+      "loss": 1.3914,
+      "step": 5377
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.08078118508279136,
+      "learning_rate": 0.0003985091750945233,
+      "loss": 1.3834,
+      "step": 5378
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.08079701422701666,
+      "learning_rate": 0.0003983386983936031,
+      "loss": 1.3173,
+      "step": 5379
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.07174180247292715,
+      "learning_rate": 0.0003981682340194661,
+      "loss": 1.6485,
+      "step": 5380
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.08102967123260112,
+      "learning_rate": 0.00039799778199278176,
+      "loss": 1.4069,
+      "step": 5381
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.08146881326489869,
+      "learning_rate": 0.0003978273423342179,
+      "loss": 1.3601,
+      "step": 5382
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.07227396353091131,
+      "learning_rate": 0.00039765691506444113,
+      "loss": 1.5108,
+      "step": 5383
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.07519482559814357,
+      "learning_rate": 0.00039748650020411625,
+      "loss": 1.3298,
+      "step": 5384
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.06780383496089704,
+      "learning_rate": 0.00039731609777390647,
+      "loss": 1.4425,
+      "step": 5385
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.06902960397709067,
+      "learning_rate": 0.000397145707794474,
+      "loss": 1.3888,
+      "step": 5386
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.06986637026932582,
+      "learning_rate": 0.0003969753302864791,
+      "loss": 1.3675,
+      "step": 5387
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.0691980178959223,
+      "learning_rate": 0.00039680496527058054,
+      "loss": 1.3784,
+      "step": 5388
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.08156045851765133,
+      "learning_rate": 0.00039663461276743557,
+      "loss": 1.4175,
+      "step": 5389
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.0704137566422376,
+      "learning_rate": 0.00039646427279770035,
+      "loss": 1.4708,
+      "step": 5390
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.06468213372241707,
+      "learning_rate": 0.000396293945382029,
+      "loss": 1.3275,
+      "step": 5391
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.07230454038232008,
+      "learning_rate": 0.00039612363054107423,
+      "loss": 1.4686,
+      "step": 5392
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.06080188424261516,
+      "learning_rate": 0.00039595332829548736,
+      "loss": 1.3506,
+      "step": 5393
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.08481811022519306,
+      "learning_rate": 0.00039578303866591804,
+      "loss": 1.4125,
+      "step": 5394
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.07164820257648974,
+      "learning_rate": 0.00039561276167301475,
+      "loss": 1.398,
+      "step": 5395
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.07295058056378592,
+      "learning_rate": 0.0003954424973374239,
+      "loss": 1.4414,
+      "step": 5396
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.07617537325382101,
+      "learning_rate": 0.0003952722456797906,
+      "loss": 1.5086,
+      "step": 5397
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.0723477940325076,
+      "learning_rate": 0.0003951020067207587,
+      "loss": 1.4446,
+      "step": 5398
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.08723825363550239,
+      "learning_rate": 0.0003949317804809701,
+      "loss": 1.3639,
+      "step": 5399
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.06936075965065762,
+      "learning_rate": 0.00039476156698106554,
+      "loss": 1.3929,
+      "step": 5400
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.07934707933846326,
+      "learning_rate": 0.0003945913662416836,
+      "loss": 1.3018,
+      "step": 5401
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.07516365654439944,
+      "learning_rate": 0.0003944211782834618,
+      "loss": 1.4766,
+      "step": 5402
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.07511497051173129,
+      "learning_rate": 0.00039425100312703626,
+      "loss": 1.446,
+      "step": 5403
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.07399958481633905,
+      "learning_rate": 0.00039408084079304114,
+      "loss": 1.4446,
+      "step": 5404
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.07322395813105882,
+      "learning_rate": 0.0003939106913021091,
+      "loss": 1.4424,
+      "step": 5405
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.08006397128046831,
+      "learning_rate": 0.00039374055467487163,
+      "loss": 1.4953,
+      "step": 5406
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.08113213797120367,
+      "learning_rate": 0.00039357043093195824,
+      "loss": 1.4554,
+      "step": 5407
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.08034824832299563,
+      "learning_rate": 0.0003934003200939969,
+      "loss": 1.2838,
+      "step": 5408
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.07582701660283922,
+      "learning_rate": 0.00039323022218161436,
+      "loss": 1.3473,
+      "step": 5409
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.07262946774124066,
+      "learning_rate": 0.0003930601372154355,
+      "loss": 1.4852,
+      "step": 5410
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.0816848756450506,
+      "learning_rate": 0.00039289006521608374,
+      "loss": 1.3998,
+      "step": 5411
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.08026040007883284,
+      "learning_rate": 0.00039272000620418077,
+      "loss": 1.3904,
+      "step": 5412
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.08005161253565878,
+      "learning_rate": 0.0003925499602003467,
+      "loss": 1.3568,
+      "step": 5413
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.08971571870062349,
+      "learning_rate": 0.00039237992722520064,
+      "loss": 1.3311,
+      "step": 5414
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.0753830626442861,
+      "learning_rate": 0.00039220990729935937,
+      "loss": 1.4081,
+      "step": 5415
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.08683335516423724,
+      "learning_rate": 0.0003920399004434383,
+      "loss": 1.4205,
+      "step": 5416
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.08114337234622578,
+      "learning_rate": 0.0003918699066780517,
+      "loss": 1.4885,
+      "step": 5417
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.07527136672140505,
+      "learning_rate": 0.0003916999260238115,
+      "loss": 1.3366,
+      "step": 5418
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.07529646089819533,
+      "learning_rate": 0.0003915299585013288,
+      "loss": 1.4195,
+      "step": 5419
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.08565027017623841,
+      "learning_rate": 0.00039136000413121246,
+      "loss": 1.4588,
+      "step": 5420
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.07864361897315884,
+      "learning_rate": 0.00039119006293407025,
+      "loss": 1.4351,
+      "step": 5421
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.07915941892313927,
+      "learning_rate": 0.0003910201349305081,
+      "loss": 1.3672,
+      "step": 5422
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.07156264135526873,
+      "learning_rate": 0.0003908502201411303,
+      "loss": 1.4315,
+      "step": 5423
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.07618419764564519,
+      "learning_rate": 0.0003906803185865395,
+      "loss": 1.5126,
+      "step": 5424
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.0814665018963305,
+      "learning_rate": 0.0003905104302873369,
+      "loss": 1.467,
+      "step": 5425
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.06795989374856158,
+      "learning_rate": 0.0003903405552641222,
+      "loss": 1.4223,
+      "step": 5426
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.07673542464248548,
+      "learning_rate": 0.00039017069353749324,
+      "loss": 1.3443,
+      "step": 5427
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.07803117959879066,
+      "learning_rate": 0.00039000084512804615,
+      "loss": 1.4296,
+      "step": 5428
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.07549394873728246,
+      "learning_rate": 0.000389831010056376,
+      "loss": 1.4681,
+      "step": 5429
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.06315628829607588,
+      "learning_rate": 0.00038966118834307564,
+      "loss": 1.4331,
+      "step": 5430
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.07803465268274827,
+      "learning_rate": 0.0003894913800087365,
+      "loss": 1.3954,
+      "step": 5431
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.0775945281904031,
+      "learning_rate": 0.00038932158507394885,
+      "loss": 1.5557,
+      "step": 5432
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.07374023221278247,
+      "learning_rate": 0.0003891518035593002,
+      "loss": 1.4375,
+      "step": 5433
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.06745322580856865,
+      "learning_rate": 0.00038898203548537767,
+      "loss": 1.4085,
+      "step": 5434
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.07268110095101987,
+      "learning_rate": 0.0003888122808727661,
+      "loss": 1.2798,
+      "step": 5435
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.06780737778028098,
+      "learning_rate": 0.00038864253974204866,
+      "loss": 1.42,
+      "step": 5436
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.06232751201032706,
+      "learning_rate": 0.0003884728121138073,
+      "loss": 1.368,
+      "step": 5437
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.07425964477371508,
+      "learning_rate": 0.00038830309800862194,
+      "loss": 1.3845,
+      "step": 5438
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.0696927104312303,
+      "learning_rate": 0.00038813339744707095,
+      "loss": 1.5026,
+      "step": 5439
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.08829232052578292,
+      "learning_rate": 0.0003879637104497313,
+      "loss": 1.459,
+      "step": 5440
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.0800197842374922,
+      "learning_rate": 0.00038779403703717797,
+      "loss": 1.3374,
+      "step": 5441
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.08461196112841203,
+      "learning_rate": 0.0003876243772299843,
+      "loss": 1.4803,
+      "step": 5442
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07591094969169548,
+      "learning_rate": 0.00038745473104872277,
+      "loss": 1.4347,
+      "step": 5443
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.06951993132870146,
+      "learning_rate": 0.00038728509851396274,
+      "loss": 1.3894,
+      "step": 5444
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.0709955709325333,
+      "learning_rate": 0.0003871154796462732,
+      "loss": 1.4263,
+      "step": 5445
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07217695702237546,
+      "learning_rate": 0.00038694587446622093,
+      "loss": 1.4771,
+      "step": 5446
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07624851858056204,
+      "learning_rate": 0.000386776282994371,
+      "loss": 1.5054,
+      "step": 5447
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.0805471238267498,
+      "learning_rate": 0.0003866067052512872,
+      "loss": 1.4343,
+      "step": 5448
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07193135020200961,
+      "learning_rate": 0.0003864371412575314,
+      "loss": 1.4548,
+      "step": 5449
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07357020438933615,
+      "learning_rate": 0.00038626759103366367,
+      "loss": 1.4873,
+      "step": 5450
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.08159699049650707,
+      "learning_rate": 0.0003860980546002425,
+      "loss": 1.4111,
+      "step": 5451
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.08174122049659675,
+      "learning_rate": 0.00038592853197782503,
+      "loss": 1.3474,
+      "step": 5452
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07962770380661251,
+      "learning_rate": 0.0003857590231869664,
+      "loss": 1.3244,
+      "step": 5453
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07448555257556383,
+      "learning_rate": 0.00038558952824821994,
+      "loss": 1.4534,
+      "step": 5454
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07277064642875133,
+      "learning_rate": 0.0003854200471821376,
+      "loss": 1.4508,
+      "step": 5455
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.08151762819863291,
+      "learning_rate": 0.0003852505800092695,
+      "loss": 1.3559,
+      "step": 5456
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.08049679696768172,
+      "learning_rate": 0.0003850811267501643,
+      "loss": 1.3082,
+      "step": 5457
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.06983411242009606,
+      "learning_rate": 0.0003849116874253686,
+      "loss": 1.4051,
+      "step": 5458
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07159626918035832,
+      "learning_rate": 0.0003847422620554276,
+      "loss": 1.4298,
+      "step": 5459
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.08503770511535662,
+      "learning_rate": 0.00038457285066088463,
+      "loss": 1.3354,
+      "step": 5460
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.0790162568405594,
+      "learning_rate": 0.00038440345326228156,
+      "loss": 1.3853,
+      "step": 5461
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.0782423315532537,
+      "learning_rate": 0.0003842340698801581,
+      "loss": 1.4141,
+      "step": 5462
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07828919322816909,
+      "learning_rate": 0.0003840647005350529,
+      "loss": 1.3817,
+      "step": 5463
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07866537991857481,
+      "learning_rate": 0.00038389534524750266,
+      "loss": 1.2861,
+      "step": 5464
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07442515151685897,
+      "learning_rate": 0.00038372600403804185,
+      "loss": 1.4331,
+      "step": 5465
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.08202431853846086,
+      "learning_rate": 0.000383556676927204,
+      "loss": 1.3602,
+      "step": 5466
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.08207556685486656,
+      "learning_rate": 0.0003833873639355203,
+      "loss": 1.4111,
+      "step": 5467
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07216498981129676,
+      "learning_rate": 0.000383218065083521,
+      "loss": 1.3615,
+      "step": 5468
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07135005849767023,
+      "learning_rate": 0.0003830487803917338,
+      "loss": 1.4001,
+      "step": 5469
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07557193610406127,
+      "learning_rate": 0.00038287950988068507,
+      "loss": 1.4354,
+      "step": 5470
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07651272647178654,
+      "learning_rate": 0.0003827102535708997,
+      "loss": 1.4316,
+      "step": 5471
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07205282353825856,
+      "learning_rate": 0.00038254101148290043,
+      "loss": 1.6366,
+      "step": 5472
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.08323470409611829,
+      "learning_rate": 0.00038237178363720835,
+      "loss": 1.413,
+      "step": 5473
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.06608968374186841,
+      "learning_rate": 0.00038220257005434314,
+      "loss": 1.3479,
+      "step": 5474
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.08621171799268833,
+      "learning_rate": 0.00038203337075482245,
+      "loss": 1.4297,
+      "step": 5475
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.06968222782720418,
+      "learning_rate": 0.00038186418575916227,
+      "loss": 1.3863,
+      "step": 5476
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.08547720995611516,
+      "learning_rate": 0.0003816950150878768,
+      "loss": 1.4868,
+      "step": 5477
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07746098530633695,
+      "learning_rate": 0.0003815258587614785,
+      "loss": 1.3578,
+      "step": 5478
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07851996940602321,
+      "learning_rate": 0.00038135671680047836,
+      "loss": 1.3269,
+      "step": 5479
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.0671031884146776,
+      "learning_rate": 0.00038118758922538534,
+      "loss": 1.3328,
+      "step": 5480
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07086563839326887,
+      "learning_rate": 0.0003810184760567066,
+      "loss": 1.3801,
+      "step": 5481
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.06829227406937718,
+      "learning_rate": 0.00038084937731494795,
+      "loss": 1.3466,
+      "step": 5482
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07644260647676682,
+      "learning_rate": 0.0003806802930206131,
+      "loss": 1.463,
+      "step": 5483
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07327912352959912,
+      "learning_rate": 0.00038051122319420405,
+      "loss": 1.5011,
+      "step": 5484
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07474131903712519,
+      "learning_rate": 0.00038034216785622126,
+      "loss": 1.4398,
+      "step": 5485
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.06789310016224517,
+      "learning_rate": 0.0003801731270271629,
+      "loss": 1.4834,
+      "step": 5486
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07304893429551268,
+      "learning_rate": 0.0003800041007275261,
+      "loss": 1.4962,
+      "step": 5487
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.06520647255990622,
+      "learning_rate": 0.00037983508897780583,
+      "loss": 1.3565,
+      "step": 5488
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.06924010423700398,
+      "learning_rate": 0.00037966609179849526,
+      "loss": 1.3497,
+      "step": 5489
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07253469655081643,
+      "learning_rate": 0.0003794971092100858,
+      "loss": 1.363,
+      "step": 5490
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.06834466384222009,
+      "learning_rate": 0.00037932814123306735,
+      "loss": 1.4292,
+      "step": 5491
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07204827925530317,
+      "learning_rate": 0.00037915918788792793,
+      "loss": 1.5018,
+      "step": 5492
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07344716634947808,
+      "learning_rate": 0.00037899024919515334,
+      "loss": 1.38,
+      "step": 5493
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07802721400721604,
+      "learning_rate": 0.0003788213251752284,
+      "loss": 1.4881,
+      "step": 5494
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07188704680601309,
+      "learning_rate": 0.00037865241584863547,
+      "loss": 1.3633,
+      "step": 5495
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.06597149494787036,
+      "learning_rate": 0.0003784835212358556,
+      "loss": 1.4876,
+      "step": 5496
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07149421228145474,
+      "learning_rate": 0.00037831464135736764,
+      "loss": 1.4034,
+      "step": 5497
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07525825575415172,
+      "learning_rate": 0.00037814577623364884,
+      "loss": 1.2948,
+      "step": 5498
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.08479029685348394,
+      "learning_rate": 0.00037797692588517487,
+      "loss": 1.4546,
+      "step": 5499
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07051587577136423,
+      "learning_rate": 0.00037780809033241926,
+      "loss": 1.5106,
+      "step": 5500
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07380593539742977,
+      "learning_rate": 0.00037763926959585387,
+      "loss": 1.3892,
+      "step": 5501
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.06881571512378142,
+      "learning_rate": 0.000377470463695949,
+      "loss": 1.4178,
+      "step": 5502
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07028405289575519,
+      "learning_rate": 0.0003773016726531728,
+      "loss": 1.3714,
+      "step": 5503
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.0666545859612911,
+      "learning_rate": 0.00037713289648799177,
+      "loss": 1.4037,
+      "step": 5504
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.06767247145129425,
+      "learning_rate": 0.00037696413522087067,
+      "loss": 1.2454,
+      "step": 5505
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07342524302568901,
+      "learning_rate": 0.00037679538887227246,
+      "loss": 1.3988,
+      "step": 5506
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07264381277887294,
+      "learning_rate": 0.000376626657462658,
+      "loss": 1.5017,
+      "step": 5507
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.06619640197119608,
+      "learning_rate": 0.00037645794101248665,
+      "loss": 1.3296,
+      "step": 5508
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.08066585644997017,
+      "learning_rate": 0.00037628923954221574,
+      "loss": 1.4821,
+      "step": 5509
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.0779357570182373,
+      "learning_rate": 0.00037612055307230114,
+      "loss": 1.3817,
+      "step": 5510
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.08662510505643409,
+      "learning_rate": 0.0003759518816231966,
+      "loss": 1.2358,
+      "step": 5511
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.06984981321199445,
+      "learning_rate": 0.0003757832252153539,
+      "loss": 1.2271,
+      "step": 5512
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07931974426531138,
+      "learning_rate": 0.00037561458386922356,
+      "loss": 1.4736,
+      "step": 5513
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.08119674665971423,
+      "learning_rate": 0.0003754459576052537,
+      "loss": 1.399,
+      "step": 5514
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07910822175099083,
+      "learning_rate": 0.0003752773464438909,
+      "loss": 1.2959,
+      "step": 5515
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07694259855361689,
+      "learning_rate": 0.0003751087504055797,
+      "loss": 1.4247,
+      "step": 5516
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07256522449985622,
+      "learning_rate": 0.0003749401695107634,
+      "loss": 1.397,
+      "step": 5517
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.08607831710188796,
+      "learning_rate": 0.00037477160377988246,
+      "loss": 1.5007,
+      "step": 5518
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07415689538573572,
+      "learning_rate": 0.0003746030532333763,
+      "loss": 1.4612,
+      "step": 5519
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07758296899328197,
+      "learning_rate": 0.0003744345178916823,
+      "loss": 1.3624,
+      "step": 5520
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.08082820232713231,
+      "learning_rate": 0.00037426599777523573,
+      "loss": 1.4306,
+      "step": 5521
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07400004575923684,
+      "learning_rate": 0.0003740974929044706,
+      "loss": 1.3545,
+      "step": 5522
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07599643071172713,
+      "learning_rate": 0.00037392900329981843,
+      "loss": 1.3584,
+      "step": 5523
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07895934453748053,
+      "learning_rate": 0.00037376052898170913,
+      "loss": 1.4622,
+      "step": 5524
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.0668384332444495,
+      "learning_rate": 0.00037359206997057107,
+      "loss": 1.5848,
+      "step": 5525
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07121712503947927,
+      "learning_rate": 0.0003734236262868303,
+      "loss": 1.3717,
+      "step": 5526
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.0732117144452143,
+      "learning_rate": 0.00037325519795091127,
+      "loss": 1.5614,
+      "step": 5527
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.08251413632425897,
+      "learning_rate": 0.00037308678498323647,
+      "loss": 1.3482,
+      "step": 5528
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07988610447393804,
+      "learning_rate": 0.00037291838740422623,
+      "loss": 1.441,
+      "step": 5529
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07343160375426805,
+      "learning_rate": 0.0003727500052342999,
+      "loss": 1.4126,
+      "step": 5530
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.0750280771624314,
+      "learning_rate": 0.00037258163849387407,
+      "loss": 1.3603,
+      "step": 5531
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.0699588962331154,
+      "learning_rate": 0.0003724132872033638,
+      "loss": 1.4122,
+      "step": 5532
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.08618571728366133,
+      "learning_rate": 0.0003722449513831823,
+      "loss": 1.3199,
+      "step": 5533
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.0752992863299164,
+      "learning_rate": 0.00037207663105374094,
+      "loss": 1.3892,
+      "step": 5534
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.0756774846041659,
+      "learning_rate": 0.00037190832623544903,
+      "loss": 1.4252,
+      "step": 5535
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.0755846484269021,
+      "learning_rate": 0.0003717400369487142,
+      "loss": 1.4436,
+      "step": 5536
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.07183736040402938,
+      "learning_rate": 0.0003715717632139421,
+      "loss": 1.2719,
+      "step": 5537
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.07633084282336201,
+      "learning_rate": 0.0003714035050515366,
+      "loss": 1.4026,
+      "step": 5538
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.07675837417693088,
+      "learning_rate": 0.0003712352624818993,
+      "loss": 1.2591,
+      "step": 5539
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.07676462023153384,
+      "learning_rate": 0.00037106703552543024,
+      "loss": 1.3136,
+      "step": 5540
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.07689170885756248,
+      "learning_rate": 0.0003708988242025277,
+      "loss": 1.3616,
+      "step": 5541
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.07474625872774958,
+      "learning_rate": 0.0003707306285335879,
+      "loss": 1.4878,
+      "step": 5542
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.08136785858544164,
+      "learning_rate": 0.0003705624485390049,
+      "loss": 1.3011,
+      "step": 5543
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.08019714407726865,
+      "learning_rate": 0.0003703942842391714,
+      "loss": 1.4349,
+      "step": 5544
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.06931800845539068,
+      "learning_rate": 0.0003702261356544777,
+      "loss": 1.336,
+      "step": 5545
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.06865776028434824,
+      "learning_rate": 0.00037005800280531245,
+      "loss": 1.3829,
+      "step": 5546
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.07110683293273809,
+      "learning_rate": 0.0003698898857120622,
+      "loss": 1.3449,
+      "step": 5547
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.06752517336614956,
+      "learning_rate": 0.00036972178439511206,
+      "loss": 1.4429,
+      "step": 5548
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.07369625011178438,
+      "learning_rate": 0.00036955369887484477,
+      "loss": 1.3135,
+      "step": 5549
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.08217739104019585,
+      "learning_rate": 0.00036938562917164107,
+      "loss": 1.44,
+      "step": 5550
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.0679176740185262,
+      "learning_rate": 0.00036921757530588016,
+      "loss": 1.4109,
+      "step": 5551
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.08183989458372486,
+      "learning_rate": 0.000369049537297939,
+      "loss": 1.5463,
+      "step": 5552
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.08068697173407725,
+      "learning_rate": 0.00036888151516819304,
+      "loss": 1.3758,
+      "step": 5553
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.0916066824594436,
+      "learning_rate": 0.0003687135089370153,
+      "loss": 1.4083,
+      "step": 5554
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.07203913084540307,
+      "learning_rate": 0.0003685455186247772,
+      "loss": 1.3663,
+      "step": 5555
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.07810221503711978,
+      "learning_rate": 0.0003683775442518482,
+      "loss": 1.337,
+      "step": 5556
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.07378044550397307,
+      "learning_rate": 0.0003682095858385958,
+      "loss": 1.4417,
+      "step": 5557
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.07662962879711044,
+      "learning_rate": 0.0003680416434053854,
+      "loss": 1.4428,
+      "step": 5558
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.07391624067869233,
+      "learning_rate": 0.00036787371697258087,
+      "loss": 1.5021,
+      "step": 5559
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.08018611758276063,
+      "learning_rate": 0.0003677058065605434,
+      "loss": 1.4745,
+      "step": 5560
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.07141145070323125,
+      "learning_rate": 0.00036753791218963305,
+      "loss": 1.5691,
+      "step": 5561
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.07200974918384846,
+      "learning_rate": 0.00036737003388020756,
+      "loss": 1.3799,
+      "step": 5562
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.07893684170114153,
+      "learning_rate": 0.0003672021716526226,
+      "loss": 1.4226,
+      "step": 5563
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.08330290597126691,
+      "learning_rate": 0.0003670343255272322,
+      "loss": 1.556,
+      "step": 5564
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.07058247997252769,
+      "learning_rate": 0.00036686649552438827,
+      "loss": 1.4039,
+      "step": 5565
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.07577260612687944,
+      "learning_rate": 0.00036669868166444065,
+      "loss": 1.3556,
+      "step": 5566
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.07442078259679677,
+      "learning_rate": 0.0003665308839677375,
+      "loss": 1.4111,
+      "step": 5567
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.09062648799718757,
+      "learning_rate": 0.00036636310245462484,
+      "loss": 1.3685,
+      "step": 5568
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.07404072406825808,
+      "learning_rate": 0.00036619533714544664,
+      "loss": 1.2866,
+      "step": 5569
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.07341209572066199,
+      "learning_rate": 0.00036602758806054535,
+      "loss": 1.3663,
+      "step": 5570
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.08367994768627389,
+      "learning_rate": 0.00036585985522026067,
+      "loss": 1.3816,
+      "step": 5571
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.06888392014168349,
+      "learning_rate": 0.000365692138644931,
+      "loss": 1.4769,
+      "step": 5572
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.07446670991451924,
+      "learning_rate": 0.0003655244383548926,
+      "loss": 1.4686,
+      "step": 5573
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.07197126237743647,
+      "learning_rate": 0.00036535675437047955,
+      "loss": 1.3757,
+      "step": 5574
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.07647613896270093,
+      "learning_rate": 0.00036518908671202426,
+      "loss": 1.5409,
+      "step": 5575
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.06842253526012758,
+      "learning_rate": 0.00036502143539985705,
+      "loss": 1.3703,
+      "step": 5576
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.07301037980837188,
+      "learning_rate": 0.00036485380045430597,
+      "loss": 1.4469,
+      "step": 5577
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.07007991133973272,
+      "learning_rate": 0.0003646861818956977,
+      "loss": 1.4505,
+      "step": 5578
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.06574887026402759,
+      "learning_rate": 0.0003645185797443563,
+      "loss": 1.2006,
+      "step": 5579
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.07588016531671705,
+      "learning_rate": 0.0003643509940206043,
+      "loss": 1.3754,
+      "step": 5580
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.06515398225680037,
+      "learning_rate": 0.00036418342474476184,
+      "loss": 1.368,
+      "step": 5581
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.07544085970434565,
+      "learning_rate": 0.00036401587193714724,
+      "loss": 1.5863,
+      "step": 5582
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.07153631958666246,
+      "learning_rate": 0.00036384833561807706,
+      "loss": 1.4028,
+      "step": 5583
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.06872766857481105,
+      "learning_rate": 0.0003636808158078656,
+      "loss": 1.3447,
+      "step": 5584
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.08987840989416099,
+      "learning_rate": 0.00036351331252682515,
+      "loss": 1.4119,
+      "step": 5585
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.06816450340215846,
+      "learning_rate": 0.0003633458257952661,
+      "loss": 1.4812,
+      "step": 5586
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.0737825306184352,
+      "learning_rate": 0.0003631783556334968,
+      "loss": 1.4151,
+      "step": 5587
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.07504967730605744,
+      "learning_rate": 0.00036301090206182366,
+      "loss": 1.4566,
+      "step": 5588
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.08464044928871382,
+      "learning_rate": 0.00036284346510055064,
+      "loss": 1.3867,
+      "step": 5589
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.07468777452003433,
+      "learning_rate": 0.0003626760447699806,
+      "loss": 1.5272,
+      "step": 5590
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.07626542283005774,
+      "learning_rate": 0.0003625086410904136,
+      "loss": 1.3815,
+      "step": 5591
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.06636087185478588,
+      "learning_rate": 0.0003623412540821478,
+      "loss": 1.2893,
+      "step": 5592
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.06748116009133115,
+      "learning_rate": 0.0003621738837654795,
+      "loss": 1.4307,
+      "step": 5593
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.08014808265191209,
+      "learning_rate": 0.00036200653016070283,
+      "loss": 1.2923,
+      "step": 5594
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.07630050936118145,
+      "learning_rate": 0.0003618391932881102,
+      "loss": 1.5177,
+      "step": 5595
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.08900857906661022,
+      "learning_rate": 0.0003616718731679918,
+      "loss": 1.421,
+      "step": 5596
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.07500532846595925,
+      "learning_rate": 0.0003615045698206355,
+      "loss": 1.3891,
+      "step": 5597
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.07810709867424619,
+      "learning_rate": 0.00036133728326632764,
+      "loss": 1.4612,
+      "step": 5598
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.06859701862319084,
+      "learning_rate": 0.0003611700135253523,
+      "loss": 1.3761,
+      "step": 5599
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.07414205880381541,
+      "learning_rate": 0.0003610027606179913,
+      "loss": 1.4306,
+      "step": 5600
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.06934922386528766,
+      "learning_rate": 0.00036083552456452484,
+      "loss": 1.463,
+      "step": 5601
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.0737091137736752,
+      "learning_rate": 0.00036066830538523096,
+      "loss": 1.3899,
+      "step": 5602
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.07854348024840235,
+      "learning_rate": 0.0003605011031003853,
+      "loss": 1.348,
+      "step": 5603
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.08047178520802252,
+      "learning_rate": 0.0003603339177302618,
+      "loss": 1.5244,
+      "step": 5604
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.07429033202974966,
+      "learning_rate": 0.00036016674929513216,
+      "loss": 1.4901,
+      "step": 5605
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.07099137795592182,
+      "learning_rate": 0.00035999959781526637,
+      "loss": 1.4711,
+      "step": 5606
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.07344959139200635,
+      "learning_rate": 0.00035983246331093197,
+      "loss": 1.4571,
+      "step": 5607
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.08400809429915737,
+      "learning_rate": 0.00035966534580239454,
+      "loss": 1.2612,
+      "step": 5608
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.08538467421828809,
+      "learning_rate": 0.0003594982453099178,
+      "loss": 1.3764,
+      "step": 5609
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.07206064549666952,
+      "learning_rate": 0.00035933116185376325,
+      "loss": 1.4262,
+      "step": 5610
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.07958752096332246,
+      "learning_rate": 0.00035916409545419027,
+      "loss": 1.3995,
+      "step": 5611
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.07919682179856065,
+      "learning_rate": 0.00035899704613145635,
+      "loss": 1.5493,
+      "step": 5612
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.07986030378083334,
+      "learning_rate": 0.00035883001390581647,
+      "loss": 1.4705,
+      "step": 5613
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.07799866116922917,
+      "learning_rate": 0.0003586629987975243,
+      "loss": 1.4285,
+      "step": 5614
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.07339482852672152,
+      "learning_rate": 0.00035849600082683066,
+      "loss": 1.4108,
+      "step": 5615
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.07719796136377283,
+      "learning_rate": 0.00035832902001398483,
+      "loss": 1.5008,
+      "step": 5616
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.08039852916277493,
+      "learning_rate": 0.0003581620563792336,
+      "loss": 1.4794,
+      "step": 5617
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.08287709546388787,
+      "learning_rate": 0.00035799510994282216,
+      "loss": 1.3035,
+      "step": 5618
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.07496917924553367,
+      "learning_rate": 0.0003578281807249931,
+      "loss": 1.4986,
+      "step": 5619
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.06504525578999058,
+      "learning_rate": 0.0003576612687459873,
+      "loss": 1.6326,
+      "step": 5620
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.07543080011759193,
+      "learning_rate": 0.00035749437402604346,
+      "loss": 1.4127,
+      "step": 5621
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.06900592630400436,
+      "learning_rate": 0.00035732749658539797,
+      "loss": 1.4264,
+      "step": 5622
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.0689469569944877,
+      "learning_rate": 0.00035716063644428565,
+      "loss": 1.2965,
+      "step": 5623
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.07147758440559308,
+      "learning_rate": 0.00035699379362293836,
+      "loss": 1.3985,
+      "step": 5624
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.07056189553778609,
+      "learning_rate": 0.00035682696814158657,
+      "loss": 1.333,
+      "step": 5625
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.07538426411309675,
+      "learning_rate": 0.00035666016002045854,
+      "loss": 1.3421,
+      "step": 5626
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.06937029736035574,
+      "learning_rate": 0.0003564933692797803,
+      "loss": 1.4397,
+      "step": 5627
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.06838536960364613,
+      "learning_rate": 0.0003563265959397757,
+      "loss": 1.4764,
+      "step": 5628
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.06858569045988618,
+      "learning_rate": 0.0003561598400206667,
+      "loss": 1.3769,
+      "step": 5629
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.0821419094374213,
+      "learning_rate": 0.00035599310154267307,
+      "loss": 1.4455,
+      "step": 5630
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.07978858517243766,
+      "learning_rate": 0.00035582638052601223,
+      "loss": 1.4309,
+      "step": 5631
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.07710280608060323,
+      "learning_rate": 0.00035565967699089984,
+      "loss": 1.412,
+      "step": 5632
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.07006371462775733,
+      "learning_rate": 0.00035549299095754937,
+      "loss": 1.3961,
+      "step": 5633
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.07673372395440631,
+      "learning_rate": 0.0003553263224461718,
+      "loss": 1.2666,
+      "step": 5634
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.07617606707799483,
+      "learning_rate": 0.00035515967147697647,
+      "loss": 1.3182,
+      "step": 5635
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.07299159334036456,
+      "learning_rate": 0.0003549930380701701,
+      "loss": 1.3535,
+      "step": 5636
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.07548869479560834,
+      "learning_rate": 0.000354826422245958,
+      "loss": 1.3972,
+      "step": 5637
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.07528600549226311,
+      "learning_rate": 0.00035465982402454267,
+      "loss": 1.3542,
+      "step": 5638
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.0761655413610927,
+      "learning_rate": 0.0003544932434261246,
+      "loss": 1.5579,
+      "step": 5639
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.07328638809355852,
+      "learning_rate": 0.0003543266804709026,
+      "loss": 1.4611,
+      "step": 5640
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.0684025976932677,
+      "learning_rate": 0.00035416013517907264,
+      "loss": 1.3723,
+      "step": 5641
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.0709444446488586,
+      "learning_rate": 0.00035399360757082913,
+      "loss": 1.3982,
+      "step": 5642
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.07902829727346929,
+      "learning_rate": 0.000353827097666364,
+      "loss": 1.3444,
+      "step": 5643
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.07138351900963495,
+      "learning_rate": 0.00035366060548586744,
+      "loss": 1.3536,
+      "step": 5644
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.06664446570362624,
+      "learning_rate": 0.00035349413104952685,
+      "loss": 1.4987,
+      "step": 5645
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.06565668811270635,
+      "learning_rate": 0.0003533276743775279,
+      "loss": 1.3504,
+      "step": 5646
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.06867503022652852,
+      "learning_rate": 0.00035316123549005416,
+      "loss": 1.3162,
+      "step": 5647
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.08049531458815742,
+      "learning_rate": 0.0003529948144072867,
+      "loss": 1.3726,
+      "step": 5648
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.06401788263416966,
+      "learning_rate": 0.0003528284111494049,
+      "loss": 1.3049,
+      "step": 5649
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.08055355952171174,
+      "learning_rate": 0.00035266202573658567,
+      "loss": 1.5669,
+      "step": 5650
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.07453158758644808,
+      "learning_rate": 0.00035249565818900366,
+      "loss": 1.4813,
+      "step": 5651
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.07887708489653913,
+      "learning_rate": 0.00035232930852683165,
+      "loss": 1.4236,
+      "step": 5652
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.07109618467705087,
+      "learning_rate": 0.0003521629767702401,
+      "loss": 1.2988,
+      "step": 5653
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.06710743003825625,
+      "learning_rate": 0.00035199666293939726,
+      "loss": 1.4079,
+      "step": 5654
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.07155522294010239,
+      "learning_rate": 0.0003518303670544696,
+      "loss": 1.4105,
+      "step": 5655
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.07196979769357402,
+      "learning_rate": 0.00035166408913562034,
+      "loss": 1.4136,
+      "step": 5656
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.07745496675046756,
+      "learning_rate": 0.00035149782920301174,
+      "loss": 1.4632,
+      "step": 5657
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.07447542178913955,
+      "learning_rate": 0.0003513315872768035,
+      "loss": 1.4284,
+      "step": 5658
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.07836930328423564,
+      "learning_rate": 0.00035116536337715255,
+      "loss": 1.4194,
+      "step": 5659
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.08966749202790163,
+      "learning_rate": 0.00035099915752421465,
+      "loss": 1.4606,
+      "step": 5660
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.06903241945632177,
+      "learning_rate": 0.0003508329697381425,
+      "loss": 1.4138,
+      "step": 5661
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.08028060766204789,
+      "learning_rate": 0.0003506668000390869,
+      "loss": 1.4581,
+      "step": 5662
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.07138617804983953,
+      "learning_rate": 0.00035050064844719677,
+      "loss": 1.3281,
+      "step": 5663
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.09922730576053841,
+      "learning_rate": 0.0003503345149826185,
+      "loss": 1.343,
+      "step": 5664
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.07817632658073502,
+      "learning_rate": 0.00035016839966549627,
+      "loss": 1.5559,
+      "step": 5665
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.07861593935008956,
+      "learning_rate": 0.000350002302515972,
+      "loss": 1.3796,
+      "step": 5666
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.07251164645828607,
+      "learning_rate": 0.0003498362235541856,
+      "loss": 1.4127,
+      "step": 5667
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.06707603220109654,
+      "learning_rate": 0.0003496701628002749,
+      "loss": 1.3143,
+      "step": 5668
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.06988053951437045,
+      "learning_rate": 0.00034950412027437525,
+      "loss": 1.4305,
+      "step": 5669
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.07867145829080913,
+      "learning_rate": 0.0003493380959966197,
+      "loss": 1.2806,
+      "step": 5670
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.07429734651699726,
+      "learning_rate": 0.00034917208998713956,
+      "loss": 1.4044,
+      "step": 5671
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.07446972862489869,
+      "learning_rate": 0.0003490061022660634,
+      "loss": 1.4015,
+      "step": 5672
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.07585596482806031,
+      "learning_rate": 0.00034884013285351796,
+      "loss": 1.4996,
+      "step": 5673
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.0765452008993846,
+      "learning_rate": 0.0003486741817696275,
+      "loss": 1.3428,
+      "step": 5674
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.07743629189736943,
+      "learning_rate": 0.0003485082490345143,
+      "loss": 1.3471,
+      "step": 5675
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.07962414669280672,
+      "learning_rate": 0.00034834233466829825,
+      "loss": 1.5078,
+      "step": 5676
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.06823095793714402,
+      "learning_rate": 0.0003481764386910968,
+      "loss": 1.2538,
+      "step": 5677
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.07430885125530548,
+      "learning_rate": 0.00034801056112302554,
+      "loss": 1.5397,
+      "step": 5678
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.07967890542328465,
+      "learning_rate": 0.0003478447019841978,
+      "loss": 1.4244,
+      "step": 5679
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.08165222831567905,
+      "learning_rate": 0.00034767886129472453,
+      "loss": 1.4542,
+      "step": 5680
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.08037753942606386,
+      "learning_rate": 0.0003475130390747144,
+      "loss": 1.5213,
+      "step": 5681
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.07534567859413022,
+      "learning_rate": 0.000347347235344274,
+      "loss": 1.3967,
+      "step": 5682
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.08780878813456622,
+      "learning_rate": 0.0003471814501235076,
+      "loss": 1.2945,
+      "step": 5683
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.07526528057162522,
+      "learning_rate": 0.00034701568343251723,
+      "loss": 1.4243,
+      "step": 5684
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.0877269611831134,
+      "learning_rate": 0.0003468499352914026,
+      "loss": 1.2562,
+      "step": 5685
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.07211662248051867,
+      "learning_rate": 0.00034668420572026156,
+      "loss": 1.3204,
+      "step": 5686
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.0785050832257113,
+      "learning_rate": 0.00034651849473918883,
+      "loss": 1.3089,
+      "step": 5687
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.0733152327279244,
+      "learning_rate": 0.00034635280236827785,
+      "loss": 1.4598,
+      "step": 5688
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.07122292618843426,
+      "learning_rate": 0.0003461871286276194,
+      "loss": 1.4084,
+      "step": 5689
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.07521042733808453,
+      "learning_rate": 0.0003460214735373016,
+      "loss": 1.2403,
+      "step": 5690
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.08669744108413878,
+      "learning_rate": 0.00034585583711741114,
+      "loss": 1.3522,
+      "step": 5691
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.07735992970438209,
+      "learning_rate": 0.00034569021938803184,
+      "loss": 1.3652,
+      "step": 5692
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.08321736056147003,
+      "learning_rate": 0.0003455246203692454,
+      "loss": 1.4371,
+      "step": 5693
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.07046318299439341,
+      "learning_rate": 0.0003453590400811313,
+      "loss": 1.2962,
+      "step": 5694
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.08303814502094009,
+      "learning_rate": 0.0003451934785437668,
+      "loss": 1.3947,
+      "step": 5695
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.0778596641434968,
+      "learning_rate": 0.00034502793577722657,
+      "loss": 1.4846,
+      "step": 5696
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.07305095025031742,
+      "learning_rate": 0.00034486241180158375,
+      "loss": 1.4616,
+      "step": 5697
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.07711619596739502,
+      "learning_rate": 0.00034469690663690793,
+      "loss": 1.244,
+      "step": 5698
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.08701736627740074,
+      "learning_rate": 0.0003445314203032678,
+      "loss": 1.5445,
+      "step": 5699
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.08825112294005741,
+      "learning_rate": 0.0003443659528207289,
+      "loss": 1.2481,
+      "step": 5700
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.08846932338421327,
+      "learning_rate": 0.0003442005042093547,
+      "loss": 1.3274,
+      "step": 5701
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.07205817936514314,
+      "learning_rate": 0.00034403507448920655,
+      "loss": 1.433,
+      "step": 5702
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.07069231060608115,
+      "learning_rate": 0.00034386966368034333,
+      "loss": 1.2406,
+      "step": 5703
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.07557644180661727,
+      "learning_rate": 0.0003437042718028215,
+      "loss": 1.4779,
+      "step": 5704
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.07339080355677123,
+      "learning_rate": 0.00034353889887669574,
+      "loss": 1.4528,
+      "step": 5705
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.06957718608936084,
+      "learning_rate": 0.00034337354492201784,
+      "loss": 1.5388,
+      "step": 5706
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.07314609581938167,
+      "learning_rate": 0.0003432082099588377,
+      "loss": 1.349,
+      "step": 5707
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.0818507302919073,
+      "learning_rate": 0.0003430428940072026,
+      "loss": 1.3628,
+      "step": 5708
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.0648690866111131,
+      "learning_rate": 0.0003428775970871575,
+      "loss": 1.4247,
+      "step": 5709
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.06938785217931223,
+      "learning_rate": 0.0003427123192187456,
+      "loss": 1.5578,
+      "step": 5710
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.07535029207051505,
+      "learning_rate": 0.00034254706042200725,
+      "loss": 1.3668,
+      "step": 5711
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.07028672868584891,
+      "learning_rate": 0.00034238182071698065,
+      "loss": 1.4355,
+      "step": 5712
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.08299347809832416,
+      "learning_rate": 0.0003422166001237016,
+      "loss": 1.4188,
+      "step": 5713
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.06277772175495448,
+      "learning_rate": 0.0003420513986622038,
+      "loss": 1.2739,
+      "step": 5714
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.06737514638829561,
+      "learning_rate": 0.0003418862163525185,
+      "loss": 1.3447,
+      "step": 5715
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.07349604640587126,
+      "learning_rate": 0.0003417210532146744,
+      "loss": 1.3473,
+      "step": 5716
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.0861956852391659,
+      "learning_rate": 0.00034155590926869837,
+      "loss": 1.1675,
+      "step": 5717
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.06819916582304868,
+      "learning_rate": 0.0003413907845346147,
+      "loss": 1.418,
+      "step": 5718
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.07113808982321314,
+      "learning_rate": 0.0003412256790324452,
+      "loss": 1.3437,
+      "step": 5719
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.08113355490818232,
+      "learning_rate": 0.00034106059278220935,
+      "loss": 1.3868,
+      "step": 5720
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.09610388012345854,
+      "learning_rate": 0.00034089552580392456,
+      "loss": 1.4229,
+      "step": 5721
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07795794477824468,
+      "learning_rate": 0.00034073047811760586,
+      "loss": 1.4547,
+      "step": 5722
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.09967082596439263,
+      "learning_rate": 0.0003405654497432658,
+      "loss": 1.4848,
+      "step": 5723
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.06998698314030954,
+      "learning_rate": 0.0003404004407009145,
+      "loss": 1.4644,
+      "step": 5724
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.08424361964264927,
+      "learning_rate": 0.0003402354510105601,
+      "loss": 1.468,
+      "step": 5725
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07588249803140369,
+      "learning_rate": 0.00034007048069220803,
+      "loss": 1.2903,
+      "step": 5726
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.0718749317137767,
+      "learning_rate": 0.00033990552976586143,
+      "loss": 1.4871,
+      "step": 5727
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.06527144687140929,
+      "learning_rate": 0.0003397405982515214,
+      "loss": 1.3258,
+      "step": 5728
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.08734474868450842,
+      "learning_rate": 0.0003395756861691864,
+      "loss": 1.3422,
+      "step": 5729
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07129891373227273,
+      "learning_rate": 0.0003394107935388525,
+      "loss": 1.5485,
+      "step": 5730
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.0688040837256214,
+      "learning_rate": 0.00033924592038051346,
+      "loss": 1.4304,
+      "step": 5731
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07444968184588249,
+      "learning_rate": 0.0003390810667141606,
+      "loss": 1.3824,
+      "step": 5732
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07877965685992819,
+      "learning_rate": 0.0003389162325597834,
+      "loss": 1.4078,
+      "step": 5733
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07229117315627312,
+      "learning_rate": 0.0003387514179373683,
+      "loss": 1.3603,
+      "step": 5734
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07622358969159525,
+      "learning_rate": 0.00033858662286689954,
+      "loss": 1.4646,
+      "step": 5735
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07074157195742369,
+      "learning_rate": 0.0003384218473683594,
+      "loss": 1.4831,
+      "step": 5736
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07407245287057423,
+      "learning_rate": 0.0003382570914617273,
+      "loss": 1.519,
+      "step": 5737
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07466035597162812,
+      "learning_rate": 0.0003380923551669804,
+      "loss": 1.228,
+      "step": 5738
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.0683061545829768,
+      "learning_rate": 0.0003379276385040938,
+      "loss": 1.4295,
+      "step": 5739
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07171623879287245,
+      "learning_rate": 0.00033776294149303953,
+      "loss": 1.5231,
+      "step": 5740
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07017306983889705,
+      "learning_rate": 0.0003375982641537881,
+      "loss": 1.2489,
+      "step": 5741
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.06908949184279577,
+      "learning_rate": 0.0003374336065063069,
+      "loss": 1.5955,
+      "step": 5742
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07588952795946573,
+      "learning_rate": 0.0003372689685705614,
+      "loss": 1.5576,
+      "step": 5743
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.0807730430692002,
+      "learning_rate": 0.00033710435036651446,
+      "loss": 1.422,
+      "step": 5744
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07615707250575719,
+      "learning_rate": 0.0003369397519141267,
+      "loss": 1.4909,
+      "step": 5745
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07415125469436308,
+      "learning_rate": 0.00033677517323335614,
+      "loss": 1.4658,
+      "step": 5746
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07714066587006227,
+      "learning_rate": 0.00033661061434415844,
+      "loss": 1.423,
+      "step": 5747
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.06975556296096408,
+      "learning_rate": 0.00033644607526648717,
+      "loss": 1.4794,
+      "step": 5748
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07700941531030474,
+      "learning_rate": 0.0003362815560202931,
+      "loss": 1.4553,
+      "step": 5749
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07152268833698286,
+      "learning_rate": 0.00033611705662552494,
+      "loss": 1.4017,
+      "step": 5750
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07145733194757649,
+      "learning_rate": 0.0003359525771021285,
+      "loss": 1.3974,
+      "step": 5751
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07285528831823655,
+      "learning_rate": 0.0003357881174700476,
+      "loss": 1.3365,
+      "step": 5752
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07203425116451612,
+      "learning_rate": 0.0003356236777492236,
+      "loss": 1.4425,
+      "step": 5753
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07216624969608204,
+      "learning_rate": 0.00033545925795959544,
+      "loss": 1.4025,
+      "step": 5754
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07809662601678845,
+      "learning_rate": 0.00033529485812109933,
+      "loss": 1.4063,
+      "step": 5755
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07661840340526746,
+      "learning_rate": 0.0003351304782536697,
+      "loss": 1.4225,
+      "step": 5756
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.0756729672262736,
+      "learning_rate": 0.0003349661183772379,
+      "loss": 1.3925,
+      "step": 5757
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.0788412033149621,
+      "learning_rate": 0.0003348017785117331,
+      "loss": 1.5799,
+      "step": 5758
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07957797898161738,
+      "learning_rate": 0.0003346374586770823,
+      "loss": 1.3948,
+      "step": 5759
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07042844223390973,
+      "learning_rate": 0.00033447315889320985,
+      "loss": 1.2962,
+      "step": 5760
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07623292390015404,
+      "learning_rate": 0.0003343088791800374,
+      "loss": 1.4195,
+      "step": 5761
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07479667738963493,
+      "learning_rate": 0.0003341446195574846,
+      "loss": 1.5432,
+      "step": 5762
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.09130577475740195,
+      "learning_rate": 0.0003339803800454684,
+      "loss": 1.3322,
+      "step": 5763
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.0830672023590636,
+      "learning_rate": 0.00033381616066390353,
+      "loss": 1.3991,
+      "step": 5764
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07167158802173361,
+      "learning_rate": 0.00033365196143270203,
+      "loss": 1.5648,
+      "step": 5765
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.08324039636311231,
+      "learning_rate": 0.0003334877823717737,
+      "loss": 1.1953,
+      "step": 5766
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07364194494627035,
+      "learning_rate": 0.0003333236235010259,
+      "loss": 1.3725,
+      "step": 5767
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.08352686220297945,
+      "learning_rate": 0.0003331594848403634,
+      "loss": 1.2813,
+      "step": 5768
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.0782586913324151,
+      "learning_rate": 0.0003329953664096884,
+      "loss": 1.4421,
+      "step": 5769
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07344286447573398,
+      "learning_rate": 0.00033283126822890107,
+      "loss": 1.4148,
+      "step": 5770
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07640927867015894,
+      "learning_rate": 0.00033266719031789895,
+      "loss": 1.3939,
+      "step": 5771
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07444837850894832,
+      "learning_rate": 0.0003325031326965767,
+      "loss": 1.344,
+      "step": 5772
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07453338538015851,
+      "learning_rate": 0.0003323390953848271,
+      "loss": 1.5043,
+      "step": 5773
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07563824140788851,
+      "learning_rate": 0.0003321750784025401,
+      "loss": 1.4378,
+      "step": 5774
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07733704126648222,
+      "learning_rate": 0.00033201108176960346,
+      "loss": 1.3634,
+      "step": 5775
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07477323086087968,
+      "learning_rate": 0.00033184710550590243,
+      "loss": 1.3377,
+      "step": 5776
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.0673227413865871,
+      "learning_rate": 0.00033168314963131953,
+      "loss": 1.3578,
+      "step": 5777
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07476933458770825,
+      "learning_rate": 0.00033151921416573485,
+      "loss": 1.3904,
+      "step": 5778
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.06899281005764235,
+      "learning_rate": 0.0003313552991290264,
+      "loss": 1.3827,
+      "step": 5779
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07602161019541753,
+      "learning_rate": 0.0003311914045410694,
+      "loss": 1.4342,
+      "step": 5780
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.06935786602316016,
+      "learning_rate": 0.00033102753042173644,
+      "loss": 1.2842,
+      "step": 5781
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07221675602626097,
+      "learning_rate": 0.00033086367679089826,
+      "loss": 1.3586,
+      "step": 5782
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.08951430196528462,
+      "learning_rate": 0.00033069984366842207,
+      "loss": 1.3451,
+      "step": 5783
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.06737689928701268,
+      "learning_rate": 0.00033053603107417365,
+      "loss": 1.3806,
+      "step": 5784
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.0926594824874406,
+      "learning_rate": 0.0003303722390280156,
+      "loss": 1.3205,
+      "step": 5785
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.06461257789980361,
+      "learning_rate": 0.00033020846754980824,
+      "loss": 1.3715,
+      "step": 5786
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.08945684857302136,
+      "learning_rate": 0.0003300447166594097,
+      "loss": 1.4229,
+      "step": 5787
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.06326462043703998,
+      "learning_rate": 0.0003298809863766752,
+      "loss": 1.3268,
+      "step": 5788
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07313020967684405,
+      "learning_rate": 0.0003297172767214576,
+      "loss": 1.4367,
+      "step": 5789
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07901312968141494,
+      "learning_rate": 0.0003295535877136072,
+      "loss": 1.3237,
+      "step": 5790
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07933674541674322,
+      "learning_rate": 0.00032938991937297193,
+      "loss": 1.3319,
+      "step": 5791
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.06740982353930541,
+      "learning_rate": 0.0003292262717193972,
+      "loss": 1.3509,
+      "step": 5792
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07754239914727032,
+      "learning_rate": 0.00032906264477272575,
+      "loss": 1.3317,
+      "step": 5793
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07521240547901578,
+      "learning_rate": 0.0003288990385527978,
+      "loss": 1.4327,
+      "step": 5794
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.0807145346435453,
+      "learning_rate": 0.00032873545307945143,
+      "loss": 1.4121,
+      "step": 5795
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.0873071269093502,
+      "learning_rate": 0.0003285718883725217,
+      "loss": 1.346,
+      "step": 5796
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07939894137269075,
+      "learning_rate": 0.00032840834445184154,
+      "loss": 1.4587,
+      "step": 5797
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.08248088922264714,
+      "learning_rate": 0.0003282448213372412,
+      "loss": 1.3405,
+      "step": 5798
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.08439168880904907,
+      "learning_rate": 0.0003280813190485484,
+      "loss": 1.3494,
+      "step": 5799
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.08123391440953981,
+      "learning_rate": 0.00032791783760558835,
+      "loss": 1.4026,
+      "step": 5800
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07908771534976788,
+      "learning_rate": 0.00032775437702818377,
+      "loss": 1.5283,
+      "step": 5801
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07532811513448483,
+      "learning_rate": 0.0003275909373361548,
+      "loss": 1.4235,
+      "step": 5802
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07166058099901026,
+      "learning_rate": 0.0003274275185493192,
+      "loss": 1.4699,
+      "step": 5803
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.08297468674472518,
+      "learning_rate": 0.0003272641206874918,
+      "loss": 1.4871,
+      "step": 5804
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07243487395107884,
+      "learning_rate": 0.0003271007437704852,
+      "loss": 1.3383,
+      "step": 5805
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.0805152716343893,
+      "learning_rate": 0.00032693738781810957,
+      "loss": 1.482,
+      "step": 5806
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.09221296792088401,
+      "learning_rate": 0.00032677405285017226,
+      "loss": 1.4894,
+      "step": 5807
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07764307331973927,
+      "learning_rate": 0.0003266107388864783,
+      "loss": 1.4943,
+      "step": 5808
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.0771348222959278,
+      "learning_rate": 0.0003264474459468299,
+      "loss": 1.395,
+      "step": 5809
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07434205741667296,
+      "learning_rate": 0.00032628417405102704,
+      "loss": 1.3975,
+      "step": 5810
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.08459239119791959,
+      "learning_rate": 0.0003261209232188671,
+      "loss": 1.4852,
+      "step": 5811
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.08710954280161412,
+      "learning_rate": 0.00032595769347014446,
+      "loss": 1.2992,
+      "step": 5812
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07702098501175848,
+      "learning_rate": 0.0003257944848246519,
+      "loss": 1.384,
+      "step": 5813
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.0729400817671885,
+      "learning_rate": 0.00032563129730217826,
+      "loss": 1.3524,
+      "step": 5814
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07949284921131701,
+      "learning_rate": 0.0003254681309225111,
+      "loss": 1.4447,
+      "step": 5815
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.08413069338126254,
+      "learning_rate": 0.00032530498570543476,
+      "loss": 1.5163,
+      "step": 5816
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07414571013437149,
+      "learning_rate": 0.000325141861670731,
+      "loss": 1.2736,
+      "step": 5817
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.0709935359677738,
+      "learning_rate": 0.0003249787588381795,
+      "loss": 1.4867,
+      "step": 5818
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.08271727344452846,
+      "learning_rate": 0.0003248156772275569,
+      "loss": 1.4364,
+      "step": 5819
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07447507008238005,
+      "learning_rate": 0.00032465261685863723,
+      "loss": 1.3646,
+      "step": 5820
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07074186981804943,
+      "learning_rate": 0.0003244895777511925,
+      "loss": 1.2683,
+      "step": 5821
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.08099285892882745,
+      "learning_rate": 0.0003243265599249914,
+      "loss": 1.3554,
+      "step": 5822
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.0766455672893839,
+      "learning_rate": 0.00032416356339980053,
+      "loss": 1.3961,
+      "step": 5823
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07939556315878743,
+      "learning_rate": 0.00032400058819538407,
+      "loss": 1.3909,
+      "step": 5824
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07740960550571895,
+      "learning_rate": 0.00032383763433150274,
+      "loss": 1.5367,
+      "step": 5825
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07881442246987015,
+      "learning_rate": 0.0003236747018279157,
+      "loss": 1.6059,
+      "step": 5826
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07457194940604828,
+      "learning_rate": 0.000323511790704379,
+      "loss": 1.4712,
+      "step": 5827
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07818352176339892,
+      "learning_rate": 0.00032334890098064593,
+      "loss": 1.3353,
+      "step": 5828
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07577655989104184,
+      "learning_rate": 0.00032318603267646775,
+      "loss": 1.4101,
+      "step": 5829
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.08063222118709179,
+      "learning_rate": 0.0003230231858115927,
+      "loss": 1.4651,
+      "step": 5830
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07871062903426618,
+      "learning_rate": 0.0003228603604057664,
+      "loss": 1.54,
+      "step": 5831
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07753991143364851,
+      "learning_rate": 0.00032269755647873217,
+      "loss": 1.4145,
+      "step": 5832
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07769271192950013,
+      "learning_rate": 0.00032253477405023046,
+      "loss": 1.2744,
+      "step": 5833
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07149459574934523,
+      "learning_rate": 0.00032237201313999926,
+      "loss": 1.4112,
+      "step": 5834
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.06967019720018568,
+      "learning_rate": 0.0003222092737677739,
+      "loss": 1.4218,
+      "step": 5835
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07366431502138024,
+      "learning_rate": 0.0003220465559532869,
+      "loss": 1.3923,
+      "step": 5836
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.0774173907387302,
+      "learning_rate": 0.0003218838597162685,
+      "loss": 1.3827,
+      "step": 5837
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.09568525653553758,
+      "learning_rate": 0.0003217211850764462,
+      "loss": 1.4002,
+      "step": 5838
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.06820415005290484,
+      "learning_rate": 0.0003215585320535449,
+      "loss": 1.5259,
+      "step": 5839
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07099253896133205,
+      "learning_rate": 0.0003213959006672866,
+      "loss": 1.4814,
+      "step": 5840
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.10048590704557143,
+      "learning_rate": 0.0003212332909373912,
+      "loss": 1.3643,
+      "step": 5841
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.0868215430127107,
+      "learning_rate": 0.0003210707028835755,
+      "loss": 1.2828,
+      "step": 5842
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07111722341191054,
+      "learning_rate": 0.00032090813652555395,
+      "loss": 1.2882,
+      "step": 5843
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.072419215989912,
+      "learning_rate": 0.00032074559188303834,
+      "loss": 1.4059,
+      "step": 5844
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07626815207305306,
+      "learning_rate": 0.00032058306897573787,
+      "loss": 1.3845,
+      "step": 5845
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07593411084596641,
+      "learning_rate": 0.0003204205678233586,
+      "loss": 1.2355,
+      "step": 5846
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07671474226352402,
+      "learning_rate": 0.00032025808844560465,
+      "loss": 1.2905,
+      "step": 5847
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07194390806509542,
+      "learning_rate": 0.000320095630862177,
+      "loss": 1.3861,
+      "step": 5848
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.0882906222537737,
+      "learning_rate": 0.0003199331950927745,
+      "loss": 1.4485,
+      "step": 5849
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07800170587048255,
+      "learning_rate": 0.0003197707811570928,
+      "loss": 1.5706,
+      "step": 5850
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.0798895187757382,
+      "learning_rate": 0.0003196083890748252,
+      "loss": 1.3286,
+      "step": 5851
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.08321227249490748,
+      "learning_rate": 0.0003194460188656624,
+      "loss": 1.3371,
+      "step": 5852
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07770505707512657,
+      "learning_rate": 0.0003192836705492923,
+      "loss": 1.5892,
+      "step": 5853
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.08257359787550428,
+      "learning_rate": 0.00031912134414540007,
+      "loss": 1.4664,
+      "step": 5854
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07573547661484602,
+      "learning_rate": 0.00031895903967366844,
+      "loss": 1.3355,
+      "step": 5855
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07977638149685118,
+      "learning_rate": 0.00031879675715377765,
+      "loss": 1.5056,
+      "step": 5856
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.0700197292003201,
+      "learning_rate": 0.00031863449660540457,
+      "loss": 1.4674,
+      "step": 5857
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07645749377104584,
+      "learning_rate": 0.00031847225804822405,
+      "loss": 1.4547,
+      "step": 5858
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07689507595670635,
+      "learning_rate": 0.00031831004150190796,
+      "loss": 1.4252,
+      "step": 5859
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.06998294349242434,
+      "learning_rate": 0.0003181478469861259,
+      "loss": 1.4104,
+      "step": 5860
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07176385236960926,
+      "learning_rate": 0.00031798567452054414,
+      "loss": 1.4554,
+      "step": 5861
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07609553056940516,
+      "learning_rate": 0.0003178235241248269,
+      "loss": 1.3244,
+      "step": 5862
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07696676290756585,
+      "learning_rate": 0.0003176613958186355,
+      "loss": 1.3908,
+      "step": 5863
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.08404653627393327,
+      "learning_rate": 0.00031749928962162844,
+      "loss": 1.4277,
+      "step": 5864
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07363258110080857,
+      "learning_rate": 0.00031733720555346157,
+      "loss": 1.5204,
+      "step": 5865
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07463614990723165,
+      "learning_rate": 0.0003171751436337886,
+      "loss": 1.4818,
+      "step": 5866
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07347277265972793,
+      "learning_rate": 0.00031701310388225945,
+      "loss": 1.5231,
+      "step": 5867
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07243901069441412,
+      "learning_rate": 0.0003168510863185224,
+      "loss": 1.368,
+      "step": 5868
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07148451498935104,
+      "learning_rate": 0.00031668909096222255,
+      "loss": 1.4055,
+      "step": 5869
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07683946813743962,
+      "learning_rate": 0.0003165271178330023,
+      "loss": 1.304,
+      "step": 5870
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.08296014033921659,
+      "learning_rate": 0.00031636516695050164,
+      "loss": 1.3915,
+      "step": 5871
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07000354267739249,
+      "learning_rate": 0.0003162032383343576,
+      "loss": 1.3169,
+      "step": 5872
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07802272002326763,
+      "learning_rate": 0.0003160413320042045,
+      "loss": 1.47,
+      "step": 5873
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.0732789511602166,
+      "learning_rate": 0.000315879447979674,
+      "loss": 1.3223,
+      "step": 5874
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07027730292619676,
+      "learning_rate": 0.0003157175862803953,
+      "loss": 1.3835,
+      "step": 5875
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07682569968171925,
+      "learning_rate": 0.0003155557469259946,
+      "loss": 1.359,
+      "step": 5876
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.08350094563625435,
+      "learning_rate": 0.00031539392993609554,
+      "loss": 1.3508,
+      "step": 5877
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.08295722149013737,
+      "learning_rate": 0.0003152321353303188,
+      "loss": 1.4293,
+      "step": 5878
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07254010025148988,
+      "learning_rate": 0.0003150703631282826,
+      "loss": 1.3245,
+      "step": 5879
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07751813189871887,
+      "learning_rate": 0.00031490861334960247,
+      "loss": 1.2882,
+      "step": 5880
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07379579097767187,
+      "learning_rate": 0.00031474688601389113,
+      "loss": 1.5361,
+      "step": 5881
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.06670928281533853,
+      "learning_rate": 0.00031458518114075836,
+      "loss": 1.3023,
+      "step": 5882
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.06800302796922687,
+      "learning_rate": 0.00031442349874981167,
+      "loss": 1.4567,
+      "step": 5883
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07874237147537874,
+      "learning_rate": 0.0003142618388606556,
+      "loss": 1.3324,
+      "step": 5884
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07217062766502823,
+      "learning_rate": 0.0003141002014928918,
+      "loss": 1.4494,
+      "step": 5885
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07671074428154255,
+      "learning_rate": 0.0003139385866661196,
+      "loss": 1.4268,
+      "step": 5886
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.08152511966680674,
+      "learning_rate": 0.0003137769943999352,
+      "loss": 1.3887,
+      "step": 5887
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.0857558783134707,
+      "learning_rate": 0.00031361542471393226,
+      "loss": 1.373,
+      "step": 5888
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.08265072365117745,
+      "learning_rate": 0.00031345387762770163,
+      "loss": 1.4538,
+      "step": 5889
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.08161194391076494,
+      "learning_rate": 0.00031329235316083126,
+      "loss": 1.3585,
+      "step": 5890
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07593695011204729,
+      "learning_rate": 0.00031313085133290695,
+      "loss": 1.4407,
+      "step": 5891
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07210716065409313,
+      "learning_rate": 0.00031296937216351113,
+      "loss": 1.4562,
+      "step": 5892
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.0850661492304202,
+      "learning_rate": 0.0003128079156722236,
+      "loss": 1.572,
+      "step": 5893
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.09330568226061443,
+      "learning_rate": 0.0003126464818786218,
+      "loss": 1.3642,
+      "step": 5894
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.0737289442763714,
+      "learning_rate": 0.0003124850708022799,
+      "loss": 1.3718,
+      "step": 5895
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.08218771057410153,
+      "learning_rate": 0.00031232368246276955,
+      "loss": 1.4054,
+      "step": 5896
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07967506567629058,
+      "learning_rate": 0.00031216231687965977,
+      "loss": 1.4413,
+      "step": 5897
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.079496248186044,
+      "learning_rate": 0.00031200097407251687,
+      "loss": 1.5249,
+      "step": 5898
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07693598790618682,
+      "learning_rate": 0.0003118396540609038,
+      "loss": 1.4197,
+      "step": 5899
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.08543054581020419,
+      "learning_rate": 0.0003116783568643814,
+      "loss": 1.3692,
+      "step": 5900
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.0774464397249776,
+      "learning_rate": 0.00031151708250250735,
+      "loss": 1.349,
+      "step": 5901
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.06921752872907357,
+      "learning_rate": 0.000311355830994837,
+      "loss": 1.4459,
+      "step": 5902
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07514722249298769,
+      "learning_rate": 0.00031119460236092247,
+      "loss": 1.4358,
+      "step": 5903
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.08242771723457627,
+      "learning_rate": 0.00031103339662031325,
+      "loss": 1.3992,
+      "step": 5904
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.0757734027343892,
+      "learning_rate": 0.00031087221379255616,
+      "loss": 1.3503,
+      "step": 5905
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07672806483686527,
+      "learning_rate": 0.0003107110538971952,
+      "loss": 1.4053,
+      "step": 5906
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.08680615262647434,
+      "learning_rate": 0.00031054991695377156,
+      "loss": 1.5851,
+      "step": 5907
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.08822808648220977,
+      "learning_rate": 0.0003103888029818235,
+      "loss": 1.4323,
+      "step": 5908
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07106740585924771,
+      "learning_rate": 0.00031022771200088706,
+      "loss": 1.4376,
+      "step": 5909
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07779044400772091,
+      "learning_rate": 0.0003100666440304946,
+      "loss": 1.3308,
+      "step": 5910
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07194910372531178,
+      "learning_rate": 0.00030990559909017636,
+      "loss": 1.4664,
+      "step": 5911
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07745473453560854,
+      "learning_rate": 0.00030974457719945954,
+      "loss": 1.37,
+      "step": 5912
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.0743549948835792,
+      "learning_rate": 0.00030958357837786854,
+      "loss": 1.5214,
+      "step": 5913
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07378607983212841,
+      "learning_rate": 0.00030942260264492534,
+      "loss": 1.59,
+      "step": 5914
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.08354354717365467,
+      "learning_rate": 0.0003092616500201485,
+      "loss": 1.3002,
+      "step": 5915
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.0759387023706967,
+      "learning_rate": 0.0003091007205230541,
+      "loss": 1.3571,
+      "step": 5916
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.06915812880321831,
+      "learning_rate": 0.00030893981417315553,
+      "loss": 1.4155,
+      "step": 5917
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.08602623259978609,
+      "learning_rate": 0.00030877893098996324,
+      "loss": 1.2942,
+      "step": 5918
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07686753605428095,
+      "learning_rate": 0.0003086180709929849,
+      "loss": 1.2931,
+      "step": 5919
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.0743510939165553,
+      "learning_rate": 0.0003084572342017251,
+      "loss": 1.3802,
+      "step": 5920
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.08302234915061892,
+      "learning_rate": 0.00030829642063568595,
+      "loss": 1.4475,
+      "step": 5921
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.08139252633012196,
+      "learning_rate": 0.00030813563031436674,
+      "loss": 1.3602,
+      "step": 5922
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07148548640627475,
+      "learning_rate": 0.0003079748632572639,
+      "loss": 1.3305,
+      "step": 5923
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.0792567664336374,
+      "learning_rate": 0.0003078141194838707,
+      "loss": 1.3468,
+      "step": 5924
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07455553638782846,
+      "learning_rate": 0.00030765339901367824,
+      "loss": 1.4425,
+      "step": 5925
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.0762782607168731,
+      "learning_rate": 0.00030749270186617426,
+      "loss": 1.4555,
+      "step": 5926
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07274576676044348,
+      "learning_rate": 0.0003073320280608437,
+      "loss": 1.4174,
+      "step": 5927
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.0684112547807106,
+      "learning_rate": 0.00030717137761716916,
+      "loss": 1.539,
+      "step": 5928
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07220831738070271,
+      "learning_rate": 0.0003070107505546298,
+      "loss": 1.2818,
+      "step": 5929
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.0973184931653835,
+      "learning_rate": 0.00030685014689270243,
+      "loss": 1.4565,
+      "step": 5930
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07731877758281028,
+      "learning_rate": 0.0003066895666508605,
+      "loss": 1.3536,
+      "step": 5931
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07415858374192567,
+      "learning_rate": 0.0003065290098485749,
+      "loss": 1.4543,
+      "step": 5932
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07258684303449475,
+      "learning_rate": 0.000306368476505314,
+      "loss": 1.3418,
+      "step": 5933
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.0692859825650207,
+      "learning_rate": 0.0003062079666405429,
+      "loss": 1.4166,
+      "step": 5934
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07029306454398922,
+      "learning_rate": 0.00030604748027372394,
+      "loss": 1.4104,
+      "step": 5935
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.0733708023287787,
+      "learning_rate": 0.0003058870174243165,
+      "loss": 1.368,
+      "step": 5936
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07324475363368584,
+      "learning_rate": 0.00030572657811177756,
+      "loss": 1.5348,
+      "step": 5937
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07459659298410051,
+      "learning_rate": 0.0003055661623555608,
+      "loss": 1.5159,
+      "step": 5938
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07507360349738781,
+      "learning_rate": 0.000305405770175117,
+      "loss": 1.4729,
+      "step": 5939
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.0633251934623615,
+      "learning_rate": 0.0003052454015898948,
+      "loss": 1.3679,
+      "step": 5940
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.0793842984745501,
+      "learning_rate": 0.00030508505661933874,
+      "loss": 1.3561,
+      "step": 5941
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.06819289917335115,
+      "learning_rate": 0.0003049247352828917,
+      "loss": 1.4199,
+      "step": 5942
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.06660602563715051,
+      "learning_rate": 0.00030476443759999293,
+      "loss": 1.4692,
+      "step": 5943
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.0742100503490787,
+      "learning_rate": 0.00030460416359007913,
+      "loss": 1.3488,
+      "step": 5944
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07296538978034156,
+      "learning_rate": 0.00030444391327258424,
+      "loss": 1.4411,
+      "step": 5945
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07363651934129214,
+      "learning_rate": 0.00030428368666693905,
+      "loss": 1.4599,
+      "step": 5946
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.08639448348123929,
+      "learning_rate": 0.0003041234837925715,
+      "loss": 1.3221,
+      "step": 5947
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.0759787051112038,
+      "learning_rate": 0.0003039633046689069,
+      "loss": 1.4504,
+      "step": 5948
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.08473623915638286,
+      "learning_rate": 0.0003038031493153675,
+      "loss": 1.4055,
+      "step": 5949
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.09560174827623097,
+      "learning_rate": 0.00030364301775137245,
+      "loss": 1.4658,
+      "step": 5950
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.06961875392765221,
+      "learning_rate": 0.00030348290999633875,
+      "loss": 1.3609,
+      "step": 5951
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.0714538707164538,
+      "learning_rate": 0.0003033228260696795,
+      "loss": 1.3151,
+      "step": 5952
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07808335058336303,
+      "learning_rate": 0.00030316276599080566,
+      "loss": 1.4946,
+      "step": 5953
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.08282668322888084,
+      "learning_rate": 0.0003030027297791251,
+      "loss": 1.3759,
+      "step": 5954
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.06800510962454319,
+      "learning_rate": 0.00030284271745404257,
+      "loss": 1.4049,
+      "step": 5955
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07475879845301409,
+      "learning_rate": 0.00030268272903496036,
+      "loss": 1.4277,
+      "step": 5956
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07642167347690558,
+      "learning_rate": 0.0003025227645412775,
+      "loss": 1.3954,
+      "step": 5957
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07643834144422731,
+      "learning_rate": 0.0003023628239923902,
+      "loss": 1.3185,
+      "step": 5958
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07685637721938932,
+      "learning_rate": 0.0003022029074076919,
+      "loss": 1.4371,
+      "step": 5959
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07909995683759365,
+      "learning_rate": 0.000302043014806573,
+      "loss": 1.4715,
+      "step": 5960
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07349604482190411,
+      "learning_rate": 0.00030188314620842094,
+      "loss": 1.3706,
+      "step": 5961
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07233447503471593,
+      "learning_rate": 0.0003017233016326207,
+      "loss": 1.3405,
+      "step": 5962
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.0720093425925684,
+      "learning_rate": 0.0003015634810985534,
+      "loss": 1.5037,
+      "step": 5963
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.0726984904284892,
+      "learning_rate": 0.0003014036846255982,
+      "loss": 1.1959,
+      "step": 5964
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07403128407492846,
+      "learning_rate": 0.0003012439122331309,
+      "loss": 1.4965,
+      "step": 5965
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07709803247439713,
+      "learning_rate": 0.0003010841639405244,
+      "loss": 1.5168,
+      "step": 5966
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07602607776353369,
+      "learning_rate": 0.0003009244397671489,
+      "loss": 1.5096,
+      "step": 5967
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07978329522099706,
+      "learning_rate": 0.0003007647397323714,
+      "loss": 1.5163,
+      "step": 5968
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.08888699106540596,
+      "learning_rate": 0.00030060506385555595,
+      "loss": 1.479,
+      "step": 5969
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07168143621723064,
+      "learning_rate": 0.000300445412156064,
+      "loss": 1.3272,
+      "step": 5970
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07329373929619436,
+      "learning_rate": 0.0003002857846532538,
+      "loss": 1.3277,
+      "step": 5971
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07654288914163611,
+      "learning_rate": 0.00030012618136648087,
+      "loss": 1.3359,
+      "step": 5972
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07354952702168815,
+      "learning_rate": 0.0002999666023150974,
+      "loss": 1.4215,
+      "step": 5973
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.06871219507940302,
+      "learning_rate": 0.000299807047518453,
+      "loss": 1.3821,
+      "step": 5974
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.06497905023784828,
+      "learning_rate": 0.00029964751699589423,
+      "loss": 1.403,
+      "step": 5975
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07652963092713426,
+      "learning_rate": 0.00029948801076676475,
+      "loss": 1.3289,
+      "step": 5976
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.071199491483903,
+      "learning_rate": 0.00029932852885040527,
+      "loss": 1.3271,
+      "step": 5977
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.0745442586868254,
+      "learning_rate": 0.0002991690712661534,
+      "loss": 1.4917,
+      "step": 5978
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07328338974119428,
+      "learning_rate": 0.00029900963803334413,
+      "loss": 1.5249,
+      "step": 5979
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07679011633383984,
+      "learning_rate": 0.0002988502291713091,
+      "loss": 1.4964,
+      "step": 5980
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.06510401298584284,
+      "learning_rate": 0.0002986908446993772,
+      "loss": 1.4489,
+      "step": 5981
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07088074772151598,
+      "learning_rate": 0.0002985314846368745,
+      "loss": 1.4252,
+      "step": 5982
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.08126290418323827,
+      "learning_rate": 0.00029837214900312395,
+      "loss": 1.3525,
+      "step": 5983
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.06975498384895287,
+      "learning_rate": 0.00029821283781744537,
+      "loss": 1.5507,
+      "step": 5984
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07945659664328315,
+      "learning_rate": 0.0002980535510991558,
+      "loss": 1.3528,
+      "step": 5985
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07123058023593087,
+      "learning_rate": 0.0002978942888675693,
+      "loss": 1.3405,
+      "step": 5986
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07698558265674868,
+      "learning_rate": 0.0002977350511419972,
+      "loss": 1.432,
+      "step": 5987
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.06750769980073211,
+      "learning_rate": 0.0002975758379417474,
+      "loss": 1.5532,
+      "step": 5988
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07374373847083314,
+      "learning_rate": 0.00029741664928612503,
+      "loss": 1.3748,
+      "step": 5989
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.06784693456068619,
+      "learning_rate": 0.00029725748519443247,
+      "loss": 1.3238,
+      "step": 5990
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.0844336706395431,
+      "learning_rate": 0.0002970983456859687,
+      "loss": 1.3847,
+      "step": 5991
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07079001459931357,
+      "learning_rate": 0.00029693923078003004,
+      "loss": 1.2528,
+      "step": 5992
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07149040354390375,
+      "learning_rate": 0.00029678014049590995,
+      "loss": 1.3724,
+      "step": 5993
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.06824266996087566,
+      "learning_rate": 0.00029662107485289814,
+      "loss": 1.4637,
+      "step": 5994
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.0774125335188962,
+      "learning_rate": 0.0002964620338702823,
+      "loss": 1.38,
+      "step": 5995
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07150515374202804,
+      "learning_rate": 0.00029630301756734643,
+      "loss": 1.524,
+      "step": 5996
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07467047959655457,
+      "learning_rate": 0.000296144025963372,
+      "loss": 1.3389,
+      "step": 5997
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07005420069909853,
+      "learning_rate": 0.00029598505907763717,
+      "loss": 1.2697,
+      "step": 5998
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07941624953171082,
+      "learning_rate": 0.00029582611692941733,
+      "loss": 1.5686,
+      "step": 5999
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.08267345828706943,
+      "learning_rate": 0.0002956671995379847,
+      "loss": 1.4291,
+      "step": 6000
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.09815036580786829,
+      "learning_rate": 0.0002955083069226086,
+      "loss": 1.3402,
+      "step": 6001
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07314143721806532,
+      "learning_rate": 0.0002953494391025552,
+      "loss": 1.3167,
+      "step": 6002
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07718320633480043,
+      "learning_rate": 0.0002951905960970879,
+      "loss": 1.3187,
+      "step": 6003
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07022848258771092,
+      "learning_rate": 0.000295031777925467,
+      "loss": 1.388,
+      "step": 6004
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.0880795025536751,
+      "learning_rate": 0.0002948729846069496,
+      "loss": 1.3403,
+      "step": 6005
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.0701067624869438,
+      "learning_rate": 0.00029471421616078985,
+      "loss": 1.2425,
+      "step": 6006
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.09666753220487535,
+      "learning_rate": 0.0002945554726062392,
+      "loss": 1.4792,
+      "step": 6007
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.08171114169072571,
+      "learning_rate": 0.0002943967539625458,
+      "loss": 1.5026,
+      "step": 6008
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.11522314251288446,
+      "learning_rate": 0.0002942380602489547,
+      "loss": 1.5163,
+      "step": 6009
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.0769067408158239,
+      "learning_rate": 0.0002940793914847083,
+      "loss": 1.3969,
+      "step": 6010
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07605757451663893,
+      "learning_rate": 0.0002939207476890456,
+      "loss": 1.3257,
+      "step": 6011
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.08573115295190666,
+      "learning_rate": 0.00029376212888120257,
+      "loss": 1.4362,
+      "step": 6012
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.09276575884914588,
+      "learning_rate": 0.00029360353508041257,
+      "loss": 1.503,
+      "step": 6013
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07993298408856686,
+      "learning_rate": 0.00029344496630590543,
+      "loss": 1.4253,
+      "step": 6014
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.08552281071719056,
+      "learning_rate": 0.00029328642257690833,
+      "loss": 1.3724,
+      "step": 6015
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07851945938289058,
+      "learning_rate": 0.0002931279039126451,
+      "loss": 1.314,
+      "step": 6016
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07843057357746579,
+      "learning_rate": 0.0002929694103323366,
+      "loss": 1.475,
+      "step": 6017
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.08650642474253452,
+      "learning_rate": 0.00029281094185520094,
+      "loss": 1.4325,
+      "step": 6018
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07755758493084858,
+      "learning_rate": 0.00029265249850045285,
+      "loss": 1.4469,
+      "step": 6019
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.08462709644270011,
+      "learning_rate": 0.00029249408028730406,
+      "loss": 1.4838,
+      "step": 6020
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.08276090223466218,
+      "learning_rate": 0.0002923356872349635,
+      "loss": 1.4411,
+      "step": 6021
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.08773444524959222,
+      "learning_rate": 0.00029217731936263683,
+      "loss": 1.3605,
+      "step": 6022
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.08218699242748785,
+      "learning_rate": 0.0002920189766895265,
+      "loss": 1.3351,
+      "step": 6023
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07040731094288417,
+      "learning_rate": 0.00029186065923483243,
+      "loss": 1.3763,
+      "step": 6024
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.0740188467934851,
+      "learning_rate": 0.0002917023670177511,
+      "loss": 1.3715,
+      "step": 6025
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07362205859410367,
+      "learning_rate": 0.00029154410005747584,
+      "loss": 1.3693,
+      "step": 6026
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07984353221296256,
+      "learning_rate": 0.0002913858583731969,
+      "loss": 1.4406,
+      "step": 6027
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07519072202814801,
+      "learning_rate": 0.0002912276419841019,
+      "loss": 1.4988,
+      "step": 6028
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07631248401472579,
+      "learning_rate": 0.0002910694509093752,
+      "loss": 1.4252,
+      "step": 6029
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.08027426610180229,
+      "learning_rate": 0.00029091128516819763,
+      "loss": 1.2927,
+      "step": 6030
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.10664126095564695,
+      "learning_rate": 0.00029075314477974764,
+      "loss": 1.3158,
+      "step": 6031
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07485328411133679,
+      "learning_rate": 0.0002905950297632004,
+      "loss": 1.3941,
+      "step": 6032
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.08601281945619063,
+      "learning_rate": 0.0002904369401377275,
+      "loss": 1.4172,
+      "step": 6033
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07099566029910874,
+      "learning_rate": 0.00029027887592249815,
+      "loss": 1.3107,
+      "step": 6034
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07243682422839058,
+      "learning_rate": 0.00029012083713667833,
+      "loss": 1.3992,
+      "step": 6035
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07818980658292189,
+      "learning_rate": 0.0002899628237994305,
+      "loss": 1.3995,
+      "step": 6036
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07345641595125825,
+      "learning_rate": 0.00028980483592991427,
+      "loss": 1.4564,
+      "step": 6037
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07378658076318882,
+      "learning_rate": 0.0002896468735472863,
+      "loss": 1.3621,
+      "step": 6038
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07730154495877888,
+      "learning_rate": 0.0002894889366707002,
+      "loss": 1.3741,
+      "step": 6039
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07867744297535843,
+      "learning_rate": 0.0002893310253193062,
+      "loss": 1.4205,
+      "step": 6040
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.0806104655157845,
+      "learning_rate": 0.0002891731395122516,
+      "loss": 1.3632,
+      "step": 6041
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.09214380580505434,
+      "learning_rate": 0.0002890152792686809,
+      "loss": 1.7369,
+      "step": 6042
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.09003660892734845,
+      "learning_rate": 0.0002888574446077348,
+      "loss": 1.3157,
+      "step": 6043
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.08107202269847597,
+      "learning_rate": 0.0002886996355485514,
+      "loss": 1.3031,
+      "step": 6044
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07863112211660843,
+      "learning_rate": 0.0002885418521102658,
+      "loss": 1.4732,
+      "step": 6045
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.08105230040654554,
+      "learning_rate": 0.00028838409431200974,
+      "loss": 1.2731,
+      "step": 6046
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.08080853361750891,
+      "learning_rate": 0.0002882263621729116,
+      "loss": 1.4831,
+      "step": 6047
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.06849354451778983,
+      "learning_rate": 0.0002880686557120973,
+      "loss": 1.3394,
+      "step": 6048
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.0843146464274864,
+      "learning_rate": 0.00028791097494868895,
+      "loss": 1.3068,
+      "step": 6049
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07860656896213726,
+      "learning_rate": 0.0002877533199018061,
+      "loss": 1.4533,
+      "step": 6050
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07718407963474753,
+      "learning_rate": 0.00028759569059056507,
+      "loss": 1.4552,
+      "step": 6051
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07045024717197383,
+      "learning_rate": 0.00028743808703407867,
+      "loss": 1.4094,
+      "step": 6052
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.08392898263167461,
+      "learning_rate": 0.000287280509251457,
+      "loss": 1.4199,
+      "step": 6053
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07135753984364539,
+      "learning_rate": 0.00028712295726180715,
+      "loss": 1.3966,
+      "step": 6054
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07971585134038697,
+      "learning_rate": 0.00028696543108423247,
+      "loss": 1.3936,
+      "step": 6055
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07298002690430236,
+      "learning_rate": 0.0002868079307378336,
+      "loss": 1.4275,
+      "step": 6056
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07606117292151965,
+      "learning_rate": 0.0002866504562417086,
+      "loss": 1.4009,
+      "step": 6057
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07673914296478794,
+      "learning_rate": 0.0002864930076149509,
+      "loss": 1.4826,
+      "step": 6058
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07935188856224078,
+      "learning_rate": 0.00028633558487665213,
+      "loss": 1.3802,
+      "step": 6059
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07771790506095733,
+      "learning_rate": 0.00028617818804590046,
+      "loss": 1.3848,
+      "step": 6060
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.08481704582222165,
+      "learning_rate": 0.0002860208171417804,
+      "loss": 1.4468,
+      "step": 6061
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.08220689565548146,
+      "learning_rate": 0.0002858634721833741,
+      "loss": 1.4044,
+      "step": 6062
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.08349805126907142,
+      "learning_rate": 0.00028570615318976013,
+      "loss": 1.4339,
+      "step": 6063
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.0782694576272315,
+      "learning_rate": 0.0002855488601800137,
+      "loss": 1.377,
+      "step": 6064
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07431275992276852,
+      "learning_rate": 0.0002853915931732073,
+      "loss": 1.3656,
+      "step": 6065
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.09453173639235407,
+      "learning_rate": 0.00028523435218841026,
+      "loss": 1.3901,
+      "step": 6066
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07031269040346856,
+      "learning_rate": 0.00028507713724468844,
+      "loss": 1.4042,
+      "step": 6067
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.0882233685502084,
+      "learning_rate": 0.0002849199483611045,
+      "loss": 1.3748,
+      "step": 6068
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07604641853251956,
+      "learning_rate": 0.0002847627855567183,
+      "loss": 1.36,
+      "step": 6069
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.08338095020434803,
+      "learning_rate": 0.0002846056488505866,
+      "loss": 1.4305,
+      "step": 6070
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07695158760839388,
+      "learning_rate": 0.0002844485382617624,
+      "loss": 1.3133,
+      "step": 6071
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07082354574245085,
+      "learning_rate": 0.000284291453809296,
+      "loss": 1.3982,
+      "step": 6072
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07190282632214344,
+      "learning_rate": 0.0002841343955122347,
+      "loss": 1.3965,
+      "step": 6073
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.08003261627937998,
+      "learning_rate": 0.00028397736338962187,
+      "loss": 1.4751,
+      "step": 6074
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07217113093699555,
+      "learning_rate": 0.0002838203574604986,
+      "loss": 1.4419,
+      "step": 6075
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.09568634177982653,
+      "learning_rate": 0.00028366337774390237,
+      "loss": 1.3784,
+      "step": 6076
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.08143677906916136,
+      "learning_rate": 0.0002835064242588673,
+      "loss": 1.3997,
+      "step": 6077
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.06714182759584644,
+      "learning_rate": 0.00028334949702442483,
+      "loss": 1.3418,
+      "step": 6078
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07817525588333529,
+      "learning_rate": 0.0002831925960596027,
+      "loss": 1.4515,
+      "step": 6079
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07766148184272453,
+      "learning_rate": 0.0002830357213834256,
+      "loss": 1.4837,
+      "step": 6080
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.08646731671686747,
+      "learning_rate": 0.0002828788730149152,
+      "loss": 1.4054,
+      "step": 6081
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.08325901309195903,
+      "learning_rate": 0.0002827220509730903,
+      "loss": 1.3141,
+      "step": 6082
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07714246322339714,
+      "learning_rate": 0.0002825652552769655,
+      "loss": 1.3832,
+      "step": 6083
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.08684031433009541,
+      "learning_rate": 0.0002824084859455531,
+      "loss": 1.3973,
+      "step": 6084
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.06920637783187342,
+      "learning_rate": 0.0002822517429978622,
+      "loss": 1.3132,
+      "step": 6085
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07746410524265271,
+      "learning_rate": 0.0002820950264528979,
+      "loss": 1.3978,
+      "step": 6086
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07727541915447587,
+      "learning_rate": 0.00028193833632966293,
+      "loss": 1.3852,
+      "step": 6087
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07721133474886398,
+      "learning_rate": 0.0002817816726471566,
+      "loss": 1.4555,
+      "step": 6088
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07974176592718676,
+      "learning_rate": 0.0002816250354243748,
+      "loss": 1.4801,
+      "step": 6089
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.06787223694888275,
+      "learning_rate": 0.00028146842468031007,
+      "loss": 1.4043,
+      "step": 6090
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.08216216293404127,
+      "learning_rate": 0.00028131184043395254,
+      "loss": 1.3816,
+      "step": 6091
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07311183455847245,
+      "learning_rate": 0.000281155282704288,
+      "loss": 1.5151,
+      "step": 6092
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.08228569645390815,
+      "learning_rate": 0.00028099875151029987,
+      "loss": 1.4022,
+      "step": 6093
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08468156268757646,
+      "learning_rate": 0.0002808422468709684,
+      "loss": 1.3635,
+      "step": 6094
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.084115782220857,
+      "learning_rate": 0.00028068576880526985,
+      "loss": 1.5106,
+      "step": 6095
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.0939884815741877,
+      "learning_rate": 0.0002805293173321779,
+      "loss": 1.4746,
+      "step": 6096
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08867327787743164,
+      "learning_rate": 0.00028037289247066305,
+      "loss": 1.3817,
+      "step": 6097
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.0810061217356107,
+      "learning_rate": 0.000280216494239692,
+      "loss": 1.3271,
+      "step": 6098
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08379496101255957,
+      "learning_rate": 0.0002800601226582289,
+      "loss": 1.3737,
+      "step": 6099
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08031308779705486,
+      "learning_rate": 0.00027990377774523396,
+      "loss": 1.5601,
+      "step": 6100
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.07753119254590604,
+      "learning_rate": 0.00027974745951966494,
+      "loss": 1.3183,
+      "step": 6101
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08531571995244437,
+      "learning_rate": 0.00027959116800047566,
+      "loss": 1.3538,
+      "step": 6102
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.0776622313440044,
+      "learning_rate": 0.00027943490320661704,
+      "loss": 1.4598,
+      "step": 6103
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.09492248497676697,
+      "learning_rate": 0.000279278665157037,
+      "loss": 1.3709,
+      "step": 6104
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08248818947863316,
+      "learning_rate": 0.0002791224538706797,
+      "loss": 1.3884,
+      "step": 6105
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.07468677694976233,
+      "learning_rate": 0.0002789662693664862,
+      "loss": 1.3133,
+      "step": 6106
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08673493539274889,
+      "learning_rate": 0.00027881011166339483,
+      "loss": 1.3215,
+      "step": 6107
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08082725029580298,
+      "learning_rate": 0.00027865398078033977,
+      "loss": 1.4312,
+      "step": 6108
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08379905097619719,
+      "learning_rate": 0.0002784978767362528,
+      "loss": 1.419,
+      "step": 6109
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.07778141120838956,
+      "learning_rate": 0.0002783417995500619,
+      "loss": 1.4632,
+      "step": 6110
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.07634657455999405,
+      "learning_rate": 0.00027818574924069186,
+      "loss": 1.3575,
+      "step": 6111
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08029962788812359,
+      "learning_rate": 0.0002780297258270643,
+      "loss": 1.3785,
+      "step": 6112
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08130136262534324,
+      "learning_rate": 0.00027787372932809806,
+      "loss": 1.5693,
+      "step": 6113
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.0788973332314687,
+      "learning_rate": 0.0002777177597627077,
+      "loss": 1.3804,
+      "step": 6114
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.09548515561076097,
+      "learning_rate": 0.00027756181714980524,
+      "loss": 1.3901,
+      "step": 6115
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.07356430507555382,
+      "learning_rate": 0.0002774059015082996,
+      "loss": 1.3691,
+      "step": 6116
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.0895079629964563,
+      "learning_rate": 0.0002772500128570955,
+      "loss": 1.3651,
+      "step": 6117
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.0756789159867376,
+      "learning_rate": 0.00027709415121509527,
+      "loss": 1.3808,
+      "step": 6118
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08174272585535419,
+      "learning_rate": 0.00027693831660119796,
+      "loss": 1.4582,
+      "step": 6119
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.07347513506484865,
+      "learning_rate": 0.00027678250903429876,
+      "loss": 1.3427,
+      "step": 6120
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.07628590655891802,
+      "learning_rate": 0.0002766267285332896,
+      "loss": 1.3316,
+      "step": 6121
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.0805266209828289,
+      "learning_rate": 0.00027647097511706,
+      "loss": 1.4766,
+      "step": 6122
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.07804202650616353,
+      "learning_rate": 0.0002763152488044951,
+      "loss": 1.3473,
+      "step": 6123
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.07563813059212712,
+      "learning_rate": 0.0002761595496144773,
+      "loss": 1.4365,
+      "step": 6124
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.07205259007008297,
+      "learning_rate": 0.00027600387756588615,
+      "loss": 1.4753,
+      "step": 6125
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.0768814719100409,
+      "learning_rate": 0.00027584823267759684,
+      "loss": 1.2608,
+      "step": 6126
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.07656294739845501,
+      "learning_rate": 0.000275692614968482,
+      "loss": 1.3841,
+      "step": 6127
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.0785779592286055,
+      "learning_rate": 0.00027553702445741126,
+      "loss": 1.5065,
+      "step": 6128
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.07947757326516629,
+      "learning_rate": 0.0002753814611632498,
+      "loss": 1.4559,
+      "step": 6129
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.07434673630299542,
+      "learning_rate": 0.0002752259251048606,
+      "loss": 1.3781,
+      "step": 6130
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08122447339583906,
+      "learning_rate": 0.00027507041630110305,
+      "loss": 1.5396,
+      "step": 6131
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.07766472263427851,
+      "learning_rate": 0.0002749149347708331,
+      "loss": 1.3898,
+      "step": 6132
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.0875456038166925,
+      "learning_rate": 0.000274759480532903,
+      "loss": 1.3486,
+      "step": 6133
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08330838816033348,
+      "learning_rate": 0.00027460405360616244,
+      "loss": 1.37,
+      "step": 6134
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.07743418564873557,
+      "learning_rate": 0.00027444865400945765,
+      "loss": 1.2946,
+      "step": 6135
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.07452191764881552,
+      "learning_rate": 0.0002742932817616309,
+      "loss": 1.4033,
+      "step": 6136
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08104030355111651,
+      "learning_rate": 0.0002741379368815219,
+      "loss": 1.3705,
+      "step": 6137
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08512209537619303,
+      "learning_rate": 0.0002739826193879669,
+      "loss": 1.4029,
+      "step": 6138
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08401661215344652,
+      "learning_rate": 0.0002738273292997983,
+      "loss": 1.3075,
+      "step": 6139
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08496479321642669,
+      "learning_rate": 0.00027367206663584574,
+      "loss": 1.412,
+      "step": 6140
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.07883745123661512,
+      "learning_rate": 0.00027351683141493567,
+      "loss": 1.3398,
+      "step": 6141
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.07946198965375008,
+      "learning_rate": 0.00027336162365589045,
+      "loss": 1.3921,
+      "step": 6142
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.06745687764508318,
+      "learning_rate": 0.00027320644337752964,
+      "loss": 1.4423,
+      "step": 6143
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08892866120135398,
+      "learning_rate": 0.00027305129059866954,
+      "loss": 1.2575,
+      "step": 6144
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.06875335376318778,
+      "learning_rate": 0.00027289616533812276,
+      "loss": 1.4766,
+      "step": 6145
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08030815203424742,
+      "learning_rate": 0.00027274106761469884,
+      "loss": 1.3601,
+      "step": 6146
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.07478088113386552,
+      "learning_rate": 0.00027258599744720414,
+      "loss": 1.4551,
+      "step": 6147
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.07733691783191772,
+      "learning_rate": 0.000272430954854441,
+      "loss": 1.4606,
+      "step": 6148
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.07488938900948164,
+      "learning_rate": 0.0002722759398552093,
+      "loss": 1.315,
+      "step": 6149
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.07264299787776964,
+      "learning_rate": 0.00027212095246830506,
+      "loss": 1.4117,
+      "step": 6150
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08146802376307108,
+      "learning_rate": 0.00027196599271252074,
+      "loss": 1.3426,
+      "step": 6151
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.07965692665262414,
+      "learning_rate": 0.00027181106060664617,
+      "loss": 1.5066,
+      "step": 6152
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08155946793072419,
+      "learning_rate": 0.0002716561561694673,
+      "loss": 1.3994,
+      "step": 6153
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.07316359754028685,
+      "learning_rate": 0.00027150127941976635,
+      "loss": 1.395,
+      "step": 6154
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08202388565748747,
+      "learning_rate": 0.0002713464303763231,
+      "loss": 1.3386,
+      "step": 6155
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.07804125380153318,
+      "learning_rate": 0.0002711916090579137,
+      "loss": 1.4255,
+      "step": 6156
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.07175148851830035,
+      "learning_rate": 0.00027103681548331027,
+      "loss": 1.3847,
+      "step": 6157
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.07387603354527797,
+      "learning_rate": 0.00027088204967128235,
+      "loss": 1.4942,
+      "step": 6158
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.06952552748421588,
+      "learning_rate": 0.00027072731164059594,
+      "loss": 1.4904,
+      "step": 6159
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.07546195447701311,
+      "learning_rate": 0.00027057260141001327,
+      "loss": 1.3508,
+      "step": 6160
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.07874233894984566,
+      "learning_rate": 0.0002704179189982936,
+      "loss": 1.3431,
+      "step": 6161
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.07561417364598048,
+      "learning_rate": 0.00027026326442419296,
+      "loss": 1.3988,
+      "step": 6162
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.06969834031693835,
+      "learning_rate": 0.00027010863770646356,
+      "loss": 1.4481,
+      "step": 6163
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.06957347570850005,
+      "learning_rate": 0.0002699540388638542,
+      "loss": 1.5234,
+      "step": 6164
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08021592576713846,
+      "learning_rate": 0.0002697994679151106,
+      "loss": 1.4236,
+      "step": 6165
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.07249343492297182,
+      "learning_rate": 0.0002696449248789754,
+      "loss": 1.4235,
+      "step": 6166
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08347570438059976,
+      "learning_rate": 0.000269490409774187,
+      "loss": 1.4639,
+      "step": 6167
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.06813592066830779,
+      "learning_rate": 0.0002693359226194812,
+      "loss": 1.4131,
+      "step": 6168
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.088520289977426,
+      "learning_rate": 0.0002691814634335902,
+      "loss": 1.4653,
+      "step": 6169
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.07927725480495933,
+      "learning_rate": 0.00026902703223524217,
+      "loss": 1.3032,
+      "step": 6170
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.07476030815511266,
+      "learning_rate": 0.0002688726290431629,
+      "loss": 1.4893,
+      "step": 6171
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.07592562154301802,
+      "learning_rate": 0.0002687182538760743,
+      "loss": 1.547,
+      "step": 6172
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.07741661330328563,
+      "learning_rate": 0.0002685639067526948,
+      "loss": 1.432,
+      "step": 6173
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.07284139646621345,
+      "learning_rate": 0.0002684095876917393,
+      "loss": 1.4776,
+      "step": 6174
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08675181858083143,
+      "learning_rate": 0.00026825529671191986,
+      "loss": 1.3249,
+      "step": 6175
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08411753433275078,
+      "learning_rate": 0.0002681010338319444,
+      "loss": 1.4295,
+      "step": 6176
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.07837071521163565,
+      "learning_rate": 0.00026794679907051817,
+      "loss": 1.4625,
+      "step": 6177
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.07019930824849117,
+      "learning_rate": 0.0002677925924463427,
+      "loss": 1.3814,
+      "step": 6178
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.07544575050760528,
+      "learning_rate": 0.00026763841397811573,
+      "loss": 1.6822,
+      "step": 6179
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.0711653064704804,
+      "learning_rate": 0.0002674842636845322,
+      "loss": 1.2985,
+      "step": 6180
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08226687927820503,
+      "learning_rate": 0.0002673301415842835,
+      "loss": 1.2929,
+      "step": 6181
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08082723301222589,
+      "learning_rate": 0.00026717604769605695,
+      "loss": 1.4702,
+      "step": 6182
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.0771809400828876,
+      "learning_rate": 0.0002670219820385374,
+      "loss": 1.4699,
+      "step": 6183
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.07712962428203328,
+      "learning_rate": 0.0002668679446304061,
+      "loss": 1.3207,
+      "step": 6184
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.07212775681925493,
+      "learning_rate": 0.00026671393549033983,
+      "loss": 1.4173,
+      "step": 6185
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08585221084325963,
+      "learning_rate": 0.0002665599546370131,
+      "loss": 1.4432,
+      "step": 6186
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.07161579555511431,
+      "learning_rate": 0.0002664060020890969,
+      "loss": 1.3093,
+      "step": 6187
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.07166138838974309,
+      "learning_rate": 0.00026625207786525805,
+      "loss": 1.3763,
+      "step": 6188
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08179686758556896,
+      "learning_rate": 0.0002660981819841607,
+      "loss": 1.3632,
+      "step": 6189
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.07687496958224152,
+      "learning_rate": 0.00026594431446446526,
+      "loss": 1.2778,
+      "step": 6190
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08429720606973103,
+      "learning_rate": 0.00026579047532482845,
+      "loss": 1.4235,
+      "step": 6191
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.07874039330715418,
+      "learning_rate": 0.00026563666458390394,
+      "loss": 1.468,
+      "step": 6192
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.0849153270188568,
+      "learning_rate": 0.00026548288226034203,
+      "loss": 1.3426,
+      "step": 6193
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.07719925522577803,
+      "learning_rate": 0.00026532912837278893,
+      "loss": 1.3349,
+      "step": 6194
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.07244606222045334,
+      "learning_rate": 0.0002651754029398883,
+      "loss": 1.4874,
+      "step": 6195
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.07435186661080587,
+      "learning_rate": 0.0002650217059802794,
+      "loss": 1.4572,
+      "step": 6196
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.07718559468867621,
+      "learning_rate": 0.0002648680375125989,
+      "loss": 1.4856,
+      "step": 6197
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08202214417553845,
+      "learning_rate": 0.0002647143975554793,
+      "loss": 1.3457,
+      "step": 6198
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.0760250326747222,
+      "learning_rate": 0.0002645607861275502,
+      "loss": 1.439,
+      "step": 6199
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.07603471814525921,
+      "learning_rate": 0.00026440720324743763,
+      "loss": 1.3743,
+      "step": 6200
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.0828742228552649,
+      "learning_rate": 0.0002642536489337636,
+      "loss": 1.4984,
+      "step": 6201
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08352744199976005,
+      "learning_rate": 0.00026410012320514744,
+      "loss": 1.5592,
+      "step": 6202
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.07460991618324651,
+      "learning_rate": 0.0002639466260802048,
+      "loss": 1.4728,
+      "step": 6203
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08349369977006431,
+      "learning_rate": 0.0002637931575775474,
+      "loss": 1.3287,
+      "step": 6204
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08921844889478096,
+      "learning_rate": 0.00026363971771578413,
+      "loss": 1.4241,
+      "step": 6205
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.07896415452446431,
+      "learning_rate": 0.00026348630651352,
+      "loss": 1.3908,
+      "step": 6206
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08815108621019546,
+      "learning_rate": 0.00026333292398935635,
+      "loss": 1.4188,
+      "step": 6207
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.0785662676339109,
+      "learning_rate": 0.00026317957016189155,
+      "loss": 1.4875,
+      "step": 6208
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.07435226352357563,
+      "learning_rate": 0.0002630262450497205,
+      "loss": 1.4019,
+      "step": 6209
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.07869236534132566,
+      "learning_rate": 0.0002628729486714341,
+      "loss": 1.5194,
+      "step": 6210
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08111495196318634,
+      "learning_rate": 0.0002627196810456201,
+      "loss": 1.4171,
+      "step": 6211
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.0752417116249629,
+      "learning_rate": 0.000262566442190863,
+      "loss": 1.1826,
+      "step": 6212
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.07090451927739674,
+      "learning_rate": 0.0002624132321257432,
+      "loss": 1.4379,
+      "step": 6213
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.072209257295845,
+      "learning_rate": 0.00026226005086883806,
+      "loss": 1.2822,
+      "step": 6214
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.0775073866558998,
+      "learning_rate": 0.0002621068984387216,
+      "loss": 1.6329,
+      "step": 6215
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.07276027526755828,
+      "learning_rate": 0.00026195377485396375,
+      "loss": 1.4094,
+      "step": 6216
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.07660187872068078,
+      "learning_rate": 0.0002618006801331313,
+      "loss": 1.4048,
+      "step": 6217
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.07561214597012043,
+      "learning_rate": 0.00026164761429478767,
+      "loss": 1.3556,
+      "step": 6218
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.07608383237940829,
+      "learning_rate": 0.00026149457735749235,
+      "loss": 1.4997,
+      "step": 6219
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08117197065650657,
+      "learning_rate": 0.00026134156933980187,
+      "loss": 1.5299,
+      "step": 6220
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.07667811374754589,
+      "learning_rate": 0.000261188590260269,
+      "loss": 1.5114,
+      "step": 6221
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.07236230591516954,
+      "learning_rate": 0.0002610356401374427,
+      "loss": 1.4631,
+      "step": 6222
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.0729862517775749,
+      "learning_rate": 0.0002608827189898688,
+      "loss": 1.4073,
+      "step": 6223
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08487372745622469,
+      "learning_rate": 0.0002607298268360899,
+      "loss": 1.5601,
+      "step": 6224
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.0895758449211697,
+      "learning_rate": 0.0002605769636946441,
+      "loss": 1.2656,
+      "step": 6225
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.07973668237947351,
+      "learning_rate": 0.00026042412958406715,
+      "loss": 1.2544,
+      "step": 6226
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.07568869419754443,
+      "learning_rate": 0.00026027132452289013,
+      "loss": 1.4048,
+      "step": 6227
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.0744278379703313,
+      "learning_rate": 0.0002601185485296418,
+      "loss": 1.4452,
+      "step": 6228
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.07385615820726067,
+      "learning_rate": 0.00025996580162284614,
+      "loss": 1.3883,
+      "step": 6229
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.07289747818726597,
+      "learning_rate": 0.0002598130838210246,
+      "loss": 1.5465,
+      "step": 6230
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.07867781806652333,
+      "learning_rate": 0.0002596603951426949,
+      "loss": 1.4572,
+      "step": 6231
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.07007763050297192,
+      "learning_rate": 0.00025950773560637076,
+      "loss": 1.5115,
+      "step": 6232
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.0781571173215587,
+      "learning_rate": 0.00025935510523056267,
+      "loss": 1.2784,
+      "step": 6233
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08274854533054618,
+      "learning_rate": 0.00025920250403377787,
+      "loss": 1.4113,
+      "step": 6234
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.0769569700994162,
+      "learning_rate": 0.0002590499320345195,
+      "loss": 1.4623,
+      "step": 6235
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.07322203756357455,
+      "learning_rate": 0.0002588973892512875,
+      "loss": 1.3969,
+      "step": 6236
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.07285908482504222,
+      "learning_rate": 0.00025874487570257864,
+      "loss": 1.3235,
+      "step": 6237
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.0824245182399957,
+      "learning_rate": 0.0002585923914068851,
+      "loss": 1.3926,
+      "step": 6238
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08538154374456672,
+      "learning_rate": 0.00025843993638269616,
+      "loss": 1.3852,
+      "step": 6239
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08303180796543792,
+      "learning_rate": 0.000258287510648498,
+      "loss": 1.291,
+      "step": 6240
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.0744008611060297,
+      "learning_rate": 0.00025813511422277224,
+      "loss": 1.4407,
+      "step": 6241
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.07491379607406161,
+      "learning_rate": 0.0002579827471239978,
+      "loss": 1.5728,
+      "step": 6242
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.07224802257116374,
+      "learning_rate": 0.0002578304093706497,
+      "loss": 1.4437,
+      "step": 6243
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08108850715756206,
+      "learning_rate": 0.00025767810098119927,
+      "loss": 1.527,
+      "step": 6244
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.07409277926113383,
+      "learning_rate": 0.00025752582197411446,
+      "loss": 1.5389,
+      "step": 6245
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08538872621157978,
+      "learning_rate": 0.00025737357236785984,
+      "loss": 1.3672,
+      "step": 6246
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08357544942655332,
+      "learning_rate": 0.0002572213521808959,
+      "loss": 1.5069,
+      "step": 6247
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.07945973711646714,
+      "learning_rate": 0.0002570691614316802,
+      "loss": 1.368,
+      "step": 6248
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08676847576563627,
+      "learning_rate": 0.00025691700013866616,
+      "loss": 1.3731,
+      "step": 6249
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08102362106338495,
+      "learning_rate": 0.00025676486832030366,
+      "loss": 1.4159,
+      "step": 6250
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08972372454953455,
+      "learning_rate": 0.00025661276599503945,
+      "loss": 1.3462,
+      "step": 6251
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08439968518474841,
+      "learning_rate": 0.0002564606931813166,
+      "loss": 1.3473,
+      "step": 6252
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.07589797298359928,
+      "learning_rate": 0.0002563086498975742,
+      "loss": 1.4392,
+      "step": 6253
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.07431464211955395,
+      "learning_rate": 0.00025615663616224805,
+      "loss": 1.3391,
+      "step": 6254
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.07807770877253503,
+      "learning_rate": 0.00025600465199377066,
+      "loss": 1.3193,
+      "step": 6255
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.07428484878269796,
+      "learning_rate": 0.00025585269741057026,
+      "loss": 1.4478,
+      "step": 6256
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.07081123718795393,
+      "learning_rate": 0.00025570077243107194,
+      "loss": 1.4212,
+      "step": 6257
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08312259339625878,
+      "learning_rate": 0.00025554887707369733,
+      "loss": 1.3672,
+      "step": 6258
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.09350706453268828,
+      "learning_rate": 0.0002553970113568642,
+      "loss": 1.5117,
+      "step": 6259
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.07624662839361142,
+      "learning_rate": 0.0002552451752989865,
+      "loss": 1.4194,
+      "step": 6260
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08381231697762678,
+      "learning_rate": 0.0002550933689184751,
+      "loss": 1.4533,
+      "step": 6261
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.0744415792079469,
+      "learning_rate": 0.00025494159223373727,
+      "loss": 1.4775,
+      "step": 6262
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.07773631383404493,
+      "learning_rate": 0.00025478984526317597,
+      "loss": 1.3141,
+      "step": 6263
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.07366855260516596,
+      "learning_rate": 0.0002546381280251913,
+      "loss": 1.4199,
+      "step": 6264
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.0801259194771756,
+      "learning_rate": 0.0002544864405381797,
+      "loss": 1.2572,
+      "step": 6265
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.07759226269944282,
+      "learning_rate": 0.0002543347828205335,
+      "loss": 1.3919,
+      "step": 6266
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.07901655886822204,
+      "learning_rate": 0.0002541831548906417,
+      "loss": 1.4355,
+      "step": 6267
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08200143516361498,
+      "learning_rate": 0.0002540315567668901,
+      "loss": 1.3433,
+      "step": 6268
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.0860479874628173,
+      "learning_rate": 0.00025387998846766014,
+      "loss": 1.3361,
+      "step": 6269
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.0750332981935864,
+      "learning_rate": 0.00025372845001132995,
+      "loss": 1.39,
+      "step": 6270
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08091321341472489,
+      "learning_rate": 0.0002535769414162744,
+      "loss": 1.3943,
+      "step": 6271
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08131512343994637,
+      "learning_rate": 0.0002534254627008641,
+      "loss": 1.3901,
+      "step": 6272
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08340487879640397,
+      "learning_rate": 0.0002532740138834665,
+      "loss": 1.4814,
+      "step": 6273
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.0823885510183785,
+      "learning_rate": 0.00025312259498244547,
+      "loss": 1.4846,
+      "step": 6274
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08451185254200341,
+      "learning_rate": 0.00025297120601616073,
+      "loss": 1.2977,
+      "step": 6275
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08059700760409108,
+      "learning_rate": 0.0002528198470029689,
+      "loss": 1.3714,
+      "step": 6276
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08329022693965066,
+      "learning_rate": 0.00025266851796122296,
+      "loss": 1.3995,
+      "step": 6277
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.07771378342062599,
+      "learning_rate": 0.0002525172189092717,
+      "loss": 1.4348,
+      "step": 6278
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08460287024642418,
+      "learning_rate": 0.00025236594986546113,
+      "loss": 1.4236,
+      "step": 6279
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.07704392579140935,
+      "learning_rate": 0.00025221471084813275,
+      "loss": 1.3786,
+      "step": 6280
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08446989165175656,
+      "learning_rate": 0.0002520635018756248,
+      "loss": 1.3692,
+      "step": 6281
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08354359803524838,
+      "learning_rate": 0.00025191232296627197,
+      "loss": 1.569,
+      "step": 6282
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08001104967843899,
+      "learning_rate": 0.0002517611741384055,
+      "loss": 1.3756,
+      "step": 6283
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.0799008608461087,
+      "learning_rate": 0.0002516100554103523,
+      "loss": 1.4743,
+      "step": 6284
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08895399103852052,
+      "learning_rate": 0.0002514589668004362,
+      "loss": 1.4824,
+      "step": 6285
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08389049233996097,
+      "learning_rate": 0.00025130790832697737,
+      "loss": 1.4368,
+      "step": 6286
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08167409917869137,
+      "learning_rate": 0.0002511568800082919,
+      "loss": 1.4084,
+      "step": 6287
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.0820738335616238,
+      "learning_rate": 0.00025100588186269257,
+      "loss": 1.272,
+      "step": 6288
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.07262380780001264,
+      "learning_rate": 0.0002508549139084887,
+      "loss": 1.5007,
+      "step": 6289
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08250263142633718,
+      "learning_rate": 0.00025070397616398543,
+      "loss": 1.291,
+      "step": 6290
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.07324180126088345,
+      "learning_rate": 0.00025055306864748433,
+      "loss": 1.3554,
+      "step": 6291
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.07491530886846616,
+      "learning_rate": 0.0002504021913772836,
+      "loss": 1.3406,
+      "step": 6292
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.07998014625838289,
+      "learning_rate": 0.00025025134437167793,
+      "loss": 1.522,
+      "step": 6293
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08076320367760288,
+      "learning_rate": 0.0002501005276489575,
+      "loss": 1.3689,
+      "step": 6294
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.0781089542719641,
+      "learning_rate": 0.00024994974122740965,
+      "loss": 1.4426,
+      "step": 6295
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.07999448507874525,
+      "learning_rate": 0.00024979898512531793,
+      "loss": 1.4671,
+      "step": 6296
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.07443936163831828,
+      "learning_rate": 0.00024964825936096163,
+      "loss": 1.4709,
+      "step": 6297
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.07752927506986254,
+      "learning_rate": 0.00024949756395261703,
+      "loss": 1.4224,
+      "step": 6298
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.07297892064768649,
+      "learning_rate": 0.00024934689891855657,
+      "loss": 1.4106,
+      "step": 6299
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.0705016410192572,
+      "learning_rate": 0.00024919626427704874,
+      "loss": 1.4373,
+      "step": 6300
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.06998077427911564,
+      "learning_rate": 0.0002490456600463583,
+      "loss": 1.3348,
+      "step": 6301
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08936338441734996,
+      "learning_rate": 0.0002488950862447469,
+      "loss": 1.4434,
+      "step": 6302
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08152473768427583,
+      "learning_rate": 0.00024874454289047196,
+      "loss": 1.3254,
+      "step": 6303
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.0811328676071586,
+      "learning_rate": 0.0002485940300017873,
+      "loss": 1.2318,
+      "step": 6304
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.07536323950532108,
+      "learning_rate": 0.00024844354759694337,
+      "loss": 1.3747,
+      "step": 6305
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08189574613290967,
+      "learning_rate": 0.0002482930956941865,
+      "loss": 1.3824,
+      "step": 6306
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08466969026863551,
+      "learning_rate": 0.0002481426743117594,
+      "loss": 1.4511,
+      "step": 6307
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.07700482866848125,
+      "learning_rate": 0.00024799228346790154,
+      "loss": 1.3756,
+      "step": 6308
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.07394423034126002,
+      "learning_rate": 0.00024784192318084796,
+      "loss": 1.3997,
+      "step": 6309
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08163148393060057,
+      "learning_rate": 0.0002476915934688305,
+      "loss": 1.3569,
+      "step": 6310
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08245747676419808,
+      "learning_rate": 0.00024754129435007756,
+      "loss": 1.4508,
+      "step": 6311
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08316884689132627,
+      "learning_rate": 0.00024739102584281266,
+      "loss": 1.5506,
+      "step": 6312
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08981091998302579,
+      "learning_rate": 0.0002472407879652567,
+      "loss": 1.3847,
+      "step": 6313
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.07069410558333868,
+      "learning_rate": 0.00024709058073562684,
+      "loss": 1.3207,
+      "step": 6314
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.07217099140842033,
+      "learning_rate": 0.0002469404041721358,
+      "loss": 1.2995,
+      "step": 6315
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.0819390025189319,
+      "learning_rate": 0.0002467902582929931,
+      "loss": 1.3276,
+      "step": 6316
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.09764517482898454,
+      "learning_rate": 0.0002466401431164048,
+      "loss": 1.3504,
+      "step": 6317
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08663621031922736,
+      "learning_rate": 0.0002464900586605724,
+      "loss": 1.4582,
+      "step": 6318
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.07135499230948997,
+      "learning_rate": 0.0002463400049436944,
+      "loss": 1.3456,
+      "step": 6319
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.07806358456612215,
+      "learning_rate": 0.0002461899819839654,
+      "loss": 1.4174,
+      "step": 6320
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08049798140545715,
+      "learning_rate": 0.000246039989799576,
+      "loss": 1.3761,
+      "step": 6321
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08031384734125688,
+      "learning_rate": 0.00024589002840871355,
+      "loss": 1.374,
+      "step": 6322
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.07909678000250613,
+      "learning_rate": 0.00024574009782956096,
+      "loss": 1.6276,
+      "step": 6323
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.07584212527403288,
+      "learning_rate": 0.0002455901980802983,
+      "loss": 1.428,
+      "step": 6324
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08462804630783761,
+      "learning_rate": 0.000245440329179101,
+      "loss": 1.3807,
+      "step": 6325
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.09435281069530776,
+      "learning_rate": 0.0002452904911441414,
+      "loss": 1.3654,
+      "step": 6326
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08026059661730983,
+      "learning_rate": 0.0002451406839935881,
+      "loss": 1.2393,
+      "step": 6327
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.09463728059653936,
+      "learning_rate": 0.00024499090774560524,
+      "loss": 1.3196,
+      "step": 6328
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08561772489183278,
+      "learning_rate": 0.00024484116241835403,
+      "loss": 1.4476,
+      "step": 6329
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08344976913598492,
+      "learning_rate": 0.00024469144802999164,
+      "loss": 1.3839,
+      "step": 6330
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.0855239337159454,
+      "learning_rate": 0.0002445417645986713,
+      "loss": 1.5262,
+      "step": 6331
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08926849614318069,
+      "learning_rate": 0.00024439211214254277,
+      "loss": 1.3539,
+      "step": 6332
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08822555963908468,
+      "learning_rate": 0.0002442424906797519,
+      "loss": 1.3936,
+      "step": 6333
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08578391466405638,
+      "learning_rate": 0.0002440929002284406,
+      "loss": 1.4229,
+      "step": 6334
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.086223295049479,
+      "learning_rate": 0.00024394334080674734,
+      "loss": 1.3489,
+      "step": 6335
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.10149756749252606,
+      "learning_rate": 0.00024379381243280703,
+      "loss": 1.5394,
+      "step": 6336
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08212011780136454,
+      "learning_rate": 0.00024364431512475,
+      "loss": 1.4554,
+      "step": 6337
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.09010571016682901,
+      "learning_rate": 0.00024349484890070355,
+      "loss": 1.4116,
+      "step": 6338
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.10287841772251785,
+      "learning_rate": 0.00024334541377879116,
+      "loss": 1.4102,
+      "step": 6339
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08646469129338254,
+      "learning_rate": 0.00024319600977713203,
+      "loss": 1.3129,
+      "step": 6340
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.07477706387936801,
+      "learning_rate": 0.00024304663691384205,
+      "loss": 1.3683,
+      "step": 6341
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.0742561087884464,
+      "learning_rate": 0.00024289729520703335,
+      "loss": 1.4076,
+      "step": 6342
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.0727554495538966,
+      "learning_rate": 0.00024274798467481396,
+      "loss": 1.391,
+      "step": 6343
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.09096214796504422,
+      "learning_rate": 0.00024259870533528815,
+      "loss": 1.3449,
+      "step": 6344
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.07170963499794096,
+      "learning_rate": 0.00024244945720655693,
+      "loss": 1.3732,
+      "step": 6345
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.07006967689524883,
+      "learning_rate": 0.0002423002403067167,
+      "loss": 1.5636,
+      "step": 6346
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.07720905952192304,
+      "learning_rate": 0.00024215105465386078,
+      "loss": 1.2022,
+      "step": 6347
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.07092099425334585,
+      "learning_rate": 0.0002420019002660786,
+      "loss": 1.3869,
+      "step": 6348
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.07731471823323678,
+      "learning_rate": 0.0002418527771614553,
+      "loss": 1.375,
+      "step": 6349
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.0670229260877233,
+      "learning_rate": 0.00024170368535807274,
+      "loss": 1.3853,
+      "step": 6350
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.07526592555668116,
+      "learning_rate": 0.00024155462487400898,
+      "loss": 1.4127,
+      "step": 6351
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.0824513831130822,
+      "learning_rate": 0.00024140559572733778,
+      "loss": 1.4122,
+      "step": 6352
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.06022553289796199,
+      "learning_rate": 0.0002412565979361298,
+      "loss": 1.4234,
+      "step": 6353
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.06856769172839336,
+      "learning_rate": 0.00024110763151845112,
+      "loss": 1.3135,
+      "step": 6354
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.07706925472267376,
+      "learning_rate": 0.00024095869649236491,
+      "loss": 1.295,
+      "step": 6355
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08128807781499314,
+      "learning_rate": 0.00024080979287592952,
+      "loss": 1.4854,
+      "step": 6356
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.07997548107576101,
+      "learning_rate": 0.0002406609206872004,
+      "loss": 1.4402,
+      "step": 6357
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.0812986548596618,
+      "learning_rate": 0.00024051207994422885,
+      "loss": 1.3249,
+      "step": 6358
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.07668872546913742,
+      "learning_rate": 0.00024036327066506203,
+      "loss": 1.4075,
+      "step": 6359
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.07576029951246802,
+      "learning_rate": 0.00024021449286774378,
+      "loss": 1.3767,
+      "step": 6360
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.0705989781571333,
+      "learning_rate": 0.00024006574657031406,
+      "loss": 1.4834,
+      "step": 6361
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08410816163192895,
+      "learning_rate": 0.00023991703179080848,
+      "loss": 1.3311,
+      "step": 6362
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.0744909882298907,
+      "learning_rate": 0.00023976834854725943,
+      "loss": 1.4475,
+      "step": 6363
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.07417287239926322,
+      "learning_rate": 0.00023961969685769568,
+      "loss": 1.4801,
+      "step": 6364
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08508771858692672,
+      "learning_rate": 0.00023947107674014096,
+      "loss": 1.3167,
+      "step": 6365
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08704235220539557,
+      "learning_rate": 0.00023932248821261637,
+      "loss": 1.4221,
+      "step": 6366
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.06998956050843806,
+      "learning_rate": 0.00023917393129313896,
+      "loss": 1.4133,
+      "step": 6367
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.07324472581915419,
+      "learning_rate": 0.00023902540599972144,
+      "loss": 1.378,
+      "step": 6368
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.07294277583373579,
+      "learning_rate": 0.00023887691235037313,
+      "loss": 1.4382,
+      "step": 6369
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08534875004191578,
+      "learning_rate": 0.0002387284503630996,
+      "loss": 1.4318,
+      "step": 6370
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.07220491552722837,
+      "learning_rate": 0.00023858002005590202,
+      "loss": 1.4859,
+      "step": 6371
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08203133231113481,
+      "learning_rate": 0.00023843162144677828,
+      "loss": 1.4108,
+      "step": 6372
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.07399688753489576,
+      "learning_rate": 0.00023828325455372236,
+      "loss": 1.4796,
+      "step": 6373
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08070783051161995,
+      "learning_rate": 0.00023813491939472393,
+      "loss": 1.344,
+      "step": 6374
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08277727679169404,
+      "learning_rate": 0.0002379866159877695,
+      "loss": 1.2994,
+      "step": 6375
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.07044262154113139,
+      "learning_rate": 0.00023783834435084117,
+      "loss": 1.3152,
+      "step": 6376
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.07933013616591182,
+      "learning_rate": 0.0002376901045019172,
+      "loss": 1.4583,
+      "step": 6377
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08308805766312702,
+      "learning_rate": 0.0002375418964589724,
+      "loss": 1.2326,
+      "step": 6378
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08249281840241653,
+      "learning_rate": 0.00023739372023997763,
+      "loss": 1.4402,
+      "step": 6379
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.09484665425518647,
+      "learning_rate": 0.00023724557586289946,
+      "loss": 1.1794,
+      "step": 6380
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.0739011722621357,
+      "learning_rate": 0.0002370974633457011,
+      "loss": 1.3369,
+      "step": 6381
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.07765413859608447,
+      "learning_rate": 0.00023694938270634188,
+      "loss": 1.3658,
+      "step": 6382
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08727919704600484,
+      "learning_rate": 0.0002368013339627766,
+      "loss": 1.3588,
+      "step": 6383
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.07282288372917742,
+      "learning_rate": 0.00023665331713295702,
+      "loss": 1.3357,
+      "step": 6384
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.07995318342105713,
+      "learning_rate": 0.00023650533223483085,
+      "loss": 1.4342,
+      "step": 6385
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.07350102385403665,
+      "learning_rate": 0.00023635737928634148,
+      "loss": 1.4507,
+      "step": 6386
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.07565174355512315,
+      "learning_rate": 0.0002362094583054286,
+      "loss": 1.4482,
+      "step": 6387
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.07992476149419737,
+      "learning_rate": 0.00023606156931002838,
+      "loss": 1.4335,
+      "step": 6388
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08573538817338362,
+      "learning_rate": 0.00023591371231807296,
+      "loss": 1.397,
+      "step": 6389
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08571710420595538,
+      "learning_rate": 0.00023576588734749022,
+      "loss": 1.4763,
+      "step": 6390
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.0731916157652849,
+      "learning_rate": 0.00023561809441620458,
+      "loss": 1.3818,
+      "step": 6391
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08771184646456208,
+      "learning_rate": 0.00023547033354213658,
+      "loss": 1.3579,
+      "step": 6392
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.077798234525272,
+      "learning_rate": 0.0002353226047432025,
+      "loss": 1.3273,
+      "step": 6393
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.07815607317243291,
+      "learning_rate": 0.00023517490803731507,
+      "loss": 1.4548,
+      "step": 6394
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08263717397214161,
+      "learning_rate": 0.00023502724344238318,
+      "loss": 1.4002,
+      "step": 6395
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08569366911481649,
+      "learning_rate": 0.00023487961097631155,
+      "loss": 1.3903,
+      "step": 6396
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.07121023055466971,
+      "learning_rate": 0.0002347320106570009,
+      "loss": 1.3606,
+      "step": 6397
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08475099857112675,
+      "learning_rate": 0.00023458444250234868,
+      "loss": 1.461,
+      "step": 6398
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08412237390244266,
+      "learning_rate": 0.00023443690653024763,
+      "loss": 1.3305,
+      "step": 6399
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08095037113122881,
+      "learning_rate": 0.0002342894027585872,
+      "loss": 1.2791,
+      "step": 6400
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.06950633340592942,
+      "learning_rate": 0.0002341419312052529,
+      "loss": 1.2907,
+      "step": 6401
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08264516966426638,
+      "learning_rate": 0.00023399449188812584,
+      "loss": 1.2311,
+      "step": 6402
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08734564167313208,
+      "learning_rate": 0.00023384708482508364,
+      "loss": 1.4628,
+      "step": 6403
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08726129525601546,
+      "learning_rate": 0.0002336997100340002,
+      "loss": 1.3995,
+      "step": 6404
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.07588282327103045,
+      "learning_rate": 0.00023355236753274484,
+      "loss": 1.4151,
+      "step": 6405
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08370642493906598,
+      "learning_rate": 0.00023340505733918365,
+      "loss": 1.4016,
+      "step": 6406
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08108362721094939,
+      "learning_rate": 0.0002332577794711783,
+      "loss": 1.598,
+      "step": 6407
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08135817340718471,
+      "learning_rate": 0.0002331105339465867,
+      "loss": 1.2881,
+      "step": 6408
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.09012388270227195,
+      "learning_rate": 0.00023296332078326292,
+      "loss": 1.3892,
+      "step": 6409
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.07793560241657278,
+      "learning_rate": 0.00023281613999905732,
+      "loss": 1.4816,
+      "step": 6410
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08867833354510514,
+      "learning_rate": 0.00023266899161181564,
+      "loss": 1.4143,
+      "step": 6411
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.07688612344754879,
+      "learning_rate": 0.00023252187563938044,
+      "loss": 1.4396,
+      "step": 6412
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.07695243857532634,
+      "learning_rate": 0.00023237479209959006,
+      "loss": 1.4831,
+      "step": 6413
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.10597533406334113,
+      "learning_rate": 0.0002322277410102786,
+      "loss": 1.3586,
+      "step": 6414
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08486673254253517,
+      "learning_rate": 0.00023208072238927675,
+      "loss": 1.4086,
+      "step": 6415
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.09352301771338838,
+      "learning_rate": 0.00023193373625441112,
+      "loss": 1.3706,
+      "step": 6416
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.09267212329917103,
+      "learning_rate": 0.0002317867826235041,
+      "loss": 1.4183,
+      "step": 6417
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08024382166748611,
+      "learning_rate": 0.00023163986151437423,
+      "loss": 1.4594,
+      "step": 6418
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.07312178808860083,
+      "learning_rate": 0.0002314929729448363,
+      "loss": 1.3003,
+      "step": 6419
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1003750783018416,
+      "learning_rate": 0.0002313461169327013,
+      "loss": 1.3099,
+      "step": 6420
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08298443226564493,
+      "learning_rate": 0.00023119929349577555,
+      "loss": 1.4246,
+      "step": 6421
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08834680875376852,
+      "learning_rate": 0.00023105250265186223,
+      "loss": 1.5498,
+      "step": 6422
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08112147320182733,
+      "learning_rate": 0.00023090574441876033,
+      "loss": 1.531,
+      "step": 6423
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08887189548723178,
+      "learning_rate": 0.00023075901881426447,
+      "loss": 1.4392,
+      "step": 6424
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.07775527371255965,
+      "learning_rate": 0.00023061232585616577,
+      "loss": 1.2903,
+      "step": 6425
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.07631985491448418,
+      "learning_rate": 0.00023046566556225145,
+      "loss": 1.366,
+      "step": 6426
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.07884367654719883,
+      "learning_rate": 0.0002303190379503043,
+      "loss": 1.4406,
+      "step": 6427
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.07647140083236327,
+      "learning_rate": 0.00023017244303810363,
+      "loss": 1.5203,
+      "step": 6428
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.0755226169127557,
+      "learning_rate": 0.0002300258808434245,
+      "loss": 1.4164,
+      "step": 6429
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.07622607302032058,
+      "learning_rate": 0.00022987935138403786,
+      "loss": 1.4841,
+      "step": 6430
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.0824646838031537,
+      "learning_rate": 0.00022973285467771116,
+      "loss": 1.5279,
+      "step": 6431
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.0886875422328421,
+      "learning_rate": 0.0002295863907422077,
+      "loss": 1.545,
+      "step": 6432
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08848464683625346,
+      "learning_rate": 0.00022943995959528652,
+      "loss": 1.3529,
+      "step": 6433
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08132183333693678,
+      "learning_rate": 0.00022929356125470297,
+      "loss": 1.36,
+      "step": 6434
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.0879109504827401,
+      "learning_rate": 0.0002291471957382085,
+      "loss": 1.4193,
+      "step": 6435
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.07435601021309238,
+      "learning_rate": 0.00022900086306355022,
+      "loss": 1.4843,
+      "step": 6436
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.07947785083610934,
+      "learning_rate": 0.00022885456324847153,
+      "loss": 1.4676,
+      "step": 6437
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08881617486248049,
+      "learning_rate": 0.00022870829631071204,
+      "loss": 1.385,
+      "step": 6438
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08314651513325921,
+      "learning_rate": 0.00022856206226800686,
+      "loss": 1.4345,
+      "step": 6439
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08039152643497237,
+      "learning_rate": 0.00022841586113808726,
+      "loss": 1.4642,
+      "step": 6440
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08254667264623641,
+      "learning_rate": 0.00022826969293868098,
+      "loss": 1.3568,
+      "step": 6441
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.0850751659769357,
+      "learning_rate": 0.00022812355768751102,
+      "loss": 1.4961,
+      "step": 6442
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.07569920317359677,
+      "learning_rate": 0.00022797745540229704,
+      "loss": 1.3281,
+      "step": 6443
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08527834837940627,
+      "learning_rate": 0.00022783138610075454,
+      "loss": 1.5263,
+      "step": 6444
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.09203558938704211,
+      "learning_rate": 0.00022768534980059464,
+      "loss": 1.3848,
+      "step": 6445
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08002778832140865,
+      "learning_rate": 0.00022753934651952484,
+      "loss": 1.4289,
+      "step": 6446
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.07691283557052742,
+      "learning_rate": 0.0002273933762752488,
+      "loss": 1.4839,
+      "step": 6447
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08555435524639886,
+      "learning_rate": 0.00022724743908546552,
+      "loss": 1.3288,
+      "step": 6448
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.07359411407844321,
+      "learning_rate": 0.00022710153496787074,
+      "loss": 1.3386,
+      "step": 6449
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08535186262318026,
+      "learning_rate": 0.0002269556639401555,
+      "loss": 1.3845,
+      "step": 6450
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08081806643440836,
+      "learning_rate": 0.00022680982602000748,
+      "loss": 1.3093,
+      "step": 6451
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08042330242595397,
+      "learning_rate": 0.00022666402122510976,
+      "loss": 1.2788,
+      "step": 6452
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.0770073548322996,
+      "learning_rate": 0.00022651824957314176,
+      "loss": 1.3039,
+      "step": 6453
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.07566086212385112,
+      "learning_rate": 0.00022637251108177902,
+      "loss": 1.4038,
+      "step": 6454
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.09810806066509348,
+      "learning_rate": 0.0002262268057686925,
+      "loss": 1.393,
+      "step": 6455
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.07640142270989021,
+      "learning_rate": 0.0002260811336515496,
+      "loss": 1.5287,
+      "step": 6456
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08288360938472895,
+      "learning_rate": 0.00022593549474801377,
+      "loss": 1.3465,
+      "step": 6457
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08416800605991567,
+      "learning_rate": 0.00022578988907574388,
+      "loss": 1.375,
+      "step": 6458
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08118995759959549,
+      "learning_rate": 0.00022564431665239544,
+      "loss": 1.4158,
+      "step": 6459
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.09327858003512501,
+      "learning_rate": 0.00022549877749561943,
+      "loss": 1.4409,
+      "step": 6460
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08104347386981663,
+      "learning_rate": 0.00022535327162306285,
+      "loss": 1.3011,
+      "step": 6461
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08079496316961529,
+      "learning_rate": 0.00022520779905236892,
+      "loss": 1.3472,
+      "step": 6462
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.09930812169108252,
+      "learning_rate": 0.00022506235980117697,
+      "loss": 1.3556,
+      "step": 6463
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.081827074933239,
+      "learning_rate": 0.00022491695388712147,
+      "loss": 1.4348,
+      "step": 6464
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08327321510981332,
+      "learning_rate": 0.0002247715813278337,
+      "loss": 1.2857,
+      "step": 6465
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.07800325727916337,
+      "learning_rate": 0.00022462624214094075,
+      "loss": 1.4488,
+      "step": 6466
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.09429196943285079,
+      "learning_rate": 0.00022448093634406507,
+      "loss": 1.483,
+      "step": 6467
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.09258873648328031,
+      "learning_rate": 0.00022433566395482573,
+      "loss": 1.3856,
+      "step": 6468
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08195780122541721,
+      "learning_rate": 0.0002241904249908377,
+      "loss": 1.288,
+      "step": 6469
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.07988131648227399,
+      "learning_rate": 0.0002240452194697115,
+      "loss": 1.4128,
+      "step": 6470
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.0894159323508057,
+      "learning_rate": 0.00022390004740905362,
+      "loss": 1.5256,
+      "step": 6471
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08880130190892369,
+      "learning_rate": 0.000223754908826467,
+      "loss": 1.2916,
+      "step": 6472
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.09106245882443116,
+      "learning_rate": 0.00022360980373954987,
+      "loss": 1.4184,
+      "step": 6473
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08358404758333761,
+      "learning_rate": 0.0002234647321658969,
+      "loss": 1.4161,
+      "step": 6474
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08990776461075271,
+      "learning_rate": 0.00022331969412309877,
+      "loss": 1.3321,
+      "step": 6475
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08028040754728914,
+      "learning_rate": 0.00022317468962874132,
+      "loss": 1.3376,
+      "step": 6476
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.09033607654277113,
+      "learning_rate": 0.00022302971870040718,
+      "loss": 1.4686,
+      "step": 6477
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.09633659162771828,
+      "learning_rate": 0.00022288478135567465,
+      "loss": 1.4402,
+      "step": 6478
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.07054694812027362,
+      "learning_rate": 0.00022273987761211756,
+      "loss": 1.4753,
+      "step": 6479
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.07971226310945649,
+      "learning_rate": 0.00022259500748730637,
+      "loss": 1.4936,
+      "step": 6480
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08270479882148428,
+      "learning_rate": 0.00022245017099880665,
+      "loss": 1.3579,
+      "step": 6481
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.0943209601689459,
+      "learning_rate": 0.0002223053681641808,
+      "loss": 1.4191,
+      "step": 6482
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08834209518549241,
+      "learning_rate": 0.00022216059900098624,
+      "loss": 1.445,
+      "step": 6483
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08388454303429033,
+      "learning_rate": 0.00022201586352677688,
+      "loss": 1.4445,
+      "step": 6484
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.07909092713902374,
+      "learning_rate": 0.0002218711617591026,
+      "loss": 1.3958,
+      "step": 6485
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.06978073299876707,
+      "learning_rate": 0.00022172649371550863,
+      "loss": 1.4058,
+      "step": 6486
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08679179635542227,
+      "learning_rate": 0.0002215818594135367,
+      "loss": 1.3989,
+      "step": 6487
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.07993985077411585,
+      "learning_rate": 0.0002214372588707243,
+      "loss": 1.4374,
+      "step": 6488
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.07427798279179676,
+      "learning_rate": 0.00022129269210460444,
+      "loss": 1.4993,
+      "step": 6489
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08155449054888644,
+      "learning_rate": 0.00022114815913270653,
+      "loss": 1.4904,
+      "step": 6490
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.07970364164040267,
+      "learning_rate": 0.00022100365997255601,
+      "loss": 1.2852,
+      "step": 6491
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08682486023092911,
+      "learning_rate": 0.00022085919464167326,
+      "loss": 1.6039,
+      "step": 6492
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.07331684440190417,
+      "learning_rate": 0.00022071476315757544,
+      "loss": 1.3076,
+      "step": 6493
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.0757239710866081,
+      "learning_rate": 0.00022057036553777565,
+      "loss": 1.3685,
+      "step": 6494
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.07585935488051394,
+      "learning_rate": 0.00022042600179978216,
+      "loss": 1.4274,
+      "step": 6495
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08444732219970214,
+      "learning_rate": 0.00022028167196109981,
+      "loss": 1.4211,
+      "step": 6496
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.07518887569695178,
+      "learning_rate": 0.0002201373760392293,
+      "loss": 1.3419,
+      "step": 6497
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.07301302504211518,
+      "learning_rate": 0.0002199931140516665,
+      "loss": 1.5196,
+      "step": 6498
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.07598122555463341,
+      "learning_rate": 0.00021984888601590404,
+      "loss": 1.4913,
+      "step": 6499
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.07442679390168067,
+      "learning_rate": 0.0002197046919494301,
+      "loss": 1.4397,
+      "step": 6500
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.10135433982914047,
+      "learning_rate": 0.0002195605318697284,
+      "loss": 1.3109,
+      "step": 6501
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.07208256394451946,
+      "learning_rate": 0.00021941640579427928,
+      "loss": 1.4048,
+      "step": 6502
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.0787725168679611,
+      "learning_rate": 0.00021927231374055823,
+      "loss": 1.4708,
+      "step": 6503
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08301111523284674,
+      "learning_rate": 0.00021912825572603678,
+      "loss": 1.2329,
+      "step": 6504
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08558231206577362,
+      "learning_rate": 0.00021898423176818266,
+      "loss": 1.5087,
+      "step": 6505
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08352462729960797,
+      "learning_rate": 0.0002188402418844594,
+      "loss": 1.3946,
+      "step": 6506
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.07363711061252484,
+      "learning_rate": 0.00021869628609232596,
+      "loss": 1.3682,
+      "step": 6507
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.0805971289346462,
+      "learning_rate": 0.0002185523644092376,
+      "loss": 1.3311,
+      "step": 6508
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.07032069350557803,
+      "learning_rate": 0.00021840847685264555,
+      "loss": 1.3778,
+      "step": 6509
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.0744760329733795,
+      "learning_rate": 0.00021826462343999627,
+      "loss": 1.4149,
+      "step": 6510
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.09067490689524893,
+      "learning_rate": 0.0002181208041887327,
+      "loss": 1.442,
+      "step": 6511
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.0837501444944338,
+      "learning_rate": 0.0002179770191162936,
+      "loss": 1.4889,
+      "step": 6512
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08007588034484431,
+      "learning_rate": 0.00021783326824011324,
+      "loss": 1.4214,
+      "step": 6513
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.0799023651814438,
+      "learning_rate": 0.00021768955157762165,
+      "loss": 1.4904,
+      "step": 6514
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.09065512534636933,
+      "learning_rate": 0.00021754586914624524,
+      "loss": 1.4548,
+      "step": 6515
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08477293649906326,
+      "learning_rate": 0.0002174022209634061,
+      "loss": 1.4235,
+      "step": 6516
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.0779585839119935,
+      "learning_rate": 0.0002172586070465218,
+      "loss": 1.3522,
+      "step": 6517
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.07615724676838401,
+      "learning_rate": 0.0002171150274130061,
+      "loss": 1.4947,
+      "step": 6518
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.07558094107947531,
+      "learning_rate": 0.0002169714820802688,
+      "loss": 1.5059,
+      "step": 6519
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.09027457729867946,
+      "learning_rate": 0.0002168279710657149,
+      "loss": 1.4962,
+      "step": 6520
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.07694453291924619,
+      "learning_rate": 0.0002166844943867457,
+      "loss": 1.5161,
+      "step": 6521
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08320596409149107,
+      "learning_rate": 0.00021654105206075848,
+      "loss": 1.3651,
+      "step": 6522
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.09290290757932727,
+      "learning_rate": 0.0002163976441051459,
+      "loss": 1.3206,
+      "step": 6523
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.07738617365116893,
+      "learning_rate": 0.00021625427053729656,
+      "loss": 1.3863,
+      "step": 6524
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08379385652143403,
+      "learning_rate": 0.0002161109313745953,
+      "loss": 1.3935,
+      "step": 6525
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08494754800851972,
+      "learning_rate": 0.00021596762663442215,
+      "loss": 1.4045,
+      "step": 6526
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08370244101996513,
+      "learning_rate": 0.0002158243563341535,
+      "loss": 1.3563,
+      "step": 6527
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08251734877414564,
+      "learning_rate": 0.00021568112049116152,
+      "loss": 1.1581,
+      "step": 6528
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.07399029848646638,
+      "learning_rate": 0.00021553791912281369,
+      "loss": 1.2488,
+      "step": 6529
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08615102440250381,
+      "learning_rate": 0.00021539475224647382,
+      "loss": 1.3296,
+      "step": 6530
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08760288742801585,
+      "learning_rate": 0.00021525161987950164,
+      "loss": 1.4621,
+      "step": 6531
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08548065697127558,
+      "learning_rate": 0.00021510852203925206,
+      "loss": 1.4295,
+      "step": 6532
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08830861858417012,
+      "learning_rate": 0.0002149654587430765,
+      "loss": 1.3621,
+      "step": 6533
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.07901395402583786,
+      "learning_rate": 0.00021482243000832158,
+      "loss": 1.3396,
+      "step": 6534
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.07521402912545687,
+      "learning_rate": 0.00021467943585233036,
+      "loss": 1.4412,
+      "step": 6535
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.0757548604259532,
+      "learning_rate": 0.000214536476292441,
+      "loss": 1.35,
+      "step": 6536
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08152821401632664,
+      "learning_rate": 0.0002143935513459882,
+      "loss": 1.4062,
+      "step": 6537
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.09936944029433495,
+      "learning_rate": 0.0002142506610303017,
+      "loss": 1.4571,
+      "step": 6538
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.07181111096163625,
+      "learning_rate": 0.00021410780536270779,
+      "loss": 1.3778,
+      "step": 6539
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.07568757215933702,
+      "learning_rate": 0.00021396498436052826,
+      "loss": 1.3805,
+      "step": 6540
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08252191822278969,
+      "learning_rate": 0.0002138221980410802,
+      "loss": 1.4755,
+      "step": 6541
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.0965810583333691,
+      "learning_rate": 0.00021367944642167736,
+      "loss": 1.4406,
+      "step": 6542
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08218602214053825,
+      "learning_rate": 0.00021353672951962888,
+      "loss": 1.3727,
+      "step": 6543
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.07903607477438594,
+      "learning_rate": 0.0002133940473522395,
+      "loss": 1.493,
+      "step": 6544
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.09434836139967438,
+      "learning_rate": 0.00021325139993680982,
+      "loss": 1.489,
+      "step": 6545
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.09415498817850494,
+      "learning_rate": 0.00021310878729063643,
+      "loss": 1.4413,
+      "step": 6546
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.0837639102330669,
+      "learning_rate": 0.00021296620943101185,
+      "loss": 1.4946,
+      "step": 6547
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08977086282659386,
+      "learning_rate": 0.00021282366637522378,
+      "loss": 1.3019,
+      "step": 6548
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.0778087682911158,
+      "learning_rate": 0.00021268115814055617,
+      "loss": 1.333,
+      "step": 6549
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.0750368541827475,
+      "learning_rate": 0.0002125386847442889,
+      "loss": 1.5137,
+      "step": 6550
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.07990480483989432,
+      "learning_rate": 0.00021239624620369692,
+      "loss": 1.4183,
+      "step": 6551
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08691069242823043,
+      "learning_rate": 0.00021225384253605156,
+      "loss": 1.3846,
+      "step": 6552
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08922151650826089,
+      "learning_rate": 0.00021211147375862005,
+      "loss": 1.5065,
+      "step": 6553
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08532353831500455,
+      "learning_rate": 0.00021196913988866467,
+      "loss": 1.2723,
+      "step": 6554
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08102466561548363,
+      "learning_rate": 0.00021182684094344422,
+      "loss": 1.3856,
+      "step": 6555
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.07457371890788876,
+      "learning_rate": 0.00021168457694021282,
+      "loss": 1.3585,
+      "step": 6556
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.07968114957329238,
+      "learning_rate": 0.00021154234789622024,
+      "loss": 1.4162,
+      "step": 6557
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08118810789452972,
+      "learning_rate": 0.00021140015382871248,
+      "loss": 1.4492,
+      "step": 6558
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.0803685591148541,
+      "learning_rate": 0.00021125799475493114,
+      "loss": 1.3808,
+      "step": 6559
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08168756775809427,
+      "learning_rate": 0.00021111587069211324,
+      "loss": 1.306,
+      "step": 6560
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08662437531664363,
+      "learning_rate": 0.00021097378165749192,
+      "loss": 1.3639,
+      "step": 6561
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07911911522419161,
+      "learning_rate": 0.00021083172766829623,
+      "loss": 1.4021,
+      "step": 6562
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08142696269384453,
+      "learning_rate": 0.00021068970874175025,
+      "loss": 1.3587,
+      "step": 6563
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08014034479548356,
+      "learning_rate": 0.00021054772489507452,
+      "loss": 1.3174,
+      "step": 6564
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08382442401757927,
+      "learning_rate": 0.00021040577614548522,
+      "loss": 1.4604,
+      "step": 6565
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07901248508081173,
+      "learning_rate": 0.0002102638625101939,
+      "loss": 1.4022,
+      "step": 6566
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08222010138581641,
+      "learning_rate": 0.00021012198400640803,
+      "loss": 1.4083,
+      "step": 6567
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08063227244132427,
+      "learning_rate": 0.0002099801406513311,
+      "loss": 1.3365,
+      "step": 6568
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08005206480715796,
+      "learning_rate": 0.00020983833246216178,
+      "loss": 1.421,
+      "step": 6569
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.0734211309304561,
+      "learning_rate": 0.00020969655945609494,
+      "loss": 1.299,
+      "step": 6570
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07864468033355397,
+      "learning_rate": 0.00020955482165032136,
+      "loss": 1.4178,
+      "step": 6571
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07951495019839461,
+      "learning_rate": 0.0002094131190620267,
+      "loss": 1.2936,
+      "step": 6572
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.0797927878475767,
+      "learning_rate": 0.00020927145170839323,
+      "loss": 1.4509,
+      "step": 6573
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07822452330414903,
+      "learning_rate": 0.00020912981960659872,
+      "loss": 1.4728,
+      "step": 6574
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.09082590234638262,
+      "learning_rate": 0.00020898822277381612,
+      "loss": 1.57,
+      "step": 6575
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07487868864298572,
+      "learning_rate": 0.00020884666122721502,
+      "loss": 1.4102,
+      "step": 6576
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08487261642892924,
+      "learning_rate": 0.00020870513498395978,
+      "loss": 1.3731,
+      "step": 6577
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07665208224280295,
+      "learning_rate": 0.00020856364406121136,
+      "loss": 1.4139,
+      "step": 6578
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.0736129009710678,
+      "learning_rate": 0.0002084221884761257,
+      "loss": 1.4054,
+      "step": 6579
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08456473210977278,
+      "learning_rate": 0.00020828076824585485,
+      "loss": 1.2447,
+      "step": 6580
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07277321479264692,
+      "learning_rate": 0.0002081393833875468,
+      "loss": 1.5035,
+      "step": 6581
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08135520280681077,
+      "learning_rate": 0.00020799803391834443,
+      "loss": 1.3139,
+      "step": 6582
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08097736134436409,
+      "learning_rate": 0.00020785671985538724,
+      "loss": 1.3363,
+      "step": 6583
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08425630719100732,
+      "learning_rate": 0.00020771544121581003,
+      "loss": 1.4049,
+      "step": 6584
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08781705002009797,
+      "learning_rate": 0.0002075741980167431,
+      "loss": 1.4498,
+      "step": 6585
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07521461031735634,
+      "learning_rate": 0.000207432990275313,
+      "loss": 1.4064,
+      "step": 6586
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08087613506815886,
+      "learning_rate": 0.00020729181800864145,
+      "loss": 1.4177,
+      "step": 6587
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08082117128976071,
+      "learning_rate": 0.00020715068123384588,
+      "loss": 1.4106,
+      "step": 6588
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.0746232658644306,
+      "learning_rate": 0.00020700957996803982,
+      "loss": 1.4639,
+      "step": 6589
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07719550175977072,
+      "learning_rate": 0.00020686851422833247,
+      "loss": 1.4654,
+      "step": 6590
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08352707823350032,
+      "learning_rate": 0.00020672748403182818,
+      "loss": 1.4165,
+      "step": 6591
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08250365019800493,
+      "learning_rate": 0.00020658648939562752,
+      "loss": 1.4926,
+      "step": 6592
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08343810025009954,
+      "learning_rate": 0.00020644553033682673,
+      "loss": 1.4649,
+      "step": 6593
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08558698583672085,
+      "learning_rate": 0.00020630460687251723,
+      "loss": 1.4754,
+      "step": 6594
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.0862044285372647,
+      "learning_rate": 0.0002061637190197867,
+      "loss": 1.4983,
+      "step": 6595
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08385121801130181,
+      "learning_rate": 0.0002060228667957184,
+      "loss": 1.4094,
+      "step": 6596
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08458614443447766,
+      "learning_rate": 0.000205882050217391,
+      "loss": 1.4314,
+      "step": 6597
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.0940062094037127,
+      "learning_rate": 0.00020574126930187882,
+      "loss": 1.3113,
+      "step": 6598
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08044626858810763,
+      "learning_rate": 0.00020560052406625235,
+      "loss": 1.2947,
+      "step": 6599
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.10262807774306441,
+      "learning_rate": 0.00020545981452757718,
+      "loss": 1.4088,
+      "step": 6600
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08041627362609009,
+      "learning_rate": 0.00020531914070291485,
+      "loss": 1.3947,
+      "step": 6601
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08017300590515987,
+      "learning_rate": 0.00020517850260932285,
+      "loss": 1.3487,
+      "step": 6602
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.09120867994933178,
+      "learning_rate": 0.00020503790026385365,
+      "loss": 1.3817,
+      "step": 6603
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08010760497374898,
+      "learning_rate": 0.00020489733368355588,
+      "loss": 1.4757,
+      "step": 6604
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07122969441559135,
+      "learning_rate": 0.00020475680288547398,
+      "loss": 1.2778,
+      "step": 6605
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08403794664807253,
+      "learning_rate": 0.00020461630788664743,
+      "loss": 1.5009,
+      "step": 6606
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08299400551352507,
+      "learning_rate": 0.0002044758487041119,
+      "loss": 1.3146,
+      "step": 6607
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07976714523904847,
+      "learning_rate": 0.0002043354253548987,
+      "loss": 1.4699,
+      "step": 6608
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08026151620792248,
+      "learning_rate": 0.00020419503785603445,
+      "loss": 1.3158,
+      "step": 6609
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08797413778999093,
+      "learning_rate": 0.00020405468622454155,
+      "loss": 1.3516,
+      "step": 6610
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07741596819135142,
+      "learning_rate": 0.00020391437047743817,
+      "loss": 1.4267,
+      "step": 6611
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.0867998747324431,
+      "learning_rate": 0.0002037740906317383,
+      "loss": 1.3826,
+      "step": 6612
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.09645908165108294,
+      "learning_rate": 0.000203633846704451,
+      "loss": 1.4439,
+      "step": 6613
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07676929477174242,
+      "learning_rate": 0.0002034936387125816,
+      "loss": 1.3473,
+      "step": 6614
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08455784449172907,
+      "learning_rate": 0.0002033534666731308,
+      "loss": 1.2772,
+      "step": 6615
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08107012147918617,
+      "learning_rate": 0.00020321333060309478,
+      "loss": 1.286,
+      "step": 6616
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08464678556549356,
+      "learning_rate": 0.00020307323051946553,
+      "loss": 1.4273,
+      "step": 6617
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08124259630730062,
+      "learning_rate": 0.0002029331664392311,
+      "loss": 1.2747,
+      "step": 6618
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.09432309785978557,
+      "learning_rate": 0.0002027931383793741,
+      "loss": 1.4427,
+      "step": 6619
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07536259226364693,
+      "learning_rate": 0.0002026531463568736,
+      "loss": 1.3773,
+      "step": 6620
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.10086324922011485,
+      "learning_rate": 0.00020251319038870448,
+      "loss": 1.5472,
+      "step": 6621
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.09653867084761454,
+      "learning_rate": 0.0002023732704918364,
+      "loss": 1.5599,
+      "step": 6622
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.084699031110335,
+      "learning_rate": 0.00020223338668323533,
+      "loss": 1.4765,
+      "step": 6623
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07554415347563273,
+      "learning_rate": 0.0002020935389798629,
+      "loss": 1.4258,
+      "step": 6624
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07818526536382409,
+      "learning_rate": 0.00020195372739867569,
+      "loss": 1.3709,
+      "step": 6625
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08850803610170471,
+      "learning_rate": 0.00020181395195662655,
+      "loss": 1.5076,
+      "step": 6626
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.0876713653675245,
+      "learning_rate": 0.00020167421267066392,
+      "loss": 1.3318,
+      "step": 6627
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08688217151281227,
+      "learning_rate": 0.0002015345095577314,
+      "loss": 1.4213,
+      "step": 6628
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08836116928923096,
+      "learning_rate": 0.00020139484263476866,
+      "loss": 1.2419,
+      "step": 6629
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.0895517491670006,
+      "learning_rate": 0.00020125521191871054,
+      "loss": 1.433,
+      "step": 6630
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08227120369012335,
+      "learning_rate": 0.00020111561742648803,
+      "loss": 1.4532,
+      "step": 6631
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08709107947772152,
+      "learning_rate": 0.0002009760591750272,
+      "loss": 1.4795,
+      "step": 6632
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08139579162806306,
+      "learning_rate": 0.00020083653718125027,
+      "loss": 1.4581,
+      "step": 6633
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08475824957944199,
+      "learning_rate": 0.0002006970514620744,
+      "loss": 1.4174,
+      "step": 6634
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08621342413318868,
+      "learning_rate": 0.00020055760203441288,
+      "loss": 1.4104,
+      "step": 6635
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08288494862554391,
+      "learning_rate": 0.0002004181889151746,
+      "loss": 1.3434,
+      "step": 6636
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08305143444288314,
+      "learning_rate": 0.0002002788121212636,
+      "loss": 1.3572,
+      "step": 6637
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07601786290155091,
+      "learning_rate": 0.00020013947166957992,
+      "loss": 1.3279,
+      "step": 6638
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07653362087750513,
+      "learning_rate": 0.00020000016757701922,
+      "loss": 1.4184,
+      "step": 6639
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07932924099857178,
+      "learning_rate": 0.00019986089986047246,
+      "loss": 1.4887,
+      "step": 6640
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07901853454265233,
+      "learning_rate": 0.00019972166853682617,
+      "loss": 1.4399,
+      "step": 6641
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.06553771047838028,
+      "learning_rate": 0.00019958247362296278,
+      "loss": 1.3047,
+      "step": 6642
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07917463053808547,
+      "learning_rate": 0.00019944331513576036,
+      "loss": 1.5311,
+      "step": 6643
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.0745452152182659,
+      "learning_rate": 0.00019930419309209196,
+      "loss": 1.4733,
+      "step": 6644
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07259087329080621,
+      "learning_rate": 0.00019916510750882683,
+      "loss": 1.3976,
+      "step": 6645
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08174348493303095,
+      "learning_rate": 0.00019902605840282968,
+      "loss": 1.3222,
+      "step": 6646
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07719419124593137,
+      "learning_rate": 0.0001988870457909604,
+      "loss": 1.4798,
+      "step": 6647
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.09086005817838735,
+      "learning_rate": 0.00019874806969007492,
+      "loss": 1.4071,
+      "step": 6648
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07637012035806491,
+      "learning_rate": 0.00019860913011702475,
+      "loss": 1.4735,
+      "step": 6649
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08008763937136829,
+      "learning_rate": 0.0001984702270886566,
+      "loss": 1.4013,
+      "step": 6650
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07633349271039644,
+      "learning_rate": 0.0001983313606218128,
+      "loss": 1.4385,
+      "step": 6651
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.0922809276923969,
+      "learning_rate": 0.00019819253073333166,
+      "loss": 1.5445,
+      "step": 6652
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08152360957588603,
+      "learning_rate": 0.0001980537374400465,
+      "loss": 1.3338,
+      "step": 6653
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.0858820904568059,
+      "learning_rate": 0.00019791498075878662,
+      "loss": 1.2776,
+      "step": 6654
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07880393375507283,
+      "learning_rate": 0.000197776260706377,
+      "loss": 1.3785,
+      "step": 6655
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07789571614399886,
+      "learning_rate": 0.00019763757729963745,
+      "loss": 1.2272,
+      "step": 6656
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08571963301433981,
+      "learning_rate": 0.0001974989305553841,
+      "loss": 1.3977,
+      "step": 6657
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08812834195849592,
+      "learning_rate": 0.00019736032049042846,
+      "loss": 1.3836,
+      "step": 6658
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08479265010733626,
+      "learning_rate": 0.00019722174712157714,
+      "loss": 1.4593,
+      "step": 6659
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08609566303687616,
+      "learning_rate": 0.00019708321046563298,
+      "loss": 1.2265,
+      "step": 6660
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.0765427963656903,
+      "learning_rate": 0.0001969447105393937,
+      "loss": 1.3816,
+      "step": 6661
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07992363192285609,
+      "learning_rate": 0.00019680624735965324,
+      "loss": 1.3869,
+      "step": 6662
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08216831699566847,
+      "learning_rate": 0.00019666782094320042,
+      "loss": 1.3774,
+      "step": 6663
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08439452976795346,
+      "learning_rate": 0.00019652943130682015,
+      "loss": 1.385,
+      "step": 6664
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08737819487197597,
+      "learning_rate": 0.00019639107846729242,
+      "loss": 1.5652,
+      "step": 6665
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08022829112347864,
+      "learning_rate": 0.00019625276244139317,
+      "loss": 1.3677,
+      "step": 6666
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.09083698232520176,
+      "learning_rate": 0.00019611448324589377,
+      "loss": 1.3274,
+      "step": 6667
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.09574330590908853,
+      "learning_rate": 0.00019597624089756076,
+      "loss": 1.3743,
+      "step": 6668
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07621982941746845,
+      "learning_rate": 0.0001958380354131567,
+      "loss": 1.4548,
+      "step": 6669
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08335567959546465,
+      "learning_rate": 0.00019569986680943957,
+      "loss": 1.381,
+      "step": 6670
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08064847869524633,
+      "learning_rate": 0.00019556173510316271,
+      "loss": 1.4698,
+      "step": 6671
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08671777975446042,
+      "learning_rate": 0.00019542364031107485,
+      "loss": 1.3249,
+      "step": 6672
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.0753294442038951,
+      "learning_rate": 0.0001952855824499206,
+      "loss": 1.3026,
+      "step": 6673
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.0810641059229283,
+      "learning_rate": 0.00019514756153644027,
+      "loss": 1.4282,
+      "step": 6674
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08289829139487187,
+      "learning_rate": 0.0001950095775873688,
+      "loss": 1.4603,
+      "step": 6675
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.09110856841308793,
+      "learning_rate": 0.00019487163061943758,
+      "loss": 1.3966,
+      "step": 6676
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08034250256655166,
+      "learning_rate": 0.00019473372064937323,
+      "loss": 1.4295,
+      "step": 6677
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08049880862150381,
+      "learning_rate": 0.0001945958476938975,
+      "loss": 1.426,
+      "step": 6678
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08222994978626211,
+      "learning_rate": 0.0001944580117697281,
+      "loss": 1.3235,
+      "step": 6679
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.09559967814473865,
+      "learning_rate": 0.00019432021289357833,
+      "loss": 1.4478,
+      "step": 6680
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08036374699327789,
+      "learning_rate": 0.00019418245108215637,
+      "loss": 1.422,
+      "step": 6681
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08060853174720173,
+      "learning_rate": 0.00019404472635216674,
+      "loss": 1.4803,
+      "step": 6682
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08451954416817149,
+      "learning_rate": 0.00019390703872030886,
+      "loss": 1.3489,
+      "step": 6683
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.0786560087602854,
+      "learning_rate": 0.0001937693882032776,
+      "loss": 1.4111,
+      "step": 6684
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.09707523238495702,
+      "learning_rate": 0.00019363177481776373,
+      "loss": 1.5404,
+      "step": 6685
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08754859574303579,
+      "learning_rate": 0.0001934941985804536,
+      "loss": 1.3307,
+      "step": 6686
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07550925047472697,
+      "learning_rate": 0.0001933566595080284,
+      "loss": 1.3903,
+      "step": 6687
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.09006506803931079,
+      "learning_rate": 0.00019321915761716534,
+      "loss": 1.5876,
+      "step": 6688
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08481310995440604,
+      "learning_rate": 0.00019308169292453725,
+      "loss": 1.4452,
+      "step": 6689
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08923095224891851,
+      "learning_rate": 0.00019294426544681182,
+      "loss": 1.384,
+      "step": 6690
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07069536368701973,
+      "learning_rate": 0.00019280687520065282,
+      "loss": 1.3548,
+      "step": 6691
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07602239730273473,
+      "learning_rate": 0.00019266952220271937,
+      "loss": 1.3895,
+      "step": 6692
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07398666072791947,
+      "learning_rate": 0.00019253220646966597,
+      "loss": 1.2796,
+      "step": 6693
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08773848956447536,
+      "learning_rate": 0.0001923949280181423,
+      "loss": 1.4594,
+      "step": 6694
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07650409822997449,
+      "learning_rate": 0.00019225768686479428,
+      "loss": 1.2462,
+      "step": 6695
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07586459619579579,
+      "learning_rate": 0.0001921204830262625,
+      "loss": 1.4254,
+      "step": 6696
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08147345080849903,
+      "learning_rate": 0.0001919833165191836,
+      "loss": 1.4551,
+      "step": 6697
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07772700515033043,
+      "learning_rate": 0.0001918461873601896,
+      "loss": 1.2775,
+      "step": 6698
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07878468764229728,
+      "learning_rate": 0.0001917090955659076,
+      "loss": 1.4245,
+      "step": 6699
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.0856986791366906,
+      "learning_rate": 0.0001915720411529606,
+      "loss": 1.4059,
+      "step": 6700
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08232357102950424,
+      "learning_rate": 0.0001914350241379671,
+      "loss": 1.3602,
+      "step": 6701
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08037236283367581,
+      "learning_rate": 0.00019129804453754052,
+      "loss": 1.3657,
+      "step": 6702
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08306951753793287,
+      "learning_rate": 0.0001911611023682905,
+      "loss": 1.2233,
+      "step": 6703
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.0731465844221435,
+      "learning_rate": 0.00019102419764682133,
+      "loss": 1.4167,
+      "step": 6704
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07859228132072026,
+      "learning_rate": 0.0001908873303897336,
+      "loss": 1.3531,
+      "step": 6705
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07670616178704094,
+      "learning_rate": 0.00019075050061362252,
+      "loss": 1.4396,
+      "step": 6706
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07452515984590537,
+      "learning_rate": 0.00019061370833507946,
+      "loss": 1.4069,
+      "step": 6707
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.09091438656192163,
+      "learning_rate": 0.000190476953570691,
+      "loss": 1.357,
+      "step": 6708
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.0885594352489358,
+      "learning_rate": 0.00019034023633703883,
+      "loss": 1.4248,
+      "step": 6709
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07470022744007065,
+      "learning_rate": 0.0001902035566507006,
+      "loss": 1.4364,
+      "step": 6710
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07785651744361902,
+      "learning_rate": 0.00019006691452824932,
+      "loss": 1.4123,
+      "step": 6711
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08260984881365113,
+      "learning_rate": 0.00018993030998625294,
+      "loss": 1.4369,
+      "step": 6712
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08785121882037714,
+      "learning_rate": 0.00018979374304127567,
+      "loss": 1.4798,
+      "step": 6713
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08974738369160999,
+      "learning_rate": 0.00018965721370987649,
+      "loss": 1.3725,
+      "step": 6714
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07439597985191031,
+      "learning_rate": 0.00018952072200860987,
+      "loss": 1.3802,
+      "step": 6715
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07989917071160745,
+      "learning_rate": 0.00018938426795402614,
+      "loss": 1.4656,
+      "step": 6716
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07990885134710889,
+      "learning_rate": 0.00018924785156267088,
+      "loss": 1.3247,
+      "step": 6717
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.0920282425858599,
+      "learning_rate": 0.0001891114728510848,
+      "loss": 1.5226,
+      "step": 6718
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08098133988211124,
+      "learning_rate": 0.00018897513183580445,
+      "loss": 1.3289,
+      "step": 6719
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08236995237832075,
+      "learning_rate": 0.00018883882853336183,
+      "loss": 1.4082,
+      "step": 6720
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08109450367713297,
+      "learning_rate": 0.00018870256296028376,
+      "loss": 1.3274,
+      "step": 6721
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08854925614558622,
+      "learning_rate": 0.00018856633513309313,
+      "loss": 1.4633,
+      "step": 6722
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08776537527339491,
+      "learning_rate": 0.00018843014506830823,
+      "loss": 1.4247,
+      "step": 6723
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08403123751330256,
+      "learning_rate": 0.0001882939927824424,
+      "loss": 1.3254,
+      "step": 6724
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07484992936906583,
+      "learning_rate": 0.00018815787829200436,
+      "loss": 1.4866,
+      "step": 6725
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08493873152531263,
+      "learning_rate": 0.0001880218016134987,
+      "loss": 1.2764,
+      "step": 6726
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08016073948899083,
+      "learning_rate": 0.00018788576276342528,
+      "loss": 1.5368,
+      "step": 6727
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08021313491786318,
+      "learning_rate": 0.00018774976175827895,
+      "loss": 1.4752,
+      "step": 6728
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08468752473482426,
+      "learning_rate": 0.00018761379861455076,
+      "loss": 1.4151,
+      "step": 6729
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07521467300223297,
+      "learning_rate": 0.00018747787334872618,
+      "loss": 1.4668,
+      "step": 6730
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07685387563084291,
+      "learning_rate": 0.00018734198597728698,
+      "loss": 1.4331,
+      "step": 6731
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07505611286286995,
+      "learning_rate": 0.00018720613651670997,
+      "loss": 1.4281,
+      "step": 6732
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07502769080720458,
+      "learning_rate": 0.0001870703249834671,
+      "loss": 1.3482,
+      "step": 6733
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07203431096058605,
+      "learning_rate": 0.0001869345513940262,
+      "loss": 1.3121,
+      "step": 6734
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07909327996419319,
+      "learning_rate": 0.00018679881576485043,
+      "loss": 1.4207,
+      "step": 6735
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.09281906210695522,
+      "learning_rate": 0.00018666311811239795,
+      "loss": 1.3021,
+      "step": 6736
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.10364312066748778,
+      "learning_rate": 0.00018652745845312252,
+      "loss": 1.3919,
+      "step": 6737
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07723214415961016,
+      "learning_rate": 0.00018639183680347344,
+      "loss": 1.4411,
+      "step": 6738
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07938462403198333,
+      "learning_rate": 0.00018625625317989547,
+      "loss": 1.5324,
+      "step": 6739
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.0817394915072993,
+      "learning_rate": 0.00018612070759882826,
+      "loss": 1.4159,
+      "step": 6740
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07546117695110775,
+      "learning_rate": 0.00018598520007670743,
+      "loss": 1.4377,
+      "step": 6741
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07981149695058505,
+      "learning_rate": 0.00018584973062996379,
+      "loss": 1.3169,
+      "step": 6742
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.09914212403092261,
+      "learning_rate": 0.0001857142992750232,
+      "loss": 1.3235,
+      "step": 6743
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08197515988292878,
+      "learning_rate": 0.0001855789060283073,
+      "loss": 1.4592,
+      "step": 6744
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.10267448225101261,
+      "learning_rate": 0.00018544355090623337,
+      "loss": 1.3162,
+      "step": 6745
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07850174588448537,
+      "learning_rate": 0.00018530823392521302,
+      "loss": 1.3222,
+      "step": 6746
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08075899888732171,
+      "learning_rate": 0.00018517295510165417,
+      "loss": 1.4816,
+      "step": 6747
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08032012562669273,
+      "learning_rate": 0.00018503771445196006,
+      "loss": 1.3729,
+      "step": 6748
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07918002511033893,
+      "learning_rate": 0.00018490251199252878,
+      "loss": 1.267,
+      "step": 6749
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07957204447098715,
+      "learning_rate": 0.0001847673477397542,
+      "loss": 1.4327,
+      "step": 6750
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07659661414310863,
+      "learning_rate": 0.00018463222171002564,
+      "loss": 1.4241,
+      "step": 6751
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08910841479830099,
+      "learning_rate": 0.00018449713391972732,
+      "loss": 1.4116,
+      "step": 6752
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.0708037129156938,
+      "learning_rate": 0.00018436208438523915,
+      "loss": 1.3844,
+      "step": 6753
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.080526008014301,
+      "learning_rate": 0.00018422707312293663,
+      "loss": 1.4397,
+      "step": 6754
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.09423939921649865,
+      "learning_rate": 0.00018409210014918992,
+      "loss": 1.3331,
+      "step": 6755
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08920464006608342,
+      "learning_rate": 0.0001839571654803654,
+      "loss": 1.43,
+      "step": 6756
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08378333881364608,
+      "learning_rate": 0.0001838222691328239,
+      "loss": 1.4928,
+      "step": 6757
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07805182934549852,
+      "learning_rate": 0.00018368741112292252,
+      "loss": 1.3854,
+      "step": 6758
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.0854097806654915,
+      "learning_rate": 0.0001835525914670128,
+      "loss": 1.3806,
+      "step": 6759
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08040018273693553,
+      "learning_rate": 0.00018341781018144253,
+      "loss": 1.3825,
+      "step": 6760
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.0936813686907315,
+      "learning_rate": 0.00018328306728255405,
+      "loss": 1.5222,
+      "step": 6761
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07344171050358542,
+      "learning_rate": 0.00018314836278668557,
+      "loss": 1.3782,
+      "step": 6762
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07684534780565848,
+      "learning_rate": 0.00018301369671017058,
+      "loss": 1.4264,
+      "step": 6763
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07826953273917311,
+      "learning_rate": 0.00018287906906933754,
+      "loss": 1.3789,
+      "step": 6764
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08928722109970684,
+      "learning_rate": 0.00018274447988051064,
+      "loss": 1.3112,
+      "step": 6765
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08674179911378348,
+      "learning_rate": 0.00018260992916000946,
+      "loss": 1.4021,
+      "step": 6766
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08089963246558474,
+      "learning_rate": 0.00018247541692414858,
+      "loss": 1.489,
+      "step": 6767
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.09059576891894354,
+      "learning_rate": 0.00018234094318923794,
+      "loss": 1.2786,
+      "step": 6768
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.0848582732695057,
+      "learning_rate": 0.000182206507971583,
+      "loss": 1.3481,
+      "step": 6769
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07842148758676074,
+      "learning_rate": 0.0001820721112874848,
+      "loss": 1.3778,
+      "step": 6770
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08192393685903324,
+      "learning_rate": 0.000181937753153239,
+      "loss": 1.431,
+      "step": 6771
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08794094605478588,
+      "learning_rate": 0.00018180343358513713,
+      "loss": 1.5262,
+      "step": 6772
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07989683360465649,
+      "learning_rate": 0.00018166915259946616,
+      "loss": 1.4531,
+      "step": 6773
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07895430276687004,
+      "learning_rate": 0.00018153491021250762,
+      "loss": 1.3966,
+      "step": 6774
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07984445806504739,
+      "learning_rate": 0.0001814007064405392,
+      "loss": 1.3677,
+      "step": 6775
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08715921957027024,
+      "learning_rate": 0.00018126654129983367,
+      "loss": 1.3904,
+      "step": 6776
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.0923525775887198,
+      "learning_rate": 0.00018113241480665883,
+      "loss": 1.4055,
+      "step": 6777
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08056941965998873,
+      "learning_rate": 0.00018099832697727786,
+      "loss": 1.3418,
+      "step": 6778
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07920305560868664,
+      "learning_rate": 0.00018086427782794962,
+      "loss": 1.3279,
+      "step": 6779
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07615284040014661,
+      "learning_rate": 0.00018073026737492782,
+      "loss": 1.3265,
+      "step": 6780
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07508902500142603,
+      "learning_rate": 0.00018059629563446173,
+      "loss": 1.4519,
+      "step": 6781
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.09785702332279825,
+      "learning_rate": 0.00018046236262279615,
+      "loss": 1.4866,
+      "step": 6782
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.09289516193260337,
+      "learning_rate": 0.00018032846835617055,
+      "loss": 1.1696,
+      "step": 6783
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.09300543988745133,
+      "learning_rate": 0.00018019461285082023,
+      "loss": 1.4885,
+      "step": 6784
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.0857508580574828,
+      "learning_rate": 0.00018006079612297582,
+      "loss": 1.3232,
+      "step": 6785
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07557541283565075,
+      "learning_rate": 0.0001799270181888627,
+      "loss": 1.3176,
+      "step": 6786
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.0937229554586484,
+      "learning_rate": 0.00017979327906470204,
+      "loss": 1.4221,
+      "step": 6787
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08422026801156707,
+      "learning_rate": 0.00017965957876671046,
+      "loss": 1.3148,
+      "step": 6788
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.0870791699339773,
+      "learning_rate": 0.0001795259173110993,
+      "loss": 1.1682,
+      "step": 6789
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08035752064799632,
+      "learning_rate": 0.0001793922947140753,
+      "loss": 1.381,
+      "step": 6790
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07755396439121148,
+      "learning_rate": 0.0001792587109918411,
+      "loss": 1.3554,
+      "step": 6791
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07669787633054774,
+      "learning_rate": 0.00017912516616059378,
+      "loss": 1.347,
+      "step": 6792
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.0838443620302552,
+      "learning_rate": 0.00017899166023652624,
+      "loss": 1.3293,
+      "step": 6793
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08594325030118602,
+      "learning_rate": 0.0001788581932358268,
+      "loss": 1.5661,
+      "step": 6794
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.0741443300137945,
+      "learning_rate": 0.0001787247651746784,
+      "loss": 1.3332,
+      "step": 6795
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08201836586996483,
+      "learning_rate": 0.0001785913760692598,
+      "loss": 1.4351,
+      "step": 6796
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08571123279116978,
+      "learning_rate": 0.0001784580259357451,
+      "loss": 1.3921,
+      "step": 6797
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08420236015565297,
+      "learning_rate": 0.00017832471479030328,
+      "loss": 1.4287,
+      "step": 6798
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08673883035937369,
+      "learning_rate": 0.0001781914426490986,
+      "loss": 1.3989,
+      "step": 6799
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08328479946706482,
+      "learning_rate": 0.00017805820952829094,
+      "loss": 1.3287,
+      "step": 6800
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.0911264911348588,
+      "learning_rate": 0.00017792501544403546,
+      "loss": 1.2763,
+      "step": 6801
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07562969503527785,
+      "learning_rate": 0.00017779186041248202,
+      "loss": 1.3719,
+      "step": 6802
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08436317460320578,
+      "learning_rate": 0.00017765874444977637,
+      "loss": 1.5092,
+      "step": 6803
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08201803089533123,
+      "learning_rate": 0.00017752566757205934,
+      "loss": 1.3735,
+      "step": 6804
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08212245968024402,
+      "learning_rate": 0.0001773926297954667,
+      "loss": 1.437,
+      "step": 6805
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08936482521224416,
+      "learning_rate": 0.00017725963113612996,
+      "loss": 1.3278,
+      "step": 6806
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08005483146062434,
+      "learning_rate": 0.0001771266716101757,
+      "loss": 1.5125,
+      "step": 6807
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08323718244792362,
+      "learning_rate": 0.00017699375123372553,
+      "loss": 1.3158,
+      "step": 6808
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08374696265235664,
+      "learning_rate": 0.0001768608700228967,
+      "loss": 1.4074,
+      "step": 6809
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08846882203118381,
+      "learning_rate": 0.0001767280279938014,
+      "loss": 1.465,
+      "step": 6810
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08056026493132028,
+      "learning_rate": 0.00017659522516254707,
+      "loss": 1.3597,
+      "step": 6811
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.09657117745998345,
+      "learning_rate": 0.0001764624615452366,
+      "loss": 1.4254,
+      "step": 6812
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08434230471203434,
+      "learning_rate": 0.00017632973715796824,
+      "loss": 1.3593,
+      "step": 6813
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07279166739385243,
+      "learning_rate": 0.00017619705201683494,
+      "loss": 1.5091,
+      "step": 6814
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08449448457832491,
+      "learning_rate": 0.00017606440613792546,
+      "loss": 1.4328,
+      "step": 6815
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.0848450788789903,
+      "learning_rate": 0.00017593179953732363,
+      "loss": 1.4954,
+      "step": 6816
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08389551707213994,
+      "learning_rate": 0.00017579923223110815,
+      "loss": 1.5587,
+      "step": 6817
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08551329662455956,
+      "learning_rate": 0.00017566670423535346,
+      "loss": 1.4431,
+      "step": 6818
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.0845378310088562,
+      "learning_rate": 0.00017553421556612924,
+      "loss": 1.4607,
+      "step": 6819
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08248996628222881,
+      "learning_rate": 0.0001754017662395,
+      "loss": 1.35,
+      "step": 6820
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.09292398961034305,
+      "learning_rate": 0.00017526935627152542,
+      "loss": 1.354,
+      "step": 6821
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.09073753070746698,
+      "learning_rate": 0.00017513698567826096,
+      "loss": 1.3038,
+      "step": 6822
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.09641388034069218,
+      "learning_rate": 0.0001750046544757571,
+      "loss": 1.3099,
+      "step": 6823
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08378939811932765,
+      "learning_rate": 0.00017487236268005918,
+      "loss": 1.4058,
+      "step": 6824
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08011359281231487,
+      "learning_rate": 0.00017474011030720832,
+      "loss": 1.3095,
+      "step": 6825
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07936773141110776,
+      "learning_rate": 0.00017460789737324024,
+      "loss": 1.4565,
+      "step": 6826
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08898734259408628,
+      "learning_rate": 0.00017447572389418643,
+      "loss": 1.4878,
+      "step": 6827
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08102931349353243,
+      "learning_rate": 0.0001743435898860735,
+      "loss": 1.3488,
+      "step": 6828
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08001499042331985,
+      "learning_rate": 0.00017421149536492282,
+      "loss": 1.4098,
+      "step": 6829
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08157457664209204,
+      "learning_rate": 0.0001740794403467517,
+      "loss": 1.3322,
+      "step": 6830
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07722036111521809,
+      "learning_rate": 0.00017394742484757187,
+      "loss": 1.3963,
+      "step": 6831
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08352673258439874,
+      "learning_rate": 0.000173815448883391,
+      "loss": 1.2971,
+      "step": 6832
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07859349149790394,
+      "learning_rate": 0.00017368351247021136,
+      "loss": 1.4829,
+      "step": 6833
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07105629459606275,
+      "learning_rate": 0.00017355161562403076,
+      "loss": 1.304,
+      "step": 6834
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08610324427789169,
+      "learning_rate": 0.00017341975836084245,
+      "loss": 1.2876,
+      "step": 6835
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.0754237529009342,
+      "learning_rate": 0.0001732879406966341,
+      "loss": 1.3283,
+      "step": 6836
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07922581197977313,
+      "learning_rate": 0.0001731561626473893,
+      "loss": 1.3629,
+      "step": 6837
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08558328737606986,
+      "learning_rate": 0.00017302442422908676,
+      "loss": 1.5443,
+      "step": 6838
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07826006509172885,
+      "learning_rate": 0.00017289272545769986,
+      "loss": 1.3043,
+      "step": 6839
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.09022868404772783,
+      "learning_rate": 0.00017276106634919774,
+      "loss": 1.4351,
+      "step": 6840
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.09189036230143136,
+      "learning_rate": 0.0001726294469195448,
+      "loss": 1.2811,
+      "step": 6841
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08050200243446126,
+      "learning_rate": 0.00017249786718469967,
+      "loss": 1.5167,
+      "step": 6842
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07399654613968369,
+      "learning_rate": 0.00017236632716061728,
+      "loss": 1.4766,
+      "step": 6843
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.0806654913246944,
+      "learning_rate": 0.00017223482686324736,
+      "loss": 1.4564,
+      "step": 6844
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.09296418958185575,
+      "learning_rate": 0.0001721033663085345,
+      "loss": 1.4995,
+      "step": 6845
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08405262508188,
+      "learning_rate": 0.00017197194551241897,
+      "loss": 1.4281,
+      "step": 6846
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.09050133536256577,
+      "learning_rate": 0.00017184056449083603,
+      "loss": 1.452,
+      "step": 6847
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.10443701802852487,
+      "learning_rate": 0.00017170922325971584,
+      "loss": 1.5564,
+      "step": 6848
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08562641911936063,
+      "learning_rate": 0.00017157792183498412,
+      "loss": 1.2908,
+      "step": 6849
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07636600524567706,
+      "learning_rate": 0.00017144666023256178,
+      "loss": 1.3667,
+      "step": 6850
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08478261331322622,
+      "learning_rate": 0.00017131543846836457,
+      "loss": 1.3854,
+      "step": 6851
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08098833976518655,
+      "learning_rate": 0.00017118425655830344,
+      "loss": 1.5091,
+      "step": 6852
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08677585467877516,
+      "learning_rate": 0.0001710531145182848,
+      "loss": 1.5628,
+      "step": 6853
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07471384846368484,
+      "learning_rate": 0.0001709220123642103,
+      "loss": 1.4041,
+      "step": 6854
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.09541745514219427,
+      "learning_rate": 0.00017079095011197608,
+      "loss": 1.4971,
+      "step": 6855
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08623794035116948,
+      "learning_rate": 0.0001706599277774743,
+      "loss": 1.3316,
+      "step": 6856
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08054776192726237,
+      "learning_rate": 0.00017052894537659147,
+      "loss": 1.43,
+      "step": 6857
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07521154256399776,
+      "learning_rate": 0.00017039800292520995,
+      "loss": 1.3866,
+      "step": 6858
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.09193438771336371,
+      "learning_rate": 0.00017026710043920702,
+      "loss": 1.3083,
+      "step": 6859
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08977963139511945,
+      "learning_rate": 0.0001701362379344547,
+      "loss": 1.3806,
+      "step": 6860
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07698085395458001,
+      "learning_rate": 0.00017000541542682086,
+      "loss": 1.4348,
+      "step": 6861
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08198947779570069,
+      "learning_rate": 0.00016987463293216814,
+      "loss": 1.4138,
+      "step": 6862
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07855413193058092,
+      "learning_rate": 0.0001697438904663543,
+      "loss": 1.4153,
+      "step": 6863
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08950865101868916,
+      "learning_rate": 0.00016961318804523218,
+      "loss": 1.5416,
+      "step": 6864
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.0831439080639515,
+      "learning_rate": 0.00016948252568465,
+      "loss": 1.4049,
+      "step": 6865
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08513958963759437,
+      "learning_rate": 0.0001693519034004512,
+      "loss": 1.4206,
+      "step": 6866
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07840923965012238,
+      "learning_rate": 0.00016922132120847388,
+      "loss": 1.3569,
+      "step": 6867
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07532709899874017,
+      "learning_rate": 0.0001690907791245518,
+      "loss": 1.2623,
+      "step": 6868
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08043137997388435,
+      "learning_rate": 0.00016896027716451363,
+      "loss": 1.3917,
+      "step": 6869
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07859670126561226,
+      "learning_rate": 0.00016882981534418302,
+      "loss": 1.495,
+      "step": 6870
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08174506145645243,
+      "learning_rate": 0.00016869939367937896,
+      "loss": 1.447,
+      "step": 6871
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07206520273087716,
+      "learning_rate": 0.00016856901218591596,
+      "loss": 1.5492,
+      "step": 6872
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08019987669464876,
+      "learning_rate": 0.00016843867087960252,
+      "loss": 1.4361,
+      "step": 6873
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07809580316950983,
+      "learning_rate": 0.00016830836977624325,
+      "loss": 1.2727,
+      "step": 6874
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07459830181974891,
+      "learning_rate": 0.00016817810889163787,
+      "loss": 1.425,
+      "step": 6875
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08806181104637921,
+      "learning_rate": 0.00016804788824158057,
+      "loss": 1.5296,
+      "step": 6876
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07800614211582017,
+      "learning_rate": 0.00016791770784186128,
+      "loss": 1.5628,
+      "step": 6877
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08281656842314608,
+      "learning_rate": 0.00016778756770826486,
+      "loss": 1.3046,
+      "step": 6878
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08666547575815937,
+      "learning_rate": 0.00016765746785657104,
+      "loss": 1.4894,
+      "step": 6879
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07603211449722727,
+      "learning_rate": 0.00016752740830255504,
+      "loss": 1.4852,
+      "step": 6880
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08116015501105417,
+      "learning_rate": 0.0001673973890619871,
+      "loss": 1.5455,
+      "step": 6881
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08366331426875613,
+      "learning_rate": 0.00016726741015063223,
+      "loss": 1.4234,
+      "step": 6882
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.09604489598470241,
+      "learning_rate": 0.00016713747158425118,
+      "loss": 1.3777,
+      "step": 6883
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08394195627592127,
+      "learning_rate": 0.00016700757337859907,
+      "loss": 1.4061,
+      "step": 6884
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08751800146513225,
+      "learning_rate": 0.00016687771554942688,
+      "loss": 1.4041,
+      "step": 6885
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08575353248678774,
+      "learning_rate": 0.00016674789811247992,
+      "loss": 1.3437,
+      "step": 6886
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08504085235306681,
+      "learning_rate": 0.0001666181210834994,
+      "loss": 1.4633,
+      "step": 6887
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.09564690373909444,
+      "learning_rate": 0.00016648838447822084,
+      "loss": 1.4409,
+      "step": 6888
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08571014076959972,
+      "learning_rate": 0.0001663586883123755,
+      "loss": 1.3986,
+      "step": 6889
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07663223113335221,
+      "learning_rate": 0.00016622903260168955,
+      "loss": 1.4524,
+      "step": 6890
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07946817775922309,
+      "learning_rate": 0.00016609941736188394,
+      "loss": 1.3571,
+      "step": 6891
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08637460712979757,
+      "learning_rate": 0.00016596984260867516,
+      "loss": 1.4576,
+      "step": 6892
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07903291672994443,
+      "learning_rate": 0.00016584030835777464,
+      "loss": 1.4884,
+      "step": 6893
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07466860169220521,
+      "learning_rate": 0.00016571081462488874,
+      "loss": 1.5506,
+      "step": 6894
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.06936360532492965,
+      "learning_rate": 0.0001655813614257189,
+      "loss": 1.3753,
+      "step": 6895
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.09451161532387511,
+      "learning_rate": 0.00016545194877596193,
+      "loss": 1.5499,
+      "step": 6896
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08750817216023275,
+      "learning_rate": 0.00016532257669130967,
+      "loss": 1.4725,
+      "step": 6897
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08644183462372003,
+      "learning_rate": 0.00016519324518744866,
+      "loss": 1.4153,
+      "step": 6898
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07856152367805538,
+      "learning_rate": 0.00016506395428006088,
+      "loss": 1.4271,
+      "step": 6899
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08313932519241506,
+      "learning_rate": 0.00016493470398482352,
+      "loss": 1.4276,
+      "step": 6900
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07758330661107049,
+      "learning_rate": 0.00016480549431740831,
+      "loss": 1.4259,
+      "step": 6901
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07890737430673259,
+      "learning_rate": 0.00016467632529348247,
+      "loss": 1.5194,
+      "step": 6902
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.09159297058951059,
+      "learning_rate": 0.0001645471969287084,
+      "loss": 1.487,
+      "step": 6903
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.09770573554325404,
+      "learning_rate": 0.00016441810923874318,
+      "loss": 1.4068,
+      "step": 6904
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.09916837391473986,
+      "learning_rate": 0.00016428906223923902,
+      "loss": 1.3563,
+      "step": 6905
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08312448828582723,
+      "learning_rate": 0.00016416005594584355,
+      "loss": 1.4346,
+      "step": 6906
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.0777976180163894,
+      "learning_rate": 0.00016403109037419893,
+      "loss": 1.3848,
+      "step": 6907
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07890686457460733,
+      "learning_rate": 0.00016390216553994292,
+      "loss": 1.3671,
+      "step": 6908
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07973944756595773,
+      "learning_rate": 0.00016377328145870823,
+      "loss": 1.3963,
+      "step": 6909
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07806363228733787,
+      "learning_rate": 0.00016364443814612207,
+      "loss": 1.3893,
+      "step": 6910
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07209223731006845,
+      "learning_rate": 0.00016351563561780742,
+      "loss": 1.4204,
+      "step": 6911
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07909441403477985,
+      "learning_rate": 0.00016338687388938217,
+      "loss": 1.3775,
+      "step": 6912
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07972099827853533,
+      "learning_rate": 0.00016325815297645873,
+      "loss": 1.5274,
+      "step": 6913
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08052821554655319,
+      "learning_rate": 0.00016312947289464518,
+      "loss": 1.5595,
+      "step": 6914
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08023420829022035,
+      "learning_rate": 0.0001630008336595446,
+      "loss": 1.3303,
+      "step": 6915
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.09799506539759276,
+      "learning_rate": 0.00016287223528675476,
+      "loss": 1.2331,
+      "step": 6916
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08491064095612154,
+      "learning_rate": 0.00016274367779186844,
+      "loss": 1.3602,
+      "step": 6917
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08807116482937358,
+      "learning_rate": 0.00016261516119047393,
+      "loss": 1.4781,
+      "step": 6918
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.10870127860147914,
+      "learning_rate": 0.00016248668549815443,
+      "loss": 1.5202,
+      "step": 6919
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.10562154840968066,
+      "learning_rate": 0.0001623582507304877,
+      "loss": 1.5252,
+      "step": 6920
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08298775907519637,
+      "learning_rate": 0.0001622298569030473,
+      "loss": 1.4743,
+      "step": 6921
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08203093210794524,
+      "learning_rate": 0.000162101504031401,
+      "loss": 1.4706,
+      "step": 6922
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08380641907542187,
+      "learning_rate": 0.00016197319213111233,
+      "loss": 1.335,
+      "step": 6923
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08283572450475699,
+      "learning_rate": 0.0001618449212177396,
+      "loss": 1.3608,
+      "step": 6924
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.09123974004311679,
+      "learning_rate": 0.00016171669130683592,
+      "loss": 1.305,
+      "step": 6925
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.09077853862321274,
+      "learning_rate": 0.00016158850241394958,
+      "loss": 1.4741,
+      "step": 6926
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08266347964611022,
+      "learning_rate": 0.00016146035455462393,
+      "loss": 1.4007,
+      "step": 6927
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07728062995794181,
+      "learning_rate": 0.0001613322477443976,
+      "loss": 1.3946,
+      "step": 6928
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.09516214966689458,
+      "learning_rate": 0.00016120418199880367,
+      "loss": 1.4611,
+      "step": 6929
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08415646166248951,
+      "learning_rate": 0.0001610761573333706,
+      "loss": 1.3602,
+      "step": 6930
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08683899922320816,
+      "learning_rate": 0.00016094817376362215,
+      "loss": 1.4251,
+      "step": 6931
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08047309008486432,
+      "learning_rate": 0.00016082023130507627,
+      "loss": 1.3802,
+      "step": 6932
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08534330070542163,
+      "learning_rate": 0.0001606923299732468,
+      "loss": 1.347,
+      "step": 6933
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08466164032891992,
+      "learning_rate": 0.00016056446978364214,
+      "loss": 1.3549,
+      "step": 6934
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.10321861631811807,
+      "learning_rate": 0.00016043665075176562,
+      "loss": 1.4485,
+      "step": 6935
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.0927153279870653,
+      "learning_rate": 0.000160308872893116,
+      "loss": 1.3688,
+      "step": 6936
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.09696757316712741,
+      "learning_rate": 0.00016018113622318664,
+      "loss": 1.3439,
+      "step": 6937
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07852402945748814,
+      "learning_rate": 0.00016005344075746585,
+      "loss": 1.4648,
+      "step": 6938
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07922286942853338,
+      "learning_rate": 0.0001599257865114374,
+      "loss": 1.233,
+      "step": 6939
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.0913938351396083,
+      "learning_rate": 0.0001597981735005799,
+      "loss": 1.4028,
+      "step": 6940
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08728248486685027,
+      "learning_rate": 0.0001596706017403665,
+      "loss": 1.3228,
+      "step": 6941
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08034521864864075,
+      "learning_rate": 0.000159543071246266,
+      "loss": 1.3307,
+      "step": 6942
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08437406293251533,
+      "learning_rate": 0.00015941558203374197,
+      "loss": 1.3317,
+      "step": 6943
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07271689555275264,
+      "learning_rate": 0.00015928813411825266,
+      "loss": 1.388,
+      "step": 6944
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.09175529197602372,
+      "learning_rate": 0.00015916072751525167,
+      "loss": 1.5004,
+      "step": 6945
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08084840838382006,
+      "learning_rate": 0.0001590333622401877,
+      "loss": 1.3225,
+      "step": 6946
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.0839105076082594,
+      "learning_rate": 0.00015890603830850402,
+      "loss": 1.4301,
+      "step": 6947
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.10145040433206401,
+      "learning_rate": 0.000158778755735639,
+      "loss": 1.4455,
+      "step": 6948
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07785525394316473,
+      "learning_rate": 0.0001586515145370262,
+      "loss": 1.3215,
+      "step": 6949
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07590586417847282,
+      "learning_rate": 0.00015852431472809426,
+      "loss": 1.4979,
+      "step": 6950
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07980851983434818,
+      "learning_rate": 0.0001583971563242662,
+      "loss": 1.3844,
+      "step": 6951
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07922936787588304,
+      "learning_rate": 0.0001582700393409608,
+      "loss": 1.4856,
+      "step": 6952
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08281630274525655,
+      "learning_rate": 0.00015814296379359106,
+      "loss": 1.3173,
+      "step": 6953
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08092482775318731,
+      "learning_rate": 0.00015801592969756555,
+      "loss": 1.3588,
+      "step": 6954
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07710543321208346,
+      "learning_rate": 0.00015788893706828773,
+      "loss": 1.4724,
+      "step": 6955
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07537722490115532,
+      "learning_rate": 0.00015776198592115553,
+      "loss": 1.4065,
+      "step": 6956
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07354587379308565,
+      "learning_rate": 0.00015763507627156265,
+      "loss": 1.4229,
+      "step": 6957
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08239350486018063,
+      "learning_rate": 0.00015750820813489685,
+      "loss": 1.3505,
+      "step": 6958
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08386481585345223,
+      "learning_rate": 0.00015738138152654175,
+      "loss": 1.3849,
+      "step": 6959
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.09620737533851652,
+      "learning_rate": 0.00015725459646187518,
+      "loss": 1.3966,
+      "step": 6960
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08601988560556079,
+      "learning_rate": 0.00015712785295627035,
+      "loss": 1.4484,
+      "step": 6961
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07897105164507949,
+      "learning_rate": 0.0001570011510250956,
+      "loss": 1.4333,
+      "step": 6962
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07358390406887073,
+      "learning_rate": 0.00015687449068371367,
+      "loss": 1.429,
+      "step": 6963
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.0924152537075702,
+      "learning_rate": 0.00015674787194748264,
+      "loss": 1.3773,
+      "step": 6964
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08221188593201989,
+      "learning_rate": 0.00015662129483175568,
+      "loss": 1.3499,
+      "step": 6965
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08749139899956103,
+      "learning_rate": 0.00015649475935188033,
+      "loss": 1.4698,
+      "step": 6966
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08393889404422417,
+      "learning_rate": 0.00015636826552319972,
+      "loss": 1.4097,
+      "step": 6967
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07418914646855072,
+      "learning_rate": 0.00015624181336105187,
+      "loss": 1.2507,
+      "step": 6968
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08998017862926071,
+      "learning_rate": 0.0001561154028807689,
+      "loss": 1.4847,
+      "step": 6969
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07646907923162406,
+      "learning_rate": 0.00015598903409767896,
+      "loss": 1.3316,
+      "step": 6970
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.10029053480803479,
+      "learning_rate": 0.00015586270702710474,
+      "loss": 1.5359,
+      "step": 6971
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08308434884812765,
+      "learning_rate": 0.00015573642168436358,
+      "loss": 1.504,
+      "step": 6972
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.10198312095249795,
+      "learning_rate": 0.00015561017808476813,
+      "loss": 1.4175,
+      "step": 6973
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.0819281330153054,
+      "learning_rate": 0.00015548397624362605,
+      "loss": 1.371,
+      "step": 6974
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08491787970373423,
+      "learning_rate": 0.0001553578161762394,
+      "loss": 1.4177,
+      "step": 6975
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08401926419336586,
+      "learning_rate": 0.00015523169789790576,
+      "loss": 1.3111,
+      "step": 6976
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08999804666215713,
+      "learning_rate": 0.00015510562142391742,
+      "loss": 1.3743,
+      "step": 6977
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08523286268070919,
+      "learning_rate": 0.0001549795867695616,
+      "loss": 1.4312,
+      "step": 6978
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08185701371966983,
+      "learning_rate": 0.00015485359395012011,
+      "loss": 1.3798,
+      "step": 6979
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.0701710856775418,
+      "learning_rate": 0.00015472764298087027,
+      "loss": 1.4199,
+      "step": 6980
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08314221104473174,
+      "learning_rate": 0.00015460173387708427,
+      "loss": 1.3538,
+      "step": 6981
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08420556211767127,
+      "learning_rate": 0.00015447586665402857,
+      "loss": 1.391,
+      "step": 6982
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07528567612802567,
+      "learning_rate": 0.00015435004132696546,
+      "loss": 1.4486,
+      "step": 6983
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.09387052135404927,
+      "learning_rate": 0.0001542242579111513,
+      "loss": 1.3717,
+      "step": 6984
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07264079271114803,
+      "learning_rate": 0.0001540985164218379,
+      "loss": 1.4776,
+      "step": 6985
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.0687592679653029,
+      "learning_rate": 0.0001539728168742721,
+      "loss": 1.4902,
+      "step": 6986
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08720354161857613,
+      "learning_rate": 0.00015384715928369502,
+      "loss": 1.45,
+      "step": 6987
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08356950973283157,
+      "learning_rate": 0.00015372154366534324,
+      "loss": 1.4589,
+      "step": 6988
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08640405818330486,
+      "learning_rate": 0.00015359597003444824,
+      "loss": 1.4614,
+      "step": 6989
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07591193807593741,
+      "learning_rate": 0.00015347043840623615,
+      "loss": 1.3445,
+      "step": 6990
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07509872422742482,
+      "learning_rate": 0.00015334494879592787,
+      "loss": 1.5149,
+      "step": 6991
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08250599953346692,
+      "learning_rate": 0.00015321950121873967,
+      "loss": 1.3254,
+      "step": 6992
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07975103285921857,
+      "learning_rate": 0.00015309409568988263,
+      "loss": 1.3806,
+      "step": 6993
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08522385169066489,
+      "learning_rate": 0.0001529687322245623,
+      "loss": 1.3766,
+      "step": 6994
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08078623532738036,
+      "learning_rate": 0.0001528434108379796,
+      "loss": 1.3841,
+      "step": 6995
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.09287538197642288,
+      "learning_rate": 0.00015271813154533033,
+      "loss": 1.397,
+      "step": 6996
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07227532782369564,
+      "learning_rate": 0.00015259289436180467,
+      "loss": 1.3578,
+      "step": 6997
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08275454676238209,
+      "learning_rate": 0.0001524676993025883,
+      "loss": 1.4225,
+      "step": 6998
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.0787200646752553,
+      "learning_rate": 0.00015234254638286183,
+      "loss": 1.433,
+      "step": 6999
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08559670118586739,
+      "learning_rate": 0.00015221743561779987,
+      "loss": 1.4289,
+      "step": 7000
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08685366803895994,
+      "learning_rate": 0.00015209236702257278,
+      "loss": 1.4789,
+      "step": 7001
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08398812238516672,
+      "learning_rate": 0.0001519673406123458,
+      "loss": 1.3169,
+      "step": 7002
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07951981075675485,
+      "learning_rate": 0.00015184235640227845,
+      "loss": 1.46,
+      "step": 7003
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07797865721296189,
+      "learning_rate": 0.00015171741440752568,
+      "loss": 1.5121,
+      "step": 7004
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08033960308925113,
+      "learning_rate": 0.00015159251464323732,
+      "loss": 1.4325,
+      "step": 7005
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07822679950394763,
+      "learning_rate": 0.00015146765712455745,
+      "loss": 1.4887,
+      "step": 7006
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07221658273320822,
+      "learning_rate": 0.00015134284186662584,
+      "loss": 1.5121,
+      "step": 7007
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.0851519454515554,
+      "learning_rate": 0.00015121806888457673,
+      "loss": 1.4916,
+      "step": 7008
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07553393733953745,
+      "learning_rate": 0.00015109333819353905,
+      "loss": 1.3452,
+      "step": 7009
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08710936906626386,
+      "learning_rate": 0.00015096864980863718,
+      "loss": 1.3347,
+      "step": 7010
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08725477793257659,
+      "learning_rate": 0.00015084400374498964,
+      "loss": 1.412,
+      "step": 7011
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08923549232165805,
+      "learning_rate": 0.0001507194000177105,
+      "loss": 1.2951,
+      "step": 7012
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.0796244829660177,
+      "learning_rate": 0.00015059483864190816,
+      "loss": 1.2852,
+      "step": 7013
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08487704673509294,
+      "learning_rate": 0.00015047031963268614,
+      "loss": 1.4351,
+      "step": 7014
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08299755706265552,
+      "learning_rate": 0.0001503458430051431,
+      "loss": 1.2572,
+      "step": 7015
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.09293429207949322,
+      "learning_rate": 0.00015022140877437185,
+      "loss": 1.4802,
+      "step": 7016
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07406152365690835,
+      "learning_rate": 0.0001500970169554608,
+      "loss": 1.3775,
+      "step": 7017
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08114305241044793,
+      "learning_rate": 0.00014997266756349264,
+      "loss": 1.3368,
+      "step": 7018
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07629530714726956,
+      "learning_rate": 0.00014984836061354524,
+      "loss": 1.4962,
+      "step": 7019
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08980752121203324,
+      "learning_rate": 0.00014972409612069138,
+      "loss": 1.3905,
+      "step": 7020
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08575581179496382,
+      "learning_rate": 0.00014959987409999853,
+      "loss": 1.3924,
+      "step": 7021
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08198613603708774,
+      "learning_rate": 0.00014947569456652876,
+      "loss": 1.339,
+      "step": 7022
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.09454705124999215,
+      "learning_rate": 0.00014935155753533947,
+      "loss": 1.4268,
+      "step": 7023
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.09143985910869612,
+      "learning_rate": 0.00014922746302148282,
+      "loss": 1.5904,
+      "step": 7024
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.08156496168263828,
+      "learning_rate": 0.00014910341104000546,
+      "loss": 1.4352,
+      "step": 7025
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.08359413915614876,
+      "learning_rate": 0.00014897940160594925,
+      "loss": 1.4386,
+      "step": 7026
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.08953163494531653,
+      "learning_rate": 0.00014885543473435088,
+      "loss": 1.4843,
+      "step": 7027
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07495695660289363,
+      "learning_rate": 0.00014873151044024146,
+      "loss": 1.2414,
+      "step": 7028
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.10177289279918775,
+      "learning_rate": 0.00014860762873864742,
+      "loss": 1.4096,
+      "step": 7029
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07368892484521711,
+      "learning_rate": 0.00014848378964458999,
+      "loss": 1.3566,
+      "step": 7030
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.08689439609994665,
+      "learning_rate": 0.0001483599931730849,
+      "loss": 1.4542,
+      "step": 7031
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.08443024292684892,
+      "learning_rate": 0.00014823623933914276,
+      "loss": 1.3873,
+      "step": 7032
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.08822916976981438,
+      "learning_rate": 0.00014811252815776955,
+      "loss": 1.5633,
+      "step": 7033
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.0887025471292702,
+      "learning_rate": 0.0001479888596439652,
+      "loss": 1.4117,
+      "step": 7034
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.08246689872933267,
+      "learning_rate": 0.0001478652338127252,
+      "loss": 1.4453,
+      "step": 7035
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07093731196022528,
+      "learning_rate": 0.00014774165067903982,
+      "loss": 1.3772,
+      "step": 7036
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.08363424689449472,
+      "learning_rate": 0.00014761811025789352,
+      "loss": 1.327,
+      "step": 7037
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07842351016801305,
+      "learning_rate": 0.00014749461256426615,
+      "loss": 1.3652,
+      "step": 7038
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.0824205239126465,
+      "learning_rate": 0.00014737115761313246,
+      "loss": 1.4238,
+      "step": 7039
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.08687263476544752,
+      "learning_rate": 0.00014724774541946146,
+      "loss": 1.3724,
+      "step": 7040
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.0920673100075393,
+      "learning_rate": 0.00014712437599821742,
+      "loss": 1.2506,
+      "step": 7041
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.09222049188447305,
+      "learning_rate": 0.00014700104936435953,
+      "loss": 1.4085,
+      "step": 7042
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07736454871275107,
+      "learning_rate": 0.00014687776553284137,
+      "loss": 1.4622,
+      "step": 7043
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.08255409541057177,
+      "learning_rate": 0.00014675452451861138,
+      "loss": 1.5405,
+      "step": 7044
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.08612416866292971,
+      "learning_rate": 0.00014663132633661313,
+      "loss": 1.417,
+      "step": 7045
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.09432241903711268,
+      "learning_rate": 0.00014650817100178492,
+      "loss": 1.4246,
+      "step": 7046
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.11601460726127812,
+      "learning_rate": 0.00014638505852905954,
+      "loss": 1.3404,
+      "step": 7047
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.0966656462751512,
+      "learning_rate": 0.00014626198893336506,
+      "loss": 1.461,
+      "step": 7048
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07949808600283965,
+      "learning_rate": 0.00014613896222962375,
+      "loss": 1.3932,
+      "step": 7049
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.09250961738605959,
+      "learning_rate": 0.00014601597843275327,
+      "loss": 1.4166,
+      "step": 7050
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07636058338566189,
+      "learning_rate": 0.00014589303755766587,
+      "loss": 1.4146,
+      "step": 7051
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.08105055959370543,
+      "learning_rate": 0.0001457701396192685,
+      "loss": 1.3308,
+      "step": 7052
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.08548635220894996,
+      "learning_rate": 0.00014564728463246275,
+      "loss": 1.3975,
+      "step": 7053
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07965255750151616,
+      "learning_rate": 0.00014552447261214534,
+      "loss": 1.4911,
+      "step": 7054
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.08537985574986948,
+      "learning_rate": 0.00014540170357320786,
+      "loss": 1.4963,
+      "step": 7055
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.09078007237282826,
+      "learning_rate": 0.0001452789775305362,
+      "loss": 1.4235,
+      "step": 7056
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.08402632342143757,
+      "learning_rate": 0.0001451562944990114,
+      "loss": 1.4207,
+      "step": 7057
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.08301706262546693,
+      "learning_rate": 0.00014503365449350936,
+      "loss": 1.2685,
+      "step": 7058
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.08203445439664891,
+      "learning_rate": 0.00014491105752890033,
+      "loss": 1.3933,
+      "step": 7059
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.09063623215892826,
+      "learning_rate": 0.00014478850362004974,
+      "loss": 1.486,
+      "step": 7060
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.09475208951182909,
+      "learning_rate": 0.00014466599278181787,
+      "loss": 1.4044,
+      "step": 7061
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.09177737333236773,
+      "learning_rate": 0.0001445435250290592,
+      "loss": 1.4223,
+      "step": 7062
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.09652109094250604,
+      "learning_rate": 0.00014442110037662375,
+      "loss": 1.4139,
+      "step": 7063
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.09218566946256986,
+      "learning_rate": 0.00014429871883935575,
+      "loss": 1.2943,
+      "step": 7064
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.08224360335861525,
+      "learning_rate": 0.0001441763804320942,
+      "loss": 1.4995,
+      "step": 7065
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.08621259301000005,
+      "learning_rate": 0.00014405408516967328,
+      "loss": 1.381,
+      "step": 7066
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.08733699795943922,
+      "learning_rate": 0.00014393183306692176,
+      "loss": 1.4354,
+      "step": 7067
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.0859822893466842,
+      "learning_rate": 0.00014380962413866288,
+      "loss": 1.4137,
+      "step": 7068
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.08013451149466999,
+      "learning_rate": 0.00014368745839971509,
+      "loss": 1.2975,
+      "step": 7069
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07670849214756494,
+      "learning_rate": 0.00014356533586489152,
+      "loss": 1.4086,
+      "step": 7070
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.08379391904018833,
+      "learning_rate": 0.00014344325654899964,
+      "loss": 1.5488,
+      "step": 7071
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.08446287225279143,
+      "learning_rate": 0.0001433212204668421,
+      "loss": 1.3633,
+      "step": 7072
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.08655495607730566,
+      "learning_rate": 0.00014319922763321642,
+      "loss": 1.4289,
+      "step": 7073
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.09076922400021534,
+      "learning_rate": 0.0001430772780629145,
+      "loss": 1.3839,
+      "step": 7074
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07944001534849933,
+      "learning_rate": 0.00014295537177072288,
+      "loss": 1.3809,
+      "step": 7075
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.09329551688513704,
+      "learning_rate": 0.00014283350877142343,
+      "loss": 1.3861,
+      "step": 7076
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07042640493257381,
+      "learning_rate": 0.00014271168907979248,
+      "loss": 1.3829,
+      "step": 7077
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07897438905434498,
+      "learning_rate": 0.00014258991271060085,
+      "loss": 1.4558,
+      "step": 7078
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.09126203160664402,
+      "learning_rate": 0.00014246817967861463,
+      "loss": 1.3902,
+      "step": 7079
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.09630423913710359,
+      "learning_rate": 0.00014234648999859412,
+      "loss": 1.3629,
+      "step": 7080
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.08349720636262284,
+      "learning_rate": 0.0001422248436852947,
+      "loss": 1.3694,
+      "step": 7081
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.08293949990701834,
+      "learning_rate": 0.00014210324075346654,
+      "loss": 1.3749,
+      "step": 7082
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07570143074146496,
+      "learning_rate": 0.00014198168121785416,
+      "loss": 1.4489,
+      "step": 7083
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.08376068026895099,
+      "learning_rate": 0.0001418601650931974,
+      "loss": 1.3698,
+      "step": 7084
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.08051201255768421,
+      "learning_rate": 0.0001417386923942301,
+      "loss": 1.3801,
+      "step": 7085
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.08361829214712761,
+      "learning_rate": 0.00014161726313568162,
+      "loss": 1.3365,
+      "step": 7086
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07539311701974652,
+      "learning_rate": 0.00014149587733227543,
+      "loss": 1.5533,
+      "step": 7087
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07636646791153252,
+      "learning_rate": 0.00014137453499873,
+      "loss": 1.3016,
+      "step": 7088
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07726830815322598,
+      "learning_rate": 0.00014125323614975878,
+      "loss": 1.4055,
+      "step": 7089
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07852283896570228,
+      "learning_rate": 0.00014113198080006927,
+      "loss": 1.3658,
+      "step": 7090
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.08455046614695923,
+      "learning_rate": 0.00014101076896436428,
+      "loss": 1.4775,
+      "step": 7091
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.0870258879015629,
+      "learning_rate": 0.00014088960065734136,
+      "loss": 1.4282,
+      "step": 7092
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.08600593545883693,
+      "learning_rate": 0.00014076847589369223,
+      "loss": 1.4818,
+      "step": 7093
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.0742503862653404,
+      "learning_rate": 0.00014064739468810388,
+      "loss": 1.4268,
+      "step": 7094
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.08703341486513806,
+      "learning_rate": 0.00014052635705525814,
+      "loss": 1.3743,
+      "step": 7095
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.0916770967788742,
+      "learning_rate": 0.00014040536300983052,
+      "loss": 1.4698,
+      "step": 7096
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.0784670687340215,
+      "learning_rate": 0.00014028441256649238,
+      "loss": 1.4073,
+      "step": 7097
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.08774542735944259,
+      "learning_rate": 0.00014016350573990948,
+      "loss": 1.4194,
+      "step": 7098
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.09906949969586148,
+      "learning_rate": 0.0001400426425447419,
+      "loss": 1.3304,
+      "step": 7099
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.09401233175045126,
+      "learning_rate": 0.00013992182299564493,
+      "loss": 1.4606,
+      "step": 7100
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07694095237545967,
+      "learning_rate": 0.00013980104710726844,
+      "loss": 1.4595,
+      "step": 7101
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1054928297352277,
+      "learning_rate": 0.00013968031489425658,
+      "loss": 1.3676,
+      "step": 7102
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.08347109928024152,
+      "learning_rate": 0.0001395596263712488,
+      "loss": 1.3819,
+      "step": 7103
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07712119034782984,
+      "learning_rate": 0.00013943898155287904,
+      "loss": 1.3646,
+      "step": 7104
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.08194542905625728,
+      "learning_rate": 0.00013931838045377586,
+      "loss": 1.4872,
+      "step": 7105
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.08432223869123434,
+      "learning_rate": 0.00013919782308856232,
+      "loss": 1.4337,
+      "step": 7106
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.09573328871170617,
+      "learning_rate": 0.00013907730947185665,
+      "loss": 1.4155,
+      "step": 7107
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.08869671452675124,
+      "learning_rate": 0.00013895683961827167,
+      "loss": 1.3511,
+      "step": 7108
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07279819542175914,
+      "learning_rate": 0.00013883641354241438,
+      "loss": 1.4694,
+      "step": 7109
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07300922537855549,
+      "learning_rate": 0.00013871603125888704,
+      "loss": 1.3933,
+      "step": 7110
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.0893723766297854,
+      "learning_rate": 0.00013859569278228668,
+      "loss": 1.2866,
+      "step": 7111
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07385807095451379,
+      "learning_rate": 0.00013847539812720435,
+      "loss": 1.3117,
+      "step": 7112
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.08154523440885406,
+      "learning_rate": 0.00013835514730822646,
+      "loss": 1.4402,
+      "step": 7113
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07631693608060588,
+      "learning_rate": 0.00013823494033993362,
+      "loss": 1.2876,
+      "step": 7114
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07501813680429896,
+      "learning_rate": 0.00013811477723690147,
+      "loss": 1.5077,
+      "step": 7115
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.09582028831990151,
+      "learning_rate": 0.0001379946580137003,
+      "loss": 1.1807,
+      "step": 7116
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.08195412649055485,
+      "learning_rate": 0.0001378745826848949,
+      "loss": 1.3206,
+      "step": 7117
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.08577541487377897,
+      "learning_rate": 0.00013775455126504465,
+      "loss": 1.4594,
+      "step": 7118
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.08293162734543053,
+      "learning_rate": 0.00013763456376870387,
+      "loss": 1.3694,
+      "step": 7119
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.08782077768957836,
+      "learning_rate": 0.00013751462021042166,
+      "loss": 1.5043,
+      "step": 7120
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07999103741806815,
+      "learning_rate": 0.0001373947206047413,
+      "loss": 1.292,
+      "step": 7121
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.08470840165153588,
+      "learning_rate": 0.00013727486496620112,
+      "loss": 1.376,
+      "step": 7122
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.08698455156845768,
+      "learning_rate": 0.00013715505330933427,
+      "loss": 1.4579,
+      "step": 7123
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.08819929465794887,
+      "learning_rate": 0.00013703528564866792,
+      "loss": 1.3595,
+      "step": 7124
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.0860304979498625,
+      "learning_rate": 0.0001369155619987245,
+      "loss": 1.3969,
+      "step": 7125
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.08037149086224887,
+      "learning_rate": 0.0001367958823740213,
+      "loss": 1.5538,
+      "step": 7126
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07933503079166238,
+      "learning_rate": 0.0001366762467890692,
+      "loss": 1.3866,
+      "step": 7127
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.0927014586146838,
+      "learning_rate": 0.00013655665525837474,
+      "loss": 1.2759,
+      "step": 7128
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.08431397953477916,
+      "learning_rate": 0.00013643710779643892,
+      "loss": 1.4357,
+      "step": 7129
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.09579874268870016,
+      "learning_rate": 0.00013631760441775703,
+      "loss": 1.4489,
+      "step": 7130
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.08347123456391652,
+      "learning_rate": 0.00013619814513681945,
+      "loss": 1.3502,
+      "step": 7131
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.08037094478894655,
+      "learning_rate": 0.0001360787299681111,
+      "loss": 1.4554,
+      "step": 7132
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.08713009992952041,
+      "learning_rate": 0.00013595935892611122,
+      "loss": 1.3501,
+      "step": 7133
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.079963043140441,
+      "learning_rate": 0.00013584003202529415,
+      "loss": 1.3216,
+      "step": 7134
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.08654644726903905,
+      "learning_rate": 0.00013572074928012878,
+      "loss": 1.3006,
+      "step": 7135
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.08684088272894674,
+      "learning_rate": 0.00013560151070507825,
+      "loss": 1.4253,
+      "step": 7136
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.08305476767947234,
+      "learning_rate": 0.00013548231631460095,
+      "loss": 1.3423,
+      "step": 7137
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07840906222655904,
+      "learning_rate": 0.00013536316612314936,
+      "loss": 1.316,
+      "step": 7138
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.08868368034508109,
+      "learning_rate": 0.00013524406014517115,
+      "loss": 1.4035,
+      "step": 7139
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.0825493183209432,
+      "learning_rate": 0.00013512499839510794,
+      "loss": 1.4327,
+      "step": 7140
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07142852180236557,
+      "learning_rate": 0.00013500598088739664,
+      "loss": 1.2882,
+      "step": 7141
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07839573055539671,
+      "learning_rate": 0.00013488700763646862,
+      "loss": 1.3449,
+      "step": 7142
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.0762490819632461,
+      "learning_rate": 0.0001347680786567495,
+      "loss": 1.446,
+      "step": 7143
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.08253038492998804,
+      "learning_rate": 0.00013464919396266017,
+      "loss": 1.3892,
+      "step": 7144
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.08080981648970774,
+      "learning_rate": 0.00013453035356861544,
+      "loss": 1.4137,
+      "step": 7145
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07895680489359269,
+      "learning_rate": 0.00013441155748902535,
+      "loss": 1.3648,
+      "step": 7146
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.09131009564719973,
+      "learning_rate": 0.00013429280573829438,
+      "loss": 1.3353,
+      "step": 7147
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.08148140496357519,
+      "learning_rate": 0.00013417409833082155,
+      "loss": 1.3986,
+      "step": 7148
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.09545610989391617,
+      "learning_rate": 0.0001340554352810003,
+      "loss": 1.5625,
+      "step": 7149
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.09783883668758428,
+      "learning_rate": 0.00013393681660321915,
+      "loss": 1.439,
+      "step": 7150
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07799550589752935,
+      "learning_rate": 0.00013381824231186113,
+      "loss": 1.4207,
+      "step": 7151
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.0795862542261532,
+      "learning_rate": 0.00013369971242130352,
+      "loss": 1.404,
+      "step": 7152
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.08548654694614052,
+      "learning_rate": 0.00013358122694591862,
+      "loss": 1.3907,
+      "step": 7153
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.08134899653490377,
+      "learning_rate": 0.00013346278590007334,
+      "loss": 1.3212,
+      "step": 7154
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.08104358016743386,
+      "learning_rate": 0.0001333443892981287,
+      "loss": 1.4092,
+      "step": 7155
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.08339197096891865,
+      "learning_rate": 0.00013322603715444097,
+      "loss": 1.4359,
+      "step": 7156
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07981284471191327,
+      "learning_rate": 0.00013310772948336085,
+      "loss": 1.512,
+      "step": 7157
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07761084452308815,
+      "learning_rate": 0.00013298946629923335,
+      "loss": 1.3671,
+      "step": 7158
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.0854348960861969,
+      "learning_rate": 0.00013287124761639823,
+      "loss": 1.494,
+      "step": 7159
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.09161357129893245,
+      "learning_rate": 0.0001327530734491902,
+      "loss": 1.4476,
+      "step": 7160
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.08312868545947096,
+      "learning_rate": 0.0001326349438119379,
+      "loss": 1.2806,
+      "step": 7161
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.09452434467284239,
+      "learning_rate": 0.0001325168587189652,
+      "loss": 1.5183,
+      "step": 7162
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.08587057449972695,
+      "learning_rate": 0.00013239881818459043,
+      "loss": 1.4184,
+      "step": 7163
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.08793194779558704,
+      "learning_rate": 0.0001322808222231261,
+      "loss": 1.386,
+      "step": 7164
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.08722167548904788,
+      "learning_rate": 0.00013216287084887984,
+      "loss": 1.3514,
+      "step": 7165
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.08853779388704727,
+      "learning_rate": 0.00013204496407615373,
+      "loss": 1.4281,
+      "step": 7166
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.09532120705919452,
+      "learning_rate": 0.00013192710191924412,
+      "loss": 1.4111,
+      "step": 7167
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.09657694916039453,
+      "learning_rate": 0.00013180928439244233,
+      "loss": 1.3998,
+      "step": 7168
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07677970663766127,
+      "learning_rate": 0.00013169151151003433,
+      "loss": 1.2778,
+      "step": 7169
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.09092227790364336,
+      "learning_rate": 0.00013157378328630025,
+      "loss": 1.4506,
+      "step": 7170
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.09163414959090961,
+      "learning_rate": 0.00013145609973551502,
+      "loss": 1.4888,
+      "step": 7171
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.089725966578204,
+      "learning_rate": 0.00013133846087194824,
+      "loss": 1.4996,
+      "step": 7172
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.08845121892261157,
+      "learning_rate": 0.0001312208667098642,
+      "loss": 1.4485,
+      "step": 7173
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.08463233673164806,
+      "learning_rate": 0.00013110331726352132,
+      "loss": 1.4001,
+      "step": 7174
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.08616388370330381,
+      "learning_rate": 0.00013098581254717312,
+      "loss": 1.4804,
+      "step": 7175
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.08580067427689526,
+      "learning_rate": 0.00013086835257506717,
+      "loss": 1.4569,
+      "step": 7176
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.08164815485831177,
+      "learning_rate": 0.00013075093736144612,
+      "loss": 1.4586,
+      "step": 7177
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.09324023988053676,
+      "learning_rate": 0.000130633566920547,
+      "loss": 1.4063,
+      "step": 7178
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.08436554667047103,
+      "learning_rate": 0.00013051624126660132,
+      "loss": 1.4858,
+      "step": 7179
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.08268273314661022,
+      "learning_rate": 0.00013039896041383504,
+      "loss": 1.4565,
+      "step": 7180
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.08402105295671775,
+      "learning_rate": 0.000130281724376469,
+      "loss": 1.4243,
+      "step": 7181
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.0714282284349483,
+      "learning_rate": 0.00013016453316871867,
+      "loss": 1.3911,
+      "step": 7182
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.08425509199398158,
+      "learning_rate": 0.00013004738680479354,
+      "loss": 1.3734,
+      "step": 7183
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.09272718405902941,
+      "learning_rate": 0.00012993028529889816,
+      "loss": 1.4259,
+      "step": 7184
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.0790776375510767,
+      "learning_rate": 0.00012981322866523171,
+      "loss": 1.3094,
+      "step": 7185
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07733944744848811,
+      "learning_rate": 0.00012969621691798734,
+      "loss": 1.3611,
+      "step": 7186
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07745412806827585,
+      "learning_rate": 0.0001295792500713533,
+      "loss": 1.2361,
+      "step": 7187
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.08395765539333197,
+      "learning_rate": 0.00012946232813951236,
+      "loss": 1.4544,
+      "step": 7188
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.08265912082983201,
+      "learning_rate": 0.00012934545113664142,
+      "loss": 1.4492,
+      "step": 7189
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.08664702529416787,
+      "learning_rate": 0.00012922861907691257,
+      "loss": 1.3912,
+      "step": 7190
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.08828986408607989,
+      "learning_rate": 0.00012911183197449183,
+      "loss": 1.3833,
+      "step": 7191
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.0783447810072099,
+      "learning_rate": 0.00012899508984354002,
+      "loss": 1.5175,
+      "step": 7192
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.09375213580877549,
+      "learning_rate": 0.00012887839269821262,
+      "loss": 1.3619,
+      "step": 7193
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.09220600644539853,
+      "learning_rate": 0.00012876174055265966,
+      "loss": 1.4331,
+      "step": 7194
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07795700921437147,
+      "learning_rate": 0.00012864513342102535,
+      "loss": 1.4297,
+      "step": 7195
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.08299473491200346,
+      "learning_rate": 0.00012852857131744888,
+      "loss": 1.469,
+      "step": 7196
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.09896839999352933,
+      "learning_rate": 0.00012841205425606395,
+      "loss": 1.3864,
+      "step": 7197
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.0817748274050442,
+      "learning_rate": 0.00012829558225099834,
+      "loss": 1.4143,
+      "step": 7198
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.12473542368934352,
+      "learning_rate": 0.0001281791553163749,
+      "loss": 1.3604,
+      "step": 7199
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.09029268877199315,
+      "learning_rate": 0.00012806277346631085,
+      "loss": 1.4519,
+      "step": 7200
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.09018864980005668,
+      "learning_rate": 0.0001279464367149178,
+      "loss": 1.3292,
+      "step": 7201
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07386791132315604,
+      "learning_rate": 0.00012783014507630175,
+      "loss": 1.5217,
+      "step": 7202
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.08689566481818276,
+      "learning_rate": 0.00012771389856456371,
+      "loss": 1.5659,
+      "step": 7203
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07796769099863722,
+      "learning_rate": 0.0001275976971937991,
+      "loss": 1.3743,
+      "step": 7204
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.08347089844634027,
+      "learning_rate": 0.00012748154097809745,
+      "loss": 1.403,
+      "step": 7205
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.08194139361623523,
+      "learning_rate": 0.00012736542993154316,
+      "loss": 1.5831,
+      "step": 7206
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.08814386072544067,
+      "learning_rate": 0.00012724936406821537,
+      "loss": 1.4637,
+      "step": 7207
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.09305664303495668,
+      "learning_rate": 0.000127133343402187,
+      "loss": 1.3209,
+      "step": 7208
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.09063551416325358,
+      "learning_rate": 0.00012701736794752644,
+      "loss": 1.3946,
+      "step": 7209
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.09015549472437932,
+      "learning_rate": 0.0001269014377182957,
+      "loss": 1.3488,
+      "step": 7210
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.0804346799482095,
+      "learning_rate": 0.00012678555272855208,
+      "loss": 1.4834,
+      "step": 7211
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07872396579005872,
+      "learning_rate": 0.00012666971299234668,
+      "loss": 1.3482,
+      "step": 7212
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.08578567070288019,
+      "learning_rate": 0.00012655391852372584,
+      "loss": 1.3363,
+      "step": 7213
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.08786717737868875,
+      "learning_rate": 0.00012643816933672968,
+      "loss": 1.3873,
+      "step": 7214
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.10304717691086623,
+      "learning_rate": 0.00012632246544539333,
+      "loss": 1.4498,
+      "step": 7215
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.08021184395395602,
+      "learning_rate": 0.00012620680686374646,
+      "loss": 1.3443,
+      "step": 7216
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.10865624373995214,
+      "learning_rate": 0.00012609119360581277,
+      "loss": 1.4498,
+      "step": 7217
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07778178613046131,
+      "learning_rate": 0.00012597562568561095,
+      "loss": 1.5317,
+      "step": 7218
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07738530338405038,
+      "learning_rate": 0.00012586010311715408,
+      "loss": 1.4558,
+      "step": 7219
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.08931510175039695,
+      "learning_rate": 0.0001257446259144494,
+      "loss": 1.503,
+      "step": 7220
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.09255034325625278,
+      "learning_rate": 0.00012562919409149915,
+      "loss": 1.3232,
+      "step": 7221
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.079332226562438,
+      "learning_rate": 0.00012551380766230003,
+      "loss": 1.4021,
+      "step": 7222
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.08365566359520343,
+      "learning_rate": 0.00012539846664084248,
+      "loss": 1.4243,
+      "step": 7223
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.09729578487033033,
+      "learning_rate": 0.00012528317104111225,
+      "loss": 1.3776,
+      "step": 7224
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.08689847549229128,
+      "learning_rate": 0.00012516792087708962,
+      "loss": 1.4569,
+      "step": 7225
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.08986204287640165,
+      "learning_rate": 0.0001250527161627486,
+      "loss": 1.3315,
+      "step": 7226
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.08902530175684555,
+      "learning_rate": 0.00012493755691205845,
+      "loss": 1.3933,
+      "step": 7227
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07783856035461831,
+      "learning_rate": 0.00012482244313898267,
+      "loss": 1.3352,
+      "step": 7228
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07579194127448878,
+      "learning_rate": 0.0001247073748574789,
+      "loss": 1.3036,
+      "step": 7229
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07552103458438016,
+      "learning_rate": 0.00012459235208149984,
+      "loss": 1.4446,
+      "step": 7230
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07717100398872971,
+      "learning_rate": 0.00012447737482499245,
+      "loss": 1.4675,
+      "step": 7231
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.08179823386986428,
+      "learning_rate": 0.000124362443101898,
+      "loss": 1.2272,
+      "step": 7232
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.08450782049483559,
+      "learning_rate": 0.00012424755692615213,
+      "loss": 1.4358,
+      "step": 7233
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.08644532397598073,
+      "learning_rate": 0.00012413271631168545,
+      "loss": 1.4029,
+      "step": 7234
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07850539908313922,
+      "learning_rate": 0.00012401792127242285,
+      "loss": 1.2961,
+      "step": 7235
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.0896043126589486,
+      "learning_rate": 0.00012390317182228334,
+      "loss": 1.405,
+      "step": 7236
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07587555720819515,
+      "learning_rate": 0.0001237884679751809,
+      "loss": 1.4865,
+      "step": 7237
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.08624531471226798,
+      "learning_rate": 0.00012367380974502383,
+      "loss": 1.4715,
+      "step": 7238
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.08904974913087878,
+      "learning_rate": 0.00012355919714571458,
+      "loss": 1.5391,
+      "step": 7239
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.09677429940240297,
+      "learning_rate": 0.0001234446301911506,
+      "loss": 1.3389,
+      "step": 7240
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.09073539193834754,
+      "learning_rate": 0.00012333010889522327,
+      "loss": 1.4164,
+      "step": 7241
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.08293798819734584,
+      "learning_rate": 0.00012321563327181883,
+      "loss": 1.3885,
+      "step": 7242
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07925689642566201,
+      "learning_rate": 0.00012310120333481795,
+      "loss": 1.3638,
+      "step": 7243
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.0848046515098787,
+      "learning_rate": 0.0001229868190980955,
+      "loss": 1.3759,
+      "step": 7244
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.09296734522311166,
+      "learning_rate": 0.00012287248057552092,
+      "loss": 1.3025,
+      "step": 7245
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.08184586378841201,
+      "learning_rate": 0.0001227581877809582,
+      "loss": 1.4481,
+      "step": 7246
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07882572428683679,
+      "learning_rate": 0.0001226439407282659,
+      "loss": 1.3656,
+      "step": 7247
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.09208007004167677,
+      "learning_rate": 0.0001225297394312966,
+      "loss": 1.4169,
+      "step": 7248
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.08227884585316046,
+      "learning_rate": 0.00012241558390389769,
+      "loss": 1.4895,
+      "step": 7249
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.0897848075866568,
+      "learning_rate": 0.00012230147415991116,
+      "loss": 1.4633,
+      "step": 7250
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07979829089445617,
+      "learning_rate": 0.0001221874102131728,
+      "loss": 1.39,
+      "step": 7251
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.08325287183551135,
+      "learning_rate": 0.00012207339207751354,
+      "loss": 1.3288,
+      "step": 7252
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.08921350734918522,
+      "learning_rate": 0.00012195941976675867,
+      "loss": 1.3689,
+      "step": 7253
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.09011140635333832,
+      "learning_rate": 0.00012184549329472717,
+      "loss": 1.4063,
+      "step": 7254
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.08625806008476061,
+      "learning_rate": 0.0001217316126752333,
+      "loss": 1.445,
+      "step": 7255
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07562931560165986,
+      "learning_rate": 0.00012161777792208562,
+      "loss": 1.5065,
+      "step": 7256
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.08207305626274926,
+      "learning_rate": 0.00012150398904908672,
+      "loss": 1.2579,
+      "step": 7257
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.08553307505785233,
+      "learning_rate": 0.00012139024607003402,
+      "loss": 1.3923,
+      "step": 7258
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07733318986471323,
+      "learning_rate": 0.00012127654899871936,
+      "loss": 1.439,
+      "step": 7259
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.09377725522902389,
+      "learning_rate": 0.0001211628978489287,
+      "loss": 1.348,
+      "step": 7260
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.09011338811288713,
+      "learning_rate": 0.00012104929263444269,
+      "loss": 1.4032,
+      "step": 7261
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.09418809745410185,
+      "learning_rate": 0.00012093573336903651,
+      "loss": 1.3855,
+      "step": 7262
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.09075476347236389,
+      "learning_rate": 0.00012082222006647942,
+      "loss": 1.4853,
+      "step": 7263
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.08586052880258192,
+      "learning_rate": 0.0001207087527405355,
+      "loss": 1.4392,
+      "step": 7264
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.09258361143240627,
+      "learning_rate": 0.00012059533140496276,
+      "loss": 1.4554,
+      "step": 7265
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.09197353044257664,
+      "learning_rate": 0.0001204819560735142,
+      "loss": 1.4223,
+      "step": 7266
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.08122398178814307,
+      "learning_rate": 0.00012036862675993677,
+      "loss": 1.4581,
+      "step": 7267
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.08569140445620367,
+      "learning_rate": 0.00012025534347797212,
+      "loss": 1.557,
+      "step": 7268
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.08442336639975227,
+      "learning_rate": 0.00012014210624135641,
+      "loss": 1.3068,
+      "step": 7269
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.09359124396229755,
+      "learning_rate": 0.00012002891506381974,
+      "loss": 1.4421,
+      "step": 7270
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.09664875074993715,
+      "learning_rate": 0.0001199157699590872,
+      "loss": 1.51,
+      "step": 7271
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.08256153160798978,
+      "learning_rate": 0.00011980267094087777,
+      "loss": 1.4581,
+      "step": 7272
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.08356266546750146,
+      "learning_rate": 0.00011968961802290523,
+      "loss": 1.3983,
+      "step": 7273
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07715962500880093,
+      "learning_rate": 0.0001195766112188778,
+      "loss": 1.3396,
+      "step": 7274
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07775746783057781,
+      "learning_rate": 0.00011946365054249775,
+      "loss": 1.41,
+      "step": 7275
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.08050408481614375,
+      "learning_rate": 0.00011935073600746182,
+      "loss": 1.4984,
+      "step": 7276
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.08051706956456427,
+      "learning_rate": 0.00011923786762746148,
+      "loss": 1.3049,
+      "step": 7277
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.09187755184619545,
+      "learning_rate": 0.00011912504541618252,
+      "loss": 1.3603,
+      "step": 7278
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.0945150395456591,
+      "learning_rate": 0.00011901226938730469,
+      "loss": 1.5084,
+      "step": 7279
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07896225931399797,
+      "learning_rate": 0.00011889953955450272,
+      "loss": 1.5249,
+      "step": 7280
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07849848551878082,
+      "learning_rate": 0.00011878685593144556,
+      "loss": 1.4698,
+      "step": 7281
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.08101879037358578,
+      "learning_rate": 0.00011867421853179622,
+      "loss": 1.3817,
+      "step": 7282
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.10727296608575486,
+      "learning_rate": 0.0001185616273692125,
+      "loss": 1.2219,
+      "step": 7283
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.08454663697420552,
+      "learning_rate": 0.00011844908245734659,
+      "loss": 1.396,
+      "step": 7284
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.09072240043086682,
+      "learning_rate": 0.00011833658380984491,
+      "loss": 1.2478,
+      "step": 7285
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.08550914434220919,
+      "learning_rate": 0.00011822413144034805,
+      "loss": 1.4036,
+      "step": 7286
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.08168062993549902,
+      "learning_rate": 0.0001181117253624916,
+      "loss": 1.3707,
+      "step": 7287
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.08925027465919966,
+      "learning_rate": 0.00011799936558990482,
+      "loss": 1.3877,
+      "step": 7288
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.09254528526528016,
+      "learning_rate": 0.00011788705213621198,
+      "loss": 1.4509,
+      "step": 7289
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.11415106012704615,
+      "learning_rate": 0.00011777478501503152,
+      "loss": 1.4539,
+      "step": 7290
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.0838403721688183,
+      "learning_rate": 0.00011766256423997602,
+      "loss": 1.2205,
+      "step": 7291
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07991379938968937,
+      "learning_rate": 0.00011755038982465266,
+      "loss": 1.4549,
+      "step": 7292
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.08065051687823484,
+      "learning_rate": 0.0001174382617826632,
+      "loss": 1.3445,
+      "step": 7293
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.08678497683129696,
+      "learning_rate": 0.00011732618012760327,
+      "loss": 1.492,
+      "step": 7294
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.08293945639486879,
+      "learning_rate": 0.00011721414487306326,
+      "loss": 1.3688,
+      "step": 7295
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.08227444119004665,
+      "learning_rate": 0.00011710215603262797,
+      "loss": 1.3773,
+      "step": 7296
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.08266583007232876,
+      "learning_rate": 0.00011699021361987633,
+      "loss": 1.3837,
+      "step": 7297
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.09906936808598413,
+      "learning_rate": 0.00011687831764838158,
+      "loss": 1.429,
+      "step": 7298
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.08195428558569683,
+      "learning_rate": 0.00011676646813171166,
+      "loss": 1.4555,
+      "step": 7299
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.08268397919617312,
+      "learning_rate": 0.00011665466508342876,
+      "loss": 1.358,
+      "step": 7300
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.08430853926059426,
+      "learning_rate": 0.0001165429085170892,
+      "loss": 1.3842,
+      "step": 7301
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.09381193980846146,
+      "learning_rate": 0.000116431198446244,
+      "loss": 1.365,
+      "step": 7302
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08280165821756859,
+      "learning_rate": 0.00011631953488443847,
+      "loss": 1.3367,
+      "step": 7303
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.09674682116615263,
+      "learning_rate": 0.00011620791784521195,
+      "loss": 1.5064,
+      "step": 7304
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.0895578939531404,
+      "learning_rate": 0.00011609634734209867,
+      "loss": 1.3573,
+      "step": 7305
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.09195433336912709,
+      "learning_rate": 0.00011598482338862676,
+      "loss": 1.316,
+      "step": 7306
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.0850041748062091,
+      "learning_rate": 0.00011587334599831877,
+      "loss": 1.3207,
+      "step": 7307
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.09066164990814109,
+      "learning_rate": 0.00011576191518469192,
+      "loss": 1.4068,
+      "step": 7308
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08737458422241964,
+      "learning_rate": 0.0001156505309612576,
+      "loss": 1.3028,
+      "step": 7309
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.0818355985921247,
+      "learning_rate": 0.00011553919334152135,
+      "loss": 1.4178,
+      "step": 7310
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.09412720167056654,
+      "learning_rate": 0.00011542790233898331,
+      "loss": 1.4519,
+      "step": 7311
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.07815617669426021,
+      "learning_rate": 0.00011531665796713814,
+      "loss": 1.4109,
+      "step": 7312
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08461422716107501,
+      "learning_rate": 0.0001152054602394742,
+      "loss": 1.4821,
+      "step": 7313
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.09249266123132792,
+      "learning_rate": 0.00011509430916947483,
+      "loss": 1.4866,
+      "step": 7314
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08558705248643976,
+      "learning_rate": 0.00011498320477061758,
+      "loss": 1.462,
+      "step": 7315
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08300356854080107,
+      "learning_rate": 0.00011487214705637395,
+      "loss": 1.2952,
+      "step": 7316
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08004935332127502,
+      "learning_rate": 0.00011476113604021038,
+      "loss": 1.3786,
+      "step": 7317
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.07554824940526937,
+      "learning_rate": 0.00011465017173558717,
+      "loss": 1.4424,
+      "step": 7318
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08169386017284468,
+      "learning_rate": 0.00011453925415595901,
+      "loss": 1.4019,
+      "step": 7319
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08875579848412374,
+      "learning_rate": 0.00011442838331477513,
+      "loss": 1.4455,
+      "step": 7320
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08377514812003137,
+      "learning_rate": 0.00011431755922547921,
+      "loss": 1.3962,
+      "step": 7321
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08782221330386418,
+      "learning_rate": 0.00011420678190150879,
+      "loss": 1.4341,
+      "step": 7322
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.07954298875210211,
+      "learning_rate": 0.00011409605135629603,
+      "loss": 1.5656,
+      "step": 7323
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.0961529478173863,
+      "learning_rate": 0.00011398536760326761,
+      "loss": 1.3828,
+      "step": 7324
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.0792818090967388,
+      "learning_rate": 0.00011387473065584403,
+      "loss": 1.3704,
+      "step": 7325
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.09990077427687467,
+      "learning_rate": 0.00011376414052744055,
+      "loss": 1.4352,
+      "step": 7326
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.09129573736611461,
+      "learning_rate": 0.00011365359723146673,
+      "loss": 1.3839,
+      "step": 7327
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08876545291340723,
+      "learning_rate": 0.00011354310078132618,
+      "loss": 1.2871,
+      "step": 7328
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.0826376515378444,
+      "learning_rate": 0.00011343265119041685,
+      "loss": 1.4464,
+      "step": 7329
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08897338013882695,
+      "learning_rate": 0.00011332224847213124,
+      "loss": 1.4832,
+      "step": 7330
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08554350555655972,
+      "learning_rate": 0.00011321189263985621,
+      "loss": 1.3448,
+      "step": 7331
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08963501695769718,
+      "learning_rate": 0.00011310158370697254,
+      "loss": 1.4981,
+      "step": 7332
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08891139005332287,
+      "learning_rate": 0.00011299132168685566,
+      "loss": 1.5208,
+      "step": 7333
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.07985887227024256,
+      "learning_rate": 0.00011288110659287543,
+      "loss": 1.2565,
+      "step": 7334
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08249018502823845,
+      "learning_rate": 0.00011277093843839548,
+      "loss": 1.3601,
+      "step": 7335
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.0882205268492248,
+      "learning_rate": 0.00011266081723677434,
+      "loss": 1.343,
+      "step": 7336
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08568573371670497,
+      "learning_rate": 0.00011255074300136437,
+      "loss": 1.3917,
+      "step": 7337
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.07816830269456879,
+      "learning_rate": 0.00011244071574551268,
+      "loss": 1.3174,
+      "step": 7338
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08957362122918865,
+      "learning_rate": 0.00011233073548256018,
+      "loss": 1.354,
+      "step": 7339
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08878251511658643,
+      "learning_rate": 0.00011222080222584263,
+      "loss": 1.3677,
+      "step": 7340
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08375678858047989,
+      "learning_rate": 0.00011211091598868956,
+      "loss": 1.4903,
+      "step": 7341
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.0887016822378597,
+      "learning_rate": 0.00011200107678442517,
+      "loss": 1.2804,
+      "step": 7342
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.0869215161095677,
+      "learning_rate": 0.000111891284626368,
+      "loss": 1.3929,
+      "step": 7343
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08806012815888258,
+      "learning_rate": 0.00011178153952783043,
+      "loss": 1.3918,
+      "step": 7344
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08770599045346396,
+      "learning_rate": 0.00011167184150211962,
+      "loss": 1.3323,
+      "step": 7345
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08611854804897985,
+      "learning_rate": 0.00011156219056253691,
+      "loss": 1.3703,
+      "step": 7346
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08968488237086665,
+      "learning_rate": 0.00011145258672237762,
+      "loss": 1.4883,
+      "step": 7347
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.09790986626447239,
+      "learning_rate": 0.00011134302999493173,
+      "loss": 1.3519,
+      "step": 7348
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08092278640901887,
+      "learning_rate": 0.0001112335203934836,
+      "loss": 1.278,
+      "step": 7349
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.07777105411353137,
+      "learning_rate": 0.00011112405793131114,
+      "loss": 1.3237,
+      "step": 7350
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.09088764719653418,
+      "learning_rate": 0.00011101464262168731,
+      "loss": 1.4631,
+      "step": 7351
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08000448230902753,
+      "learning_rate": 0.00011090527447787924,
+      "loss": 1.3473,
+      "step": 7352
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08583186983779009,
+      "learning_rate": 0.00011079595351314791,
+      "loss": 1.4034,
+      "step": 7353
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.0960634586034175,
+      "learning_rate": 0.00011068667974074903,
+      "loss": 1.4432,
+      "step": 7354
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08076190348815085,
+      "learning_rate": 0.00011057745317393252,
+      "loss": 1.334,
+      "step": 7355
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.09802013313693342,
+      "learning_rate": 0.00011046827382594227,
+      "loss": 1.3139,
+      "step": 7356
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08870505251066106,
+      "learning_rate": 0.00011035914171001665,
+      "loss": 1.369,
+      "step": 7357
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.10760046353869862,
+      "learning_rate": 0.0001102500568393886,
+      "loss": 1.4748,
+      "step": 7358
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08482909523650954,
+      "learning_rate": 0.0001101410192272848,
+      "loss": 1.309,
+      "step": 7359
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08967836852723367,
+      "learning_rate": 0.00011003202888692632,
+      "loss": 1.3521,
+      "step": 7360
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08971932596246342,
+      "learning_rate": 0.00010992308583152877,
+      "loss": 1.3055,
+      "step": 7361
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08965578968467429,
+      "learning_rate": 0.00010981419007430199,
+      "loss": 1.496,
+      "step": 7362
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08924314469159167,
+      "learning_rate": 0.00010970534162844975,
+      "loss": 1.495,
+      "step": 7363
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08709222636542864,
+      "learning_rate": 0.00010959654050717032,
+      "loss": 1.2642,
+      "step": 7364
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08930747249209584,
+      "learning_rate": 0.00010948778672365644,
+      "loss": 1.3918,
+      "step": 7365
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08647523289589773,
+      "learning_rate": 0.00010937908029109461,
+      "loss": 1.4539,
+      "step": 7366
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08336240679126437,
+      "learning_rate": 0.00010927042122266611,
+      "loss": 1.3636,
+      "step": 7367
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.07905964491305716,
+      "learning_rate": 0.00010916180953154592,
+      "loss": 1.3862,
+      "step": 7368
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.09007504611985878,
+      "learning_rate": 0.00010905324523090377,
+      "loss": 1.4469,
+      "step": 7369
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08221245517814316,
+      "learning_rate": 0.00010894472833390357,
+      "loss": 1.4572,
+      "step": 7370
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08384087412285039,
+      "learning_rate": 0.00010883625885370319,
+      "loss": 1.5368,
+      "step": 7371
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.09317832358790074,
+      "learning_rate": 0.00010872783680345489,
+      "loss": 1.4228,
+      "step": 7372
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.10753007879502102,
+      "learning_rate": 0.00010861946219630525,
+      "loss": 1.3968,
+      "step": 7373
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.09197775161563887,
+      "learning_rate": 0.00010851113504539528,
+      "loss": 1.3543,
+      "step": 7374
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08071650914079001,
+      "learning_rate": 0.00010840285536385968,
+      "loss": 1.421,
+      "step": 7375
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.09167372528944248,
+      "learning_rate": 0.00010829462316482797,
+      "loss": 1.5266,
+      "step": 7376
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.09098539935181792,
+      "learning_rate": 0.00010818643846142373,
+      "loss": 1.3709,
+      "step": 7377
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08429483207935262,
+      "learning_rate": 0.00010807830126676443,
+      "loss": 1.3496,
+      "step": 7378
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08474915241120738,
+      "learning_rate": 0.0001079702115939623,
+      "loss": 1.3521,
+      "step": 7379
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08776848704159057,
+      "learning_rate": 0.0001078621694561237,
+      "loss": 1.3723,
+      "step": 7380
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08454882765125812,
+      "learning_rate": 0.00010775417486634893,
+      "loss": 1.4085,
+      "step": 7381
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08768259170723725,
+      "learning_rate": 0.00010764622783773259,
+      "loss": 1.5293,
+      "step": 7382
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08057887240979167,
+      "learning_rate": 0.00010753832838336397,
+      "loss": 1.4617,
+      "step": 7383
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08991989764932008,
+      "learning_rate": 0.00010743047651632588,
+      "loss": 1.359,
+      "step": 7384
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08712173949360097,
+      "learning_rate": 0.00010732267224969588,
+      "loss": 1.3348,
+      "step": 7385
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.09148073926024225,
+      "learning_rate": 0.00010721491559654579,
+      "loss": 1.3145,
+      "step": 7386
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.09115660354014754,
+      "learning_rate": 0.00010710720656994117,
+      "loss": 1.3405,
+      "step": 7387
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.09285950683691585,
+      "learning_rate": 0.00010699954518294224,
+      "loss": 1.351,
+      "step": 7388
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08496458358175875,
+      "learning_rate": 0.00010689193144860354,
+      "loss": 1.3887,
+      "step": 7389
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08814196543992106,
+      "learning_rate": 0.00010678436537997321,
+      "loss": 1.3929,
+      "step": 7390
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08392941262208066,
+      "learning_rate": 0.00010667684699009439,
+      "loss": 1.4186,
+      "step": 7391
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.0825169692685744,
+      "learning_rate": 0.00010656937629200369,
+      "loss": 1.5325,
+      "step": 7392
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.0969018358700212,
+      "learning_rate": 0.00010646195329873259,
+      "loss": 1.5803,
+      "step": 7393
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08785792908215653,
+      "learning_rate": 0.00010635457802330628,
+      "loss": 1.3293,
+      "step": 7394
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08970546446133174,
+      "learning_rate": 0.0001062472504787445,
+      "loss": 1.4592,
+      "step": 7395
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.09689624451528016,
+      "learning_rate": 0.0001061399706780612,
+      "loss": 1.3814,
+      "step": 7396
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.08661583703872869,
+      "learning_rate": 0.00010603273863426411,
+      "loss": 1.4196,
+      "step": 7397
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.08672305211358776,
+      "learning_rate": 0.00010592555436035573,
+      "loss": 1.4574,
+      "step": 7398
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.07671987734255735,
+      "learning_rate": 0.00010581841786933261,
+      "loss": 1.471,
+      "step": 7399
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.07563795205528116,
+      "learning_rate": 0.00010571132917418508,
+      "loss": 1.4617,
+      "step": 7400
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.09722906137813156,
+      "learning_rate": 0.00010560428828789837,
+      "loss": 1.2705,
+      "step": 7401
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.09142567018690823,
+      "learning_rate": 0.00010549729522345142,
+      "loss": 1.4441,
+      "step": 7402
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.09043229441252876,
+      "learning_rate": 0.0001053903499938173,
+      "loss": 1.4887,
+      "step": 7403
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.08808126569357126,
+      "learning_rate": 0.0001052834526119637,
+      "loss": 1.5165,
+      "step": 7404
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.10853347121607891,
+      "learning_rate": 0.00010517660309085242,
+      "loss": 1.3782,
+      "step": 7405
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.08195414140573754,
+      "learning_rate": 0.00010506980144343898,
+      "loss": 1.3016,
+      "step": 7406
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.09994610269025854,
+      "learning_rate": 0.00010496304768267373,
+      "loss": 1.3852,
+      "step": 7407
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.09477202194767692,
+      "learning_rate": 0.00010485634182150089,
+      "loss": 1.3728,
+      "step": 7408
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.08841337933183288,
+      "learning_rate": 0.00010474968387285882,
+      "loss": 1.3434,
+      "step": 7409
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.0959761214605188,
+      "learning_rate": 0.00010464307384968019,
+      "loss": 1.5121,
+      "step": 7410
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.08672942952818297,
+      "learning_rate": 0.00010453651176489198,
+      "loss": 1.4939,
+      "step": 7411
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.08689856792950482,
+      "learning_rate": 0.00010442999763141509,
+      "loss": 1.4176,
+      "step": 7412
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.08847913538326554,
+      "learning_rate": 0.00010432353146216456,
+      "loss": 1.3859,
+      "step": 7413
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.08940969838357604,
+      "learning_rate": 0.00010421711327005013,
+      "loss": 1.3339,
+      "step": 7414
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.08584386698958232,
+      "learning_rate": 0.00010411074306797502,
+      "loss": 1.2584,
+      "step": 7415
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.09041104984604122,
+      "learning_rate": 0.00010400442086883715,
+      "loss": 1.5023,
+      "step": 7416
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.08328881943694692,
+      "learning_rate": 0.0001038981466855286,
+      "loss": 1.2562,
+      "step": 7417
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.07626411694127093,
+      "learning_rate": 0.00010379192053093522,
+      "loss": 1.3377,
+      "step": 7418
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.08720457355500504,
+      "learning_rate": 0.00010368574241793738,
+      "loss": 1.364,
+      "step": 7419
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.07654734453604108,
+      "learning_rate": 0.00010357961235940977,
+      "loss": 1.3639,
+      "step": 7420
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.08657449519324938,
+      "learning_rate": 0.0001034735303682206,
+      "loss": 1.3308,
+      "step": 7421
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.08250237320381912,
+      "learning_rate": 0.00010336749645723298,
+      "loss": 1.4285,
+      "step": 7422
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.08605554787724745,
+      "learning_rate": 0.00010326151063930395,
+      "loss": 1.3979,
+      "step": 7423
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.09436166424350222,
+      "learning_rate": 0.0001031555729272845,
+      "loss": 1.4827,
+      "step": 7424
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.0873621789720717,
+      "learning_rate": 0.00010304968333401982,
+      "loss": 1.4897,
+      "step": 7425
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.08852806729006432,
+      "learning_rate": 0.00010294384187234951,
+      "loss": 1.4744,
+      "step": 7426
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.09802979552309042,
+      "learning_rate": 0.00010283804855510742,
+      "loss": 1.3337,
+      "step": 7427
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.08249683463009905,
+      "learning_rate": 0.00010273230339512102,
+      "loss": 1.3253,
+      "step": 7428
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.08737069761111654,
+      "learning_rate": 0.00010262660640521242,
+      "loss": 1.4135,
+      "step": 7429
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.07672002832446094,
+      "learning_rate": 0.00010252095759819785,
+      "loss": 1.3387,
+      "step": 7430
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.08146901689082577,
+      "learning_rate": 0.00010241535698688743,
+      "loss": 1.4587,
+      "step": 7431
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.08357617514349239,
+      "learning_rate": 0.00010230980458408573,
+      "loss": 1.3892,
+      "step": 7432
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.09688593678688902,
+      "learning_rate": 0.00010220430040259116,
+      "loss": 1.4337,
+      "step": 7433
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.0949961436992429,
+      "learning_rate": 0.00010209884445519668,
+      "loss": 1.479,
+      "step": 7434
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.09217958997274633,
+      "learning_rate": 0.00010199343675468897,
+      "loss": 1.4274,
+      "step": 7435
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.0927875344960079,
+      "learning_rate": 0.0001018880773138493,
+      "loss": 1.4548,
+      "step": 7436
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.10240453114593467,
+      "learning_rate": 0.00010178276614545267,
+      "loss": 1.3366,
+      "step": 7437
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.07930711688295726,
+      "learning_rate": 0.00010167750326226848,
+      "loss": 1.3864,
+      "step": 7438
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.08988242509021517,
+      "learning_rate": 0.00010157228867706041,
+      "loss": 1.3858,
+      "step": 7439
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.08546466142893103,
+      "learning_rate": 0.00010146712240258577,
+      "loss": 1.337,
+      "step": 7440
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.07476124167921687,
+      "learning_rate": 0.00010136200445159649,
+      "loss": 1.3619,
+      "step": 7441
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.08503840192574158,
+      "learning_rate": 0.00010125693483683862,
+      "loss": 1.5063,
+      "step": 7442
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.08121488722572119,
+      "learning_rate": 0.00010115191357105192,
+      "loss": 1.3646,
+      "step": 7443
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.08638910819398816,
+      "learning_rate": 0.00010104694066697089,
+      "loss": 1.4189,
+      "step": 7444
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.07556675074773149,
+      "learning_rate": 0.00010094201613732373,
+      "loss": 1.4163,
+      "step": 7445
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.09294527602966572,
+      "learning_rate": 0.00010083713999483269,
+      "loss": 1.3499,
+      "step": 7446
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.09451147367174063,
+      "learning_rate": 0.00010073231225221457,
+      "loss": 1.3585,
+      "step": 7447
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.09048973763769977,
+      "learning_rate": 0.00010062753292218025,
+      "loss": 1.4001,
+      "step": 7448
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.09350501774740308,
+      "learning_rate": 0.00010052280201743425,
+      "loss": 1.376,
+      "step": 7449
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.08746862295951763,
+      "learning_rate": 0.0001004181195506757,
+      "loss": 1.4711,
+      "step": 7450
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.08986402754420884,
+      "learning_rate": 0.00010031348553459785,
+      "loss": 1.5504,
+      "step": 7451
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.08757788063869364,
+      "learning_rate": 0.00010020889998188771,
+      "loss": 1.3495,
+      "step": 7452
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.0786814505136955,
+      "learning_rate": 0.00010010436290522673,
+      "loss": 1.3102,
+      "step": 7453
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.08204954091010624,
+      "learning_rate": 9.999987431729051e-05,
+      "loss": 1.6017,
+      "step": 7454
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.08699570930256927,
+      "learning_rate": 9.989543423074854e-05,
+      "loss": 1.4445,
+      "step": 7455
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.08730952432960949,
+      "learning_rate": 9.979104265826438e-05,
+      "loss": 1.409,
+      "step": 7456
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.08541075113090651,
+      "learning_rate": 9.96866996124961e-05,
+      "loss": 1.3488,
+      "step": 7457
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.09227061283218439,
+      "learning_rate": 9.958240510609568e-05,
+      "loss": 1.3261,
+      "step": 7458
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.07891602228455197,
+      "learning_rate": 9.947815915170894e-05,
+      "loss": 1.4237,
+      "step": 7459
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.10392044416795326,
+      "learning_rate": 9.937396176197621e-05,
+      "loss": 1.373,
+      "step": 7460
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.08909632543444393,
+      "learning_rate": 9.926981294953191e-05,
+      "loss": 1.4048,
+      "step": 7461
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.08711148193189107,
+      "learning_rate": 9.91657127270042e-05,
+      "loss": 1.301,
+      "step": 7462
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.09677062189294994,
+      "learning_rate": 9.906166110701587e-05,
+      "loss": 1.4211,
+      "step": 7463
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.08012618245248758,
+      "learning_rate": 9.895765810218322e-05,
+      "loss": 1.3605,
+      "step": 7464
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.07881596378929731,
+      "learning_rate": 9.885370372511727e-05,
+      "loss": 1.2357,
+      "step": 7465
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.09099731322497107,
+      "learning_rate": 9.874979798842255e-05,
+      "loss": 1.3286,
+      "step": 7466
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.08318317206545245,
+      "learning_rate": 9.86459409046983e-05,
+      "loss": 1.3099,
+      "step": 7467
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.08834905278247804,
+      "learning_rate": 9.854213248653721e-05,
+      "loss": 1.3114,
+      "step": 7468
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.08740069078768421,
+      "learning_rate": 9.843837274652667e-05,
+      "loss": 1.5023,
+      "step": 7469
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.08485296530187857,
+      "learning_rate": 9.833466169724792e-05,
+      "loss": 1.3516,
+      "step": 7470
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.09309838507104544,
+      "learning_rate": 9.823099935127605e-05,
+      "loss": 1.4752,
+      "step": 7471
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.08699771252470198,
+      "learning_rate": 9.812738572118063e-05,
+      "loss": 1.467,
+      "step": 7472
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.08672188155001476,
+      "learning_rate": 9.802382081952527e-05,
+      "loss": 1.3471,
+      "step": 7473
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.0881050276317137,
+      "learning_rate": 9.792030465886736e-05,
+      "loss": 1.3859,
+      "step": 7474
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.0867622769936228,
+      "learning_rate": 9.781683725175866e-05,
+      "loss": 1.542,
+      "step": 7475
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.08473741654209976,
+      "learning_rate": 9.771341861074523e-05,
+      "loss": 1.3428,
+      "step": 7476
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1022488836419732,
+      "learning_rate": 9.761004874836644e-05,
+      "loss": 1.3613,
+      "step": 7477
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.09639605063432713,
+      "learning_rate": 9.750672767715651e-05,
+      "loss": 1.4877,
+      "step": 7478
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.09102505887751455,
+      "learning_rate": 9.740345540964357e-05,
+      "loss": 1.3897,
+      "step": 7479
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.09344953204952636,
+      "learning_rate": 9.730023195834947e-05,
+      "loss": 1.5112,
+      "step": 7480
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.09225774918138216,
+      "learning_rate": 9.71970573357906e-05,
+      "loss": 1.2466,
+      "step": 7481
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.08174545024210766,
+      "learning_rate": 9.709393155447732e-05,
+      "loss": 1.4278,
+      "step": 7482
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.08691964908390155,
+      "learning_rate": 9.699085462691376e-05,
+      "loss": 1.3324,
+      "step": 7483
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.08499423352944359,
+      "learning_rate": 9.688782656559842e-05,
+      "loss": 1.3487,
+      "step": 7484
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1086986268584967,
+      "learning_rate": 9.6784847383024e-05,
+      "loss": 1.3336,
+      "step": 7485
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.10390079226716396,
+      "learning_rate": 9.66819170916769e-05,
+      "loss": 1.4299,
+      "step": 7486
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.08179398408151457,
+      "learning_rate": 9.657903570403769e-05,
+      "loss": 1.3214,
+      "step": 7487
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.09271074862383373,
+      "learning_rate": 9.647620323258121e-05,
+      "loss": 1.3402,
+      "step": 7488
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.09774757883056379,
+      "learning_rate": 9.637341968977636e-05,
+      "loss": 1.4131,
+      "step": 7489
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.07832701611547217,
+      "learning_rate": 9.627068508808579e-05,
+      "loss": 1.3784,
+      "step": 7490
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1041386239978518,
+      "learning_rate": 9.616799943996651e-05,
+      "loss": 1.295,
+      "step": 7491
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.08705483512520137,
+      "learning_rate": 9.606536275786965e-05,
+      "loss": 1.3471,
+      "step": 7492
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.08664530820976736,
+      "learning_rate": 9.596277505423994e-05,
+      "loss": 1.4664,
+      "step": 7493
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.10276841331323949,
+      "learning_rate": 9.586023634151674e-05,
+      "loss": 1.4516,
+      "step": 7494
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.09129386033307563,
+      "learning_rate": 9.575774663213327e-05,
+      "loss": 1.4629,
+      "step": 7495
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.09433687443593726,
+      "learning_rate": 9.565530593851656e-05,
+      "loss": 1.478,
+      "step": 7496
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.08781641900491051,
+      "learning_rate": 9.55529142730881e-05,
+      "loss": 1.2935,
+      "step": 7497
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.10812828620908439,
+      "learning_rate": 9.545057164826315e-05,
+      "loss": 1.4357,
+      "step": 7498
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.10200044996981032,
+      "learning_rate": 9.534827807645091e-05,
+      "loss": 1.229,
+      "step": 7499
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.09240236105764514,
+      "learning_rate": 9.524603357005501e-05,
+      "loss": 1.4634,
+      "step": 7500
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.09454120611142644,
+      "learning_rate": 9.514383814147309e-05,
+      "loss": 1.4214,
+      "step": 7501
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.09548820233824558,
+      "learning_rate": 9.50416918030964e-05,
+      "loss": 1.3984,
+      "step": 7502
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.07630747179285953,
+      "learning_rate": 9.493959456731072e-05,
+      "loss": 1.4153,
+      "step": 7503
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.0819944969222753,
+      "learning_rate": 9.483754644649573e-05,
+      "loss": 1.2477,
+      "step": 7504
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.08982939320379192,
+      "learning_rate": 9.473554745302492e-05,
+      "loss": 1.4237,
+      "step": 7505
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.0941844203191723,
+      "learning_rate": 9.463359759926615e-05,
+      "loss": 1.405,
+      "step": 7506
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.09325934826499595,
+      "learning_rate": 9.453169689758134e-05,
+      "loss": 1.5425,
+      "step": 7507
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.09148103195006312,
+      "learning_rate": 9.442984536032612e-05,
+      "loss": 1.4171,
+      "step": 7508
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.09186090103298966,
+      "learning_rate": 9.43280429998502e-05,
+      "loss": 1.3543,
+      "step": 7509
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.08035881723942448,
+      "learning_rate": 9.422628982849785e-05,
+      "loss": 1.4375,
+      "step": 7510
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.09344417265322892,
+      "learning_rate": 9.412458585860656e-05,
+      "loss": 1.4238,
+      "step": 7511
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.0960705210553254,
+      "learning_rate": 9.402293110250853e-05,
+      "loss": 1.5747,
+      "step": 7512
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1033151773139312,
+      "learning_rate": 9.392132557252986e-05,
+      "loss": 1.3517,
+      "step": 7513
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.09385641709122178,
+      "learning_rate": 9.381976928099029e-05,
+      "loss": 1.3943,
+      "step": 7514
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.0779487282752614,
+      "learning_rate": 9.371826224020397e-05,
+      "loss": 1.4504,
+      "step": 7515
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.081718534217225,
+      "learning_rate": 9.361680446247922e-05,
+      "loss": 1.3773,
+      "step": 7516
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.07810061118795375,
+      "learning_rate": 9.351539596011776e-05,
+      "loss": 1.336,
+      "step": 7517
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.08383963624184462,
+      "learning_rate": 9.341403674541605e-05,
+      "loss": 1.4401,
+      "step": 7518
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.08943783242052575,
+      "learning_rate": 9.331272683066399e-05,
+      "loss": 1.401,
+      "step": 7519
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.08459384581972972,
+      "learning_rate": 9.321146622814597e-05,
+      "loss": 1.6098,
+      "step": 7520
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.08989778220774186,
+      "learning_rate": 9.311025495013997e-05,
+      "loss": 1.4655,
+      "step": 7521
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.08350938557351134,
+      "learning_rate": 9.30090930089183e-05,
+      "loss": 1.4446,
+      "step": 7522
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.09453361809735741,
+      "learning_rate": 9.290798041674736e-05,
+      "loss": 1.3595,
+      "step": 7523
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.08402181500544711,
+      "learning_rate": 9.280691718588713e-05,
+      "loss": 1.4212,
+      "step": 7524
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.08162461265250069,
+      "learning_rate": 9.270590332859202e-05,
+      "loss": 1.4145,
+      "step": 7525
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.07636153927084101,
+      "learning_rate": 9.260493885711035e-05,
+      "loss": 1.3644,
+      "step": 7526
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.08638875647764933,
+      "learning_rate": 9.250402378368427e-05,
+      "loss": 1.5009,
+      "step": 7527
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.0784996066786888,
+      "learning_rate": 9.240315812055028e-05,
+      "loss": 1.3155,
+      "step": 7528
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.09020209891163362,
+      "learning_rate": 9.230234187993858e-05,
+      "loss": 1.3745,
+      "step": 7529
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.07728842140450685,
+      "learning_rate": 9.220157507407334e-05,
+      "loss": 1.473,
+      "step": 7530
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.09999693012740897,
+      "learning_rate": 9.210085771517296e-05,
+      "loss": 1.3881,
+      "step": 7531
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.07475659418648158,
+      "learning_rate": 9.200018981544999e-05,
+      "loss": 1.4635,
+      "step": 7532
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.08610114317683623,
+      "learning_rate": 9.189957138711053e-05,
+      "loss": 1.3655,
+      "step": 7533
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.08629125966598605,
+      "learning_rate": 9.179900244235489e-05,
+      "loss": 1.4349,
+      "step": 7534
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.08839588239342566,
+      "learning_rate": 9.169848299337764e-05,
+      "loss": 1.3432,
+      "step": 7535
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.09113204431680144,
+      "learning_rate": 9.159801305236687e-05,
+      "loss": 1.4708,
+      "step": 7536
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.08841340566615136,
+      "learning_rate": 9.149759263150493e-05,
+      "loss": 1.3131,
+      "step": 7537
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.09119900971580401,
+      "learning_rate": 9.139722174296838e-05,
+      "loss": 1.4045,
+      "step": 7538
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.08828684717343677,
+      "learning_rate": 9.129690039892735e-05,
+      "loss": 1.3373,
+      "step": 7539
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.09411904346839678,
+      "learning_rate": 9.119662861154598e-05,
+      "loss": 1.4342,
+      "step": 7540
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.09128939712399425,
+      "learning_rate": 9.109640639298294e-05,
+      "loss": 1.3869,
+      "step": 7541
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.0786323938770777,
+      "learning_rate": 9.099623375539017e-05,
+      "loss": 1.3981,
+      "step": 7542
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.08550467735039376,
+      "learning_rate": 9.089611071091414e-05,
+      "loss": 1.2798,
+      "step": 7543
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.0956798289127353,
+      "learning_rate": 9.079603727169521e-05,
+      "loss": 1.4181,
+      "step": 7544
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.09370505228997768,
+      "learning_rate": 9.069601344986733e-05,
+      "loss": 1.4377,
+      "step": 7545
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.09234717974288324,
+      "learning_rate": 9.059603925755894e-05,
+      "loss": 1.308,
+      "step": 7546
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.07823660374355815,
+      "learning_rate": 9.049611470689234e-05,
+      "loss": 1.5184,
+      "step": 7547
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.08104361176940131,
+      "learning_rate": 9.039623980998346e-05,
+      "loss": 1.4144,
+      "step": 7548
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.08313514255907395,
+      "learning_rate": 9.029641457894266e-05,
+      "loss": 1.558,
+      "step": 7549
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.08348448293389067,
+      "learning_rate": 9.019663902587416e-05,
+      "loss": 1.4278,
+      "step": 7550
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.09062830049572031,
+      "learning_rate": 9.009691316287599e-05,
+      "loss": 1.2734,
+      "step": 7551
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.09077517992882414,
+      "learning_rate": 8.99972370020401e-05,
+      "loss": 1.3934,
+      "step": 7552
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.08704980419159235,
+      "learning_rate": 8.989761055545276e-05,
+      "loss": 1.4072,
+      "step": 7553
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.09613577933462986,
+      "learning_rate": 8.979803383519408e-05,
+      "loss": 1.2797,
+      "step": 7554
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.08722762807612891,
+      "learning_rate": 8.969850685333786e-05,
+      "loss": 1.4651,
+      "step": 7555
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.08697344594357496,
+      "learning_rate": 8.959902962195221e-05,
+      "loss": 1.2314,
+      "step": 7556
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.08732520594131644,
+      "learning_rate": 8.949960215309921e-05,
+      "loss": 1.3147,
+      "step": 7557
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.08586624155147374,
+      "learning_rate": 8.94002244588345e-05,
+      "loss": 1.297,
+      "step": 7558
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.0888351209656821,
+      "learning_rate": 8.930089655120832e-05,
+      "loss": 1.4383,
+      "step": 7559
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.08601421622050333,
+      "learning_rate": 8.920161844226416e-05,
+      "loss": 1.454,
+      "step": 7560
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.09618707225930943,
+      "learning_rate": 8.910239014404015e-05,
+      "loss": 1.4776,
+      "step": 7561
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.08193952890730456,
+      "learning_rate": 8.900321166856773e-05,
+      "loss": 1.5756,
+      "step": 7562
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.08770806903706792,
+      "learning_rate": 8.8904083027873e-05,
+      "loss": 1.4326,
+      "step": 7563
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.084513768626814,
+      "learning_rate": 8.88050042339753e-05,
+      "loss": 1.4,
+      "step": 7564
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.09441648259709004,
+      "learning_rate": 8.870597529888847e-05,
+      "loss": 1.4048,
+      "step": 7565
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.09472993962668384,
+      "learning_rate": 8.86069962346202e-05,
+      "loss": 1.4668,
+      "step": 7566
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.08358205795408,
+      "learning_rate": 8.850806705317183e-05,
+      "loss": 1.4674,
+      "step": 7567
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.09200422224464193,
+      "learning_rate": 8.840918776653889e-05,
+      "loss": 1.3779,
+      "step": 7568
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.07949303886065458,
+      "learning_rate": 8.831035838671108e-05,
+      "loss": 1.3128,
+      "step": 7569
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.09294930528202666,
+      "learning_rate": 8.82115789256715e-05,
+      "loss": 1.4005,
+      "step": 7570
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.09418408164473525,
+      "learning_rate": 8.811284939539771e-05,
+      "loss": 1.3714,
+      "step": 7571
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.08533122362657032,
+      "learning_rate": 8.801416980786098e-05,
+      "loss": 1.3943,
+      "step": 7572
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.08479975478034155,
+      "learning_rate": 8.791554017502635e-05,
+      "loss": 1.4126,
+      "step": 7573
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.08617405006978586,
+      "learning_rate": 8.781696050885313e-05,
+      "loss": 1.2658,
+      "step": 7574
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.08868157737014712,
+      "learning_rate": 8.771843082129455e-05,
+      "loss": 1.3586,
+      "step": 7575
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.09198408870723747,
+      "learning_rate": 8.761995112429749e-05,
+      "loss": 1.3592,
+      "step": 7576
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.09868921187630147,
+      "learning_rate": 8.752152142980308e-05,
+      "loss": 1.339,
+      "step": 7577
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.0973691738128529,
+      "learning_rate": 8.74231417497463e-05,
+      "loss": 1.299,
+      "step": 7578
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.09481197124866866,
+      "learning_rate": 8.732481209605587e-05,
+      "loss": 1.3499,
+      "step": 7579
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.10127492341278146,
+      "learning_rate": 8.722653248065466e-05,
+      "loss": 1.4342,
+      "step": 7580
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.0879516972178395,
+      "learning_rate": 8.712830291545953e-05,
+      "loss": 1.3871,
+      "step": 7581
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.08331806460834346,
+      "learning_rate": 8.703012341238109e-05,
+      "loss": 1.3958,
+      "step": 7582
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.09854127039810365,
+      "learning_rate": 8.69319939833238e-05,
+      "loss": 1.4561,
+      "step": 7583
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.08925024465869981,
+      "learning_rate": 8.683391464018625e-05,
+      "loss": 1.4291,
+      "step": 7584
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.09092135858413694,
+      "learning_rate": 8.673588539486116e-05,
+      "loss": 1.4605,
+      "step": 7585
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.08755962376538613,
+      "learning_rate": 8.663790625923452e-05,
+      "loss": 1.502,
+      "step": 7586
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.08256713069946808,
+      "learning_rate": 8.653997724518681e-05,
+      "loss": 1.4174,
+      "step": 7587
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.09290712035854404,
+      "learning_rate": 8.644209836459244e-05,
+      "loss": 1.4073,
+      "step": 7588
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.0882259607720163,
+      "learning_rate": 8.634426962931924e-05,
+      "loss": 1.3124,
+      "step": 7589
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.08602143518147524,
+      "learning_rate": 8.62464910512294e-05,
+      "loss": 1.3498,
+      "step": 7590
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.08533184505192618,
+      "learning_rate": 8.614876264217913e-05,
+      "loss": 1.4247,
+      "step": 7591
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.08913080708015374,
+      "learning_rate": 8.60510844140181e-05,
+      "loss": 1.4991,
+      "step": 7592
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.09800999230765522,
+      "learning_rate": 8.595345637859009e-05,
+      "loss": 1.3599,
+      "step": 7593
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.08656762358628595,
+      "learning_rate": 8.585587854773308e-05,
+      "loss": 1.3821,
+      "step": 7594
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.07657721840844864,
+      "learning_rate": 8.575835093327844e-05,
+      "loss": 1.4007,
+      "step": 7595
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.08925546459390414,
+      "learning_rate": 8.566087354705182e-05,
+      "loss": 1.4172,
+      "step": 7596
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.09137356710770007,
+      "learning_rate": 8.556344640087288e-05,
+      "loss": 1.4559,
+      "step": 7597
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.0977920327352257,
+      "learning_rate": 8.546606950655467e-05,
+      "loss": 1.2935,
+      "step": 7598
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.09795370695417718,
+      "learning_rate": 8.536874287590468e-05,
+      "loss": 1.3887,
+      "step": 7599
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.09176662763448601,
+      "learning_rate": 8.527146652072421e-05,
+      "loss": 1.5284,
+      "step": 7600
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.08372935816698124,
+      "learning_rate": 8.517424045280808e-05,
+      "loss": 1.3019,
+      "step": 7601
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.08514073546626989,
+      "learning_rate": 8.507706468394544e-05,
+      "loss": 1.3655,
+      "step": 7602
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.09309189623499317,
+      "learning_rate": 8.497993922591934e-05,
+      "loss": 1.3931,
+      "step": 7603
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.10140416177523041,
+      "learning_rate": 8.488286409050622e-05,
+      "loss": 1.3656,
+      "step": 7604
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.09603451991030429,
+      "learning_rate": 8.478583928947692e-05,
+      "loss": 1.5697,
+      "step": 7605
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.0977670160586023,
+      "learning_rate": 8.46888648345962e-05,
+      "loss": 1.5035,
+      "step": 7606
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.08551839072309805,
+      "learning_rate": 8.459194073762227e-05,
+      "loss": 1.4294,
+      "step": 7607
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.08037591290539492,
+      "learning_rate": 8.449506701030774e-05,
+      "loss": 1.5657,
+      "step": 7608
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.09590980900355255,
+      "learning_rate": 8.439824366439885e-05,
+      "loss": 1.2933,
+      "step": 7609
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.08713666532241826,
+      "learning_rate": 8.430147071163558e-05,
+      "loss": 1.4531,
+      "step": 7610
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.08992186550468441,
+      "learning_rate": 8.420474816375212e-05,
+      "loss": 1.4069,
+      "step": 7611
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.08338253904896663,
+      "learning_rate": 8.410807603247655e-05,
+      "loss": 1.496,
+      "step": 7612
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.09154435310147199,
+      "learning_rate": 8.401145432953045e-05,
+      "loss": 1.4009,
+      "step": 7613
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.11182926553803102,
+      "learning_rate": 8.391488306662976e-05,
+      "loss": 1.4552,
+      "step": 7614
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.0862731974075149,
+      "learning_rate": 8.381836225548383e-05,
+      "loss": 1.2472,
+      "step": 7615
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.08422895721651501,
+      "learning_rate": 8.372189190779639e-05,
+      "loss": 1.439,
+      "step": 7616
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.09190535096534928,
+      "learning_rate": 8.362547203526455e-05,
+      "loss": 1.3751,
+      "step": 7617
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.08301888840775173,
+      "learning_rate": 8.352910264957969e-05,
+      "loss": 1.351,
+      "step": 7618
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.09395473446856016,
+      "learning_rate": 8.343278376242708e-05,
+      "loss": 1.426,
+      "step": 7619
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.09327444144541883,
+      "learning_rate": 8.333651538548542e-05,
+      "loss": 1.4021,
+      "step": 7620
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.09002012453941054,
+      "learning_rate": 8.324029753042773e-05,
+      "loss": 1.4536,
+      "step": 7621
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.11071663575181982,
+      "learning_rate": 8.314413020892086e-05,
+      "loss": 1.4495,
+      "step": 7622
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.0961113671788749,
+      "learning_rate": 8.304801343262525e-05,
+      "loss": 1.4342,
+      "step": 7623
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.09501998230845707,
+      "learning_rate": 8.295194721319555e-05,
+      "loss": 1.2294,
+      "step": 7624
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.08577388799800784,
+      "learning_rate": 8.285593156228005e-05,
+      "loss": 1.4448,
+      "step": 7625
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.08585778313120833,
+      "learning_rate": 8.275996649152085e-05,
+      "loss": 1.4738,
+      "step": 7626
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.09073679170236165,
+      "learning_rate": 8.266405201255411e-05,
+      "loss": 1.3423,
+      "step": 7627
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.08366332122612558,
+      "learning_rate": 8.256818813701e-05,
+      "loss": 1.3645,
+      "step": 7628
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.1010036314167468,
+      "learning_rate": 8.247237487651204e-05,
+      "loss": 1.2581,
+      "step": 7629
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.08561960263147947,
+      "learning_rate": 8.237661224267806e-05,
+      "loss": 1.3947,
+      "step": 7630
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.08579164829357397,
+      "learning_rate": 8.228090024711976e-05,
+      "loss": 1.585,
+      "step": 7631
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.07982612565350453,
+      "learning_rate": 8.218523890144225e-05,
+      "loss": 1.3644,
+      "step": 7632
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.09154428443060891,
+      "learning_rate": 8.208962821724497e-05,
+      "loss": 1.4524,
+      "step": 7633
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.10977900239432357,
+      "learning_rate": 8.199406820612116e-05,
+      "loss": 1.4807,
+      "step": 7634
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.087843255682426,
+      "learning_rate": 8.189855887965763e-05,
+      "loss": 1.3473,
+      "step": 7635
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.0944862488161355,
+      "learning_rate": 8.180310024943515e-05,
+      "loss": 1.4448,
+      "step": 7636
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.08936470015321014,
+      "learning_rate": 8.170769232702868e-05,
+      "loss": 1.4059,
+      "step": 7637
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.0942866875546708,
+      "learning_rate": 8.16123351240064e-05,
+      "loss": 1.46,
+      "step": 7638
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.09433441619949548,
+      "learning_rate": 8.15170286519309e-05,
+      "loss": 1.3914,
+      "step": 7639
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.08536000362791818,
+      "learning_rate": 8.142177292235853e-05,
+      "loss": 1.4205,
+      "step": 7640
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.1131150345766374,
+      "learning_rate": 8.132656794683912e-05,
+      "loss": 1.3373,
+      "step": 7641
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.09912843864926806,
+      "learning_rate": 8.123141373691678e-05,
+      "loss": 1.4338,
+      "step": 7642
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.09231463345028104,
+      "learning_rate": 8.113631030412932e-05,
+      "loss": 1.4434,
+      "step": 7643
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.09127964871657997,
+      "learning_rate": 8.104125766000814e-05,
+      "loss": 1.4054,
+      "step": 7644
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.09058060725166724,
+      "learning_rate": 8.094625581607901e-05,
+      "loss": 1.3209,
+      "step": 7645
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.09059867682124947,
+      "learning_rate": 8.085130478386088e-05,
+      "loss": 1.4915,
+      "step": 7646
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.08947413283209918,
+      "learning_rate": 8.075640457486722e-05,
+      "loss": 1.2977,
+      "step": 7647
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.08459648939665679,
+      "learning_rate": 8.066155520060475e-05,
+      "loss": 1.2952,
+      "step": 7648
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.08603049389030244,
+      "learning_rate": 8.05667566725744e-05,
+      "loss": 1.378,
+      "step": 7649
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.0853412982084942,
+      "learning_rate": 8.047200900227092e-05,
+      "loss": 1.3794,
+      "step": 7650
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.10203544145492183,
+      "learning_rate": 8.037731220118256e-05,
+      "loss": 1.4407,
+      "step": 7651
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.09023196890129039,
+      "learning_rate": 8.028266628079173e-05,
+      "loss": 1.4527,
+      "step": 7652
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.0808947281733063,
+      "learning_rate": 8.018807125257476e-05,
+      "loss": 1.4643,
+      "step": 7653
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.09435279095402682,
+      "learning_rate": 8.009352712800128e-05,
+      "loss": 1.4356,
+      "step": 7654
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.08363094023395468,
+      "learning_rate": 7.999903391853547e-05,
+      "loss": 1.4149,
+      "step": 7655
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.09201333593543766,
+      "learning_rate": 7.990459163563474e-05,
+      "loss": 1.4506,
+      "step": 7656
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.08267699262829642,
+      "learning_rate": 7.981020029075043e-05,
+      "loss": 1.3957,
+      "step": 7657
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.09365508246569157,
+      "learning_rate": 7.971585989532792e-05,
+      "loss": 1.2511,
+      "step": 7658
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.08658708993034162,
+      "learning_rate": 7.962157046080648e-05,
+      "loss": 1.4408,
+      "step": 7659
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.09438322216968405,
+      "learning_rate": 7.952733199861878e-05,
+      "loss": 1.439,
+      "step": 7660
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.0874301144366475,
+      "learning_rate": 7.943314452019168e-05,
+      "loss": 1.4711,
+      "step": 7661
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.08256679231136793,
+      "learning_rate": 7.933900803694576e-05,
+      "loss": 1.4718,
+      "step": 7662
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.08325944584328684,
+      "learning_rate": 7.92449225602953e-05,
+      "loss": 1.3529,
+      "step": 7663
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.08691118535626666,
+      "learning_rate": 7.915088810164855e-05,
+      "loss": 1.4396,
+      "step": 7664
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.08063120692409934,
+      "learning_rate": 7.905690467240762e-05,
+      "loss": 1.5015,
+      "step": 7665
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.080837406638755,
+      "learning_rate": 7.89629722839682e-05,
+      "loss": 1.2455,
+      "step": 7666
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.08510788864340407,
+      "learning_rate": 7.886909094771982e-05,
+      "loss": 1.3956,
+      "step": 7667
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.08059477667817433,
+      "learning_rate": 7.877526067504614e-05,
+      "loss": 1.4848,
+      "step": 7668
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.08384953530723206,
+      "learning_rate": 7.868148147732413e-05,
+      "loss": 1.3094,
+      "step": 7669
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.08851835803646642,
+      "learning_rate": 7.858775336592505e-05,
+      "loss": 1.3807,
+      "step": 7670
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.09268354025660072,
+      "learning_rate": 7.849407635221379e-05,
+      "loss": 1.4575,
+      "step": 7671
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.07953833744699162,
+      "learning_rate": 7.840045044754879e-05,
+      "loss": 1.4784,
+      "step": 7672
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.09061781935977607,
+      "learning_rate": 7.830687566328265e-05,
+      "loss": 1.493,
+      "step": 7673
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.09061908733465755,
+      "learning_rate": 7.82133520107618e-05,
+      "loss": 1.2913,
+      "step": 7674
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.08607373087290072,
+      "learning_rate": 7.811987950132599e-05,
+      "loss": 1.2954,
+      "step": 7675
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.09028855772629082,
+      "learning_rate": 7.802645814630921e-05,
+      "loss": 1.3465,
+      "step": 7676
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.08628410645460714,
+      "learning_rate": 7.793308795703929e-05,
+      "loss": 1.3811,
+      "step": 7677
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.09710158064190848,
+      "learning_rate": 7.783976894483752e-05,
+      "loss": 1.4639,
+      "step": 7678
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.08846974972943605,
+      "learning_rate": 7.774650112101911e-05,
+      "loss": 1.3841,
+      "step": 7679
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.07809311245800762,
+      "learning_rate": 7.765328449689312e-05,
+      "loss": 1.5403,
+      "step": 7680
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.08931552275490337,
+      "learning_rate": 7.756011908376254e-05,
+      "loss": 1.5256,
+      "step": 7681
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.0891463968433547,
+      "learning_rate": 7.74670048929238e-05,
+      "loss": 1.3094,
+      "step": 7682
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.08977675661275883,
+      "learning_rate": 7.737394193566744e-05,
+      "loss": 1.4532,
+      "step": 7683
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.08825867071922235,
+      "learning_rate": 7.72809302232777e-05,
+      "loss": 1.3373,
+      "step": 7684
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.09372314922944172,
+      "learning_rate": 7.71879697670324e-05,
+      "loss": 1.3349,
+      "step": 7685
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.09768865167003908,
+      "learning_rate": 7.709506057820343e-05,
+      "loss": 1.4697,
+      "step": 7686
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.08895694737402585,
+      "learning_rate": 7.700220266805647e-05,
+      "loss": 1.3731,
+      "step": 7687
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.09179791236859439,
+      "learning_rate": 7.690939604785069e-05,
+      "loss": 1.4724,
+      "step": 7688
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.0897147299222315,
+      "learning_rate": 7.681664072883914e-05,
+      "loss": 1.4539,
+      "step": 7689
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.09394049537942313,
+      "learning_rate": 7.672393672226901e-05,
+      "loss": 1.2418,
+      "step": 7690
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.08350359871804075,
+      "learning_rate": 7.663128403938063e-05,
+      "loss": 1.5139,
+      "step": 7691
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.08732287161011305,
+      "learning_rate": 7.653868269140869e-05,
+      "loss": 1.306,
+      "step": 7692
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.0896616196238533,
+      "learning_rate": 7.644613268958144e-05,
+      "loss": 1.3701,
+      "step": 7693
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.08764433658911641,
+      "learning_rate": 7.635363404512069e-05,
+      "loss": 1.4636,
+      "step": 7694
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1030645066169904,
+      "learning_rate": 7.626118676924237e-05,
+      "loss": 1.2745,
+      "step": 7695
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.09333495190715421,
+      "learning_rate": 7.616879087315614e-05,
+      "loss": 1.3622,
+      "step": 7696
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.10195538862682457,
+      "learning_rate": 7.607644636806505e-05,
+      "loss": 1.4147,
+      "step": 7697
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.08923282440460135,
+      "learning_rate": 7.598415326516638e-05,
+      "loss": 1.4762,
+      "step": 7698
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.08931760167995045,
+      "learning_rate": 7.5891911575651e-05,
+      "loss": 1.3004,
+      "step": 7699
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.08938475556208395,
+      "learning_rate": 7.579972131070334e-05,
+      "loss": 1.2865,
+      "step": 7700
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.07974469882035637,
+      "learning_rate": 7.570758248150194e-05,
+      "loss": 1.32,
+      "step": 7701
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.11124565390940556,
+      "learning_rate": 7.5615495099219e-05,
+      "loss": 1.3306,
+      "step": 7702
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.09465556704118003,
+      "learning_rate": 7.552345917502023e-05,
+      "loss": 1.4847,
+      "step": 7703
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.08633690644820705,
+      "learning_rate": 7.543147472006545e-05,
+      "loss": 1.3595,
+      "step": 7704
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.07694845369969425,
+      "learning_rate": 7.533954174550817e-05,
+      "loss": 1.4188,
+      "step": 7705
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.0888851522271176,
+      "learning_rate": 7.524766026249536e-05,
+      "loss": 1.4088,
+      "step": 7706
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.08604335226277512,
+      "learning_rate": 7.51558302821681e-05,
+      "loss": 1.3517,
+      "step": 7707
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.0916267922007084,
+      "learning_rate": 7.506405181566123e-05,
+      "loss": 1.507,
+      "step": 7708
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.0858819386378702,
+      "learning_rate": 7.497232487410299e-05,
+      "loss": 1.4484,
+      "step": 7709
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.08443971498772733,
+      "learning_rate": 7.488064946861555e-05,
+      "loss": 1.3468,
+      "step": 7710
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.08820651296219477,
+      "learning_rate": 7.478902561031503e-05,
+      "loss": 1.3766,
+      "step": 7711
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.08569244916818314,
+      "learning_rate": 7.469745331031114e-05,
+      "loss": 1.4528,
+      "step": 7712
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.09377223486281296,
+      "learning_rate": 7.460593257970716e-05,
+      "loss": 1.426,
+      "step": 7713
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.09446187498251976,
+      "learning_rate": 7.451446342960044e-05,
+      "loss": 1.3474,
+      "step": 7714
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.08327545046965788,
+      "learning_rate": 7.4423045871082e-05,
+      "loss": 1.3567,
+      "step": 7715
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.08832129555517432,
+      "learning_rate": 7.433167991523631e-05,
+      "loss": 1.4224,
+      "step": 7716
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.08522053430819553,
+      "learning_rate": 7.424036557314196e-05,
+      "loss": 1.5376,
+      "step": 7717
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.09278789708741045,
+      "learning_rate": 7.414910285587117e-05,
+      "loss": 1.4763,
+      "step": 7718
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.09264503427204793,
+      "learning_rate": 7.405789177448985e-05,
+      "loss": 1.2921,
+      "step": 7719
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.0856871779296312,
+      "learning_rate": 7.39667323400574e-05,
+      "loss": 1.3306,
+      "step": 7720
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.08823520611675048,
+      "learning_rate": 7.387562456362762e-05,
+      "loss": 1.345,
+      "step": 7721
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.08796941351682677,
+      "learning_rate": 7.378456845624725e-05,
+      "loss": 1.5243,
+      "step": 7722
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.08979097920858628,
+      "learning_rate": 7.369356402895738e-05,
+      "loss": 1.3536,
+      "step": 7723
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.08436809874166219,
+      "learning_rate": 7.360261129279272e-05,
+      "loss": 1.3238,
+      "step": 7724
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.10352532323355235,
+      "learning_rate": 7.351171025878128e-05,
+      "loss": 1.4435,
+      "step": 7725
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.07711372998127677,
+      "learning_rate": 7.342086093794531e-05,
+      "loss": 1.4107,
+      "step": 7726
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.09640807472724792,
+      "learning_rate": 7.333006334130077e-05,
+      "loss": 1.3328,
+      "step": 7727
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.08914874349546113,
+      "learning_rate": 7.32393174798568e-05,
+      "loss": 1.3379,
+      "step": 7728
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.08496885016196233,
+      "learning_rate": 7.314862336461692e-05,
+      "loss": 1.3343,
+      "step": 7729
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.07735334418737376,
+      "learning_rate": 7.305798100657823e-05,
+      "loss": 1.4266,
+      "step": 7730
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.08938125918239577,
+      "learning_rate": 7.2967390416731e-05,
+      "loss": 1.4782,
+      "step": 7731
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.09278179808639771,
+      "learning_rate": 7.287685160605995e-05,
+      "loss": 1.3485,
+      "step": 7732
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.08403083852248348,
+      "learning_rate": 7.278636458554322e-05,
+      "loss": 1.4589,
+      "step": 7733
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.09068179591128482,
+      "learning_rate": 7.26959293661525e-05,
+      "loss": 1.5567,
+      "step": 7734
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.0762056740331929,
+      "learning_rate": 7.26055459588535e-05,
+      "loss": 1.2854,
+      "step": 7735
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.07528963922741068,
+      "learning_rate": 7.251521437460562e-05,
+      "loss": 1.3909,
+      "step": 7736
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.0942168001290263,
+      "learning_rate": 7.242493462436161e-05,
+      "loss": 1.2895,
+      "step": 7737
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.10240321129164565,
+      "learning_rate": 7.233470671906844e-05,
+      "loss": 1.2489,
+      "step": 7738
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.08504284002213647,
+      "learning_rate": 7.224453066966652e-05,
+      "loss": 1.3601,
+      "step": 7739
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.09587235474791764,
+      "learning_rate": 7.215440648708982e-05,
+      "loss": 1.4656,
+      "step": 7740
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.08536226344078945,
+      "learning_rate": 7.206433418226649e-05,
+      "loss": 1.4795,
+      "step": 7741
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.09140896762573765,
+      "learning_rate": 7.197431376611785e-05,
+      "loss": 1.4325,
+      "step": 7742
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.09015041988015379,
+      "learning_rate": 7.188434524955939e-05,
+      "loss": 1.4109,
+      "step": 7743
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.08271750495957778,
+      "learning_rate": 7.179442864349988e-05,
+      "loss": 1.4846,
+      "step": 7744
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.08829217021420534,
+      "learning_rate": 7.170456395884217e-05,
+      "loss": 1.4766,
+      "step": 7745
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.08833891796417223,
+      "learning_rate": 7.161475120648276e-05,
+      "loss": 1.3419,
+      "step": 7746
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.08900580755952446,
+      "learning_rate": 7.152499039731153e-05,
+      "loss": 1.5181,
+      "step": 7747
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.0860664905985196,
+      "learning_rate": 7.143528154221235e-05,
+      "loss": 1.4304,
+      "step": 7748
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.08884608101931617,
+      "learning_rate": 7.13456246520629e-05,
+      "loss": 1.4872,
+      "step": 7749
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.08909893472267622,
+      "learning_rate": 7.125601973773416e-05,
+      "loss": 1.3473,
+      "step": 7750
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.08642569234811122,
+      "learning_rate": 7.11664668100912e-05,
+      "loss": 1.5026,
+      "step": 7751
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.0867109188311294,
+      "learning_rate": 7.107696587999263e-05,
+      "loss": 1.3332,
+      "step": 7752
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.07775756140298237,
+      "learning_rate": 7.098751695829048e-05,
+      "loss": 1.2844,
+      "step": 7753
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.08664325787505547,
+      "learning_rate": 7.089812005583096e-05,
+      "loss": 1.4412,
+      "step": 7754
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.08714207866053697,
+      "learning_rate": 7.080877518345385e-05,
+      "loss": 1.5046,
+      "step": 7755
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.09008286495715372,
+      "learning_rate": 7.071948235199228e-05,
+      "loss": 1.4347,
+      "step": 7756
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.09963396647707887,
+      "learning_rate": 7.063024157227344e-05,
+      "loss": 1.411,
+      "step": 7757
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.08125783843505932,
+      "learning_rate": 7.054105285511814e-05,
+      "loss": 1.4923,
+      "step": 7758
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.09825938535932652,
+      "learning_rate": 7.045191621134068e-05,
+      "loss": 1.4293,
+      "step": 7759
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.08176546166160557,
+      "learning_rate": 7.036283165174923e-05,
+      "loss": 1.4709,
+      "step": 7760
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.07799311174622484,
+      "learning_rate": 7.027379918714577e-05,
+      "loss": 1.3462,
+      "step": 7761
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.08215288669420732,
+      "learning_rate": 7.018481882832561e-05,
+      "loss": 1.3547,
+      "step": 7762
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.0971949111065358,
+      "learning_rate": 7.00958905860779e-05,
+      "loss": 1.3067,
+      "step": 7763
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.0810382090903955,
+      "learning_rate": 7.000701447118563e-05,
+      "loss": 1.4154,
+      "step": 7764
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.08624335475689453,
+      "learning_rate": 6.991819049442516e-05,
+      "loss": 1.4382,
+      "step": 7765
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.0923274614359002,
+      "learning_rate": 6.982941866656684e-05,
+      "loss": 1.4152,
+      "step": 7766
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.08995040104069521,
+      "learning_rate": 6.974069899837465e-05,
+      "loss": 1.5865,
+      "step": 7767
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.0899147874974617,
+      "learning_rate": 6.965203150060589e-05,
+      "loss": 1.5105,
+      "step": 7768
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.0835013673745431,
+      "learning_rate": 6.956341618401196e-05,
+      "loss": 1.3654,
+      "step": 7769
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.09350769325031699,
+      "learning_rate": 6.947485305933787e-05,
+      "loss": 1.3866,
+      "step": 7770
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.07761428841872348,
+      "learning_rate": 6.938634213732198e-05,
+      "loss": 1.3952,
+      "step": 7771
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.10393913360882924,
+      "learning_rate": 6.929788342869676e-05,
+      "loss": 1.4183,
+      "step": 7772
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.09154838170195079,
+      "learning_rate": 6.920947694418789e-05,
+      "loss": 1.2753,
+      "step": 7773
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.09637707494862725,
+      "learning_rate": 6.912112269451526e-05,
+      "loss": 1.3974,
+      "step": 7774
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.09508845670840485,
+      "learning_rate": 6.903282069039185e-05,
+      "loss": 1.4752,
+      "step": 7775
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.11155591005942683,
+      "learning_rate": 6.894457094252471e-05,
+      "loss": 1.3607,
+      "step": 7776
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.10317484976555591,
+      "learning_rate": 6.885637346161449e-05,
+      "loss": 1.3526,
+      "step": 7777
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.08389106407279763,
+      "learning_rate": 6.876822825835527e-05,
+      "loss": 1.2543,
+      "step": 7778
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.09115262148193395,
+      "learning_rate": 6.86801353434351e-05,
+      "loss": 1.422,
+      "step": 7779
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.0919312396699346,
+      "learning_rate": 6.859209472753559e-05,
+      "loss": 1.3834,
+      "step": 7780
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.0771367133563575,
+      "learning_rate": 6.850410642133176e-05,
+      "loss": 1.3485,
+      "step": 7781
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.08852606976685708,
+      "learning_rate": 6.841617043549259e-05,
+      "loss": 1.337,
+      "step": 7782
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.09033201320751114,
+      "learning_rate": 6.832828678068092e-05,
+      "loss": 1.4328,
+      "step": 7783
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.08586344860911486,
+      "learning_rate": 6.824045546755242e-05,
+      "loss": 1.5029,
+      "step": 7784
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.09016545117776988,
+      "learning_rate": 6.815267650675721e-05,
+      "loss": 1.5326,
+      "step": 7785
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.09318420049557376,
+      "learning_rate": 6.80649499089389e-05,
+      "loss": 1.3503,
+      "step": 7786
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.09915799008577442,
+      "learning_rate": 6.797727568473444e-05,
+      "loss": 1.4098,
+      "step": 7787
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.0998684888213455,
+      "learning_rate": 6.788965384477463e-05,
+      "loss": 1.3458,
+      "step": 7788
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.09354345098880852,
+      "learning_rate": 6.780208439968417e-05,
+      "loss": 1.4528,
+      "step": 7789
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.10346740823288775,
+      "learning_rate": 6.771456736008086e-05,
+      "loss": 1.4361,
+      "step": 7790
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.08885901539295292,
+      "learning_rate": 6.762710273657658e-05,
+      "loss": 1.389,
+      "step": 7791
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.09811897324623484,
+      "learning_rate": 6.753969053977683e-05,
+      "loss": 1.4192,
+      "step": 7792
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.08673617423121921,
+      "learning_rate": 6.745233078028041e-05,
+      "loss": 1.2513,
+      "step": 7793
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.0912880836237642,
+      "learning_rate": 6.736502346868018e-05,
+      "loss": 1.2771,
+      "step": 7794
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.08520826836546021,
+      "learning_rate": 6.727776861556239e-05,
+      "loss": 1.2978,
+      "step": 7795
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.09253343356505377,
+      "learning_rate": 6.719056623150687e-05,
+      "loss": 1.4179,
+      "step": 7796
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.09727907298470896,
+      "learning_rate": 6.710341632708733e-05,
+      "loss": 1.2541,
+      "step": 7797
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.10455922957410266,
+      "learning_rate": 6.701631891287108e-05,
+      "loss": 1.4294,
+      "step": 7798
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.09565500343844599,
+      "learning_rate": 6.692927399941878e-05,
+      "loss": 1.3228,
+      "step": 7799
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.09670216486916945,
+      "learning_rate": 6.684228159728501e-05,
+      "loss": 1.337,
+      "step": 7800
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.09031993166332863,
+      "learning_rate": 6.675534171701802e-05,
+      "loss": 1.4904,
+      "step": 7801
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1106867657427879,
+      "learning_rate": 6.666845436915941e-05,
+      "loss": 1.4883,
+      "step": 7802
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.08750518151251388,
+      "learning_rate": 6.658161956424457e-05,
+      "loss": 1.5734,
+      "step": 7803
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.09277619307591722,
+      "learning_rate": 6.64948373128027e-05,
+      "loss": 1.4289,
+      "step": 7804
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.09255864978715336,
+      "learning_rate": 6.640810762535637e-05,
+      "loss": 1.4721,
+      "step": 7805
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.08307761174818809,
+      "learning_rate": 6.632143051242168e-05,
+      "loss": 1.3225,
+      "step": 7806
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.08957267637082313,
+      "learning_rate": 6.623480598450866e-05,
+      "loss": 1.262,
+      "step": 7807
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.10565226536610126,
+      "learning_rate": 6.61482340521209e-05,
+      "loss": 1.3611,
+      "step": 7808
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.09842970799267935,
+      "learning_rate": 6.606171472575539e-05,
+      "loss": 1.3758,
+      "step": 7809
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.09138816156285817,
+      "learning_rate": 6.597524801590299e-05,
+      "loss": 1.3514,
+      "step": 7810
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.08304644485072452,
+      "learning_rate": 6.588883393304818e-05,
+      "loss": 1.4504,
+      "step": 7811
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.08179421892153053,
+      "learning_rate": 6.580247248766875e-05,
+      "loss": 1.4556,
+      "step": 7812
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.0905917159466832,
+      "learning_rate": 6.571616369023642e-05,
+      "loss": 1.343,
+      "step": 7813
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.08782451845628722,
+      "learning_rate": 6.562990755121661e-05,
+      "loss": 1.4379,
+      "step": 7814
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.08681317029442595,
+      "learning_rate": 6.554370408106796e-05,
+      "loss": 1.3353,
+      "step": 7815
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.09018112284327645,
+      "learning_rate": 6.545755329024295e-05,
+      "loss": 1.4735,
+      "step": 7816
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.08596032825622608,
+      "learning_rate": 6.537145518918774e-05,
+      "loss": 1.4837,
+      "step": 7817
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.0842586560156261,
+      "learning_rate": 6.528540978834191e-05,
+      "loss": 1.3523,
+      "step": 7818
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.09000204358620255,
+      "learning_rate": 6.519941709813881e-05,
+      "loss": 1.3803,
+      "step": 7819
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.09290775564210622,
+      "learning_rate": 6.511347712900545e-05,
+      "loss": 1.375,
+      "step": 7820
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.09291821777416866,
+      "learning_rate": 6.502758989136221e-05,
+      "loss": 1.3519,
+      "step": 7821
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.10079947970251664,
+      "learning_rate": 6.494175539562325e-05,
+      "loss": 1.4085,
+      "step": 7822
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.08932330902539856,
+      "learning_rate": 6.485597365219648e-05,
+      "loss": 1.4194,
+      "step": 7823
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.08012412657276556,
+      "learning_rate": 6.477024467148291e-05,
+      "loss": 1.4567,
+      "step": 7824
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.08880866528917783,
+      "learning_rate": 6.468456846387777e-05,
+      "loss": 1.3814,
+      "step": 7825
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.08893574046009894,
+      "learning_rate": 6.459894503976943e-05,
+      "loss": 1.3846,
+      "step": 7826
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.08956078096009155,
+      "learning_rate": 6.451337440953998e-05,
+      "loss": 1.3439,
+      "step": 7827
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.10529417583896636,
+      "learning_rate": 6.442785658356525e-05,
+      "loss": 1.4773,
+      "step": 7828
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.09718365997826772,
+      "learning_rate": 6.434239157221466e-05,
+      "loss": 1.3271,
+      "step": 7829
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.08864347101686207,
+      "learning_rate": 6.42569793858509e-05,
+      "loss": 1.3864,
+      "step": 7830
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.09400814744697399,
+      "learning_rate": 6.417162003483063e-05,
+      "loss": 1.5692,
+      "step": 7831
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.08719216814926357,
+      "learning_rate": 6.408631352950406e-05,
+      "loss": 1.3842,
+      "step": 7832
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.07677562030215503,
+      "learning_rate": 6.40010598802147e-05,
+      "loss": 1.3368,
+      "step": 7833
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.07755566060205164,
+      "learning_rate": 6.391585909729997e-05,
+      "loss": 1.4292,
+      "step": 7834
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.08532107826874637,
+      "learning_rate": 6.383071119109079e-05,
+      "loss": 1.4152,
+      "step": 7835
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.0802070942760828,
+      "learning_rate": 6.374561617191166e-05,
+      "loss": 1.3902,
+      "step": 7836
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.07747124635800962,
+      "learning_rate": 6.366057405008042e-05,
+      "loss": 1.1828,
+      "step": 7837
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.09197460885363941,
+      "learning_rate": 6.357558483590887e-05,
+      "loss": 1.2937,
+      "step": 7838
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.08686641796307093,
+      "learning_rate": 6.349064853970232e-05,
+      "loss": 1.4106,
+      "step": 7839
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.09014370721732225,
+      "learning_rate": 6.340576517175939e-05,
+      "loss": 1.4875,
+      "step": 7840
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.08256481623816217,
+      "learning_rate": 6.332093474237265e-05,
+      "loss": 1.3485,
+      "step": 7841
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.08533178161354073,
+      "learning_rate": 6.323615726182813e-05,
+      "loss": 1.4926,
+      "step": 7842
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.09430168667379107,
+      "learning_rate": 6.315143274040519e-05,
+      "loss": 1.4384,
+      "step": 7843
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.11100609742165347,
+      "learning_rate": 6.306676118837707e-05,
+      "loss": 1.365,
+      "step": 7844
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.08113856887278259,
+      "learning_rate": 6.29821426160106e-05,
+      "loss": 1.3599,
+      "step": 7845
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.0783171748843241,
+      "learning_rate": 6.289757703356597e-05,
+      "loss": 1.3204,
+      "step": 7846
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.10432054223441194,
+      "learning_rate": 6.281306445129697e-05,
+      "loss": 1.4898,
+      "step": 7847
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.08632182749912745,
+      "learning_rate": 6.27286048794512e-05,
+      "loss": 1.3361,
+      "step": 7848
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.09476579758108704,
+      "learning_rate": 6.264419832826945e-05,
+      "loss": 1.425,
+      "step": 7849
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.08215841019664528,
+      "learning_rate": 6.255984480798649e-05,
+      "loss": 1.4258,
+      "step": 7850
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.08563993560179116,
+      "learning_rate": 6.247554432883052e-05,
+      "loss": 1.3825,
+      "step": 7851
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.08548292176570288,
+      "learning_rate": 6.239129690102307e-05,
+      "loss": 1.3595,
+      "step": 7852
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.07864817887270605,
+      "learning_rate": 6.230710253477956e-05,
+      "loss": 1.4026,
+      "step": 7853
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.07912513593576152,
+      "learning_rate": 6.222296124030891e-05,
+      "loss": 1.4491,
+      "step": 7854
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.0775400683256103,
+      "learning_rate": 6.213887302781335e-05,
+      "loss": 1.3045,
+      "step": 7855
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.08471158998301948,
+      "learning_rate": 6.2054837907489e-05,
+      "loss": 1.3708,
+      "step": 7856
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.07866617830977912,
+      "learning_rate": 6.197085588952556e-05,
+      "loss": 1.379,
+      "step": 7857
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.08669780094760313,
+      "learning_rate": 6.188692698410575e-05,
+      "loss": 1.2851,
+      "step": 7858
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.09045076299727618,
+      "learning_rate": 6.180305120140644e-05,
+      "loss": 1.3331,
+      "step": 7859
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.0852228463127037,
+      "learning_rate": 6.1719228551598e-05,
+      "loss": 1.4925,
+      "step": 7860
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.08271086957993047,
+      "learning_rate": 6.163545904484397e-05,
+      "loss": 1.3317,
+      "step": 7861
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.07976940004119311,
+      "learning_rate": 6.155174269130182e-05,
+      "loss": 1.37,
+      "step": 7862
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.07844337767876314,
+      "learning_rate": 6.146807950112248e-05,
+      "loss": 1.5441,
+      "step": 7863
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.08755021665953842,
+      "learning_rate": 6.13844694844503e-05,
+      "loss": 1.3235,
+      "step": 7864
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.11060847383411719,
+      "learning_rate": 6.130091265142329e-05,
+      "loss": 1.424,
+      "step": 7865
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.0794293051223916,
+      "learning_rate": 6.12174090121732e-05,
+      "loss": 1.4387,
+      "step": 7866
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.08712838702078128,
+      "learning_rate": 6.113395857682485e-05,
+      "loss": 1.3756,
+      "step": 7867
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.09304835093342179,
+      "learning_rate": 6.105056135549714e-05,
+      "loss": 1.4444,
+      "step": 7868
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.08206255704216997,
+      "learning_rate": 6.096721735830202e-05,
+      "loss": 1.4388,
+      "step": 7869
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.0805377403471389,
+      "learning_rate": 6.0883926595345475e-05,
+      "loss": 1.3054,
+      "step": 7870
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.09047074750357005,
+      "learning_rate": 6.080068907672659e-05,
+      "loss": 1.3322,
+      "step": 7871
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.10712564001781039,
+      "learning_rate": 6.071750481253835e-05,
+      "loss": 1.4827,
+      "step": 7872
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.08000036424175803,
+      "learning_rate": 6.0634373812867185e-05,
+      "loss": 1.4809,
+      "step": 7873
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.08355315667819273,
+      "learning_rate": 6.055129608779275e-05,
+      "loss": 1.3978,
+      "step": 7874
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.0897445519805741,
+      "learning_rate": 6.046827164738872e-05,
+      "loss": 1.4598,
+      "step": 7875
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.08593164499884852,
+      "learning_rate": 6.03853005017222e-05,
+      "loss": 1.2729,
+      "step": 7876
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.09280406295835326,
+      "learning_rate": 6.0302382660853416e-05,
+      "loss": 1.3568,
+      "step": 7877
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.09287445817382776,
+      "learning_rate": 6.0219518134836726e-05,
+      "loss": 1.3667,
+      "step": 7878
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.08942747992632392,
+      "learning_rate": 6.0136706933719586e-05,
+      "loss": 1.3309,
+      "step": 7879
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.09044881308433124,
+      "learning_rate": 6.0053949067543077e-05,
+      "loss": 1.333,
+      "step": 7880
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.08218485919758337,
+      "learning_rate": 5.997124454634201e-05,
+      "loss": 1.5086,
+      "step": 7881
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.08150958759929276,
+      "learning_rate": 5.9888593380144574e-05,
+      "loss": 1.4831,
+      "step": 7882
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.09238224876183417,
+      "learning_rate": 5.9805995578972426e-05,
+      "loss": 1.3159,
+      "step": 7883
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.08132944733289485,
+      "learning_rate": 5.972345115284095e-05,
+      "loss": 1.4739,
+      "step": 7884
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.09200855967945543,
+      "learning_rate": 5.96409601117589e-05,
+      "loss": 1.3772,
+      "step": 7885
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.10121485467365761,
+      "learning_rate": 5.955852246572851e-05,
+      "loss": 1.317,
+      "step": 7886
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.08068215984602292,
+      "learning_rate": 5.9476138224745734e-05,
+      "loss": 1.3589,
+      "step": 7887
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.08866123553653318,
+      "learning_rate": 5.939380739880002e-05,
+      "loss": 1.4508,
+      "step": 7888
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.08589646053077514,
+      "learning_rate": 5.9311529997874095e-05,
+      "loss": 1.3438,
+      "step": 7889
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.09725805400582763,
+      "learning_rate": 5.922930603194437e-05,
+      "loss": 1.4084,
+      "step": 7890
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.09455544461619558,
+      "learning_rate": 5.9147135510981e-05,
+      "loss": 1.4086,
+      "step": 7891
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.10467580678385532,
+      "learning_rate": 5.906501844494716e-05,
+      "loss": 1.3407,
+      "step": 7892
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.07795243427227122,
+      "learning_rate": 5.898295484379995e-05,
+      "loss": 1.3612,
+      "step": 7893
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1038394375264948,
+      "learning_rate": 5.890094471749002e-05,
+      "loss": 1.4229,
+      "step": 7894
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.09720640404759297,
+      "learning_rate": 5.881898807596114e-05,
+      "loss": 1.3402,
+      "step": 7895
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.07723043481387958,
+      "learning_rate": 5.873708492915092e-05,
+      "loss": 1.4209,
+      "step": 7896
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.09067326993878297,
+      "learning_rate": 5.8655235286990513e-05,
+      "loss": 1.4408,
+      "step": 7897
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.09774190841467567,
+      "learning_rate": 5.857343915940433e-05,
+      "loss": 1.452,
+      "step": 7898
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.09164211233774362,
+      "learning_rate": 5.8491696556310535e-05,
+      "loss": 1.4819,
+      "step": 7899
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.08938979443232763,
+      "learning_rate": 5.841000748762054e-05,
+      "loss": 1.2762,
+      "step": 7900
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.08276635984258755,
+      "learning_rate": 5.832837196323964e-05,
+      "loss": 1.4339,
+      "step": 7901
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.10140530865368988,
+      "learning_rate": 5.824678999306621e-05,
+      "loss": 1.4236,
+      "step": 7902
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.08952137652557425,
+      "learning_rate": 5.816526158699243e-05,
+      "loss": 1.3607,
+      "step": 7903
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.08598514612393836,
+      "learning_rate": 5.8083786754904e-05,
+      "loss": 1.3682,
+      "step": 7904
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.08180567616052078,
+      "learning_rate": 5.800236550667992e-05,
+      "loss": 1.3125,
+      "step": 7905
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.08795388382397808,
+      "learning_rate": 5.7920997852192735e-05,
+      "loss": 1.44,
+      "step": 7906
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.07932353663760587,
+      "learning_rate": 5.7839683801308775e-05,
+      "loss": 1.3103,
+      "step": 7907
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.08941791390098239,
+      "learning_rate": 5.775842336388742e-05,
+      "loss": 1.2789,
+      "step": 7908
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.0943055727960341,
+      "learning_rate": 5.767721654978186e-05,
+      "loss": 1.4448,
+      "step": 7909
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.08605694259527719,
+      "learning_rate": 5.759606336883893e-05,
+      "loss": 1.3103,
+      "step": 7910
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.09528400159248888,
+      "learning_rate": 5.7514963830898324e-05,
+      "loss": 1.4157,
+      "step": 7911
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.0838022572347359,
+      "learning_rate": 5.7433917945793776e-05,
+      "loss": 1.3387,
+      "step": 7912
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.11424160907650843,
+      "learning_rate": 5.73529257233526e-05,
+      "loss": 1.3327,
+      "step": 7913
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.09560274070572669,
+      "learning_rate": 5.72719871733951e-05,
+      "loss": 1.4464,
+      "step": 7914
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.08509483829820473,
+      "learning_rate": 5.719110230573543e-05,
+      "loss": 1.3786,
+      "step": 7915
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.10596939753078925,
+      "learning_rate": 5.7110271130181356e-05,
+      "loss": 1.523,
+      "step": 7916
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.09238716068882286,
+      "learning_rate": 5.7029493656533646e-05,
+      "loss": 1.3509,
+      "step": 7917
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.09918986128963173,
+      "learning_rate": 5.694876989458697e-05,
+      "loss": 1.3125,
+      "step": 7918
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.09547700248487656,
+      "learning_rate": 5.6868099854129493e-05,
+      "loss": 1.4866,
+      "step": 7919
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.10238214475114867,
+      "learning_rate": 5.678748354494251e-05,
+      "loss": 1.511,
+      "step": 7920
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.09238523148440415,
+      "learning_rate": 5.6706920976801246e-05,
+      "loss": 1.4352,
+      "step": 7921
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.08533477612675754,
+      "learning_rate": 5.662641215947406e-05,
+      "loss": 1.4433,
+      "step": 7922
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.0923241212621198,
+      "learning_rate": 5.654595710272287e-05,
+      "loss": 1.4694,
+      "step": 7923
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.098711665661809,
+      "learning_rate": 5.6465555816303196e-05,
+      "loss": 1.3371,
+      "step": 7924
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.09113188080489218,
+      "learning_rate": 5.638520830996402e-05,
+      "loss": 1.5073,
+      "step": 7925
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.09294883936559913,
+      "learning_rate": 5.630491459344766e-05,
+      "loss": 1.4731,
+      "step": 7926
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.0958511718237879,
+      "learning_rate": 5.622467467649006e-05,
+      "loss": 1.5133,
+      "step": 7927
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.09150286778743173,
+      "learning_rate": 5.614448856882065e-05,
+      "loss": 1.3905,
+      "step": 7928
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.09388354098673922,
+      "learning_rate": 5.60643562801621e-05,
+      "loss": 1.433,
+      "step": 7929
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1002396530061073,
+      "learning_rate": 5.5984277820230855e-05,
+      "loss": 1.3662,
+      "step": 7930
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.10274461624460365,
+      "learning_rate": 5.5904253198736824e-05,
+      "loss": 1.5483,
+      "step": 7931
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.08530544037612882,
+      "learning_rate": 5.582428242538307e-05,
+      "loss": 1.467,
+      "step": 7932
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.09574392115798785,
+      "learning_rate": 5.574436550986628e-05,
+      "loss": 1.3484,
+      "step": 7933
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.07777008187231625,
+      "learning_rate": 5.5664502461876706e-05,
+      "loss": 1.287,
+      "step": 7934
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.09424269991976277,
+      "learning_rate": 5.558469329109822e-05,
+      "loss": 1.4066,
+      "step": 7935
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.08766616531253907,
+      "learning_rate": 5.550493800720774e-05,
+      "loss": 1.2889,
+      "step": 7936
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.08486592252459829,
+      "learning_rate": 5.5425236619875866e-05,
+      "loss": 1.3519,
+      "step": 7937
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.08143108081202038,
+      "learning_rate": 5.5345589138766864e-05,
+      "loss": 1.4174,
+      "step": 7938
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.08634015059967805,
+      "learning_rate": 5.5265995573538066e-05,
+      "loss": 1.3629,
+      "step": 7939
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.09101367306765057,
+      "learning_rate": 5.5186455933840476e-05,
+      "loss": 1.3984,
+      "step": 7940
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.08940959291798563,
+      "learning_rate": 5.5106970229318765e-05,
+      "loss": 1.3313,
+      "step": 7941
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.08414515779509103,
+      "learning_rate": 5.502753846961067e-05,
+      "loss": 1.387,
+      "step": 7942
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.08772764724912127,
+      "learning_rate": 5.494816066434749e-05,
+      "loss": 1.4057,
+      "step": 7943
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.08757481261207309,
+      "learning_rate": 5.4868836823154246e-05,
+      "loss": 1.2524,
+      "step": 7944
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.09430417292165912,
+      "learning_rate": 5.478956695564902e-05,
+      "loss": 1.4384,
+      "step": 7945
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.0910526631063559,
+      "learning_rate": 5.471035107144373e-05,
+      "loss": 1.4247,
+      "step": 7946
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.09675309496202968,
+      "learning_rate": 5.4631189180143596e-05,
+      "loss": 1.4936,
+      "step": 7947
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.07853317251477367,
+      "learning_rate": 5.4552081291347035e-05,
+      "loss": 1.4164,
+      "step": 7948
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.09942725188704955,
+      "learning_rate": 5.4473027414646384e-05,
+      "loss": 1.5146,
+      "step": 7949
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.08892933014210838,
+      "learning_rate": 5.439402755962719e-05,
+      "loss": 1.4987,
+      "step": 7950
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1075372531467069,
+      "learning_rate": 5.43150817358683e-05,
+      "loss": 1.3922,
+      "step": 7951
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.09669611531183489,
+      "learning_rate": 5.423618995294238e-05,
+      "loss": 1.2791,
+      "step": 7952
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.08838658799987743,
+      "learning_rate": 5.415735222041518e-05,
+      "loss": 1.2887,
+      "step": 7953
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.08859874586580094,
+      "learning_rate": 5.407856854784598e-05,
+      "loss": 1.4271,
+      "step": 7954
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.09346223291532334,
+      "learning_rate": 5.3999838944787713e-05,
+      "loss": 1.3762,
+      "step": 7955
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.08402497812656093,
+      "learning_rate": 5.392116342078662e-05,
+      "loss": 1.4116,
+      "step": 7956
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.08561235386678215,
+      "learning_rate": 5.3842541985382244e-05,
+      "loss": 1.3021,
+      "step": 7957
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.08016569594538953,
+      "learning_rate": 5.3763974648107786e-05,
+      "loss": 1.3236,
+      "step": 7958
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.08856641914868289,
+      "learning_rate": 5.36854614184899e-05,
+      "loss": 1.556,
+      "step": 7959
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.09532590526560956,
+      "learning_rate": 5.360700230604842e-05,
+      "loss": 1.408,
+      "step": 7960
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.08393650575881223,
+      "learning_rate": 5.3528597320296855e-05,
+      "loss": 1.2279,
+      "step": 7961
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.08808394205662512,
+      "learning_rate": 5.3450246470742204e-05,
+      "loss": 1.3444,
+      "step": 7962
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.08187518003995406,
+      "learning_rate": 5.337194976688464e-05,
+      "loss": 1.4492,
+      "step": 7963
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.09074287291952292,
+      "learning_rate": 5.3293707218217805e-05,
+      "loss": 1.3961,
+      "step": 7964
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.09922091033979089,
+      "learning_rate": 5.3215518834229036e-05,
+      "loss": 1.4014,
+      "step": 7965
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.09558377560461183,
+      "learning_rate": 5.313738462439899e-05,
+      "loss": 1.4919,
+      "step": 7966
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.07975403401656565,
+      "learning_rate": 5.3059304598201576e-05,
+      "loss": 1.374,
+      "step": 7967
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.09718667621815387,
+      "learning_rate": 5.298127876510428e-05,
+      "loss": 1.4082,
+      "step": 7968
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.0912549963123792,
+      "learning_rate": 5.290330713456809e-05,
+      "loss": 1.4846,
+      "step": 7969
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.08168961849298514,
+      "learning_rate": 5.282538971604728e-05,
+      "loss": 1.366,
+      "step": 7970
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.08900406927437467,
+      "learning_rate": 5.274752651898956e-05,
+      "loss": 1.3303,
+      "step": 7971
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.09610658260467948,
+      "learning_rate": 5.266971755283628e-05,
+      "loss": 1.5749,
+      "step": 7972
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.08865058880800436,
+      "learning_rate": 5.259196282702178e-05,
+      "loss": 1.3821,
+      "step": 7973
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.09517903021205283,
+      "learning_rate": 5.251426235097439e-05,
+      "loss": 1.4055,
+      "step": 7974
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.08493156034832718,
+      "learning_rate": 5.243661613411543e-05,
+      "loss": 1.3444,
+      "step": 7975
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.0838612191297965,
+      "learning_rate": 5.2359024185859585e-05,
+      "loss": 1.4417,
+      "step": 7976
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.09421372340720958,
+      "learning_rate": 5.2281486515615375e-05,
+      "loss": 1.4135,
+      "step": 7977
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.09153333076851533,
+      "learning_rate": 5.220400313278451e-05,
+      "loss": 1.4349,
+      "step": 7978
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.09161799711452127,
+      "learning_rate": 5.212657404676191e-05,
+      "loss": 1.4369,
+      "step": 7979
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.08816016774243185,
+      "learning_rate": 5.2049199266936355e-05,
+      "loss": 1.3312,
+      "step": 7980
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.07536067522168258,
+      "learning_rate": 5.197187880268972e-05,
+      "loss": 1.4213,
+      "step": 7981
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.0926666875343112,
+      "learning_rate": 5.1894612663397345e-05,
+      "loss": 1.3562,
+      "step": 7982
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.09777436980649994,
+      "learning_rate": 5.1817400858427956e-05,
+      "loss": 1.4365,
+      "step": 7983
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.0811904250377712,
+      "learning_rate": 5.1740243397144e-05,
+      "loss": 1.4611,
+      "step": 7984
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.08800265668953564,
+      "learning_rate": 5.16631402889009e-05,
+      "loss": 1.3533,
+      "step": 7985
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.0904440816356616,
+      "learning_rate": 5.158609154304761e-05,
+      "loss": 1.4972,
+      "step": 7986
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.08685858702675321,
+      "learning_rate": 5.150909716892671e-05,
+      "loss": 1.3354,
+      "step": 7987
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.09362833005578333,
+      "learning_rate": 5.143215717587385e-05,
+      "loss": 1.3281,
+      "step": 7988
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.07966025286356111,
+      "learning_rate": 5.13552715732184e-05,
+      "loss": 1.5392,
+      "step": 7989
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.08725214418209264,
+      "learning_rate": 5.127844037028312e-05,
+      "loss": 1.4684,
+      "step": 7990
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.09737399318771994,
+      "learning_rate": 5.120166357638378e-05,
+      "loss": 1.4776,
+      "step": 7991
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.08768025693285349,
+      "learning_rate": 5.112494120083e-05,
+      "loss": 1.3234,
+      "step": 7992
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.08323505095792275,
+      "learning_rate": 5.104827325292466e-05,
+      "loss": 1.5446,
+      "step": 7993
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.09839799519704864,
+      "learning_rate": 5.0971659741963883e-05,
+      "loss": 1.3734,
+      "step": 7994
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.0963408838494438,
+      "learning_rate": 5.0895100677237515e-05,
+      "loss": 1.374,
+      "step": 7995
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.10273257116177338,
+      "learning_rate": 5.081859606802841e-05,
+      "loss": 1.3963,
+      "step": 7996
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.09690958321962514,
+      "learning_rate": 5.074214592361315e-05,
+      "loss": 1.4474,
+      "step": 7997
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.09055382203489995,
+      "learning_rate": 5.066575025326148e-05,
+      "loss": 1.2824,
+      "step": 7998
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.09214018733037252,
+      "learning_rate": 5.05894090662366e-05,
+      "loss": 1.3172,
+      "step": 7999
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.08020416377684451,
+      "learning_rate": 5.05131223717954e-05,
+      "loss": 1.4465,
+      "step": 8000
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.10471508281926799,
+      "learning_rate": 5.043689017918757e-05,
+      "loss": 1.3915,
+      "step": 8001
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.10116000957392209,
+      "learning_rate": 5.0360712497656725e-05,
+      "loss": 1.2644,
+      "step": 8002
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.09559909241116268,
+      "learning_rate": 5.028458933643976e-05,
+      "loss": 1.3338,
+      "step": 8003
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.09002643649763474,
+      "learning_rate": 5.020852070476656e-05,
+      "loss": 1.4668,
+      "step": 8004
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.09528584400679911,
+      "learning_rate": 5.013250661186103e-05,
+      "loss": 1.33,
+      "step": 8005
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.11374398792014954,
+      "learning_rate": 5.005654706694002e-05,
+      "loss": 1.3475,
+      "step": 8006
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.09495489508659871,
+      "learning_rate": 4.9980642079213764e-05,
+      "loss": 1.4521,
+      "step": 8007
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.09302443756883642,
+      "learning_rate": 4.9904791657886085e-05,
+      "loss": 1.4601,
+      "step": 8008
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.09287663041146321,
+      "learning_rate": 4.982899581215428e-05,
+      "loss": 1.3386,
+      "step": 8009
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.08839803430724054,
+      "learning_rate": 4.975325455120855e-05,
+      "loss": 1.4648,
+      "step": 8010
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.09007925416026948,
+      "learning_rate": 4.9677567884233e-05,
+      "loss": 1.5249,
+      "step": 8011
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1024334249851111,
+      "learning_rate": 4.96019358204049e-05,
+      "loss": 1.3792,
+      "step": 8012
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.09493434008827761,
+      "learning_rate": 4.952635836889474e-05,
+      "loss": 1.4692,
+      "step": 8013
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.10177317419169603,
+      "learning_rate": 4.94508355388667e-05,
+      "loss": 1.3567,
+      "step": 8014
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.10359395185727091,
+      "learning_rate": 4.937536733947817e-05,
+      "loss": 1.4361,
+      "step": 8015
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1014764345449319,
+      "learning_rate": 4.9299953779879894e-05,
+      "loss": 1.4492,
+      "step": 8016
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.09141219987822603,
+      "learning_rate": 4.9224594869215934e-05,
+      "loss": 1.3856,
+      "step": 8017
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.10701698487475056,
+      "learning_rate": 4.914929061662399e-05,
+      "loss": 1.3098,
+      "step": 8018
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.08421051966495124,
+      "learning_rate": 4.9074041031234804e-05,
+      "loss": 1.2876,
+      "step": 8019
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.10280735588357584,
+      "learning_rate": 4.89988461221727e-05,
+      "loss": 1.2543,
+      "step": 8020
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.0871700644770041,
+      "learning_rate": 4.8923705898555496e-05,
+      "loss": 1.4419,
+      "step": 8021
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.09208405244887743,
+      "learning_rate": 4.884862036949389e-05,
+      "loss": 1.4242,
+      "step": 8022
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.09519872895834376,
+      "learning_rate": 4.877358954409245e-05,
+      "loss": 1.4272,
+      "step": 8023
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.08160045804385208,
+      "learning_rate": 4.8698613431448934e-05,
+      "loss": 1.442,
+      "step": 8024
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.08917162958217983,
+      "learning_rate": 4.862369204065437e-05,
+      "loss": 1.4528,
+      "step": 8025
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.09584081971903959,
+      "learning_rate": 4.8548825380793304e-05,
+      "loss": 1.4295,
+      "step": 8026
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.08745885865269687,
+      "learning_rate": 4.84740134609436e-05,
+      "loss": 1.4092,
+      "step": 8027
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.08685155962456509,
+      "learning_rate": 4.839925629017639e-05,
+      "loss": 1.3759,
+      "step": 8028
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.10410625722723307,
+      "learning_rate": 4.8324553877556186e-05,
+      "loss": 1.4921,
+      "step": 8029
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.09053463510410303,
+      "learning_rate": 4.8249906232140984e-05,
+      "loss": 1.3488,
+      "step": 8030
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.08590497883548834,
+      "learning_rate": 4.8175313362982195e-05,
+      "loss": 1.4228,
+      "step": 8031
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.10821065446154124,
+      "learning_rate": 4.810077527912421e-05,
+      "loss": 1.3144,
+      "step": 8032
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.10917917244641281,
+      "learning_rate": 4.8026291989605174e-05,
+      "loss": 1.3937,
+      "step": 8033
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.09423165223915081,
+      "learning_rate": 4.7951863503456485e-05,
+      "loss": 1.3646,
+      "step": 8034
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.088316397400894,
+      "learning_rate": 4.787748982970275e-05,
+      "loss": 1.3682,
+      "step": 8035
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.10998794053022858,
+      "learning_rate": 4.780317097736203e-05,
+      "loss": 1.3058,
+      "step": 8036
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.09088548506112062,
+      "learning_rate": 4.7728906955446015e-05,
+      "loss": 1.4925,
+      "step": 8037
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.08992202733297892,
+      "learning_rate": 4.765469777295906e-05,
+      "loss": 1.5035,
+      "step": 8038
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.08358952526085513,
+      "learning_rate": 4.7580543438899446e-05,
+      "loss": 1.3281,
+      "step": 8039
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.08568139006226679,
+      "learning_rate": 4.750644396225873e-05,
+      "loss": 1.5197,
+      "step": 8040
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.08993280314740992,
+      "learning_rate": 4.743239935202165e-05,
+      "loss": 1.3513,
+      "step": 8041
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.09889419546611622,
+      "learning_rate": 4.7358409617166307e-05,
+      "loss": 1.3172,
+      "step": 8042
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.09847518037940564,
+      "learning_rate": 4.728447476666442e-05,
+      "loss": 1.3748,
+      "step": 8043
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.09646547794616295,
+      "learning_rate": 4.7210594809480645e-05,
+      "loss": 1.4414,
+      "step": 8044
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.08558448303419687,
+      "learning_rate": 4.71367697545732e-05,
+      "loss": 1.3872,
+      "step": 8045
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.08697043289257039,
+      "learning_rate": 4.706299961089383e-05,
+      "loss": 1.338,
+      "step": 8046
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.08883717181982263,
+      "learning_rate": 4.698928438738714e-05,
+      "loss": 1.4657,
+      "step": 8047
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.09612420925192143,
+      "learning_rate": 4.691562409299161e-05,
+      "loss": 1.3923,
+      "step": 8048
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.0866917810947192,
+      "learning_rate": 4.684201873663868e-05,
+      "loss": 1.5458,
+      "step": 8049
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.10058692936269363,
+      "learning_rate": 4.676846832725312e-05,
+      "loss": 1.294,
+      "step": 8050
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.09684992742639126,
+      "learning_rate": 4.6694972873753296e-05,
+      "loss": 1.4402,
+      "step": 8051
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.11670958899864131,
+      "learning_rate": 4.66215323850509e-05,
+      "loss": 1.3937,
+      "step": 8052
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.08558904787337474,
+      "learning_rate": 4.6548146870050656e-05,
+      "loss": 1.4114,
+      "step": 8053
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.08350040748116296,
+      "learning_rate": 4.647481633765088e-05,
+      "loss": 1.2515,
+      "step": 8054
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.08270056183524263,
+      "learning_rate": 4.640154079674325e-05,
+      "loss": 1.3878,
+      "step": 8055
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.12140269918508494,
+      "learning_rate": 4.632832025621253e-05,
+      "loss": 1.369,
+      "step": 8056
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.0967211345492668,
+      "learning_rate": 4.625515472493697e-05,
+      "loss": 1.4863,
+      "step": 8057
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.09594188315943698,
+      "learning_rate": 4.6182044211788343e-05,
+      "loss": 1.3982,
+      "step": 8058
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.10442851702577187,
+      "learning_rate": 4.610898872563135e-05,
+      "loss": 1.4728,
+      "step": 8059
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.09089308048752692,
+      "learning_rate": 4.6035988275324236e-05,
+      "loss": 1.4235,
+      "step": 8060
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.10029383253191723,
+      "learning_rate": 4.596304286971853e-05,
+      "loss": 1.4527,
+      "step": 8061
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.09245414374315496,
+      "learning_rate": 4.5890152517659326e-05,
+      "loss": 1.4476,
+      "step": 8062
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.08749737182144246,
+      "learning_rate": 4.58173172279846e-05,
+      "loss": 1.4544,
+      "step": 8063
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.08128681609621158,
+      "learning_rate": 4.574453700952591e-05,
+      "loss": 1.315,
+      "step": 8064
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.0977477408035849,
+      "learning_rate": 4.56718118711083e-05,
+      "loss": 1.3868,
+      "step": 8065
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.08346194411051658,
+      "learning_rate": 4.559914182154967e-05,
+      "loss": 1.3092,
+      "step": 8066
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.08313207581438073,
+      "learning_rate": 4.5526526869661686e-05,
+      "loss": 1.3348,
+      "step": 8067
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.09434227073134552,
+      "learning_rate": 4.545396702424926e-05,
+      "loss": 1.2725,
+      "step": 8068
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.09172296973103769,
+      "learning_rate": 4.538146229411033e-05,
+      "loss": 1.5323,
+      "step": 8069
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.09822663625867462,
+      "learning_rate": 4.5309012688036334e-05,
+      "loss": 1.4093,
+      "step": 8070
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.0931555383209421,
+      "learning_rate": 4.5236618214812234e-05,
+      "loss": 1.307,
+      "step": 8071
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.09686713999156278,
+      "learning_rate": 4.5164278883215834e-05,
+      "loss": 1.4408,
+      "step": 8072
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.09891740183537313,
+      "learning_rate": 4.509199470201869e-05,
+      "loss": 1.3064,
+      "step": 8073
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.08416754685939414,
+      "learning_rate": 4.501976567998561e-05,
+      "loss": 1.2797,
+      "step": 8074
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.09197596824111855,
+      "learning_rate": 4.494759182587438e-05,
+      "loss": 1.4559,
+      "step": 8075
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.09727193551485228,
+      "learning_rate": 4.487547314843643e-05,
+      "loss": 1.466,
+      "step": 8076
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.09239804210902398,
+      "learning_rate": 4.480340965641655e-05,
+      "loss": 1.4517,
+      "step": 8077
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.09649305439725554,
+      "learning_rate": 4.473140135855247e-05,
+      "loss": 1.4156,
+      "step": 8078
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.09236065359889135,
+      "learning_rate": 4.4659448263575544e-05,
+      "loss": 1.4829,
+      "step": 8079
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.0994382561845817,
+      "learning_rate": 4.4587550380210284e-05,
+      "loss": 1.3684,
+      "step": 8080
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.09209726648563313,
+      "learning_rate": 4.451570771717467e-05,
+      "loss": 1.3426,
+      "step": 8081
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.09843478293680565,
+      "learning_rate": 4.444392028317967e-05,
+      "loss": 1.3716,
+      "step": 8082
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.10283688142537882,
+      "learning_rate": 4.437218808693e-05,
+      "loss": 1.3521,
+      "step": 8083
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.09645222990375216,
+      "learning_rate": 4.4300511137123186e-05,
+      "loss": 1.4416,
+      "step": 8084
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.08737555151807626,
+      "learning_rate": 4.422888944245046e-05,
+      "loss": 1.3301,
+      "step": 8085
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.09666379041694352,
+      "learning_rate": 4.415732301159625e-05,
+      "loss": 1.5381,
+      "step": 8086
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.09304119796433967,
+      "learning_rate": 4.408581185323807e-05,
+      "loss": 1.4706,
+      "step": 8087
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.11082626671337686,
+      "learning_rate": 4.401435597604697e-05,
+      "loss": 1.3994,
+      "step": 8088
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.08799149367663399,
+      "learning_rate": 4.394295538868731e-05,
+      "loss": 1.5273,
+      "step": 8089
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.09010079273368712,
+      "learning_rate": 4.387161009981661e-05,
+      "loss": 1.3633,
+      "step": 8090
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.08354802566216252,
+      "learning_rate": 4.3800320118085605e-05,
+      "loss": 1.4994,
+      "step": 8091
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.08402331275744471,
+      "learning_rate": 4.3729085452138496e-05,
+      "loss": 1.376,
+      "step": 8092
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.10391735563408136,
+      "learning_rate": 4.365790611061293e-05,
+      "loss": 1.4819,
+      "step": 8093
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.09007821896954256,
+      "learning_rate": 4.3586782102139344e-05,
+      "loss": 1.3756,
+      "step": 8094
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.08903422612703692,
+      "learning_rate": 4.3515713435342e-05,
+      "loss": 1.4106,
+      "step": 8095
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.09313008870552218,
+      "learning_rate": 4.344470011883817e-05,
+      "loss": 1.402,
+      "step": 8096
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.08543654211563571,
+      "learning_rate": 4.33737421612384e-05,
+      "loss": 1.2531,
+      "step": 8097
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.09067470275876136,
+      "learning_rate": 4.330283957114656e-05,
+      "loss": 1.3704,
+      "step": 8098
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.08563274293716956,
+      "learning_rate": 4.323199235716002e-05,
+      "loss": 1.3759,
+      "step": 8099
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.08905782530462407,
+      "learning_rate": 4.316120052786904e-05,
+      "loss": 1.3192,
+      "step": 8100
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.09857674223026314,
+      "learning_rate": 4.309046409185757e-05,
+      "loss": 1.3845,
+      "step": 8101
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.08694920472324444,
+      "learning_rate": 4.301978305770249e-05,
+      "loss": 1.4144,
+      "step": 8102
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.08527994623935795,
+      "learning_rate": 4.294915743397409e-05,
+      "loss": 1.4158,
+      "step": 8103
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.09263598147216909,
+      "learning_rate": 4.287858722923604e-05,
+      "loss": 1.5114,
+      "step": 8104
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.09202117731423061,
+      "learning_rate": 4.2808072452045374e-05,
+      "loss": 1.429,
+      "step": 8105
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.08638915131619937,
+      "learning_rate": 4.273761311095192e-05,
+      "loss": 1.4321,
+      "step": 8106
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.09522134929831542,
+      "learning_rate": 4.2667209214499346e-05,
+      "loss": 1.3787,
+      "step": 8107
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.10469829873713113,
+      "learning_rate": 4.259686077122443e-05,
+      "loss": 1.3612,
+      "step": 8108
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.0855733834471551,
+      "learning_rate": 4.2526567789656965e-05,
+      "loss": 1.4306,
+      "step": 8109
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.09614067627908375,
+      "learning_rate": 4.2456330278320245e-05,
+      "loss": 1.4499,
+      "step": 8110
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.08390308759988215,
+      "learning_rate": 4.2386148245731e-05,
+      "loss": 1.3943,
+      "step": 8111
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.09902179964016455,
+      "learning_rate": 4.2316021700398944e-05,
+      "loss": 1.4578,
+      "step": 8112
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.09883714745917478,
+      "learning_rate": 4.224595065082703e-05,
+      "loss": 1.4541,
+      "step": 8113
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.08727422039676933,
+      "learning_rate": 4.217593510551176e-05,
+      "loss": 1.4615,
+      "step": 8114
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.09334590367762645,
+      "learning_rate": 4.210597507294267e-05,
+      "loss": 1.4209,
+      "step": 8115
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.09456309243317293,
+      "learning_rate": 4.203607056160269e-05,
+      "loss": 1.3918,
+      "step": 8116
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.09896588270561626,
+      "learning_rate": 4.19662215799681e-05,
+      "loss": 1.4424,
+      "step": 8117
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.08847654465807804,
+      "learning_rate": 4.1896428136508125e-05,
+      "loss": 1.3337,
+      "step": 8118
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.11081100920646882,
+      "learning_rate": 4.182669023968561e-05,
+      "loss": 1.4485,
+      "step": 8119
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.09416340783753362,
+      "learning_rate": 4.17570078979565e-05,
+      "loss": 1.414,
+      "step": 8120
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.09604121115206532,
+      "learning_rate": 4.168738111976989e-05,
+      "loss": 1.4759,
+      "step": 8121
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.08494053518936755,
+      "learning_rate": 4.161780991356845e-05,
+      "loss": 1.2911,
+      "step": 8122
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.09232432805630465,
+      "learning_rate": 4.154829428778778e-05,
+      "loss": 1.4698,
+      "step": 8123
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.08560962063353268,
+      "learning_rate": 4.147883425085702e-05,
+      "loss": 1.3934,
+      "step": 8124
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.08548953603826323,
+      "learning_rate": 4.140942981119833e-05,
+      "loss": 1.3536,
+      "step": 8125
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.09186465089721559,
+      "learning_rate": 4.1340080977227244e-05,
+      "loss": 1.3503,
+      "step": 8126
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.09138641763789578,
+      "learning_rate": 4.127078775735266e-05,
+      "loss": 1.4755,
+      "step": 8127
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.09084712166052157,
+      "learning_rate": 4.1201550159976455e-05,
+      "loss": 1.3593,
+      "step": 8128
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.09212969623548035,
+      "learning_rate": 4.113236819349403e-05,
+      "loss": 1.5437,
+      "step": 8129
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.09237647458246168,
+      "learning_rate": 4.106324186629396e-05,
+      "loss": 1.3442,
+      "step": 8130
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.09251388456673824,
+      "learning_rate": 4.0994171186757966e-05,
+      "loss": 1.586,
+      "step": 8131
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.09511611784151426,
+      "learning_rate": 4.0925156163261256e-05,
+      "loss": 1.4247,
+      "step": 8132
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.09355509883336248,
+      "learning_rate": 4.085619680417196e-05,
+      "loss": 1.4298,
+      "step": 8133
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.08637365018956814,
+      "learning_rate": 4.0787293117851674e-05,
+      "loss": 1.3321,
+      "step": 8134
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.08396638217296636,
+      "learning_rate": 4.071844511265527e-05,
+      "loss": 1.5422,
+      "step": 8135
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.08649007647069878,
+      "learning_rate": 4.064965279693083e-05,
+      "loss": 1.3918,
+      "step": 8136
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.09236978127469905,
+      "learning_rate": 4.058091617901949e-05,
+      "loss": 1.4447,
+      "step": 8137
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.08843730477667042,
+      "learning_rate": 4.0512235267256e-05,
+      "loss": 1.3427,
+      "step": 8138
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.10048906860747468,
+      "learning_rate": 4.044361006996811e-05,
+      "loss": 1.3218,
+      "step": 8139
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.09575911752372211,
+      "learning_rate": 4.0375040595476754e-05,
+      "loss": 1.3955,
+      "step": 8140
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.09332874770342339,
+      "learning_rate": 4.0306526852096296e-05,
+      "loss": 1.2804,
+      "step": 8141
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.089037521052994,
+      "learning_rate": 4.023806884813436e-05,
+      "loss": 1.4618,
+      "step": 8142
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.08529572187836744,
+      "learning_rate": 4.016966659189158e-05,
+      "loss": 1.394,
+      "step": 8143
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.09280673609000385,
+      "learning_rate": 4.010132009166195e-05,
+      "loss": 1.281,
+      "step": 8144
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.08763093597932838,
+      "learning_rate": 4.0033029355732886e-05,
+      "loss": 1.344,
+      "step": 8145
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.09000825851517992,
+      "learning_rate": 3.996479439238465e-05,
+      "loss": 1.4322,
+      "step": 8146
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.10024569194799243,
+      "learning_rate": 3.989661520989102e-05,
+      "loss": 1.5138,
+      "step": 8147
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.0913436778896483,
+      "learning_rate": 3.982849181651915e-05,
+      "loss": 1.3464,
+      "step": 8148
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.0862936421392638,
+      "learning_rate": 3.9760424220529004e-05,
+      "loss": 1.3939,
+      "step": 8149
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.10394172085424071,
+      "learning_rate": 3.969241243017413e-05,
+      "loss": 1.5113,
+      "step": 8150
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.0917900252830298,
+      "learning_rate": 3.962445645370122e-05,
+      "loss": 1.3708,
+      "step": 8151
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.08331465806789863,
+      "learning_rate": 3.955655629935007e-05,
+      "loss": 1.2775,
+      "step": 8152
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.0838333563003349,
+      "learning_rate": 3.948871197535386e-05,
+      "loss": 1.4261,
+      "step": 8153
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.09130774844385221,
+      "learning_rate": 3.942092348993903e-05,
+      "loss": 1.384,
+      "step": 8154
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.09102809118679295,
+      "learning_rate": 3.935319085132505e-05,
+      "loss": 1.4354,
+      "step": 8155
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.09312677606946752,
+      "learning_rate": 3.928551406772468e-05,
+      "loss": 1.4494,
+      "step": 8156
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.08760962335416297,
+      "learning_rate": 3.921789314734409e-05,
+      "loss": 1.3796,
+      "step": 8157
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1140094609269149,
+      "learning_rate": 3.915032809838259e-05,
+      "loss": 1.502,
+      "step": 8158
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.08082863729081557,
+      "learning_rate": 3.908281892903254e-05,
+      "loss": 1.4221,
+      "step": 8159
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.08881742661475506,
+      "learning_rate": 3.901536564747965e-05,
+      "loss": 1.4697,
+      "step": 8160
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.09152808811324147,
+      "learning_rate": 3.8947968261903054e-05,
+      "loss": 1.3648,
+      "step": 8161
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.09256609786251456,
+      "learning_rate": 3.888062678047472e-05,
+      "loss": 1.4266,
+      "step": 8162
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.07859820908539289,
+      "learning_rate": 3.88133412113601e-05,
+      "loss": 1.3956,
+      "step": 8163
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.09774279455050242,
+      "learning_rate": 3.874611156271801e-05,
+      "loss": 1.4649,
+      "step": 8164
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.09097510907480971,
+      "learning_rate": 3.867893784269988e-05,
+      "loss": 1.4925,
+      "step": 8165
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.08421560686125693,
+      "learning_rate": 3.861182005945091e-05,
+      "loss": 1.4309,
+      "step": 8166
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.09794270345444765,
+      "learning_rate": 3.854475822110953e-05,
+      "loss": 1.3782,
+      "step": 8167
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.10010697158639269,
+      "learning_rate": 3.8477752335807027e-05,
+      "loss": 1.3843,
+      "step": 8168
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.09122969411040488,
+      "learning_rate": 3.841080241166811e-05,
+      "loss": 1.4884,
+      "step": 8169
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.0879295421905876,
+      "learning_rate": 3.8343908456810905e-05,
+      "loss": 1.4185,
+      "step": 8170
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.10266989127410892,
+      "learning_rate": 3.82770704793462e-05,
+      "loss": 1.4521,
+      "step": 8171
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.09692602731626357,
+      "learning_rate": 3.8210288487378566e-05,
+      "loss": 1.5678,
+      "step": 8172
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.0856255659833879,
+      "learning_rate": 3.8143562489005525e-05,
+      "loss": 1.4182,
+      "step": 8173
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.08637972321065973,
+      "learning_rate": 3.8076892492317713e-05,
+      "loss": 1.4223,
+      "step": 8174
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.085387579620603,
+      "learning_rate": 3.801027850539929e-05,
+      "loss": 1.3246,
+      "step": 8175
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.07966648182833903,
+      "learning_rate": 3.794372053632722e-05,
+      "loss": 1.3599,
+      "step": 8176
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.0839450905097096,
+      "learning_rate": 3.7877218593172057e-05,
+      "loss": 1.3715,
+      "step": 8177
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.08710525678299814,
+      "learning_rate": 3.781077268399724e-05,
+      "loss": 1.5138,
+      "step": 8178
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.0909527791884506,
+      "learning_rate": 3.7744382816859765e-05,
+      "loss": 1.2932,
+      "step": 8179
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.08725990644653168,
+      "learning_rate": 3.7678048999809365e-05,
+      "loss": 1.4722,
+      "step": 8180
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.08838118634523105,
+      "learning_rate": 3.761177124088943e-05,
+      "loss": 1.3409,
+      "step": 8181
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.0817359643411812,
+      "learning_rate": 3.7545549548136435e-05,
+      "loss": 1.3198,
+      "step": 8182
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.0910296821176027,
+      "learning_rate": 3.747938392957972e-05,
+      "loss": 1.406,
+      "step": 8183
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.08337072181587418,
+      "learning_rate": 3.741327439324232e-05,
+      "loss": 1.2309,
+      "step": 8184
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.08964416503075122,
+      "learning_rate": 3.734722094714027e-05,
+      "loss": 1.4342,
+      "step": 8185
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.08834670597246252,
+      "learning_rate": 3.7281223599282655e-05,
+      "loss": 1.5451,
+      "step": 8186
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.09156897718666505,
+      "learning_rate": 3.72152823576718e-05,
+      "loss": 1.3623,
+      "step": 8187
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.0897607277218215,
+      "learning_rate": 3.714939723030347e-05,
+      "loss": 1.4662,
+      "step": 8188
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.09034869839292647,
+      "learning_rate": 3.7083568225166454e-05,
+      "loss": 1.2751,
+      "step": 8189
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.09641389241523105,
+      "learning_rate": 3.701779535024269e-05,
+      "loss": 1.3636,
+      "step": 8190
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.09110058458879819,
+      "learning_rate": 3.69520786135073e-05,
+      "loss": 1.415,
+      "step": 8191
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.09185325382635562,
+      "learning_rate": 3.688641802292891e-05,
+      "loss": 1.4223,
+      "step": 8192
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.08397425268962305,
+      "learning_rate": 3.682081358646883e-05,
+      "loss": 1.4142,
+      "step": 8193
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.07977066022161716,
+      "learning_rate": 3.67552653120819e-05,
+      "loss": 1.414,
+      "step": 8194
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.07986806002290193,
+      "learning_rate": 3.668977320771616e-05,
+      "loss": 1.3467,
+      "step": 8195
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.0937341798408163,
+      "learning_rate": 3.66243372813127e-05,
+      "loss": 1.3114,
+      "step": 8196
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.09905772226978368,
+      "learning_rate": 3.655895754080579e-05,
+      "loss": 1.2811,
+      "step": 8197
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.0976774300018668,
+      "learning_rate": 3.649363399412309e-05,
+      "loss": 1.395,
+      "step": 8198
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.09179719331361488,
+      "learning_rate": 3.6428366649185084e-05,
+      "loss": 1.3635,
+      "step": 8199
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.0824569558000461,
+      "learning_rate": 3.636315551390584e-05,
+      "loss": 1.4963,
+      "step": 8200
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.08942944820307387,
+      "learning_rate": 3.6298000596192485e-05,
+      "loss": 1.4102,
+      "step": 8201
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.0899389335487092,
+      "learning_rate": 3.623290190394507e-05,
+      "loss": 1.3796,
+      "step": 8202
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.0912806741082675,
+      "learning_rate": 3.616785944505713e-05,
+      "loss": 1.3977,
+      "step": 8203
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.09622140600547263,
+      "learning_rate": 3.61028732274154e-05,
+      "loss": 1.3994,
+      "step": 8204
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.10230214814948685,
+      "learning_rate": 3.603794325889953e-05,
+      "loss": 1.3927,
+      "step": 8205
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.08549966685330587,
+      "learning_rate": 3.597306954738255e-05,
+      "loss": 1.4888,
+      "step": 8206
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.0797934896092907,
+      "learning_rate": 3.590825210073073e-05,
+      "loss": 1.3104,
+      "step": 8207
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.08264588594248158,
+      "learning_rate": 3.584349092680328e-05,
+      "loss": 1.5409,
+      "step": 8208
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.08128287757448842,
+      "learning_rate": 3.577878603345269e-05,
+      "loss": 1.3391,
+      "step": 8209
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.09632212209637667,
+      "learning_rate": 3.5714137428524754e-05,
+      "loss": 1.3682,
+      "step": 8210
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.08181637197782007,
+      "learning_rate": 3.564954511985824e-05,
+      "loss": 1.478,
+      "step": 8211
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.08614079720602566,
+      "learning_rate": 3.5585009115285226e-05,
+      "loss": 1.2792,
+      "step": 8212
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.09936056430723951,
+      "learning_rate": 3.552052942263101e-05,
+      "loss": 1.4256,
+      "step": 8213
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.08886464083868731,
+      "learning_rate": 3.545610604971383e-05,
+      "loss": 1.361,
+      "step": 8214
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.09233698701338725,
+      "learning_rate": 3.5391739004345335e-05,
+      "loss": 1.4835,
+      "step": 8215
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.08301256671081278,
+      "learning_rate": 3.5327428294330336e-05,
+      "loss": 1.4588,
+      "step": 8216
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.09239690186282472,
+      "learning_rate": 3.5263173927466584e-05,
+      "loss": 1.3954,
+      "step": 8217
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.09754474472861756,
+      "learning_rate": 3.5198975911545136e-05,
+      "loss": 1.3358,
+      "step": 8218
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.08706581069856931,
+      "learning_rate": 3.513483425435021e-05,
+      "loss": 1.3275,
+      "step": 8219
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.09382740937626291,
+      "learning_rate": 3.507074896365942e-05,
+      "loss": 1.3643,
+      "step": 8220
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.08549684695802111,
+      "learning_rate": 3.500672004724303e-05,
+      "loss": 1.3358,
+      "step": 8221
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.11311319233591514,
+      "learning_rate": 3.494274751286497e-05,
+      "loss": 1.4496,
+      "step": 8222
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.08135780208806964,
+      "learning_rate": 3.4878831368282126e-05,
+      "loss": 1.2636,
+      "step": 8223
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.09552206092091241,
+      "learning_rate": 3.4814971621244415e-05,
+      "loss": 1.4177,
+      "step": 8224
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.09716090579217876,
+      "learning_rate": 3.4751168279495095e-05,
+      "loss": 1.288,
+      "step": 8225
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.08658930830747578,
+      "learning_rate": 3.468742135077069e-05,
+      "loss": 1.3388,
+      "step": 8226
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.09649678229661654,
+      "learning_rate": 3.462373084280057e-05,
+      "loss": 1.3767,
+      "step": 8227
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.09611441101312547,
+      "learning_rate": 3.456009676330751e-05,
+      "loss": 1.4051,
+      "step": 8228
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.09029750550796355,
+      "learning_rate": 3.449651912000734e-05,
+      "loss": 1.3124,
+      "step": 8229
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.10382301198139691,
+      "learning_rate": 3.4432997920609e-05,
+      "loss": 1.4823,
+      "step": 8230
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.08153783473681955,
+      "learning_rate": 3.436953317281472e-05,
+      "loss": 1.4889,
+      "step": 8231
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.09764009699745096,
+      "learning_rate": 3.430612488431989e-05,
+      "loss": 1.3913,
+      "step": 8232
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.08631013956906346,
+      "learning_rate": 3.4242773062812815e-05,
+      "loss": 1.4285,
+      "step": 8233
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.08769483348443738,
+      "learning_rate": 3.4179477715975237e-05,
+      "loss": 1.4148,
+      "step": 8234
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.09050904791758085,
+      "learning_rate": 3.411623885148202e-05,
+      "loss": 1.2897,
+      "step": 8235
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.08907965528334316,
+      "learning_rate": 3.405305647700085e-05,
+      "loss": 1.3697,
+      "step": 8236
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.08801888670287704,
+      "learning_rate": 3.398993060019295e-05,
+      "loss": 1.3306,
+      "step": 8237
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.08602038634493521,
+      "learning_rate": 3.392686122871263e-05,
+      "loss": 1.4508,
+      "step": 8238
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.08606685174750452,
+      "learning_rate": 3.3863848370207226e-05,
+      "loss": 1.4797,
+      "step": 8239
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.10244754385585404,
+      "learning_rate": 3.380089203231712e-05,
+      "loss": 1.4539,
+      "step": 8240
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.09051226085948132,
+      "learning_rate": 3.373799222267615e-05,
+      "loss": 1.4318,
+      "step": 8241
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.10024875889437078,
+      "learning_rate": 3.3675148948911e-05,
+      "loss": 1.3614,
+      "step": 8242
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.09779553374981109,
+      "learning_rate": 3.361236221864172e-05,
+      "loss": 1.272,
+      "step": 8243
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.08717187939794446,
+      "learning_rate": 3.354963203948147e-05,
+      "loss": 1.423,
+      "step": 8244
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.09053686476661967,
+      "learning_rate": 3.348695841903637e-05,
+      "loss": 1.3593,
+      "step": 8245
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.08330610617134679,
+      "learning_rate": 3.342434136490585e-05,
+      "loss": 1.4082,
+      "step": 8246
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.08656762493107648,
+      "learning_rate": 3.3361780884682615e-05,
+      "loss": 1.3553,
+      "step": 8247
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.10053769825211262,
+      "learning_rate": 3.329927698595203e-05,
+      "loss": 1.3354,
+      "step": 8248
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.10128003625082153,
+      "learning_rate": 3.3236829676293215e-05,
+      "loss": 1.3203,
+      "step": 8249
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.08681193791874181,
+      "learning_rate": 3.3174438963277875e-05,
+      "loss": 1.3869,
+      "step": 8250
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.09209269431223971,
+      "learning_rate": 3.311210485447125e-05,
+      "loss": 1.3276,
+      "step": 8251
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.09032137653302173,
+      "learning_rate": 3.30498273574315e-05,
+      "loss": 1.3161,
+      "step": 8252
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.08830919203330095,
+      "learning_rate": 3.298760647970994e-05,
+      "loss": 1.4686,
+      "step": 8253
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.09601488594029912,
+      "learning_rate": 3.292544222885124e-05,
+      "loss": 1.4203,
+      "step": 8254
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.0936075049034287,
+      "learning_rate": 3.286333461239288e-05,
+      "loss": 1.453,
+      "step": 8255
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.08604891420228421,
+      "learning_rate": 3.280128363786561e-05,
+      "loss": 1.2659,
+      "step": 8256
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.08944258325837522,
+      "learning_rate": 3.273928931279346e-05,
+      "loss": 1.4166,
+      "step": 8257
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.09903855128508778,
+      "learning_rate": 3.267735164469332e-05,
+      "loss": 1.2721,
+      "step": 8258
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.09662137777758281,
+      "learning_rate": 3.261547064107551e-05,
+      "loss": 1.2992,
+      "step": 8259
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.08389210761394819,
+      "learning_rate": 3.255364630944313e-05,
+      "loss": 1.3329,
+      "step": 8260
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.08569831100242886,
+      "learning_rate": 3.249187865729264e-05,
+      "loss": 1.3118,
+      "step": 8261
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.0990971080529239,
+      "learning_rate": 3.24301676921136e-05,
+      "loss": 1.338,
+      "step": 8262
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.10539828259931236,
+      "learning_rate": 3.236851342138874e-05,
+      "loss": 1.3873,
+      "step": 8263
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.09089848271652821,
+      "learning_rate": 3.2306915852593713e-05,
+      "loss": 1.3069,
+      "step": 8264
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.08731495361645467,
+      "learning_rate": 3.224537499319757e-05,
+      "loss": 1.451,
+      "step": 8265
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.0963751591762757,
+      "learning_rate": 3.218389085066237e-05,
+      "loss": 1.4377,
+      "step": 8266
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.08766836169616739,
+      "learning_rate": 3.2122463432443125e-05,
+      "loss": 1.3519,
+      "step": 8267
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.0944698914392843,
+      "learning_rate": 3.206109274598817e-05,
+      "loss": 1.3163,
+      "step": 8268
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.08410850904393825,
+      "learning_rate": 3.199977879873906e-05,
+      "loss": 1.3856,
+      "step": 8269
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.08900858429825506,
+      "learning_rate": 3.193852159813021e-05,
+      "loss": 1.4729,
+      "step": 8270
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.08340149953121077,
+      "learning_rate": 3.18773211515892e-05,
+      "loss": 1.4201,
+      "step": 8271
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.08810900737089097,
+      "learning_rate": 3.181617746653687e-05,
+      "loss": 1.3035,
+      "step": 8272
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.09909487833329844,
+      "learning_rate": 3.1755090550387165e-05,
+      "loss": 1.3747,
+      "step": 8273
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.09206263771365691,
+      "learning_rate": 3.169406041054695e-05,
+      "loss": 1.4318,
+      "step": 8274
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.08936295451925626,
+      "learning_rate": 3.163308705441648e-05,
+      "loss": 1.354,
+      "step": 8275
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.10108582918313291,
+      "learning_rate": 3.157217048938882e-05,
+      "loss": 1.3638,
+      "step": 8276
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.09810973020134794,
+      "learning_rate": 3.151131072285041e-05,
+      "loss": 1.3389,
+      "step": 8277
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.09990108657933389,
+      "learning_rate": 3.145050776218078e-05,
+      "loss": 1.5351,
+      "step": 8278
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.09063262414102946,
+      "learning_rate": 3.138976161475238e-05,
+      "loss": 1.4391,
+      "step": 8279
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.08380459100263056,
+      "learning_rate": 3.132907228793086e-05,
+      "loss": 1.3581,
+      "step": 8280
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.08742457245686931,
+      "learning_rate": 3.126843978907518e-05,
+      "loss": 1.2955,
+      "step": 8281
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.09016049568453172,
+      "learning_rate": 3.1207864125537165e-05,
+      "loss": 1.2513,
+      "step": 8282
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.08375867746186022,
+      "learning_rate": 3.1147345304661734e-05,
+      "loss": 1.3385,
+      "step": 8283
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.10576397254106343,
+      "learning_rate": 3.108688333378701e-05,
+      "loss": 1.4306,
+      "step": 8284
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.10166972259158993,
+      "learning_rate": 3.10264782202444e-05,
+      "loss": 1.4401,
+      "step": 8285
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.0816827777590841,
+      "learning_rate": 3.0966129971358005e-05,
+      "loss": 1.3993,
+      "step": 8286
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.10865251945702202,
+      "learning_rate": 3.0905838594445346e-05,
+      "loss": 1.4408,
+      "step": 8287
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.09242901362551612,
+      "learning_rate": 3.0845604096817024e-05,
+      "loss": 1.4811,
+      "step": 8288
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.09310275405757654,
+      "learning_rate": 3.078542648577659e-05,
+      "loss": 1.427,
+      "step": 8289
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.07919899361904145,
+      "learning_rate": 3.072530576862081e-05,
+      "loss": 1.3695,
+      "step": 8290
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.09071826200447604,
+      "learning_rate": 3.06652419526397e-05,
+      "loss": 1.3786,
+      "step": 8291
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.09205784757595566,
+      "learning_rate": 3.060523504511587e-05,
+      "loss": 1.4101,
+      "step": 8292
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.09121367166574848,
+      "learning_rate": 3.05452850533256e-05,
+      "loss": 1.4315,
+      "step": 8293
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.08840960290934788,
+      "learning_rate": 3.048539198453798e-05,
+      "loss": 1.1353,
+      "step": 8294
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.10546852151528087,
+      "learning_rate": 3.0425555846015196e-05,
+      "loss": 1.3073,
+      "step": 8295
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.09881277238387846,
+      "learning_rate": 3.0365776645012666e-05,
+      "loss": 1.3881,
+      "step": 8296
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.0854003085075494,
+      "learning_rate": 3.0306054388778814e-05,
+      "loss": 1.3781,
+      "step": 8297
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.09904759952136048,
+      "learning_rate": 3.0246389084555127e-05,
+      "loss": 1.5183,
+      "step": 8298
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.08909127911291063,
+      "learning_rate": 3.0186780739576202e-05,
+      "loss": 1.4428,
+      "step": 8299
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.11642320664644766,
+      "learning_rate": 3.012722936106993e-05,
+      "loss": 1.3976,
+      "step": 8300
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.09675931420968514,
+      "learning_rate": 3.0067734956256863e-05,
+      "loss": 1.4935,
+      "step": 8301
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.09474977351473926,
+      "learning_rate": 3.0008297532351182e-05,
+      "loss": 1.365,
+      "step": 8302
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.10272168744723374,
+      "learning_rate": 2.9948917096559615e-05,
+      "loss": 1.2695,
+      "step": 8303
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.08298955604398278,
+      "learning_rate": 2.9889593656082404e-05,
+      "loss": 1.4956,
+      "step": 8304
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.09179600755809245,
+      "learning_rate": 2.983032721811263e-05,
+      "loss": 1.4941,
+      "step": 8305
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.10682099436596237,
+      "learning_rate": 2.977111778983671e-05,
+      "loss": 1.3633,
+      "step": 8306
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.0867515300642767,
+      "learning_rate": 2.9711965378433793e-05,
+      "loss": 1.3357,
+      "step": 8307
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.08754850931104234,
+      "learning_rate": 2.9652869991076413e-05,
+      "loss": 1.4155,
+      "step": 8308
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.09424777427849733,
+      "learning_rate": 2.9593831634930123e-05,
+      "loss": 1.3722,
+      "step": 8309
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.09294745036970158,
+      "learning_rate": 2.9534850317153415e-05,
+      "loss": 1.3322,
+      "step": 8310
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.09331836902204226,
+      "learning_rate": 2.947592604489807e-05,
+      "loss": 1.3521,
+      "step": 8311
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.09313847602316322,
+      "learning_rate": 2.9417058825308875e-05,
+      "loss": 1.3364,
+      "step": 8312
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.09053115548644763,
+      "learning_rate": 2.9358248665523667e-05,
+      "loss": 1.4487,
+      "step": 8313
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.09041572250814107,
+      "learning_rate": 2.9299495572673307e-05,
+      "loss": 1.4212,
+      "step": 8314
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.0904359258699773,
+      "learning_rate": 2.9240799553881814e-05,
+      "loss": 1.3782,
+      "step": 8315
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.08136115391502427,
+      "learning_rate": 2.918216061626644e-05,
+      "loss": 1.415,
+      "step": 8316
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.08998781295143174,
+      "learning_rate": 2.912357876693711e-05,
+      "loss": 1.3324,
+      "step": 8317
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.09399137616736783,
+      "learning_rate": 2.9065054012997305e-05,
+      "loss": 1.4648,
+      "step": 8318
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.08235405096329945,
+      "learning_rate": 2.900658636154324e-05,
+      "loss": 1.3484,
+      "step": 8319
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.10489144653783913,
+      "learning_rate": 2.8948175819664357e-05,
+      "loss": 1.5477,
+      "step": 8320
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.09116817963939151,
+      "learning_rate": 2.8889822394443043e-05,
+      "loss": 1.3189,
+      "step": 8321
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.09584664025860186,
+      "learning_rate": 2.883152609295503e-05,
+      "loss": 1.4975,
+      "step": 8322
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.08977636111226732,
+      "learning_rate": 2.8773286922268883e-05,
+      "loss": 1.3452,
+      "step": 8323
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.07654042445721576,
+      "learning_rate": 2.871510488944612e-05,
+      "loss": 1.4143,
+      "step": 8324
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.08818203775801058,
+      "learning_rate": 2.8656980001541765e-05,
+      "loss": 1.4636,
+      "step": 8325
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.08432117009970455,
+      "learning_rate": 2.859891226560346e-05,
+      "loss": 1.3565,
+      "step": 8326
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.09712799846018604,
+      "learning_rate": 2.854090168867224e-05,
+      "loss": 1.4048,
+      "step": 8327
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.10385353675645356,
+      "learning_rate": 2.848294827778214e-05,
+      "loss": 1.347,
+      "step": 8328
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.0959182760982914,
+      "learning_rate": 2.8425052039959987e-05,
+      "loss": 1.3536,
+      "step": 8329
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.09054068618274394,
+      "learning_rate": 2.836721298222611e-05,
+      "loss": 1.3895,
+      "step": 8330
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.08613128379624303,
+      "learning_rate": 2.8309431111593675e-05,
+      "loss": 1.329,
+      "step": 8331
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.09153174477614746,
+      "learning_rate": 2.8251706435068803e-05,
+      "loss": 1.3473,
+      "step": 8332
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.0895519519633704,
+      "learning_rate": 2.8194038959650892e-05,
+      "loss": 1.38,
+      "step": 8333
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.09474284361504132,
+      "learning_rate": 2.813642869233235e-05,
+      "loss": 1.3456,
+      "step": 8334
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.08310663147369579,
+      "learning_rate": 2.8078875640098646e-05,
+      "loss": 1.4512,
+      "step": 8335
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.09109971946210933,
+      "learning_rate": 2.802137980992814e-05,
+      "loss": 1.3312,
+      "step": 8336
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.08391879676313042,
+      "learning_rate": 2.796394120879259e-05,
+      "loss": 1.3833,
+      "step": 8337
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.07946902228691093,
+      "learning_rate": 2.790655984365642e-05,
+      "loss": 1.4357,
+      "step": 8338
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.10628108602411328,
+      "learning_rate": 2.7849235721477406e-05,
+      "loss": 1.3521,
+      "step": 8339
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.10121675475351374,
+      "learning_rate": 2.7791968849206427e-05,
+      "loss": 1.4077,
+      "step": 8340
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.08187516155572779,
+      "learning_rate": 2.7734759233787045e-05,
+      "loss": 1.3724,
+      "step": 8341
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.0986900611957909,
+      "learning_rate": 2.7677606882156314e-05,
+      "loss": 1.369,
+      "step": 8342
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.08405140486248891,
+      "learning_rate": 2.7620511801244143e-05,
+      "loss": 1.4385,
+      "step": 8343
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.09081596189763722,
+      "learning_rate": 2.7563473997973433e-05,
+      "loss": 1.3842,
+      "step": 8344
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.08228297439170708,
+      "learning_rate": 2.750649347926021e-05,
+      "loss": 1.3592,
+      "step": 8345
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.08391568513119509,
+      "learning_rate": 2.7449570252013556e-05,
+      "loss": 1.3608,
+      "step": 8346
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.08954952617668473,
+      "learning_rate": 2.7392704323135677e-05,
+      "loss": 1.3169,
+      "step": 8347
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.0917650342566164,
+      "learning_rate": 2.733589569952172e-05,
+      "loss": 1.3512,
+      "step": 8348
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.09179354595681569,
+      "learning_rate": 2.7279144388059896e-05,
+      "loss": 1.5651,
+      "step": 8349
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.08588776856338311,
+      "learning_rate": 2.7222450395631592e-05,
+      "loss": 1.3745,
+      "step": 8350
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.0909950628609119,
+      "learning_rate": 2.7165813729111032e-05,
+      "loss": 1.3828,
+      "step": 8351
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.10197866780298187,
+      "learning_rate": 2.7109234395365667e-05,
+      "loss": 1.364,
+      "step": 8352
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.0879616622479054,
+      "learning_rate": 2.7052712401256006e-05,
+      "loss": 1.4415,
+      "step": 8353
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.09835874727160945,
+      "learning_rate": 2.6996247753635404e-05,
+      "loss": 1.3447,
+      "step": 8354
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.0963073590732697,
+      "learning_rate": 2.6939840459350496e-05,
+      "loss": 1.339,
+      "step": 8355
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.08902332715993876,
+      "learning_rate": 2.6883490525240804e-05,
+      "loss": 1.3539,
+      "step": 8356
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.08103360913138252,
+      "learning_rate": 2.6827197958138928e-05,
+      "loss": 1.2907,
+      "step": 8357
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.09731155956807074,
+      "learning_rate": 2.677096276487062e-05,
+      "loss": 1.4633,
+      "step": 8358
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.0887681860239562,
+      "learning_rate": 2.6714784952254544e-05,
+      "loss": 1.4055,
+      "step": 8359
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.08898853045298434,
+      "learning_rate": 2.6658664527102417e-05,
+      "loss": 1.258,
+      "step": 8360
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.08523520098416243,
+      "learning_rate": 2.660260149621907e-05,
+      "loss": 1.3498,
+      "step": 8361
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.08728423609774599,
+      "learning_rate": 2.6546595866402403e-05,
+      "loss": 1.3685,
+      "step": 8362
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.08238661659369678,
+      "learning_rate": 2.6490647644443143e-05,
+      "loss": 1.4899,
+      "step": 8363
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.09252426038134721,
+      "learning_rate": 2.6434756837125317e-05,
+      "loss": 1.5046,
+      "step": 8364
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.08778490489889275,
+      "learning_rate": 2.6378923451225888e-05,
+      "loss": 1.3537,
+      "step": 8365
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.10171854887120627,
+      "learning_rate": 2.6323147493514833e-05,
+      "loss": 1.3695,
+      "step": 8366
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.09384589952775692,
+      "learning_rate": 2.6267428970755125e-05,
+      "loss": 1.3111,
+      "step": 8367
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.0795961798737711,
+      "learning_rate": 2.621176788970281e-05,
+      "loss": 1.3767,
+      "step": 8368
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.10107183568789642,
+      "learning_rate": 2.6156164257107097e-05,
+      "loss": 1.2705,
+      "step": 8369
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.09333781313598113,
+      "learning_rate": 2.6100618079710037e-05,
+      "loss": 1.37,
+      "step": 8370
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.08833430413490173,
+      "learning_rate": 2.6045129364246856e-05,
+      "loss": 1.4379,
+      "step": 8371
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.08838142607452992,
+      "learning_rate": 2.5989698117445615e-05,
+      "loss": 1.5307,
+      "step": 8372
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.09304373165717744,
+      "learning_rate": 2.593432434602766e-05,
+      "loss": 1.4593,
+      "step": 8373
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.08477827831395272,
+      "learning_rate": 2.5879008056707286e-05,
+      "loss": 1.3841,
+      "step": 8374
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.09963160622039322,
+      "learning_rate": 2.5823749256191687e-05,
+      "loss": 1.3131,
+      "step": 8375
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.09050444438851227,
+      "learning_rate": 2.5768547951181277e-05,
+      "loss": 1.4336,
+      "step": 8376
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.09180191842894916,
+      "learning_rate": 2.571340414836931e-05,
+      "loss": 1.37,
+      "step": 8377
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.0813612383349236,
+      "learning_rate": 2.565831785444228e-05,
+      "loss": 1.4256,
+      "step": 8378
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.10347883114896525,
+      "learning_rate": 2.5603289076079394e-05,
+      "loss": 1.4034,
+      "step": 8379
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.08899944676178057,
+      "learning_rate": 2.5548317819953203e-05,
+      "loss": 1.3462,
+      "step": 8380
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.09492130882777211,
+      "learning_rate": 2.5493404092729267e-05,
+      "loss": 1.4667,
+      "step": 8381
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.08012737818017668,
+      "learning_rate": 2.5438547901065866e-05,
+      "loss": 1.3719,
+      "step": 8382
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.09302578161515027,
+      "learning_rate": 2.538374925161463e-05,
+      "loss": 1.4519,
+      "step": 8383
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.08547204393981939,
+      "learning_rate": 2.5329008151020072e-05,
+      "loss": 1.5377,
+      "step": 8384
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.07712752399790215,
+      "learning_rate": 2.5274324605919664e-05,
+      "loss": 1.3414,
+      "step": 8385
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.08767049163112096,
+      "learning_rate": 2.521969862294404e-05,
+      "loss": 1.4337,
+      "step": 8386
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.08594081612171055,
+      "learning_rate": 2.5165130208716914e-05,
+      "loss": 1.3469,
+      "step": 8387
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.09468859963832922,
+      "learning_rate": 2.5110619369854594e-05,
+      "loss": 1.3351,
+      "step": 8388
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.09186319861919497,
+      "learning_rate": 2.505616611296685e-05,
+      "loss": 1.3704,
+      "step": 8389
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.08461677792305342,
+      "learning_rate": 2.5001770444656457e-05,
+      "loss": 1.4311,
+      "step": 8390
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1280313908368673,
+      "learning_rate": 2.494743237151892e-05,
+      "loss": 1.3177,
+      "step": 8391
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.11643983133341222,
+      "learning_rate": 2.4893151900142906e-05,
+      "loss": 1.3947,
+      "step": 8392
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.10473989822954327,
+      "learning_rate": 2.4838929037110268e-05,
+      "loss": 1.5387,
+      "step": 8393
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.0843967030128136,
+      "learning_rate": 2.4784763788995523e-05,
+      "loss": 1.4826,
+      "step": 8394
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.08342362278756797,
+      "learning_rate": 2.473065616236647e-05,
+      "loss": 1.3826,
+      "step": 8395
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.08230065911414464,
+      "learning_rate": 2.4676606163783978e-05,
+      "loss": 1.395,
+      "step": 8396
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.08570971292153445,
+      "learning_rate": 2.462261379980163e-05,
+      "loss": 1.4211,
+      "step": 8397
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.09686187113928392,
+      "learning_rate": 2.4568679076966194e-05,
+      "loss": 1.4408,
+      "step": 8398
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.0853459619893161,
+      "learning_rate": 2.4514802001817438e-05,
+      "loss": 1.4611,
+      "step": 8399
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.08748074962675632,
+      "learning_rate": 2.4460982580888303e-05,
+      "loss": 1.4843,
+      "step": 8400
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.09367239616088394,
+      "learning_rate": 2.4407220820704402e-05,
+      "loss": 1.371,
+      "step": 8401
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.10033912608729492,
+      "learning_rate": 2.435351672778463e-05,
+      "loss": 1.4728,
+      "step": 8402
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.08858671836179179,
+      "learning_rate": 2.4299870308640726e-05,
+      "loss": 1.4597,
+      "step": 8403
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.09235237842112576,
+      "learning_rate": 2.4246281569777485e-05,
+      "loss": 1.4919,
+      "step": 8404
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.103872125039485,
+      "learning_rate": 2.4192750517692873e-05,
+      "loss": 1.4196,
+      "step": 8405
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.09659718234578298,
+      "learning_rate": 2.4139277158877538e-05,
+      "loss": 1.3337,
+      "step": 8406
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.08205692848700044,
+      "learning_rate": 2.4085861499815398e-05,
+      "loss": 1.2939,
+      "step": 8407
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.10968793439303617,
+      "learning_rate": 2.4032503546983332e-05,
+      "loss": 1.4226,
+      "step": 8408
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.08408076692975283,
+      "learning_rate": 2.397920330685116e-05,
+      "loss": 1.5346,
+      "step": 8409
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.08412734327092307,
+      "learning_rate": 2.392596078588155e-05,
+      "loss": 1.4113,
+      "step": 8410
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.0865364726324679,
+      "learning_rate": 2.3872775990530504e-05,
+      "loss": 1.351,
+      "step": 8411
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.08882596689561814,
+      "learning_rate": 2.3819648927246916e-05,
+      "loss": 1.4098,
+      "step": 8412
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.09977260706687908,
+      "learning_rate": 2.376657960247247e-05,
+      "loss": 1.4846,
+      "step": 8413
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.08832253323200623,
+      "learning_rate": 2.371356802264202e-05,
+      "loss": 1.3652,
+      "step": 8414
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.09015004280653494,
+      "learning_rate": 2.3660614194183584e-05,
+      "loss": 1.5453,
+      "step": 8415
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.09993461728963231,
+      "learning_rate": 2.3607718123517753e-05,
+      "loss": 1.3951,
+      "step": 8416
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.09386030180420961,
+      "learning_rate": 2.3554879817058504e-05,
+      "loss": 1.2722,
+      "step": 8417
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.08852572128057404,
+      "learning_rate": 2.3502099281212774e-05,
+      "loss": 1.3115,
+      "step": 8418
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.08987717291858999,
+      "learning_rate": 2.3449376522380107e-05,
+      "loss": 1.3226,
+      "step": 8419
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.10203851051114089,
+      "learning_rate": 2.3396711546953442e-05,
+      "loss": 1.4404,
+      "step": 8420
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.09785692101309608,
+      "learning_rate": 2.3344104361318675e-05,
+      "loss": 1.2662,
+      "step": 8421
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.09629320552610036,
+      "learning_rate": 2.3291554971854477e-05,
+      "loss": 1.3061,
+      "step": 8422
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.09022413330869145,
+      "learning_rate": 2.32390633849327e-05,
+      "loss": 1.3738,
+      "step": 8423
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.09209515044576586,
+      "learning_rate": 2.3186629606918197e-05,
+      "loss": 1.4061,
+      "step": 8424
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.09556573011727089,
+      "learning_rate": 2.31342536441686e-05,
+      "loss": 1.3278,
+      "step": 8425
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.09352813983137584,
+      "learning_rate": 2.3081935503034777e-05,
+      "loss": 1.5637,
+      "step": 8426
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.09343439987112688,
+      "learning_rate": 2.3029675189860544e-05,
+      "loss": 1.4015,
+      "step": 8427
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.10169587181273342,
+      "learning_rate": 2.297747271098244e-05,
+      "loss": 1.3474,
+      "step": 8428
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.10179498660670094,
+      "learning_rate": 2.29253280727304e-05,
+      "loss": 1.4285,
+      "step": 8429
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.09330307333224806,
+      "learning_rate": 2.2873241281427038e-05,
+      "loss": 1.3877,
+      "step": 8430
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.10039279736866197,
+      "learning_rate": 2.2821212343388075e-05,
+      "loss": 1.4382,
+      "step": 8431
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.10346327438587212,
+      "learning_rate": 2.2769241264922193e-05,
+      "loss": 1.3383,
+      "step": 8432
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.08989579051513284,
+      "learning_rate": 2.2717328052331122e-05,
+      "loss": 1.3829,
+      "step": 8433
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.09469033069439596,
+      "learning_rate": 2.2665472711909385e-05,
+      "loss": 1.4191,
+      "step": 8434
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.08302585181396888,
+      "learning_rate": 2.2613675249944676e-05,
+      "loss": 1.552,
+      "step": 8435
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.07639788909277988,
+      "learning_rate": 2.256193567271775e-05,
+      "loss": 1.5736,
+      "step": 8436
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.10245714094055836,
+      "learning_rate": 2.2510253986502026e-05,
+      "loss": 1.4083,
+      "step": 8437
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.10338545883531776,
+      "learning_rate": 2.2458630197564222e-05,
+      "loss": 1.3372,
+      "step": 8438
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.10610452397828928,
+      "learning_rate": 2.2407064312163827e-05,
+      "loss": 1.4028,
+      "step": 8439
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.07244559862137315,
+      "learning_rate": 2.235555633655345e-05,
+      "loss": 1.3914,
+      "step": 8440
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.09405091787870902,
+      "learning_rate": 2.230410627697843e-05,
+      "loss": 1.3263,
+      "step": 8441
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.09015438680566427,
+      "learning_rate": 2.2252714139677444e-05,
+      "loss": 1.4548,
+      "step": 8442
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.08618967083701604,
+      "learning_rate": 2.2201379930882006e-05,
+      "loss": 1.3307,
+      "step": 8443
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.09062863447448566,
+      "learning_rate": 2.2150103656816356e-05,
+      "loss": 1.3327,
+      "step": 8444
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.08555145550793038,
+      "learning_rate": 2.2098885323698027e-05,
+      "loss": 1.3822,
+      "step": 8445
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.10137210592380712,
+      "learning_rate": 2.2047724937737546e-05,
+      "loss": 1.5497,
+      "step": 8446
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.09072992995784526,
+      "learning_rate": 2.1996622505138065e-05,
+      "loss": 1.5256,
+      "step": 8447
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.08917769482057666,
+      "learning_rate": 2.1945578032096015e-05,
+      "loss": 1.4008,
+      "step": 8448
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.08259821066802236,
+      "learning_rate": 2.1894591524800832e-05,
+      "loss": 1.3465,
+      "step": 8449
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.09017995327516512,
+      "learning_rate": 2.1843662989434688e-05,
+      "loss": 1.535,
+      "step": 8450
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.09148728765294482,
+      "learning_rate": 2.179279243217286e-05,
+      "loss": 1.2695,
+      "step": 8451
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.09181262449336586,
+      "learning_rate": 2.174197985918358e-05,
+      "loss": 1.3005,
+      "step": 8452
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.09799199206773658,
+      "learning_rate": 2.169122527662798e-05,
+      "loss": 1.4007,
+      "step": 8453
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.08347453430194898,
+      "learning_rate": 2.1640528690660298e-05,
+      "loss": 1.3682,
+      "step": 8454
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.08692980607633988,
+      "learning_rate": 2.158989010742779e-05,
+      "loss": 1.4219,
+      "step": 8455
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.08937756989442294,
+      "learning_rate": 2.1539309533070316e-05,
+      "loss": 1.3182,
+      "step": 8456
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.08836818598437016,
+      "learning_rate": 2.1488786973721085e-05,
+      "loss": 1.3727,
+      "step": 8457
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.0946721247446865,
+      "learning_rate": 2.1438322435506196e-05,
+      "loss": 1.4795,
+      "step": 8458
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.10350896930890328,
+      "learning_rate": 2.1387915924544475e-05,
+      "loss": 1.3853,
+      "step": 8459
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.10058033450733786,
+      "learning_rate": 2.133756744694798e-05,
+      "loss": 1.417,
+      "step": 8460
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.09827699073225621,
+      "learning_rate": 2.128727700882166e-05,
+      "loss": 1.4018,
+      "step": 8461
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.09041865873232152,
+      "learning_rate": 2.1237044616263412e-05,
+      "loss": 1.424,
+      "step": 8462
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.0924230597271768,
+      "learning_rate": 2.1186870275363977e-05,
+      "loss": 1.3944,
+      "step": 8463
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.0873473916377948,
+      "learning_rate": 2.1136753992207268e-05,
+      "loss": 1.3969,
+      "step": 8464
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.08325181748499406,
+      "learning_rate": 2.108669577287009e-05,
+      "loss": 1.4619,
+      "step": 8465
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.0866440412000076,
+      "learning_rate": 2.103669562342203e-05,
+      "loss": 1.4301,
+      "step": 8466
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.08572274163520682,
+      "learning_rate": 2.098675354992596e-05,
+      "loss": 1.347,
+      "step": 8467
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.11805279277002773,
+      "learning_rate": 2.093686955843732e-05,
+      "loss": 1.283,
+      "step": 8468
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.09658742123983585,
+      "learning_rate": 2.088704365500482e-05,
+      "loss": 1.2494,
+      "step": 8469
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.08822907080986658,
+      "learning_rate": 2.0837275845670135e-05,
+      "loss": 1.3896,
+      "step": 8470
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.08932206436251997,
+      "learning_rate": 2.0787566136467705e-05,
+      "loss": 1.393,
+      "step": 8471
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.09641547295931442,
+      "learning_rate": 2.0737914533424885e-05,
+      "loss": 1.3631,
+      "step": 8472
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.09249433521253747,
+      "learning_rate": 2.0688321042562186e-05,
+      "loss": 1.4559,
+      "step": 8473
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.08820267488822597,
+      "learning_rate": 2.0638785669893024e-05,
+      "loss": 1.4652,
+      "step": 8474
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.08816880733726157,
+      "learning_rate": 2.0589308421423704e-05,
+      "loss": 1.3932,
+      "step": 8475
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.09785846516443422,
+      "learning_rate": 2.053988930315348e-05,
+      "loss": 1.3651,
+      "step": 8476
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.09819695747342767,
+      "learning_rate": 2.049052832107473e-05,
+      "loss": 1.3235,
+      "step": 8477
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.09234628217230027,
+      "learning_rate": 2.0441225481172443e-05,
+      "loss": 1.4419,
+      "step": 8478
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.10311619376021568,
+      "learning_rate": 2.039198078942489e-05,
+      "loss": 1.3974,
+      "step": 8479
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.09331694810071177,
+      "learning_rate": 2.0342794251803188e-05,
+      "loss": 1.4798,
+      "step": 8480
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.09020593321897641,
+      "learning_rate": 2.029366587427123e-05,
+      "loss": 1.3993,
+      "step": 8481
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.09205109315945209,
+      "learning_rate": 2.0244595662786136e-05,
+      "loss": 1.4413,
+      "step": 8482
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.08248355323819702,
+      "learning_rate": 2.019558362329782e-05,
+      "loss": 1.4507,
+      "step": 8483
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.09469002579477122,
+      "learning_rate": 2.014662976174908e-05,
+      "loss": 1.496,
+      "step": 8484
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.09997222647885515,
+      "learning_rate": 2.0097734084075723e-05,
+      "loss": 1.3537,
+      "step": 8485
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.09214343688886159,
+      "learning_rate": 2.0048896596206677e-05,
+      "loss": 1.4185,
+      "step": 8486
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.08602997561822767,
+      "learning_rate": 2.000011730406348e-05,
+      "loss": 1.3327,
+      "step": 8487
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.08411014919679415,
+      "learning_rate": 1.99513962135609e-05,
+      "loss": 1.3495,
+      "step": 8488
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.09021699901169272,
+      "learning_rate": 1.9902733330606604e-05,
+      "loss": 1.4673,
+      "step": 8489
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.0894410838431081,
+      "learning_rate": 1.9854128661100925e-05,
+      "loss": 1.4469,
+      "step": 8490
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.09476249544207571,
+      "learning_rate": 1.980558221093748e-05,
+      "loss": 1.4729,
+      "step": 8491
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.08738490149073329,
+      "learning_rate": 1.9757093986002728e-05,
+      "loss": 1.4347,
+      "step": 8492
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.08126371645001917,
+      "learning_rate": 1.9708663992175968e-05,
+      "loss": 1.4183,
+      "step": 8493
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.09864540184993932,
+      "learning_rate": 1.9660292235329446e-05,
+      "loss": 1.4789,
+      "step": 8494
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.08836207440306233,
+      "learning_rate": 1.961197872132847e-05,
+      "loss": 1.3434,
+      "step": 8495
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.08445418161336433,
+      "learning_rate": 1.95637234560313e-05,
+      "loss": 1.2764,
+      "step": 8496
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.0985059807913652,
+      "learning_rate": 1.951552644528892e-05,
+      "loss": 1.455,
+      "step": 8497
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.08635710557765165,
+      "learning_rate": 1.946738769494555e-05,
+      "loss": 1.4062,
+      "step": 8498
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.0912158272982496,
+      "learning_rate": 1.9419307210837954e-05,
+      "loss": 1.4382,
+      "step": 8499
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.08528304563372603,
+      "learning_rate": 1.9371284998796147e-05,
+      "loss": 1.3473,
+      "step": 8500
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.09228094156924532,
+      "learning_rate": 1.9323321064643128e-05,
+      "loss": 1.5235,
+      "step": 8501
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.08233938925848544,
+      "learning_rate": 1.9275415414194476e-05,
+      "loss": 1.376,
+      "step": 8502
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.08850689703770695,
+      "learning_rate": 1.922756805325909e-05,
+      "loss": 1.4039,
+      "step": 8503
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.09958912606213574,
+      "learning_rate": 1.9179778987638508e-05,
+      "loss": 1.4,
+      "step": 8504
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.09810559605964786,
+      "learning_rate": 1.913204822312742e-05,
+      "loss": 1.3595,
+      "step": 8505
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.09463852301076034,
+      "learning_rate": 1.9084375765513197e-05,
+      "loss": 1.4617,
+      "step": 8506
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.11013757373202628,
+      "learning_rate": 1.9036761620576436e-05,
+      "loss": 1.4929,
+      "step": 8507
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.12099079593085553,
+      "learning_rate": 1.898920579409047e-05,
+      "loss": 1.4703,
+      "step": 8508
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.08528369129827018,
+      "learning_rate": 1.894170829182157e-05,
+      "loss": 1.4153,
+      "step": 8509
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.08461384742941523,
+      "learning_rate": 1.889426911952896e-05,
+      "loss": 1.3631,
+      "step": 8510
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.09677086610219493,
+      "learning_rate": 1.8846888282964937e-05,
+      "loss": 1.3896,
+      "step": 8511
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.09429266082090436,
+      "learning_rate": 1.8799565787874394e-05,
+      "loss": 1.3049,
+      "step": 8512
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.08097096508478885,
+      "learning_rate": 1.875230163999553e-05,
+      "loss": 1.2469,
+      "step": 8513
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.09900252527706288,
+      "learning_rate": 1.870509584505925e-05,
+      "loss": 1.3851,
+      "step": 8514
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.08627629957622081,
+      "learning_rate": 1.8657948408789262e-05,
+      "loss": 1.3561,
+      "step": 8515
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.10019881440339429,
+      "learning_rate": 1.8610859336902486e-05,
+      "loss": 1.3432,
+      "step": 8516
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.09158849272181173,
+      "learning_rate": 1.8563828635108692e-05,
+      "loss": 1.3539,
+      "step": 8517
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.09867557457914743,
+      "learning_rate": 1.8516856309110375e-05,
+      "loss": 1.4684,
+      "step": 8518
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.08122593100233426,
+      "learning_rate": 1.8469942364603142e-05,
+      "loss": 1.4484,
+      "step": 8519
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.0842885239004775,
+      "learning_rate": 1.8423086807275557e-05,
+      "loss": 1.3915,
+      "step": 8520
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.08729296238568429,
+      "learning_rate": 1.8376289642808854e-05,
+      "loss": 1.2997,
+      "step": 8521
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.08963895604227475,
+      "learning_rate": 1.8329550876877487e-05,
+      "loss": 1.5127,
+      "step": 8522
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.08888042005595129,
+      "learning_rate": 1.8282870515148652e-05,
+      "loss": 1.3661,
+      "step": 8523
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.08759784012714594,
+      "learning_rate": 1.8236248563282542e-05,
+      "loss": 1.4205,
+      "step": 8524
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.09167737995294478,
+      "learning_rate": 1.8189685026932136e-05,
+      "loss": 1.4306,
+      "step": 8525
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.08403460807889258,
+      "learning_rate": 1.8143179911743414e-05,
+      "loss": 1.357,
+      "step": 8526
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.08519773533614822,
+      "learning_rate": 1.8096733223355476e-05,
+      "loss": 1.3824,
+      "step": 8527
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.09879834853603861,
+      "learning_rate": 1.805034496739988e-05,
+      "loss": 1.4378,
+      "step": 8528
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.09648184208305852,
+      "learning_rate": 1.8004015149501563e-05,
+      "loss": 1.3694,
+      "step": 8529
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.08839735081030436,
+      "learning_rate": 1.795774377527809e-05,
+      "loss": 1.3559,
+      "step": 8530
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.08930280050423725,
+      "learning_rate": 1.7911530850339976e-05,
+      "loss": 1.3384,
+      "step": 8531
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.09363327856588406,
+      "learning_rate": 1.7865376380290842e-05,
+      "loss": 1.3748,
+      "step": 8532
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.08800254556910388,
+      "learning_rate": 1.7819280370726944e-05,
+      "loss": 1.3577,
+      "step": 8533
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.09220582337447861,
+      "learning_rate": 1.7773242827237634e-05,
+      "loss": 1.3874,
+      "step": 8534
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.09027578303790597,
+      "learning_rate": 1.7727263755405176e-05,
+      "loss": 1.5059,
+      "step": 8535
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.08826799704652526,
+      "learning_rate": 1.7681343160804608e-05,
+      "loss": 1.4667,
+      "step": 8536
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.0898786378584817,
+      "learning_rate": 1.7635481049003975e-05,
+      "loss": 1.4077,
+      "step": 8537
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.09251862718568461,
+      "learning_rate": 1.7589677425564222e-05,
+      "loss": 1.4512,
+      "step": 8538
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.08630638711926578,
+      "learning_rate": 1.7543932296039232e-05,
+      "loss": 1.4314,
+      "step": 8539
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.08552853445385915,
+      "learning_rate": 1.749824566597569e-05,
+      "loss": 1.3791,
+      "step": 8540
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.0998202597307536,
+      "learning_rate": 1.7452617540913264e-05,
+      "loss": 1.3047,
+      "step": 8541
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.08909459675523979,
+      "learning_rate": 1.7407047926384644e-05,
+      "loss": 1.3275,
+      "step": 8542
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.08463440139479447,
+      "learning_rate": 1.7361536827915137e-05,
+      "loss": 1.3971,
+      "step": 8543
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.094601329627551,
+      "learning_rate": 1.7316084251023213e-05,
+      "loss": 1.395,
+      "step": 8544
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.09133703125568057,
+      "learning_rate": 1.7270690201220242e-05,
+      "loss": 1.2725,
+      "step": 8545
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.11459907690898916,
+      "learning_rate": 1.7225354684010208e-05,
+      "loss": 1.395,
+      "step": 8546
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.09232465716072903,
+      "learning_rate": 1.7180077704890274e-05,
+      "loss": 1.3476,
+      "step": 8547
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.09444109804027652,
+      "learning_rate": 1.7134859269350543e-05,
+      "loss": 1.3807,
+      "step": 8548
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.09289326082336746,
+      "learning_rate": 1.7089699382873746e-05,
+      "loss": 1.4472,
+      "step": 8549
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.08862007020517378,
+      "learning_rate": 1.7044598050935724e-05,
+      "loss": 1.4563,
+      "step": 8550
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.0919549410530479,
+      "learning_rate": 1.6999555279005263e-05,
+      "loss": 1.4505,
+      "step": 8551
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.09517584637890769,
+      "learning_rate": 1.6954571072543777e-05,
+      "loss": 1.3477,
+      "step": 8552
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.0937890999439233,
+      "learning_rate": 1.6909645437005905e-05,
+      "loss": 1.3643,
+      "step": 8553
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.08987385955355238,
+      "learning_rate": 1.686477837783906e-05,
+      "loss": 1.4256,
+      "step": 8554
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.09477203791651483,
+      "learning_rate": 1.6819969900483345e-05,
+      "loss": 1.3805,
+      "step": 8555
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.09633659162771828,
+      "learning_rate": 1.6775220010372182e-05,
+      "loss": 1.4013,
+      "step": 8556
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.09878142652040257,
+      "learning_rate": 1.6730528712931404e-05,
+      "loss": 1.4194,
+      "step": 8557
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.09332498556478858,
+      "learning_rate": 1.6685896013580226e-05,
+      "loss": 1.4105,
+      "step": 8558
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.08565145674426411,
+      "learning_rate": 1.6641321917730268e-05,
+      "loss": 1.4823,
+      "step": 8559
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1172115003327829,
+      "learning_rate": 1.6596806430786425e-05,
+      "loss": 1.3143,
+      "step": 8560
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.09783235169089964,
+      "learning_rate": 1.655234955814644e-05,
+      "loss": 1.4258,
+      "step": 8561
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1005044359265623,
+      "learning_rate": 1.650795130520072e-05,
+      "loss": 1.4208,
+      "step": 8562
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.08384326929780318,
+      "learning_rate": 1.646361167733279e-05,
+      "loss": 1.2791,
+      "step": 8563
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.09126815563432995,
+      "learning_rate": 1.6419330679918855e-05,
+      "loss": 1.4265,
+      "step": 8564
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.0946576406020799,
+      "learning_rate": 1.637510831832828e-05,
+      "loss": 1.2326,
+      "step": 8565
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.0927423035932863,
+      "learning_rate": 1.633094459792317e-05,
+      "loss": 1.2565,
+      "step": 8566
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.09451555563163766,
+      "learning_rate": 1.6286839524058463e-05,
+      "loss": 1.4538,
+      "step": 8567
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.08899655462628817,
+      "learning_rate": 1.6242793102082043e-05,
+      "loss": 1.3785,
+      "step": 8568
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.09187045471105075,
+      "learning_rate": 1.6198805337334756e-05,
+      "loss": 1.3574,
+      "step": 8569
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.08860419206793055,
+      "learning_rate": 1.6154876235150273e-05,
+      "loss": 1.3431,
+      "step": 8570
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.08899383899424537,
+      "learning_rate": 1.611100580085506e-05,
+      "loss": 1.5686,
+      "step": 8571
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.10070643337059408,
+      "learning_rate": 1.6067194039768584e-05,
+      "loss": 1.2775,
+      "step": 8572
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.09644479465809108,
+      "learning_rate": 1.6023440957203262e-05,
+      "loss": 1.3781,
+      "step": 8573
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.07905816945435647,
+      "learning_rate": 1.5979746558464236e-05,
+      "loss": 1.2644,
+      "step": 8574
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.10817332670827906,
+      "learning_rate": 1.593611084884955e-05,
+      "loss": 1.4592,
+      "step": 8575
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.08316527046152339,
+      "learning_rate": 1.5892533833650356e-05,
+      "loss": 1.305,
+      "step": 8576
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.11415884426723807,
+      "learning_rate": 1.5849015518150377e-05,
+      "loss": 1.4799,
+      "step": 8577
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.083268649141457,
+      "learning_rate": 1.5805555907626334e-05,
+      "loss": 1.2927,
+      "step": 8578
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.0914170798789987,
+      "learning_rate": 1.5762155007347956e-05,
+      "loss": 1.3428,
+      "step": 8579
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.10408842767047184,
+      "learning_rate": 1.57188128225777e-05,
+      "loss": 1.4778,
+      "step": 8580
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.08966895390199751,
+      "learning_rate": 1.5675529358570916e-05,
+      "loss": 1.4545,
+      "step": 8581
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.09387839113739459,
+      "learning_rate": 1.5632304620575965e-05,
+      "loss": 1.3898,
+      "step": 8582
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.08271804328233938,
+      "learning_rate": 1.558913861383393e-05,
+      "loss": 1.2611,
+      "step": 8583
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.09070946609853438,
+      "learning_rate": 1.554603134357879e-05,
+      "loss": 1.4368,
+      "step": 8584
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.0922460182899588,
+      "learning_rate": 1.5502982815037648e-05,
+      "loss": 1.4203,
+      "step": 8585
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.09415164483339611,
+      "learning_rate": 1.545999303343004e-05,
+      "loss": 1.43,
+      "step": 8586
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.07968818770023066,
+      "learning_rate": 1.5417062003968808e-05,
+      "loss": 1.3758,
+      "step": 8587
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.0973453065604615,
+      "learning_rate": 1.537418973185939e-05,
+      "loss": 1.4526,
+      "step": 8588
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.09681877333037021,
+      "learning_rate": 1.53313762223003e-05,
+      "loss": 1.3711,
+      "step": 8589
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.10431149269563894,
+      "learning_rate": 1.5288621480482657e-05,
+      "loss": 1.3885,
+      "step": 8590
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.086045199743274,
+      "learning_rate": 1.5245925511590708e-05,
+      "loss": 1.4141,
+      "step": 8591
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.08735933257811349,
+      "learning_rate": 1.5203288320801589e-05,
+      "loss": 1.3915,
+      "step": 8592
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.09089960974137015,
+      "learning_rate": 1.5160709913284998e-05,
+      "loss": 1.3794,
+      "step": 8593
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.10346871676245731,
+      "learning_rate": 1.5118190294203916e-05,
+      "loss": 1.2933,
+      "step": 8594
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.09863144571471517,
+      "learning_rate": 1.5075729468713828e-05,
+      "loss": 1.4533,
+      "step": 8595
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.08689322248559195,
+      "learning_rate": 1.5033327441963395e-05,
+      "loss": 1.4257,
+      "step": 8596
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.09689504666913185,
+      "learning_rate": 1.4990984219093895e-05,
+      "loss": 1.308,
+      "step": 8597
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1088708301464488,
+      "learning_rate": 1.4948699805239719e-05,
+      "loss": 1.3454,
+      "step": 8598
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.09341805770506104,
+      "learning_rate": 1.490647420552782e-05,
+      "loss": 1.4011,
+      "step": 8599
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.09385136498735533,
+      "learning_rate": 1.4864307425078328e-05,
+      "loss": 1.4219,
+      "step": 8600
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.0821232435993565,
+      "learning_rate": 1.4822199469004094e-05,
+      "loss": 1.4294,
+      "step": 8601
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.08799735582406523,
+      "learning_rate": 1.4780150342410814e-05,
+      "loss": 1.395,
+      "step": 8602
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.09669252271897247,
+      "learning_rate": 1.4738160050397132e-05,
+      "loss": 1.3807,
+      "step": 8603
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.09423337439377573,
+      "learning_rate": 1.4696228598054473e-05,
+      "loss": 1.3328,
+      "step": 8604
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.087042175661016,
+      "learning_rate": 1.4654355990467216e-05,
+      "loss": 1.2981,
+      "step": 8605
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.08860134615702277,
+      "learning_rate": 1.4612542232712522e-05,
+      "loss": 1.5621,
+      "step": 8606
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.08828791523559235,
+      "learning_rate": 1.4570787329860502e-05,
+      "loss": 1.3671,
+      "step": 8607
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.08608911561738074,
+      "learning_rate": 1.4529091286973995e-05,
+      "loss": 1.2861,
+      "step": 8608
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.08437715352144702,
+      "learning_rate": 1.44874541091089e-05,
+      "loss": 1.3159,
+      "step": 8609
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.08595818844939189,
+      "learning_rate": 1.444587580131379e-05,
+      "loss": 1.4947,
+      "step": 8610
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.11629088640945023,
+      "learning_rate": 1.4404356368630133e-05,
+      "loss": 1.3563,
+      "step": 8611
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.09872012121874528,
+      "learning_rate": 1.4362895816092403e-05,
+      "loss": 1.2898,
+      "step": 8612
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.09723751176966317,
+      "learning_rate": 1.4321494148727854e-05,
+      "loss": 1.3872,
+      "step": 8613
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.08164864248624706,
+      "learning_rate": 1.4280151371556471e-05,
+      "loss": 1.3384,
+      "step": 8614
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.0825178834649068,
+      "learning_rate": 1.4238867489591301e-05,
+      "loss": 1.3563,
+      "step": 8615
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.09765191426189007,
+      "learning_rate": 1.4197642507838115e-05,
+      "loss": 1.2718,
+      "step": 8616
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.09370234390805968,
+      "learning_rate": 1.4156476431295584e-05,
+      "loss": 1.3856,
+      "step": 8617
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.08484573748097803,
+      "learning_rate": 1.4115369264955213e-05,
+      "loss": 1.4273,
+      "step": 8618
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.09953841084466902,
+      "learning_rate": 1.407432101380146e-05,
+      "loss": 1.3876,
+      "step": 8619
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.10160114969677866,
+      "learning_rate": 1.4033331682811568e-05,
+      "loss": 1.3629,
+      "step": 8620
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.08803516529663949,
+      "learning_rate": 1.3992401276955502e-05,
+      "loss": 1.4832,
+      "step": 8621
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.09318963980304762,
+      "learning_rate": 1.3951529801196294e-05,
+      "loss": 1.4404,
+      "step": 8622
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.08645141350772045,
+      "learning_rate": 1.3910717260489869e-05,
+      "loss": 1.411,
+      "step": 8623
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.0894926625646689,
+      "learning_rate": 1.3869963659784657e-05,
+      "loss": 1.3681,
+      "step": 8624
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.09056101206876106,
+      "learning_rate": 1.3829269004022427e-05,
+      "loss": 1.4809,
+      "step": 8625
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.09887299470866391,
+      "learning_rate": 1.3788633298137287e-05,
+      "loss": 1.4831,
+      "step": 8626
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.13513774910578263,
+      "learning_rate": 1.3748056547056632e-05,
+      "loss": 1.5056,
+      "step": 8627
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.10581823456619087,
+      "learning_rate": 1.3707538755700521e-05,
+      "loss": 1.4962,
+      "step": 8628
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.084364216444337,
+      "learning_rate": 1.3667079928981807e-05,
+      "loss": 1.3145,
+      "step": 8629
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.08748428403689872,
+      "learning_rate": 1.362668007180634e-05,
+      "loss": 1.4716,
+      "step": 8630
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.0791241119174475,
+      "learning_rate": 1.3586339189072649e-05,
+      "loss": 1.2485,
+      "step": 8631
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.08759161498800068,
+      "learning_rate": 1.354605728567232e-05,
+      "loss": 1.3997,
+      "step": 8632
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.09708298148600865,
+      "learning_rate": 1.35058343664895e-05,
+      "loss": 1.4377,
+      "step": 8633
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.09798526414450459,
+      "learning_rate": 1.3465670436401512e-05,
+      "loss": 1.4938,
+      "step": 8634
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.10297317798540645,
+      "learning_rate": 1.3425565500278347e-05,
+      "loss": 1.377,
+      "step": 8635
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.09540758310047924,
+      "learning_rate": 1.338551956298284e-05,
+      "loss": 1.4806,
+      "step": 8636
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.09294268328813406,
+      "learning_rate": 1.334553262937066e-05,
+      "loss": 1.5293,
+      "step": 8637
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.09074500638078706,
+      "learning_rate": 1.3305604704290431e-05,
+      "loss": 1.5192,
+      "step": 8638
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.08647215940428625,
+      "learning_rate": 1.3265735792583499e-05,
+      "loss": 1.3124,
+      "step": 8639
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.08799218164916595,
+      "learning_rate": 1.3225925899084169e-05,
+      "loss": 1.5912,
+      "step": 8640
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.09268037997654895,
+      "learning_rate": 1.3186175028619518e-05,
+      "loss": 1.4546,
+      "step": 8641
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.09057608833206314,
+      "learning_rate": 1.3146483186009417e-05,
+      "loss": 1.414,
+      "step": 8642
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.09004670263797164,
+      "learning_rate": 1.3106850376066626e-05,
+      "loss": 1.3588,
+      "step": 8643
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.11473962684426484,
+      "learning_rate": 1.3067276603596856e-05,
+      "loss": 1.5842,
+      "step": 8644
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.10836472932270103,
+      "learning_rate": 1.3027761873398436e-05,
+      "loss": 1.2715,
+      "step": 8645
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.09167158169420705,
+      "learning_rate": 1.2988306190262755e-05,
+      "loss": 1.3965,
+      "step": 8646
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.08991242846948891,
+      "learning_rate": 1.2948909558974042e-05,
+      "loss": 1.4294,
+      "step": 8647
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.09468638044701046,
+      "learning_rate": 1.2909571984309032e-05,
+      "loss": 1.4412,
+      "step": 8648
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.08676751377745191,
+      "learning_rate": 1.2870293471037741e-05,
+      "loss": 1.4942,
+      "step": 8649
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.0916862759987886,
+      "learning_rate": 1.283107402392275e-05,
+      "loss": 1.4295,
+      "step": 8650
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.09650520392353956,
+      "learning_rate": 1.2791913647719589e-05,
+      "loss": 1.3463,
+      "step": 8651
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.08393566526589911,
+      "learning_rate": 1.2752812347176512e-05,
+      "loss": 1.3341,
+      "step": 8652
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.09170932221535763,
+      "learning_rate": 1.271377012703473e-05,
+      "loss": 1.4208,
+      "step": 8653
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.08959861151947758,
+      "learning_rate": 1.2674786992028287e-05,
+      "loss": 1.3821,
+      "step": 8654
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.0848972599871631,
+      "learning_rate": 1.2635862946883957e-05,
+      "loss": 1.3334,
+      "step": 8655
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.0993068814281763,
+      "learning_rate": 1.2596997996321469e-05,
+      "loss": 1.2615,
+      "step": 8656
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.09646313429632936,
+      "learning_rate": 1.2558192145053326e-05,
+      "loss": 1.3858,
+      "step": 8657
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.09286391088063424,
+      "learning_rate": 1.2519445397784769e-05,
+      "loss": 1.3417,
+      "step": 8658
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.08902398492787911,
+      "learning_rate": 1.2480757759214145e-05,
+      "loss": 1.498,
+      "step": 8659
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.08915222091918729,
+      "learning_rate": 1.2442129234032373e-05,
+      "loss": 1.2159,
+      "step": 8660
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.10864732924798791,
+      "learning_rate": 1.2403559826923205e-05,
+      "loss": 1.448,
+      "step": 8661
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.0973476935490254,
+      "learning_rate": 1.236504954256351e-05,
+      "loss": 1.3799,
+      "step": 8662
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.08686513943637508,
+      "learning_rate": 1.2326598385622723e-05,
+      "loss": 1.4464,
+      "step": 8663
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.08930018544005622,
+      "learning_rate": 1.2288206360763055e-05,
+      "loss": 1.3085,
+      "step": 8664
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.08339424642042234,
+      "learning_rate": 1.2249873472639782e-05,
+      "loss": 1.4383,
+      "step": 8665
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.09666134438716192,
+      "learning_rate": 1.2211599725900913e-05,
+      "loss": 1.3808,
+      "step": 8666
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.09117203971501421,
+      "learning_rate": 1.2173385125187175e-05,
+      "loss": 1.2604,
+      "step": 8667
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.08962445862990973,
+      "learning_rate": 1.213522967513231e-05,
+      "loss": 1.3797,
+      "step": 8668
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.0963226720097203,
+      "learning_rate": 1.2097133380362835e-05,
+      "loss": 1.3911,
+      "step": 8669
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.08829865843109752,
+      "learning_rate": 1.2059096245497946e-05,
+      "loss": 1.5422,
+      "step": 8670
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.09592687958075631,
+      "learning_rate": 1.202111827514979e-05,
+      "loss": 1.3628,
+      "step": 8671
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.07879187528901951,
+      "learning_rate": 1.1983199473923456e-05,
+      "loss": 1.5028,
+      "step": 8672
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.10046305256324412,
+      "learning_rate": 1.1945339846416603e-05,
+      "loss": 1.5293,
+      "step": 8673
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.08752446327580514,
+      "learning_rate": 1.1907539397219835e-05,
+      "loss": 1.3339,
+      "step": 8674
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.09815530338289363,
+      "learning_rate": 1.1869798130916654e-05,
+      "loss": 1.3674,
+      "step": 8675
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.0921843523158939,
+      "learning_rate": 1.1832116052083231e-05,
+      "loss": 1.2522,
+      "step": 8676
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.09920247734847988,
+      "learning_rate": 1.1794493165288745e-05,
+      "loss": 1.4375,
+      "step": 8677
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.10373638226182667,
+      "learning_rate": 1.1756929475095102e-05,
+      "loss": 1.3585,
+      "step": 8678
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.11289756227251307,
+      "learning_rate": 1.1719424986056936e-05,
+      "loss": 1.398,
+      "step": 8679
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.08615684017355264,
+      "learning_rate": 1.168197970272189e-05,
+      "loss": 1.4001,
+      "step": 8680
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.08292246245321147,
+      "learning_rate": 1.1644593629630274e-05,
+      "loss": 1.3374,
+      "step": 8681
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.09281794835484808,
+      "learning_rate": 1.1607266771315295e-05,
+      "loss": 1.4318,
+      "step": 8682
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.09380792402917534,
+      "learning_rate": 1.1569999132302999e-05,
+      "loss": 1.3245,
+      "step": 8683
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.08317011645192439,
+      "learning_rate": 1.1532790717112162e-05,
+      "loss": 1.4646,
+      "step": 8684
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.10464656535720604,
+      "learning_rate": 1.1495641530254452e-05,
+      "loss": 1.4131,
+      "step": 8685
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.10344553432386078,
+      "learning_rate": 1.1458551576234322e-05,
+      "loss": 1.3669,
+      "step": 8686
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.08569223859657236,
+      "learning_rate": 1.1421520859549062e-05,
+      "loss": 1.386,
+      "step": 8687
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.09514608590411519,
+      "learning_rate": 1.1384549384688803e-05,
+      "loss": 1.4228,
+      "step": 8688
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.09771515023676958,
+      "learning_rate": 1.1347637156136459e-05,
+      "loss": 1.395,
+      "step": 8689
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.0772270936421878,
+      "learning_rate": 1.1310784178367729e-05,
+      "loss": 1.2789,
+      "step": 8690
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.0911160570848233,
+      "learning_rate": 1.1273990455851203e-05,
+      "loss": 1.3026,
+      "step": 8691
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.0914312510350616,
+      "learning_rate": 1.12372559930482e-05,
+      "loss": 1.4928,
+      "step": 8692
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.09252646977879138,
+      "learning_rate": 1.1200580794412995e-05,
+      "loss": 1.4267,
+      "step": 8693
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.09167015810592967,
+      "learning_rate": 1.1163964864392472e-05,
+      "loss": 1.3711,
+      "step": 8694
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.08638462067121223,
+      "learning_rate": 1.1127408207426471e-05,
+      "loss": 1.4772,
+      "step": 8695
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.08894168565959135,
+      "learning_rate": 1.109091082794761e-05,
+      "loss": 1.4653,
+      "step": 8696
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.08528838348096536,
+      "learning_rate": 1.1054472730381403e-05,
+      "loss": 1.5519,
+      "step": 8697
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.11797144062561185,
+      "learning_rate": 1.1018093919145988e-05,
+      "loss": 1.3926,
+      "step": 8698
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.09838257455982573,
+      "learning_rate": 1.0981774398652444e-05,
+      "loss": 1.3492,
+      "step": 8699
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.0912813666237488,
+      "learning_rate": 1.0945514173304693e-05,
+      "loss": 1.3605,
+      "step": 8700
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.08843196545973088,
+      "learning_rate": 1.090931324749933e-05,
+      "loss": 1.3818,
+      "step": 8701
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.09639414610841635,
+      "learning_rate": 1.0873171625625899e-05,
+      "loss": 1.5414,
+      "step": 8702
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.09300844643291478,
+      "learning_rate": 1.0837089312066728e-05,
+      "loss": 1.4713,
+      "step": 8703
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.0935356886871057,
+      "learning_rate": 1.0801066311196872e-05,
+      "loss": 1.3597,
+      "step": 8704
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.08624681569480633,
+      "learning_rate": 1.0765102627384226e-05,
+      "loss": 1.4299,
+      "step": 8705
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.09823549145427711,
+      "learning_rate": 1.0729198264989582e-05,
+      "loss": 1.4004,
+      "step": 8706
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.08638494814697543,
+      "learning_rate": 1.0693353228366342e-05,
+      "loss": 1.486,
+      "step": 8707
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.08204185326485428,
+      "learning_rate": 1.0657567521860977e-05,
+      "loss": 1.4409,
+      "step": 8708
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.09744301403720242,
+      "learning_rate": 1.0621841149812572e-05,
+      "loss": 1.3364,
+      "step": 8709
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.10000582529959977,
+      "learning_rate": 1.0586174116552993e-05,
+      "loss": 1.4559,
+      "step": 8710
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.09744949985617589,
+      "learning_rate": 1.0550566426407115e-05,
+      "loss": 1.3111,
+      "step": 8711
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.08195953839293012,
+      "learning_rate": 1.0515018083692484e-05,
+      "loss": 1.443,
+      "step": 8712
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.09149345640565289,
+      "learning_rate": 1.0479529092719375e-05,
+      "loss": 1.307,
+      "step": 8713
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.09314773244510169,
+      "learning_rate": 1.0444099457791012e-05,
+      "loss": 1.3025,
+      "step": 8714
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.07480106948838042,
+      "learning_rate": 1.0408729183203403e-05,
+      "loss": 1.3991,
+      "step": 8715
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.0877240164790989,
+      "learning_rate": 1.0373418273245228e-05,
+      "loss": 1.3505,
+      "step": 8716
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.09580152668794165,
+      "learning_rate": 1.0338166732198062e-05,
+      "loss": 1.4759,
+      "step": 8717
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.09232784222569064,
+      "learning_rate": 1.0302974564336265e-05,
+      "loss": 1.333,
+      "step": 8718
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.09517950722743784,
+      "learning_rate": 1.026784177392709e-05,
+      "loss": 1.4041,
+      "step": 8719
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.0870973350453796,
+      "learning_rate": 1.023276836523046e-05,
+      "loss": 1.4549,
+      "step": 8720
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.08866760352849476,
+      "learning_rate": 1.0197754342499199e-05,
+      "loss": 1.3589,
+      "step": 8721
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.09513227232510038,
+      "learning_rate": 1.0162799709978743e-05,
+      "loss": 1.3897,
+      "step": 8722
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.09371045515382317,
+      "learning_rate": 1.0127904471907589e-05,
+      "loss": 1.4099,
+      "step": 8723
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.08614066611489142,
+      "learning_rate": 1.0093068632516855e-05,
+      "loss": 1.3689,
+      "step": 8724
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.10384855286320732,
+      "learning_rate": 1.0058292196030549e-05,
+      "loss": 1.4494,
+      "step": 8725
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.08381647067968226,
+      "learning_rate": 1.0023575166665355e-05,
+      "loss": 1.4584,
+      "step": 8726
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.10300897484233143,
+      "learning_rate": 9.98891754863085e-06,
+      "loss": 1.3709,
+      "step": 8727
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.10475793362232905,
+      "learning_rate": 9.954319346129503e-06,
+      "loss": 1.2983,
+      "step": 8728
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.08644278946240058,
+      "learning_rate": 9.919780563356295e-06,
+      "loss": 1.3854,
+      "step": 8729
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.09194633055680919,
+      "learning_rate": 9.88530120449932e-06,
+      "loss": 1.4767,
+      "step": 8730
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.09169105507664722,
+      "learning_rate": 9.850881273739231e-06,
+      "loss": 1.5221,
+      "step": 8731
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.0917619755532548,
+      "learning_rate": 9.816520775249583e-06,
+      "loss": 1.4147,
+      "step": 8732
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.08201196705016786,
+      "learning_rate": 9.782219713196705e-06,
+      "loss": 1.329,
+      "step": 8733
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.09267725476057265,
+      "learning_rate": 9.747978091739774e-06,
+      "loss": 1.4925,
+      "step": 8734
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.08967739621306049,
+      "learning_rate": 9.713795915030577e-06,
+      "loss": 1.4223,
+      "step": 8735
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.08957521459912487,
+      "learning_rate": 9.679673187214022e-06,
+      "loss": 1.4221,
+      "step": 8736
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.09355925735987494,
+      "learning_rate": 9.64560991242741e-06,
+      "loss": 1.5056,
+      "step": 8737
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.08319093196250256,
+      "learning_rate": 9.611606094801052e-06,
+      "loss": 1.5389,
+      "step": 8738
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.07375547083039735,
+      "learning_rate": 9.577661738458143e-06,
+      "loss": 1.4192,
+      "step": 8739
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.08483871899049837,
+      "learning_rate": 9.543776847514507e-06,
+      "loss": 1.5044,
+      "step": 8740
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.09976069792945877,
+      "learning_rate": 9.509951426078745e-06,
+      "loss": 1.4148,
+      "step": 8741
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.08695437000305326,
+      "learning_rate": 9.476185478252352e-06,
+      "loss": 1.367,
+      "step": 8742
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.09822966313776028,
+      "learning_rate": 9.442479008129557e-06,
+      "loss": 1.479,
+      "step": 8743
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.0886525612017647,
+      "learning_rate": 9.40883201979742e-06,
+      "loss": 1.4471,
+      "step": 8744
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.08808052951426312,
+      "learning_rate": 9.37524451733568e-06,
+      "loss": 1.2375,
+      "step": 8745
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.09043626699549577,
+      "learning_rate": 9.341716504817021e-06,
+      "loss": 1.42,
+      "step": 8746
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.09434528279859916,
+      "learning_rate": 9.308247986306862e-06,
+      "loss": 1.4542,
+      "step": 8747
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.09964885817378788,
+      "learning_rate": 9.274838965863174e-06,
+      "loss": 1.3686,
+      "step": 8748
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.09345635978911934,
+      "learning_rate": 9.241489447537111e-06,
+      "loss": 1.4048,
+      "step": 8749
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.0984638899196532,
+      "learning_rate": 9.208199435372377e-06,
+      "loss": 1.3195,
+      "step": 8750
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.08103394530212814,
+      "learning_rate": 9.174968933405414e-06,
+      "loss": 1.4755,
+      "step": 8751
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.10063030829037273,
+      "learning_rate": 9.141797945665609e-06,
+      "loss": 1.3979,
+      "step": 8752
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.10028878781779615,
+      "learning_rate": 9.108686476175133e-06,
+      "loss": 1.5094,
+      "step": 8753
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.1027783191091369,
+      "learning_rate": 9.075634528948717e-06,
+      "loss": 1.3391,
+      "step": 8754
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.08672950066908128,
+      "learning_rate": 9.042642107994104e-06,
+      "loss": 1.397,
+      "step": 8755
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.08870879929526027,
+      "learning_rate": 9.0097092173117e-06,
+      "loss": 1.364,
+      "step": 8756
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.09341416581700006,
+      "learning_rate": 8.976835860894761e-06,
+      "loss": 1.3041,
+      "step": 8757
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.11196489344792297,
+      "learning_rate": 8.944022042729317e-06,
+      "loss": 1.4337,
+      "step": 8758
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.10451525028124231,
+      "learning_rate": 8.91126776679413e-06,
+      "loss": 1.3766,
+      "step": 8759
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.08593762869175031,
+      "learning_rate": 8.87857303706069e-06,
+      "loss": 1.4982,
+      "step": 8760
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.07904295751893473,
+      "learning_rate": 8.845937857493491e-06,
+      "loss": 1.3474,
+      "step": 8761
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.08276173777838187,
+      "learning_rate": 8.813362232049592e-06,
+      "loss": 1.4948,
+      "step": 8762
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.0953114667332252,
+      "learning_rate": 8.780846164678836e-06,
+      "loss": 1.4323,
+      "step": 8763
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.08730452540721823,
+      "learning_rate": 8.748389659324008e-06,
+      "loss": 1.4438,
+      "step": 8764
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.09130547003069196,
+      "learning_rate": 8.715992719920574e-06,
+      "loss": 1.5007,
+      "step": 8765
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.10602039868255145,
+      "learning_rate": 8.683655350396724e-06,
+      "loss": 1.4545,
+      "step": 8766
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.08236441196708506,
+      "learning_rate": 8.651377554673434e-06,
+      "loss": 1.4182,
+      "step": 8767
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.08527216278876118,
+      "learning_rate": 8.619159336664683e-06,
+      "loss": 1.4018,
+      "step": 8768
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.08511637576566533,
+      "learning_rate": 8.587000700276792e-06,
+      "loss": 1.5263,
+      "step": 8769
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.08838260890517773,
+      "learning_rate": 8.554901649409252e-06,
+      "loss": 1.4725,
+      "step": 8770
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.08090486237576237,
+      "learning_rate": 8.522862187954172e-06,
+      "loss": 1.438,
+      "step": 8771
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.09265052659328142,
+      "learning_rate": 8.490882319796389e-06,
+      "loss": 1.4177,
+      "step": 8772
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.0914917526601237,
+      "learning_rate": 8.458962048813634e-06,
+      "loss": 1.3132,
+      "step": 8773
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.08859690630090684,
+      "learning_rate": 8.427101378876367e-06,
+      "loss": 1.6076,
+      "step": 8774
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.08901781244111316,
+      "learning_rate": 8.39530031384772e-06,
+      "loss": 1.4397,
+      "step": 8775
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.10086813132432688,
+      "learning_rate": 8.36355885758372e-06,
+      "loss": 1.449,
+      "step": 8776
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.09140538883618672,
+      "learning_rate": 8.331877013933176e-06,
+      "loss": 1.1962,
+      "step": 8777
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.08401749894727048,
+      "learning_rate": 8.300254786737627e-06,
+      "loss": 1.4226,
+      "step": 8778
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.09632136067271782,
+      "learning_rate": 8.268692179831228e-06,
+      "loss": 1.3028,
+      "step": 8779
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.08786148939407268,
+      "learning_rate": 8.237189197041195e-06,
+      "loss": 1.5339,
+      "step": 8780
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.08665394568612342,
+      "learning_rate": 8.205745842187361e-06,
+      "loss": 1.4943,
+      "step": 8781
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.10345725784687435,
+      "learning_rate": 8.17436211908229e-06,
+      "loss": 1.3102,
+      "step": 8782
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.08766312295144872,
+      "learning_rate": 8.14303803153138e-06,
+      "loss": 1.5278,
+      "step": 8783
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.08302435920388034,
+      "learning_rate": 8.111773583332872e-06,
+      "loss": 1.3101,
+      "step": 8784
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.08086046970243763,
+      "learning_rate": 8.080568778277509e-06,
+      "loss": 1.3628,
+      "step": 8785
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.07696214120226756,
+      "learning_rate": 8.049423620149154e-06,
+      "loss": 1.3877,
+      "step": 8786
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.08412889864646508,
+      "learning_rate": 8.018338112724178e-06,
+      "loss": 1.3764,
+      "step": 8787
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.11455424686467591,
+      "learning_rate": 7.987312259771839e-06,
+      "loss": 1.4572,
+      "step": 8788
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.09261780785683074,
+      "learning_rate": 7.956346065054132e-06,
+      "loss": 1.5225,
+      "step": 8789
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.09033658557442632,
+      "learning_rate": 7.925439532325772e-06,
+      "loss": 1.4358,
+      "step": 8790
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.09358776230596985,
+      "learning_rate": 7.894592665334265e-06,
+      "loss": 1.4481,
+      "step": 8791
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.10293383897592415,
+      "learning_rate": 7.863805467820006e-06,
+      "loss": 1.2969,
+      "step": 8792
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.09437810328651862,
+      "learning_rate": 7.833077943515955e-06,
+      "loss": 1.3577,
+      "step": 8793
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.09550001282233132,
+      "learning_rate": 7.80241009614796e-06,
+      "loss": 1.362,
+      "step": 8794
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.103722456788033,
+      "learning_rate": 7.771801929434608e-06,
+      "loss": 1.5316,
+      "step": 8795
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.09672039793209836,
+      "learning_rate": 7.74125344708726e-06,
+      "loss": 1.4404,
+      "step": 8796
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.11381378663777032,
+      "learning_rate": 7.71076465281001e-06,
+      "loss": 1.3989,
+      "step": 8797
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.0943974351742112,
+      "learning_rate": 7.680335550299678e-06,
+      "loss": 1.5955,
+      "step": 8798
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.08554720301543454,
+      "learning_rate": 7.649966143245979e-06,
+      "loss": 1.3918,
+      "step": 8799
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.10710528712328721,
+      "learning_rate": 7.619656435331301e-06,
+      "loss": 1.494,
+      "step": 8800
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.09189467724009155,
+      "learning_rate": 7.589406430230705e-06,
+      "loss": 1.3021,
+      "step": 8801
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.10057691654011505,
+      "learning_rate": 7.559216131612256e-06,
+      "loss": 1.3472,
+      "step": 8802
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.0821028096426198,
+      "learning_rate": 7.529085543136472e-06,
+      "loss": 1.4207,
+      "step": 8803
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.0782184005263043,
+      "learning_rate": 7.499014668456872e-06,
+      "loss": 1.3374,
+      "step": 8804
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.0876206795684908,
+      "learning_rate": 7.469003511219707e-06,
+      "loss": 1.4071,
+      "step": 8805
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.08655313629027807,
+      "learning_rate": 7.4390520750638455e-06,
+      "loss": 1.3533,
+      "step": 8806
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.0862730449274471,
+      "learning_rate": 7.409160363621048e-06,
+      "loss": 1.3673,
+      "step": 8807
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.08486986348752033,
+      "learning_rate": 7.379328380515804e-06,
+      "loss": 1.3241,
+      "step": 8808
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.09760545363166366,
+      "learning_rate": 7.349556129365276e-06,
+      "loss": 1.3747,
+      "step": 8809
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.10666679386119536,
+      "learning_rate": 7.31984361377952e-06,
+      "loss": 1.3719,
+      "step": 8810
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.10084244049618121,
+      "learning_rate": 7.29019083736121e-06,
+      "loss": 1.3092,
+      "step": 8811
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.09202231083032954,
+      "learning_rate": 7.260597803705971e-06,
+      "loss": 1.2991,
+      "step": 8812
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.10290993988751386,
+      "learning_rate": 7.23106451640193e-06,
+      "loss": 1.4151,
+      "step": 8813
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.08561006665964292,
+      "learning_rate": 7.201590979030115e-06,
+      "loss": 1.4943,
+      "step": 8814
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.0865490043036091,
+      "learning_rate": 7.172177195164386e-06,
+      "loss": 1.3253,
+      "step": 8815
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.09230136981204798,
+      "learning_rate": 7.1428231683711705e-06,
+      "loss": 1.3461,
+      "step": 8816
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.09945844805144954,
+      "learning_rate": 7.113528902209787e-06,
+      "loss": 1.2755,
+      "step": 8817
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.09948172641397596,
+      "learning_rate": 7.084294400232283e-06,
+      "loss": 1.3729,
+      "step": 8818
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.0992359215039714,
+      "learning_rate": 7.055119665983378e-06,
+      "loss": 1.4723,
+      "step": 8819
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.10929611354225881,
+      "learning_rate": 7.026004703000688e-06,
+      "loss": 1.389,
+      "step": 8820
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.08294067051563485,
+      "learning_rate": 6.996949514814499e-06,
+      "loss": 1.2973,
+      "step": 8821
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.11065562311316486,
+      "learning_rate": 6.967954104947771e-06,
+      "loss": 1.4557,
+      "step": 8822
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.10295369467271562,
+      "learning_rate": 6.939018476916359e-06,
+      "loss": 1.3141,
+      "step": 8823
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.09378386174057356,
+      "learning_rate": 6.910142634228789e-06,
+      "loss": 1.3767,
+      "step": 8824
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.09977427325258244,
+      "learning_rate": 6.8813265803863715e-06,
+      "loss": 1.5256,
+      "step": 8825
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.10057069838924039,
+      "learning_rate": 6.852570318883145e-06,
+      "loss": 1.4013,
+      "step": 8826
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.09187264561880268,
+      "learning_rate": 6.8238738532059306e-06,
+      "loss": 1.4982,
+      "step": 8827
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.11291029430421218,
+      "learning_rate": 6.795237186834169e-06,
+      "loss": 1.3988,
+      "step": 8828
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.08897163991167156,
+      "learning_rate": 6.766660323240303e-06,
+      "loss": 1.5356,
+      "step": 8829
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.09496027710275812,
+      "learning_rate": 6.738143265889285e-06,
+      "loss": 1.4172,
+      "step": 8830
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.08272734550929048,
+      "learning_rate": 6.709686018238958e-06,
+      "loss": 1.3842,
+      "step": 8831
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.102791559328205,
+      "learning_rate": 6.681288583739786e-06,
+      "loss": 1.4251,
+      "step": 8832
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.108585952617785,
+      "learning_rate": 6.652950965835181e-06,
+      "loss": 1.3605,
+      "step": 8833
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.10555789963522569,
+      "learning_rate": 6.624673167961004e-06,
+      "loss": 1.4318,
+      "step": 8834
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.0852695401594645,
+      "learning_rate": 6.59645519354618e-06,
+      "loss": 1.3621,
+      "step": 8835
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.09853582405605046,
+      "learning_rate": 6.568297046012195e-06,
+      "loss": 1.3444,
+      "step": 8836
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.08919377889013902,
+      "learning_rate": 6.540198728773262e-06,
+      "loss": 1.3925,
+      "step": 8837
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.08643002957608505,
+      "learning_rate": 6.512160245236431e-06,
+      "loss": 1.3932,
+      "step": 8838
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.08706342127575102,
+      "learning_rate": 6.484181598801541e-06,
+      "loss": 1.4577,
+      "step": 8839
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.0957071682990628,
+      "learning_rate": 6.4562627928610455e-06,
+      "loss": 1.3636,
+      "step": 8840
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.09110174617138717,
+      "learning_rate": 6.4284038308001224e-06,
+      "loss": 1.3166,
+      "step": 8841
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.09374793989183004,
+      "learning_rate": 6.400604715996905e-06,
+      "loss": 1.3368,
+      "step": 8842
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.11342827940082034,
+      "learning_rate": 6.372865451822085e-06,
+      "loss": 1.2744,
+      "step": 8843
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.09499693290134015,
+      "learning_rate": 6.345186041639028e-06,
+      "loss": 1.4215,
+      "step": 8844
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.08185468984627872,
+      "learning_rate": 6.317566488804105e-06,
+      "loss": 1.4175,
+      "step": 8845
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.09568579794257903,
+      "learning_rate": 6.290006796666248e-06,
+      "loss": 1.5329,
+      "step": 8846
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.08883975464927474,
+      "learning_rate": 6.262506968567061e-06,
+      "loss": 1.3307,
+      "step": 8847
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.08740844984942242,
+      "learning_rate": 6.2350670078411555e-06,
+      "loss": 1.3654,
+      "step": 8848
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.09062562990586284,
+      "learning_rate": 6.207686917815592e-06,
+      "loss": 1.4515,
+      "step": 8849
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.08741639466532415,
+      "learning_rate": 6.1803667018103805e-06,
+      "loss": 1.4691,
+      "step": 8850
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.10116002798678095,
+      "learning_rate": 6.15310636313815e-06,
+      "loss": 1.3397,
+      "step": 8851
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.09585867618812836,
+      "learning_rate": 6.125905905104368e-06,
+      "loss": 1.3576,
+      "step": 8852
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.11018206546669015,
+      "learning_rate": 6.098765331007061e-06,
+      "loss": 1.3525,
+      "step": 8853
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.08725015747575635,
+      "learning_rate": 6.0716846441372655e-06,
+      "loss": 1.3746,
+      "step": 8854
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.09572859468974974,
+      "learning_rate": 6.04466384777852e-06,
+      "loss": 1.4461,
+      "step": 8855
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.09604637593854443,
+      "learning_rate": 6.017702945207149e-06,
+      "loss": 1.529,
+      "step": 8856
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.09035469492567368,
+      "learning_rate": 5.990801939692314e-06,
+      "loss": 1.4351,
+      "step": 8857
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.0945266969821737,
+      "learning_rate": 5.963960834495907e-06,
+      "loss": 1.3995,
+      "step": 8858
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.08392027138389029,
+      "learning_rate": 5.937179632872436e-06,
+      "loss": 1.3874,
+      "step": 8859
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.09700625587597556,
+      "learning_rate": 5.9104583380691914e-06,
+      "loss": 1.4869,
+      "step": 8860
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.08440155967829839,
+      "learning_rate": 5.883796953326359e-06,
+      "loss": 1.3654,
+      "step": 8861
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.08299105068275839,
+      "learning_rate": 5.85719548187652e-06,
+      "loss": 1.4762,
+      "step": 8862
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.09878299275346496,
+      "learning_rate": 5.8306539269453725e-06,
+      "loss": 1.4466,
+      "step": 8863
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.08433678896475492,
+      "learning_rate": 5.804172291751064e-06,
+      "loss": 1.3548,
+      "step": 8864
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.09183326720148738,
+      "learning_rate": 5.777750579504581e-06,
+      "loss": 1.3366,
+      "step": 8865
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.08709653574788266,
+      "learning_rate": 5.7513887934096954e-06,
+      "loss": 1.4318,
+      "step": 8866
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.09201497815107694,
+      "learning_rate": 5.725086936662905e-06,
+      "loss": 1.3497,
+      "step": 8867
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.11283563450980807,
+      "learning_rate": 5.698845012453324e-06,
+      "loss": 1.4845,
+      "step": 8868
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.09315366127124443,
+      "learning_rate": 5.672663023962854e-06,
+      "loss": 1.4852,
+      "step": 8869
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.08316494710549423,
+      "learning_rate": 5.646540974366287e-06,
+      "loss": 1.4266,
+      "step": 8870
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.09282115286451564,
+      "learning_rate": 5.620478866830814e-06,
+      "loss": 1.4259,
+      "step": 8871
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.10159676345240197,
+      "learning_rate": 5.594476704516738e-06,
+      "loss": 1.3527,
+      "step": 8872
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.09893435119118751,
+      "learning_rate": 5.5685344905768156e-06,
+      "loss": 1.5138,
+      "step": 8873
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.09329564540974669,
+      "learning_rate": 5.542652228156697e-06,
+      "loss": 1.464,
+      "step": 8874
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.09485592930638007,
+      "learning_rate": 5.516829920394595e-06,
+      "loss": 1.4764,
+      "step": 8875
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.09133805600317245,
+      "learning_rate": 5.4910675704216154e-06,
+      "loss": 1.5208,
+      "step": 8876
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.08495275003072635,
+      "learning_rate": 5.4653651813615366e-06,
+      "loss": 1.388,
+      "step": 8877
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.08760989309934672,
+      "learning_rate": 5.439722756330812e-06,
+      "loss": 1.4567,
+      "step": 8878
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.09048100636363733,
+      "learning_rate": 5.41414029843873e-06,
+      "loss": 1.4263,
+      "step": 8879
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.09197041411794825,
+      "learning_rate": 5.388617810787255e-06,
+      "loss": 1.3345,
+      "step": 8880
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.10020996050813048,
+      "learning_rate": 5.3631552964710784e-06,
+      "loss": 1.4928,
+      "step": 8881
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.09509674597731703,
+      "learning_rate": 5.337752758577563e-06,
+      "loss": 1.5502,
+      "step": 8882
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.09385515688158769,
+      "learning_rate": 5.312410200186857e-06,
+      "loss": 1.4702,
+      "step": 8883
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.08914837524362412,
+      "learning_rate": 5.287127624371946e-06,
+      "loss": 1.4887,
+      "step": 8884
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.09664353503901052,
+      "learning_rate": 5.2619050341982665e-06,
+      "loss": 1.5038,
+      "step": 8885
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.09159510331470239,
+      "learning_rate": 5.236742432724262e-06,
+      "loss": 1.5056,
+      "step": 8886
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.08356282150014831,
+      "learning_rate": 5.2116398230009355e-06,
+      "loss": 1.3525,
+      "step": 8887
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.08806156912501846,
+      "learning_rate": 5.1865972080720195e-06,
+      "loss": 1.4004,
+      "step": 8888
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.09457677127411936,
+      "learning_rate": 5.16161459097414e-06,
+      "loss": 1.4904,
+      "step": 8889
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.09236016076443747,
+      "learning_rate": 5.136691974736429e-06,
+      "loss": 1.4403,
+      "step": 8890
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.09057456912244687,
+      "learning_rate": 5.11182936238086e-06,
+      "loss": 1.2834,
+      "step": 8891
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.09408635409281225,
+      "learning_rate": 5.087026756922187e-06,
+      "loss": 1.4948,
+      "step": 8892
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.08715396491466379,
+      "learning_rate": 5.062284161367669e-06,
+      "loss": 1.2594,
+      "step": 8893
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.09798288199490669,
+      "learning_rate": 5.037601578717521e-06,
+      "loss": 1.3734,
+      "step": 8894
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.09873923966438046,
+      "learning_rate": 5.01297901196468e-06,
+      "loss": 1.5297,
+      "step": 8895
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.09524621935613473,
+      "learning_rate": 4.9884164640944806e-06,
+      "loss": 1.3661,
+      "step": 8896
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.08793095352213838,
+      "learning_rate": 4.963913938085374e-06,
+      "loss": 1.3534,
+      "step": 8897
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.08738144033221784,
+      "learning_rate": 4.939471436908427e-06,
+      "loss": 1.4034,
+      "step": 8898
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.08983226630169384,
+      "learning_rate": 4.915088963527214e-06,
+      "loss": 1.2978,
+      "step": 8899
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.09660666651070206,
+      "learning_rate": 4.890766520898315e-06,
+      "loss": 1.3553,
+      "step": 8900
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.0857590426342619,
+      "learning_rate": 4.866504111970871e-06,
+      "loss": 1.4055,
+      "step": 8901
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.09299660118242754,
+      "learning_rate": 4.8423017396868055e-06,
+      "loss": 1.3946,
+      "step": 8902
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.09265501470007696,
+      "learning_rate": 4.818159406980715e-06,
+      "loss": 1.424,
+      "step": 8903
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.0873024345487018,
+      "learning_rate": 4.794077116779927e-06,
+      "loss": 1.2523,
+      "step": 8904
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.0941279091102085,
+      "learning_rate": 4.770054872004548e-06,
+      "loss": 1.3808,
+      "step": 8905
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.10765797937551637,
+      "learning_rate": 4.74609267556736e-06,
+      "loss": 1.3262,
+      "step": 8906
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.10585056616671595,
+      "learning_rate": 4.72219053037376e-06,
+      "loss": 1.2806,
+      "step": 8907
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.10284106309103908,
+      "learning_rate": 4.6983484393220974e-06,
+      "loss": 1.3896,
+      "step": 8908
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.0927126007379314,
+      "learning_rate": 4.674566405303227e-06,
+      "loss": 1.3906,
+      "step": 8909
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.09260180558774815,
+      "learning_rate": 4.65084443120084e-06,
+      "loss": 1.3763,
+      "step": 8910
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.09414960587863648,
+      "learning_rate": 4.627182519891304e-06,
+      "loss": 1.4103,
+      "step": 8911
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.09675803044725435,
+      "learning_rate": 4.603580674243657e-06,
+      "loss": 1.3207,
+      "step": 8912
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.09529069911101469,
+      "learning_rate": 4.580038897119776e-06,
+      "loss": 1.5118,
+      "step": 8913
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.07988399140996678,
+      "learning_rate": 4.5565571913741e-06,
+      "loss": 1.2666,
+      "step": 8914
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.08536402815347709,
+      "learning_rate": 4.533135559853962e-06,
+      "loss": 1.4486,
+      "step": 8915
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.08185123235807608,
+      "learning_rate": 4.509774005399314e-06,
+      "loss": 1.388,
+      "step": 8916
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.09593022780045278,
+      "learning_rate": 4.486472530842723e-06,
+      "loss": 1.5757,
+      "step": 8917
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.08176639351464489,
+      "learning_rate": 4.463231139009649e-06,
+      "loss": 1.4924,
+      "step": 8918
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.07809646949759338,
+      "learning_rate": 4.440049832718174e-06,
+      "loss": 1.3543,
+      "step": 8919
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.09367404408882468,
+      "learning_rate": 4.416928614779103e-06,
+      "loss": 1.4076,
+      "step": 8920
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.08363712680509189,
+      "learning_rate": 4.3938674879959726e-06,
+      "loss": 1.364,
+      "step": 8921
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.0999487851449133,
+      "learning_rate": 4.370866455165046e-06,
+      "loss": 1.3693,
+      "step": 8922
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.09668059541295229,
+      "learning_rate": 4.3479255190752574e-06,
+      "loss": 1.3714,
+      "step": 8923
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.08567231889954466,
+      "learning_rate": 4.3250446825082166e-06,
+      "loss": 1.4526,
+      "step": 8924
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.08982678310195207,
+      "learning_rate": 4.302223948238426e-06,
+      "loss": 1.4021,
+      "step": 8925
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.09531371167655645,
+      "learning_rate": 4.279463319032894e-06,
+      "loss": 1.321,
+      "step": 8926
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.08096601838384176,
+      "learning_rate": 4.256762797651414e-06,
+      "loss": 1.2234,
+      "step": 8927
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.09513159682892831,
+      "learning_rate": 4.2341223868465615e-06,
+      "loss": 1.4223,
+      "step": 8928
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.10007520400593102,
+      "learning_rate": 4.2115420893635295e-06,
+      "loss": 1.4104,
+      "step": 8929
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.07827562474694724,
+      "learning_rate": 4.189021907940238e-06,
+      "loss": 1.3046,
+      "step": 8930
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.0964464664312945,
+      "learning_rate": 4.166561845307393e-06,
+      "loss": 1.4174,
+      "step": 8931
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.09301449302503279,
+      "learning_rate": 4.144161904188315e-06,
+      "loss": 1.4113,
+      "step": 8932
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.09578933773852154,
+      "learning_rate": 4.1218220872990546e-06,
+      "loss": 1.3976,
+      "step": 8933
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.08666688349596204,
+      "learning_rate": 4.099542397348444e-06,
+      "loss": 1.4173,
+      "step": 8934
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.07548291457470417,
+      "learning_rate": 4.077322837037933e-06,
+      "loss": 1.2473,
+      "step": 8935
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.08139822299567276,
+      "learning_rate": 4.0551634090617575e-06,
+      "loss": 1.3223,
+      "step": 8936
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.08647643102992054,
+      "learning_rate": 4.033064116106766e-06,
+      "loss": 1.3527,
+      "step": 8937
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.09598127034315897,
+      "learning_rate": 4.01102496085265e-06,
+      "loss": 1.4437,
+      "step": 8938
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.0811119581027814,
+      "learning_rate": 3.989045945971659e-06,
+      "loss": 1.3237,
+      "step": 8939
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.10572601439524844,
+      "learning_rate": 3.967127074128885e-06,
+      "loss": 1.3066,
+      "step": 8940
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.08475692276958388,
+      "learning_rate": 3.945268347981979e-06,
+      "loss": 1.2337,
+      "step": 8941
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.08313401249782255,
+      "learning_rate": 3.923469770181543e-06,
+      "loss": 1.387,
+      "step": 8942
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.08968561187112413,
+      "learning_rate": 3.901731343370574e-06,
+      "loss": 1.4705,
+      "step": 8943
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.08126894441730576,
+      "learning_rate": 3.880053070184964e-06,
+      "loss": 1.4054,
+      "step": 8944
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.08610883749970781,
+      "learning_rate": 3.858434953253332e-06,
+      "loss": 1.4202,
+      "step": 8945
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.0860092367603009,
+      "learning_rate": 3.836876995196914e-06,
+      "loss": 1.478,
+      "step": 8946
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.10382181892590997,
+      "learning_rate": 3.815379198629732e-06,
+      "loss": 1.4004,
+      "step": 8947
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.08284626907086795,
+      "learning_rate": 3.793941566158421e-06,
+      "loss": 1.3705,
+      "step": 8948
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.08119746209465538,
+      "learning_rate": 3.7725641003823476e-06,
+      "loss": 1.437,
+      "step": 8949
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.16574799900803355,
+      "learning_rate": 3.75124680389366e-06,
+      "loss": 1.4948,
+      "step": 8950
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.08457090061547087,
+      "learning_rate": 3.7299896792771236e-06,
+      "loss": 1.3338,
+      "step": 8951
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.08507021365850548,
+      "learning_rate": 3.708792729110233e-06,
+      "loss": 1.2983,
+      "step": 8952
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.09508738545345005,
+      "learning_rate": 3.687655955963154e-06,
+      "loss": 1.3152,
+      "step": 8953
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.08769842563914661,
+      "learning_rate": 3.666579362398892e-06,
+      "loss": 1.3787,
+      "step": 8954
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.09713108114963706,
+      "learning_rate": 3.645562950973014e-06,
+      "loss": 1.5562,
+      "step": 8955
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.09295817353089024,
+      "learning_rate": 3.624606724233759e-06,
+      "loss": 1.3034,
+      "step": 8956
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.09973446153914593,
+      "learning_rate": 3.6037106847223168e-06,
+      "loss": 1.4841,
+      "step": 8957
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.08914007872724189,
+      "learning_rate": 3.5828748349722164e-06,
+      "loss": 1.4014,
+      "step": 8958
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.09706649081960167,
+      "learning_rate": 3.5620991775099363e-06,
+      "loss": 1.4964,
+      "step": 8959
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.08729316376152077,
+      "learning_rate": 3.5413837148546847e-06,
+      "loss": 1.4441,
+      "step": 8960
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.07420659828466358,
+      "learning_rate": 3.520728449518118e-06,
+      "loss": 1.4565,
+      "step": 8961
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.10539629443071484,
+      "learning_rate": 3.5001333840049e-06,
+      "loss": 1.4997,
+      "step": 8962
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.08586868055644019,
+      "learning_rate": 3.4795985208121994e-06,
+      "loss": 1.3828,
+      "step": 8963
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.10015038752842724,
+      "learning_rate": 3.459123862429969e-06,
+      "loss": 1.3812,
+      "step": 8964
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.10194468458724672,
+      "learning_rate": 3.438709411340779e-06,
+      "loss": 1.3115,
+      "step": 8965
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.09206340021786165,
+      "learning_rate": 3.4183551700199823e-06,
+      "loss": 1.3946,
+      "step": 8966
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.08262825186118691,
+      "learning_rate": 3.3980611409356044e-06,
+      "loss": 1.3658,
+      "step": 8967
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.08863719060179712,
+      "learning_rate": 3.377827326548344e-06,
+      "loss": 1.5557,
+      "step": 8968
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.08816582721504075,
+      "learning_rate": 3.3576537293116825e-06,
+      "loss": 1.3183,
+      "step": 8969
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.0976051244422402,
+      "learning_rate": 3.337540351671664e-06,
+      "loss": 1.375,
+      "step": 8970
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.09067808571609655,
+      "learning_rate": 3.317487196067115e-06,
+      "loss": 1.4822,
+      "step": 8971
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.09566472825698379,
+      "learning_rate": 3.2974942649295904e-06,
+      "loss": 1.475,
+      "step": 8972
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.08622749394764569,
+      "learning_rate": 3.277561560683262e-06,
+      "loss": 1.4134,
+      "step": 8973
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.08993542829439416,
+      "learning_rate": 3.257689085745086e-06,
+      "loss": 1.3753,
+      "step": 8974
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.09586665598997336,
+      "learning_rate": 3.237876842524634e-06,
+      "loss": 1.4037,
+      "step": 8975
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.08857577090180915,
+      "learning_rate": 3.2181248334242076e-06,
+      "loss": 1.3484,
+      "step": 8976
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.09287873616748278,
+      "learning_rate": 3.1984330608387793e-06,
+      "loss": 1.344,
+      "step": 8977
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.09875558892977225,
+      "learning_rate": 3.1788015271561053e-06,
+      "loss": 1.4483,
+      "step": 8978
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.09041828728366473,
+      "learning_rate": 3.1592302347565605e-06,
+      "loss": 1.3532,
+      "step": 8979
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.07800414751561797,
+      "learning_rate": 3.1397191860132456e-06,
+      "loss": 1.3968,
+      "step": 8980
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.08232355971655188,
+      "learning_rate": 3.1202683832918797e-06,
+      "loss": 1.3541,
+      "step": 8981
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.0951746586827946,
+      "learning_rate": 3.1008778289509654e-06,
+      "loss": 1.6037,
+      "step": 8982
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.08960082679565905,
+      "learning_rate": 3.0815475253417325e-06,
+      "loss": 1.4719,
+      "step": 8983
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.11251667581254801,
+      "learning_rate": 3.0622774748079175e-06,
+      "loss": 1.3597,
+      "step": 8984
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.09527485741233266,
+      "learning_rate": 3.043067679686262e-06,
+      "loss": 1.3548,
+      "step": 8985
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.0836889698103457,
+      "learning_rate": 3.0239181423058483e-06,
+      "loss": 1.4547,
+      "step": 8986
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.0796557012099375,
+      "learning_rate": 3.004828864988707e-06,
+      "loss": 1.277,
+      "step": 8987
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.09149789312059337,
+      "learning_rate": 2.9857998500494866e-06,
+      "loss": 1.3898,
+      "step": 8988
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.08092008180043826,
+      "learning_rate": 2.9668310997955083e-06,
+      "loss": 1.3283,
+      "step": 8989
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.09915783393849893,
+      "learning_rate": 2.9479226165268215e-06,
+      "loss": 1.4008,
+      "step": 8990
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.08577745745175479,
+      "learning_rate": 2.9290744025360915e-06,
+      "loss": 1.3352,
+      "step": 8991
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.08612397942263612,
+      "learning_rate": 2.910286460108713e-06,
+      "loss": 1.4729,
+      "step": 8992
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.07870393466324724,
+      "learning_rate": 2.891558791522864e-06,
+      "loss": 1.4015,
+      "step": 8993
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.08394931460604015,
+      "learning_rate": 2.872891399049338e-06,
+      "loss": 1.2502,
+      "step": 8994
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.10427009072509341,
+      "learning_rate": 2.854284284951547e-06,
+      "loss": 1.428,
+      "step": 8995
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.08846241604213603,
+      "learning_rate": 2.8357374514856872e-06,
+      "loss": 1.2706,
+      "step": 8996
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.09327150463271378,
+      "learning_rate": 2.817250900900681e-06,
+      "loss": 1.3801,
+      "step": 8997
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.11128758227648324,
+      "learning_rate": 2.798824635438069e-06,
+      "loss": 1.3507,
+      "step": 8998
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.08923177417576252,
+      "learning_rate": 2.780458657332008e-06,
+      "loss": 1.3579,
+      "step": 8999
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.08947199378843616,
+      "learning_rate": 2.762152968809606e-06,
+      "loss": 1.4248,
+      "step": 9000
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.09648030606466207,
+      "learning_rate": 2.743907572090365e-06,
+      "loss": 1.4384,
+      "step": 9001
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.08206314363811901,
+      "learning_rate": 2.7257224693866266e-06,
+      "loss": 1.4017,
+      "step": 9002
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.09363077823329279,
+      "learning_rate": 2.7075976629033495e-06,
+      "loss": 1.3157,
+      "step": 9003
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.09298672001535016,
+      "learning_rate": 2.689533154838386e-06,
+      "loss": 1.3577,
+      "step": 9004
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.11480343177435744,
+      "learning_rate": 2.671528947381929e-06,
+      "loss": 1.5121,
+      "step": 9005
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.0905443724363777,
+      "learning_rate": 2.653585042717177e-06,
+      "loss": 1.3501,
+      "step": 9006
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.0913932632121586,
+      "learning_rate": 2.635701443019889e-06,
+      "loss": 1.2952,
+      "step": 9007
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.10332073678504636,
+      "learning_rate": 2.617878150458386e-06,
+      "loss": 1.4585,
+      "step": 9008
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.09795339200836412,
+      "learning_rate": 2.600115167193995e-06,
+      "loss": 1.4621,
+      "step": 9009
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.08852957559404895,
+      "learning_rate": 2.582412495380382e-06,
+      "loss": 1.4335,
+      "step": 9010
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.09455078692555385,
+      "learning_rate": 2.5647701371641075e-06,
+      "loss": 1.4161,
+      "step": 9011
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.09596235942780901,
+      "learning_rate": 2.547188094684405e-06,
+      "loss": 1.4151,
+      "step": 9012
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.08374721842744182,
+      "learning_rate": 2.5296663700731247e-06,
+      "loss": 1.3216,
+      "step": 9013
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.09278676414140465,
+      "learning_rate": 2.5122049654547897e-06,
+      "loss": 1.456,
+      "step": 9014
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.08964255210780839,
+      "learning_rate": 2.494803882946761e-06,
+      "loss": 1.3075,
+      "step": 9015
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.09620974578886025,
+      "learning_rate": 2.4774631246589074e-06,
+      "loss": 1.4034,
+      "step": 9016
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.11383840109527507,
+      "learning_rate": 2.4601826926938242e-06,
+      "loss": 1.5086,
+      "step": 9017
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.09144766054964178,
+      "learning_rate": 2.442962589146891e-06,
+      "loss": 1.3065,
+      "step": 9018
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.08185228768258924,
+      "learning_rate": 2.425802816106104e-06,
+      "loss": 1.3369,
+      "step": 9019
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.0840524727285538,
+      "learning_rate": 2.4087033756521328e-06,
+      "loss": 1.4063,
+      "step": 9020
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.08907504192099573,
+      "learning_rate": 2.391664269858318e-06,
+      "loss": 1.3916,
+      "step": 9021
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.09370670461297587,
+      "learning_rate": 2.374685500790785e-06,
+      "loss": 1.4576,
+      "step": 9022
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.08615449445440274,
+      "learning_rate": 2.3577670705081654e-06,
+      "loss": 1.4466,
+      "step": 9023
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.08779102881285784,
+      "learning_rate": 2.3409089810618734e-06,
+      "loss": 1.3638,
+      "step": 9024
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.08879645514225343,
+      "learning_rate": 2.324111234496107e-06,
+      "loss": 1.477,
+      "step": 9025
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.08444548732237621,
+      "learning_rate": 2.3073738328476255e-06,
+      "loss": 1.4513,
+      "step": 9026
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.09078258966657334,
+      "learning_rate": 2.2906967781458065e-06,
+      "loss": 1.5172,
+      "step": 9027
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.09488873511558187,
+      "learning_rate": 2.2740800724129206e-06,
+      "loss": 1.4049,
+      "step": 9028
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.09875341751463869,
+      "learning_rate": 2.257523717663745e-06,
+      "loss": 1.4016,
+      "step": 9029
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.11312923220297097,
+      "learning_rate": 2.2410277159057858e-06,
+      "loss": 1.2807,
+      "step": 9030
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.09081658104439949,
+      "learning_rate": 2.2245920691392753e-06,
+      "loss": 1.3431,
+      "step": 9031
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.09947730055048168,
+      "learning_rate": 2.208216779357064e-06,
+      "loss": 1.3545,
+      "step": 9032
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.0918733894258067,
+      "learning_rate": 2.1919018485446753e-06,
+      "loss": 1.4349,
+      "step": 9033
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.09406473803075574,
+      "learning_rate": 2.175647278680415e-06,
+      "loss": 1.4259,
+      "step": 9034
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.08127694221980779,
+      "learning_rate": 2.1594530717352068e-06,
+      "loss": 1.3125,
+      "step": 9035
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.09175464617559018,
+      "learning_rate": 2.1433192296725912e-06,
+      "loss": 1.3946,
+      "step": 9036
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.09641222973699672,
+      "learning_rate": 2.127245754448892e-06,
+      "loss": 1.4488,
+      "step": 9037
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.08726870720260296,
+      "learning_rate": 2.111232648013106e-06,
+      "loss": 1.3105,
+      "step": 9038
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.11137866349602377,
+      "learning_rate": 2.0952799123068466e-06,
+      "loss": 1.3899,
+      "step": 9039
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.08429603496474124,
+      "learning_rate": 2.0793875492644e-06,
+      "loss": 1.3003,
+      "step": 9040
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.10461075575496355,
+      "learning_rate": 2.0635555608128354e-06,
+      "loss": 1.3635,
+      "step": 9041
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.09558409348647286,
+      "learning_rate": 2.04778394887184e-06,
+      "loss": 1.3797,
+      "step": 9042
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.09213081325338919,
+      "learning_rate": 2.0320727153537165e-06,
+      "loss": 1.4851,
+      "step": 9043
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.10022093578171336,
+      "learning_rate": 2.016421862163498e-06,
+      "loss": 1.3951,
+      "step": 9044
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.09523644080123797,
+      "learning_rate": 2.0008313911989986e-06,
+      "loss": 1.3995,
+      "step": 9045
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.09133021078387467,
+      "learning_rate": 1.9853013043504844e-06,
+      "loss": 1.4518,
+      "step": 9046
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.08688534304999465,
+      "learning_rate": 1.969831603501171e-06,
+      "loss": 1.3965,
+      "step": 9047
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.08773971157963967,
+      "learning_rate": 1.954422290526725e-06,
+      "loss": 1.3931,
+      "step": 9048
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.09246961888491326,
+      "learning_rate": 1.9390733672955406e-06,
+      "loss": 1.3227,
+      "step": 9049
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.10713648382422544,
+      "learning_rate": 1.9237848356688514e-06,
+      "loss": 1.4001,
+      "step": 9050
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.08792536366905146,
+      "learning_rate": 1.9085566975003963e-06,
+      "loss": 1.4077,
+      "step": 9051
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.07995517162388631,
+      "learning_rate": 1.893388954636588e-06,
+      "loss": 1.5223,
+      "step": 9052
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.10201954934078061,
+      "learning_rate": 1.878281608916621e-06,
+      "loss": 1.4578,
+      "step": 9053
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.08765267905443119,
+      "learning_rate": 1.8632346621723085e-06,
+      "loss": 1.4235,
+      "step": 9054
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.09134361289135108,
+      "learning_rate": 1.8482481162280795e-06,
+      "loss": 1.3734,
+      "step": 9055
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.09517979343552131,
+      "learning_rate": 1.833321972901203e-06,
+      "loss": 1.4759,
+      "step": 9056
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.09392044720740886,
+      "learning_rate": 1.8184562340014532e-06,
+      "loss": 1.3799,
+      "step": 9057
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.08945528953340533,
+      "learning_rate": 1.8036509013313884e-06,
+      "loss": 1.4674,
+      "step": 9058
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.09660244996835152,
+      "learning_rate": 1.7889059766862392e-06,
+      "loss": 1.4577,
+      "step": 9059
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.08955683843704676,
+      "learning_rate": 1.7742214618537977e-06,
+      "loss": 1.3246,
+      "step": 9060
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.09367894047592056,
+      "learning_rate": 1.7595973586145842e-06,
+      "loss": 1.3879,
+      "step": 9061
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.09055649217132475,
+      "learning_rate": 1.7450336687420131e-06,
+      "loss": 1.3893,
+      "step": 9062
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.09954889766272536,
+      "learning_rate": 1.7305303940017836e-06,
+      "loss": 1.3906,
+      "step": 9063
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.08601609477859878,
+      "learning_rate": 1.7160875361525441e-06,
+      "loss": 1.41,
+      "step": 9064
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.09150624684421146,
+      "learning_rate": 1.701705096945505e-06,
+      "loss": 1.4518,
+      "step": 9065
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.08989981788436734,
+      "learning_rate": 1.6873830781246601e-06,
+      "loss": 1.4331,
+      "step": 9066
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.08875228716445183,
+      "learning_rate": 1.673121481426565e-06,
+      "loss": 1.374,
+      "step": 9067
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.08737529837010975,
+      "learning_rate": 1.6589203085804472e-06,
+      "loss": 1.4335,
+      "step": 9068
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.10109421859275902,
+      "learning_rate": 1.6447795613083183e-06,
+      "loss": 1.4097,
+      "step": 9069
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.09946451803794594,
+      "learning_rate": 1.6306992413247512e-06,
+      "loss": 1.3549,
+      "step": 9070
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.08896478727278188,
+      "learning_rate": 1.6166793503370469e-06,
+      "loss": 1.3679,
+      "step": 9071
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.10638462313903883,
+      "learning_rate": 1.602719890045179e-06,
+      "loss": 1.3335,
+      "step": 9072
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.089120932283575,
+      "learning_rate": 1.5888208621417376e-06,
+      "loss": 1.3602,
+      "step": 9073
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1012988002695788,
+      "learning_rate": 1.5749822683120419e-06,
+      "loss": 1.3864,
+      "step": 9074
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.10437337636041694,
+      "learning_rate": 1.561204110234138e-06,
+      "loss": 1.433,
+      "step": 9075
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.09601976601444809,
+      "learning_rate": 1.547486389578523e-06,
+      "loss": 1.5101,
+      "step": 9076
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.09522304132825528,
+      "learning_rate": 1.5338291080086441e-06,
+      "loss": 1.3575,
+      "step": 9077
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.09616477229666578,
+      "learning_rate": 1.5202322671805103e-06,
+      "loss": 1.5156,
+      "step": 9078
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.09378474058753226,
+      "learning_rate": 1.5066958687426914e-06,
+      "loss": 1.4158,
+      "step": 9079
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.08739425978756746,
+      "learning_rate": 1.493219914336541e-06,
+      "loss": 1.4265,
+      "step": 9080
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.098000838767873,
+      "learning_rate": 1.4798044055961414e-06,
+      "loss": 1.3646,
+      "step": 9081
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.09244726842890279,
+      "learning_rate": 1.4664493441480797e-06,
+      "loss": 1.3537,
+      "step": 9082
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.09582741121293425,
+      "learning_rate": 1.4531547316117832e-06,
+      "loss": 1.219,
+      "step": 9083
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.10720997957465893,
+      "learning_rate": 1.4399205695991847e-06,
+      "loss": 1.4641,
+      "step": 9084
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.08406612664504096,
+      "learning_rate": 1.4267468597150558e-06,
+      "loss": 1.3459,
+      "step": 9085
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.09440844125905276,
+      "learning_rate": 1.4136336035566744e-06,
+      "loss": 1.5147,
+      "step": 9086
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.09102998393092081,
+      "learning_rate": 1.4005808027141576e-06,
+      "loss": 1.4601,
+      "step": 9087
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.10331120527508215,
+      "learning_rate": 1.3875884587700727e-06,
+      "loss": 1.3414,
+      "step": 9088
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.09685017624372445,
+      "learning_rate": 1.3746565732999372e-06,
+      "loss": 1.4713,
+      "step": 9089
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.09074395569266269,
+      "learning_rate": 1.3617851478716637e-06,
+      "loss": 1.3245,
+      "step": 9090
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.09620486567097974,
+      "learning_rate": 1.348974184046059e-06,
+      "loss": 1.366,
+      "step": 9091
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.10708772859844796,
+      "learning_rate": 1.3362236833763808e-06,
+      "loss": 1.4079,
+      "step": 9092
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.0812319976321557,
+      "learning_rate": 1.3235336474087812e-06,
+      "loss": 1.3993,
+      "step": 9093
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.0775704183224788,
+      "learning_rate": 1.3109040776819182e-06,
+      "loss": 1.3545,
+      "step": 9094
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.0848197503815719,
+      "learning_rate": 1.2983349757271778e-06,
+      "loss": 1.3832,
+      "step": 9095
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.08434932724362751,
+      "learning_rate": 1.285826343068619e-06,
+      "loss": 1.4432,
+      "step": 9096
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.09075226208293427,
+      "learning_rate": 1.2733781812229729e-06,
+      "loss": 1.5216,
+      "step": 9097
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1065687767191086,
+      "learning_rate": 1.260990491699532e-06,
+      "loss": 1.4738,
+      "step": 9098
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.08992583345049047,
+      "learning_rate": 1.248663276000428e-06,
+      "loss": 1.4144,
+      "step": 9099
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.08435920582959416,
+      "learning_rate": 1.2363965356204099e-06,
+      "loss": 1.4694,
+      "step": 9100
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.10098857061162941,
+      "learning_rate": 1.2241902720467324e-06,
+      "loss": 1.4237,
+      "step": 9101
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.09986073903862418,
+      "learning_rate": 1.2120444867596003e-06,
+      "loss": 1.4478,
+      "step": 9102
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.08558072461227653,
+      "learning_rate": 1.1999591812316135e-06,
+      "loss": 1.5613,
+      "step": 9103
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.09939211521634202,
+      "learning_rate": 1.1879343569282109e-06,
+      "loss": 1.383,
+      "step": 9104
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.09749991658975933,
+      "learning_rate": 1.1759700153073927e-06,
+      "loss": 1.4085,
+      "step": 9105
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.08925710668430854,
+      "learning_rate": 1.1640661578199986e-06,
+      "loss": 1.3,
+      "step": 9106
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.08628724330928919,
+      "learning_rate": 1.1522227859092627e-06,
+      "loss": 1.5635,
+      "step": 9107
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.09571820500346655,
+      "learning_rate": 1.1404399010113696e-06,
+      "loss": 1.4007,
+      "step": 9108
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.08511360335320821,
+      "learning_rate": 1.1287175045548992e-06,
+      "loss": 1.3939,
+      "step": 9109
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.093149927054585,
+      "learning_rate": 1.117055597961325e-06,
+      "loss": 1.3354,
+      "step": 9110
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.08879778321147576,
+      "learning_rate": 1.105454182644683e-06,
+      "loss": 1.4454,
+      "step": 9111
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.09563179301500772,
+      "learning_rate": 1.0939132600116808e-06,
+      "loss": 1.4794,
+      "step": 9112
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.08586217819702109,
+      "learning_rate": 1.0824328314616994e-06,
+      "loss": 1.4134,
+      "step": 9113
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.08860707467489773,
+      "learning_rate": 1.0710128983867362e-06,
+      "loss": 1.4085,
+      "step": 9114
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.09383682611238597,
+      "learning_rate": 1.0596534621715171e-06,
+      "loss": 1.4335,
+      "step": 9115
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.0889687494892466,
+      "learning_rate": 1.04835452419344e-06,
+      "loss": 1.4469,
+      "step": 9116
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.09343696027937952,
+      "learning_rate": 1.037116085822576e-06,
+      "loss": 1.5305,
+      "step": 9117
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.10193965533313602,
+      "learning_rate": 1.0259381484215013e-06,
+      "loss": 1.3598,
+      "step": 9118
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.08673396632150293,
+      "learning_rate": 1.0148207133456878e-06,
+      "loss": 1.3471,
+      "step": 9119
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.09320776936813663,
+      "learning_rate": 1.0037637819431121e-06,
+      "loss": 1.1593,
+      "step": 9120
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.09137352888592001,
+      "learning_rate": 9.927673555544804e-07,
+      "loss": 1.315,
+      "step": 9121
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.08717919879493985,
+      "learning_rate": 9.818314355131697e-07,
+      "loss": 1.4096,
+      "step": 9122
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.0928947483788626,
+      "learning_rate": 9.709560231451198e-07,
+      "loss": 1.452,
+      "step": 9123
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.08813190059284774,
+      "learning_rate": 9.60141119769109e-07,
+      "loss": 1.4183,
+      "step": 9124
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.08887987124924661,
+      "learning_rate": 9.493867266964218e-07,
+      "loss": 1.4121,
+      "step": 9125
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.09395453002384421,
+      "learning_rate": 9.386928452310706e-07,
+      "loss": 1.426,
+      "step": 9126
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.08377161772511613,
+      "learning_rate": 9.280594766697959e-07,
+      "loss": 1.4216,
+      "step": 9127
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.09163942047825116,
+      "learning_rate": 9.174866223018441e-07,
+      "loss": 1.4048,
+      "step": 9128
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.08910281911488153,
+      "learning_rate": 9.069742834092454e-07,
+      "loss": 1.3467,
+      "step": 9129
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.09118313763931776,
+      "learning_rate": 8.965224612665912e-07,
+      "loss": 1.5055,
+      "step": 9130
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.09895577173828514,
+      "learning_rate": 8.861311571413122e-07,
+      "loss": 1.4136,
+      "step": 9131
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.09511323789429267,
+      "learning_rate": 8.75800372293345e-07,
+      "loss": 1.2273,
+      "step": 9132
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.09517490821029329,
+      "learning_rate": 8.655301079752987e-07,
+      "loss": 1.3734,
+      "step": 9133
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.08708118193105326,
+      "learning_rate": 8.553203654325103e-07,
+      "loss": 1.4706,
+      "step": 9134
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.09081075474479096,
+      "learning_rate": 8.451711459029343e-07,
+      "loss": 1.2992,
+      "step": 9135
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.10093740980687556,
+      "learning_rate": 8.350824506172527e-07,
+      "loss": 1.3703,
+      "step": 9136
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.080606655701831,
+      "learning_rate": 8.250542807986538e-07,
+      "loss": 1.4454,
+      "step": 9137
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.08759808465816735,
+      "learning_rate": 8.150866376631649e-07,
+      "loss": 1.4234,
+      "step": 9138
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.09199899630041176,
+      "learning_rate": 8.051795224193748e-07,
+      "loss": 1.5712,
+      "step": 9139
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.08938543020630885,
+      "learning_rate": 7.953329362685447e-07,
+      "loss": 1.4132,
+      "step": 9140
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.09203648239831115,
+      "learning_rate": 7.855468804046084e-07,
+      "loss": 1.4387,
+      "step": 9141
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.0975507858137972,
+      "learning_rate": 7.758213560141725e-07,
+      "loss": 1.3312,
+      "step": 9142
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.09787399811432872,
+      "learning_rate": 7.661563642765157e-07,
+      "loss": 1.3107,
+      "step": 9143
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.10078879497252471,
+      "learning_rate": 7.565519063634785e-07,
+      "loss": 1.3406,
+      "step": 9144
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.09294027961434816,
+      "learning_rate": 7.470079834396848e-07,
+      "loss": 1.3681,
+      "step": 9145
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.09205734553283813,
+      "learning_rate": 7.375245966623756e-07,
+      "loss": 1.3883,
+      "step": 9146
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.08730325996375386,
+      "learning_rate": 7.281017471814089e-07,
+      "loss": 1.3657,
+      "step": 9147
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.08832469615914386,
+      "learning_rate": 7.187394361393707e-07,
+      "loss": 1.3283,
+      "step": 9148
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.08891798495898455,
+      "learning_rate": 7.09437664671464e-07,
+      "loss": 1.4239,
+      "step": 9149
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.08509761132201243,
+      "learning_rate": 7.001964339055645e-07,
+      "loss": 1.2972,
+      "step": 9150
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.09860938568263267,
+      "learning_rate": 6.910157449621646e-07,
+      "loss": 1.3948,
+      "step": 9151
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.09700345844785578,
+      "learning_rate": 6.818955989545406e-07,
+      "loss": 1.4834,
+      "step": 9152
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.10282320662162178,
+      "learning_rate": 6.72835996988419e-07,
+      "loss": 1.3405,
+      "step": 9153
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.09532553767319768,
+      "learning_rate": 6.638369401624212e-07,
+      "loss": 1.3601,
+      "step": 9154
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.09944816708848722,
+      "learning_rate": 6.548984295676741e-07,
+      "loss": 1.3718,
+      "step": 9155
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.08797403192670447,
+      "learning_rate": 6.460204662879776e-07,
+      "loss": 1.467,
+      "step": 9156
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.09782593171572332,
+      "learning_rate": 6.37203051399804e-07,
+      "loss": 1.3464,
+      "step": 9157
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.09783108557900717,
+      "learning_rate": 6.284461859723534e-07,
+      "loss": 1.3776,
+      "step": 9158
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.08536515596852107,
+      "learning_rate": 6.197498710673876e-07,
+      "loss": 1.3792,
+      "step": 9159
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.08102996000876958,
+      "learning_rate": 6.111141077393967e-07,
+      "loss": 1.3457,
+      "step": 9160
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.08950933513385094,
+      "learning_rate": 6.025388970354872e-07,
+      "loss": 1.5071,
+      "step": 9161
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.09077163258171603,
+      "learning_rate": 5.940242399953832e-07,
+      "loss": 1.5238,
+      "step": 9162
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.08107218496059505,
+      "learning_rate": 5.855701376515921e-07,
+      "loss": 1.4836,
+      "step": 9163
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.09402355006480244,
+      "learning_rate": 5.771765910291271e-07,
+      "loss": 1.3273,
+      "step": 9164
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.08419332875007346,
+      "learning_rate": 5.688436011457854e-07,
+      "loss": 1.2936,
+      "step": 9165
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.09582927476497023,
+      "learning_rate": 5.605711690119808e-07,
+      "loss": 1.4482,
+      "step": 9166
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.08145256942098168,
+      "learning_rate": 5.523592956307444e-07,
+      "loss": 1.3393,
+      "step": 9167
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.08945519192985447,
+      "learning_rate": 5.442079819977797e-07,
+      "loss": 1.4419,
+      "step": 9168
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.08389369232847793,
+      "learning_rate": 5.361172291014627e-07,
+      "loss": 1.4232,
+      "step": 9169
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.09665716756692867,
+      "learning_rate": 5.280870379228975e-07,
+      "loss": 1.502,
+      "step": 9170
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.09107272005780782,
+      "learning_rate": 5.201174094356942e-07,
+      "loss": 1.374,
+      "step": 9171
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.08525633295064047,
+      "learning_rate": 5.122083446062464e-07,
+      "loss": 1.2464,
+      "step": 9172
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.09145092325910216,
+      "learning_rate": 5.043598443935094e-07,
+      "loss": 1.3063,
+      "step": 9173
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.09121429449434486,
+      "learning_rate": 4.965719097491661e-07,
+      "loss": 1.317,
+      "step": 9174
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.08106699097565444,
+      "learning_rate": 4.888445416175724e-07,
+      "loss": 1.3662,
+      "step": 9175
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.09235616001095853,
+      "learning_rate": 4.8117774093559e-07,
+      "loss": 1.4448,
+      "step": 9176
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.08632917581540081,
+      "learning_rate": 4.735715086329195e-07,
+      "loss": 1.4205,
+      "step": 9177
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.09459405410772538,
+      "learning_rate": 4.6602584563187886e-07,
+      "loss": 1.3787,
+      "step": 9178
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.08651932809943559,
+      "learning_rate": 4.5854075284729177e-07,
+      "loss": 1.4209,
+      "step": 9179
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.08827714040145299,
+      "learning_rate": 4.511162311868766e-07,
+      "loss": 1.3239,
+      "step": 9180
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.08381943113452264,
+      "learning_rate": 4.4375228155085766e-07,
+      "loss": 1.4219,
+      "step": 9181
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.08236325719668838,
+      "learning_rate": 4.3644890483202084e-07,
+      "loss": 1.3715,
+      "step": 9182
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.09243341173187043,
+      "learning_rate": 4.292061019160465e-07,
+      "loss": 1.4119,
+      "step": 9183
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.10070109950559669,
+      "learning_rate": 4.2202387368112104e-07,
+      "loss": 1.472,
+      "step": 9184
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.09517079092290456,
+      "learning_rate": 4.149022209981035e-07,
+      "loss": 1.3448,
+      "step": 9185
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.09084655013556947,
+      "learning_rate": 4.0784114473052526e-07,
+      "loss": 1.2689,
+      "step": 9186
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.08422060321187015,
+      "learning_rate": 4.008406457345903e-07,
+      "loss": 1.3204,
+      "step": 9187
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.08559323844255416,
+      "learning_rate": 3.9390072485906424e-07,
+      "loss": 1.3425,
+      "step": 9188
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.08430437320924258,
+      "learning_rate": 3.8702138294549605e-07,
+      "loss": 1.3853,
+      "step": 9189
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.09296172134493086,
+      "learning_rate": 3.8020262082799626e-07,
+      "loss": 1.3051,
+      "step": 9190
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.08668309232549491,
+      "learning_rate": 3.7344443933340355e-07,
+      "loss": 1.3997,
+      "step": 9191
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.08495748652651433,
+      "learning_rate": 3.6674683928111795e-07,
+      "loss": 1.3981,
+      "step": 9192
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.11190443758993558,
+      "learning_rate": 3.6010982148332314e-07,
+      "loss": 1.4863,
+      "step": 9193
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.08562198339929697,
+      "learning_rate": 3.535333867447088e-07,
+      "loss": 1.2775,
+      "step": 9194
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.08887863740413854,
+      "learning_rate": 3.4701753586269257e-07,
+      "loss": 1.3826,
+      "step": 9195
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.0911839138816826,
+      "learning_rate": 3.405622696274202e-07,
+      "loss": 1.4951,
+      "step": 9196
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.10007743747550958,
+      "learning_rate": 3.341675888215434e-07,
+      "loss": 1.3479,
+      "step": 9197
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.08820855203603718,
+      "learning_rate": 3.278334942204419e-07,
+      "loss": 1.3686,
+      "step": 9198
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.09489741230917534,
+      "learning_rate": 3.215599865921126e-07,
+      "loss": 1.3376,
+      "step": 9199
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.09132514257961939,
+      "learning_rate": 3.1534706669733573e-07,
+      "loss": 1.4629,
+      "step": 9200
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.09654225348968377,
+      "learning_rate": 3.0919473528939754e-07,
+      "loss": 1.3531,
+      "step": 9201
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.09488757695139786,
+      "learning_rate": 3.0310299311431256e-07,
+      "loss": 1.4028,
+      "step": 9202
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.10174093481738378,
+      "learning_rate": 2.9707184091071204e-07,
+      "loss": 1.4304,
+      "step": 9203
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.08821629876713558,
+      "learning_rate": 2.9110127940984445e-07,
+      "loss": 1.3028,
+      "step": 9204
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.08766097557322772,
+      "learning_rate": 2.851913093357417e-07,
+      "loss": 1.4125,
+      "step": 9205
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.0901572714979954,
+      "learning_rate": 2.793419314049972e-07,
+      "loss": 1.4407,
+      "step": 9206
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.08821420180605707,
+      "learning_rate": 2.73553146326766e-07,
+      "loss": 1.432,
+      "step": 9207
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.08633075490026776,
+      "learning_rate": 2.6782495480309756e-07,
+      "loss": 1.35,
+      "step": 9208
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.08272348964502109,
+      "learning_rate": 2.621573575284919e-07,
+      "loss": 1.3314,
+      "step": 9209
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.08738865558697523,
+      "learning_rate": 2.565503551901216e-07,
+      "loss": 1.3905,
+      "step": 9210
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.08583764484613975,
+      "learning_rate": 2.510039484679427e-07,
+      "loss": 1.4012,
+      "step": 9211
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.09265865578712432,
+      "learning_rate": 2.4551813803441734e-07,
+      "loss": 1.2547,
+      "step": 9212
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.07892496874291217,
+      "learning_rate": 2.400929245546801e-07,
+      "loss": 1.3251,
+      "step": 9213
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.08624414306385945,
+      "learning_rate": 2.3472830868670469e-07,
+      "loss": 1.3862,
+      "step": 9214
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.09302196091866992,
+      "learning_rate": 2.2942429108080422e-07,
+      "loss": 1.2674,
+      "step": 9215
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.08821491575497864,
+      "learning_rate": 2.2418087238018637e-07,
+      "loss": 1.4064,
+      "step": 9216
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.08949986370580101,
+      "learning_rate": 2.189980532206759e-07,
+      "loss": 1.4349,
+      "step": 9217
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.08278441450527144,
+      "learning_rate": 2.1387583423060354e-07,
+      "loss": 1.2998,
+      "step": 9218
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.0840801051048805,
+      "learning_rate": 2.0881421603113904e-07,
+      "loss": 1.263,
+      "step": 9219
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.08056200911342393,
+      "learning_rate": 2.0381319923595821e-07,
+      "loss": 1.4156,
+      "step": 9220
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.08724772973658486,
+      "learning_rate": 1.9887278445152036e-07,
+      "loss": 1.2512,
+      "step": 9221
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.09455867642372127,
+      "learning_rate": 1.9399297227684633e-07,
+      "loss": 1.4142,
+      "step": 9222
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.09343959412492665,
+      "learning_rate": 1.8917376330357395e-07,
+      "loss": 1.3958,
+      "step": 9223
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.08794548124148231,
+      "learning_rate": 1.8441515811612464e-07,
+      "loss": 1.3576,
+      "step": 9224
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.09691778279436149,
+      "learning_rate": 1.797171572914813e-07,
+      "loss": 1.5349,
+      "step": 9225
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.07983284987363899,
+      "learning_rate": 1.7507976139924385e-07,
+      "loss": 1.4171,
+      "step": 9226
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.09287129310883346,
+      "learning_rate": 1.7050297100174028e-07,
+      "loss": 1.3165,
+      "step": 9227
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.08646817433793243,
+      "learning_rate": 1.6598678665397104e-07,
+      "loss": 1.4993,
+      "step": 9228
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.07929013433839835,
+      "learning_rate": 1.6153120890344265e-07,
+      "loss": 1.5507,
+      "step": 9229
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.08850803873232053,
+      "learning_rate": 1.5713623829050062e-07,
+      "loss": 1.4476,
+      "step": 9230
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.0888883518336363,
+      "learning_rate": 1.5280187534794098e-07,
+      "loss": 1.3417,
+      "step": 9231
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.09379458112499682,
+      "learning_rate": 1.4852812060145438e-07,
+      "loss": 1.4362,
+      "step": 9232
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.08832514165492539,
+      "learning_rate": 1.4431497456918186e-07,
+      "loss": 1.4955,
+      "step": 9233
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.10270631070515934,
+      "learning_rate": 1.4016243776193706e-07,
+      "loss": 1.314,
+      "step": 9234
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.08689177286241863,
+      "learning_rate": 1.360705106833171e-07,
+      "loss": 1.4187,
+      "step": 9235
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.0904049295681052,
+      "learning_rate": 1.320391938294252e-07,
+      "loss": 1.4229,
+      "step": 9236
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.08826812497826056,
+      "learning_rate": 1.280684876890925e-07,
+      "loss": 1.3759,
+      "step": 9237
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.09454769152128116,
+      "learning_rate": 1.2415839274376728e-07,
+      "loss": 1.4044,
+      "step": 9238
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.12649764184311058,
+      "learning_rate": 1.2030890946757022e-07,
+      "loss": 1.4513,
+      "step": 9239
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.08893899583982746,
+      "learning_rate": 1.1652003832729464e-07,
+      "loss": 1.4767,
+      "step": 9240
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.09708840021285432,
+      "learning_rate": 1.1279177978229527e-07,
+      "loss": 1.2564,
+      "step": 9241
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.08160711737594363,
+      "learning_rate": 1.0912413428471046e-07,
+      "loss": 1.2285,
+      "step": 9242
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.0830542224283073,
+      "learning_rate": 1.0551710227912903e-07,
+      "loss": 1.3295,
+      "step": 9243
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.10843619771485967,
+      "learning_rate": 1.0197068420308986e-07,
+      "loss": 1.4789,
+      "step": 9244
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.08399511285269524,
+      "learning_rate": 9.848488048647131e-08,
+      "loss": 1.4328,
+      "step": 9245
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.08369868710316501,
+      "learning_rate": 9.505969155193528e-08,
+      "loss": 1.4427,
+      "step": 9246
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.08815787268710067,
+      "learning_rate": 9.169511781492724e-08,
+      "loss": 1.284,
+      "step": 9247
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.10365706578408941,
+      "learning_rate": 8.839115968328759e-08,
+      "loss": 1.4171,
+      "step": 9248
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.0868752961016487,
+      "learning_rate": 8.514781755769585e-08,
+      "loss": 1.3582,
+      "step": 9249
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.09414347762024002,
+      "learning_rate": 8.1965091831393e-08,
+      "loss": 1.3758,
+      "step": 9250
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.08954872276660882,
+      "learning_rate": 7.884298289029258e-08,
+      "loss": 1.3475,
+      "step": 9251
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.10030726142452168,
+      "learning_rate": 7.57814911129806e-08,
+      "loss": 1.3205,
+      "step": 9252
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.09762209771924428,
+      "learning_rate": 7.278061687066018e-08,
+      "loss": 1.3829,
+      "step": 9253
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.08836513485757565,
+      "learning_rate": 6.984036052720688e-08,
+      "loss": 1.3295,
+      "step": 9254
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.0871339758958895,
+      "learning_rate": 6.696072243911333e-08,
+      "loss": 1.3928,
+      "step": 9255
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.09098880914087802,
+      "learning_rate": 6.414170295560017e-08,
+      "loss": 1.357,
+      "step": 9256
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.09077999286470577,
+      "learning_rate": 6.138330241839407e-08,
+      "loss": 1.3801,
+      "step": 9257
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.10122837068892225,
+      "learning_rate": 5.868552116206072e-08,
+      "loss": 1.3623,
+      "step": 9258
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.08517700022337132,
+      "learning_rate": 5.604835951367182e-08,
+      "loss": 1.4786,
+      "step": 9259
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.08585211456743395,
+      "learning_rate": 5.3471817792971614e-08,
+      "loss": 1.4968,
+      "step": 9260
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.10044181669808067,
+      "learning_rate": 5.095589631237685e-08,
+      "loss": 1.2755,
+      "step": 9261
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.08811673378812793,
+      "learning_rate": 4.8500595376976816e-08,
+      "loss": 1.2943,
+      "step": 9262
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.08452180165981107,
+      "learning_rate": 4.610591528447783e-08,
+      "loss": 1.3082,
+      "step": 9263
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.09509764207121812,
+      "learning_rate": 4.3771856325203196e-08,
+      "loss": 1.3976,
+      "step": 9264
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.0905369920640337,
+      "learning_rate": 4.1498418782259795e-08,
+      "loss": 1.34,
+      "step": 9265
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.08517538334754213,
+      "learning_rate": 3.928560293120498e-08,
+      "loss": 1.3423,
+      "step": 9266
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.0853860335389513,
+      "learning_rate": 3.713340904043516e-08,
+      "loss": 1.4828,
+      "step": 9267
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.10931919049201298,
+      "learning_rate": 3.504183737085276e-08,
+      "loss": 1.3513,
+      "step": 9268
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.08736777021378683,
+      "learning_rate": 3.301088817608822e-08,
+      "loss": 1.4117,
+      "step": 9269
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.10028110998630704,
+      "learning_rate": 3.104056170244451e-08,
+      "loss": 1.4898,
+      "step": 9270
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.09773062139045051,
+      "learning_rate": 2.9130858188730626e-08,
+      "loss": 1.3318,
+      "step": 9271
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.08128181127641702,
+      "learning_rate": 2.7281777866594582e-08,
+      "loss": 1.4327,
+      "step": 9272
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.0943358489444862,
+      "learning_rate": 2.5493320960190413e-08,
+      "loss": 1.3043,
+      "step": 9273
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.09785602520380977,
+      "learning_rate": 2.3765487686400188e-08,
+      "loss": 1.4111,
+      "step": 9274
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.09533372939648309,
+      "learning_rate": 2.2098278254722993e-08,
+      "loss": 1.4909,
+      "step": 9275
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.0801968457699735,
+      "learning_rate": 2.0491692867330434e-08,
+      "loss": 1.4355,
+      "step": 9276
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.09429240525483946,
+      "learning_rate": 1.894573171901115e-08,
+      "loss": 1.4304,
+      "step": 9277
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.10266163176098933,
+      "learning_rate": 1.746039499717078e-08,
+      "loss": 1.4492,
+      "step": 9278
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.09383290941084661,
+      "learning_rate": 1.6035682881998527e-08,
+      "loss": 1.4451,
+      "step": 9279
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.080437028926394,
+      "learning_rate": 1.4671595546245086e-08,
+      "loss": 1.4012,
+      "step": 9280
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.08351402759362497,
+      "learning_rate": 1.3368133155222672e-08,
+      "loss": 1.3904,
+      "step": 9281
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.091617721380892,
+      "learning_rate": 1.2125295867027042e-08,
+      "loss": 1.3431,
+      "step": 9282
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.10262114214274354,
+      "learning_rate": 1.094308383237097e-08,
+      "loss": 1.5308,
+      "step": 9283
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.10109618312260171,
+      "learning_rate": 9.82149719458425e-09,
+      "loss": 1.3743,
+      "step": 9284
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.09909394437365746,
+      "learning_rate": 8.760536089724714e-09,
+      "loss": 1.435,
+      "step": 9285
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.08764241854659703,
+      "learning_rate": 7.760200646300675e-09,
+      "loss": 1.3543,
+      "step": 9286
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.10072587519427252,
+      "learning_rate": 6.820490985715022e-09,
+      "loss": 1.3164,
+      "step": 9287
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.09365884377453729,
+      "learning_rate": 5.941407221932149e-09,
+      "loss": 1.3535,
+      "step": 9288
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.09650653930095349,
+      "learning_rate": 5.122949461422444e-09,
+      "loss": 1.3542,
+      "step": 9289
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.08379342583541467,
+      "learning_rate": 4.365117803550866e-09,
+      "loss": 1.4495,
+      "step": 9290
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.09115901593403421,
+      "learning_rate": 3.667912340132862e-09,
+      "loss": 1.3161,
+      "step": 9291
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.10871669726957177,
+      "learning_rate": 3.031333155767424e-09,
+      "loss": 1.5187,
+      "step": 9292
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.08912092052721272,
+      "learning_rate": 2.455380327559542e-09,
+      "loss": 1.3403,
+      "step": 9293
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.08373270056508372,
+      "learning_rate": 1.9400539253977557e-09,
+      "loss": 1.3308,
+      "step": 9294
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.09699578700835747,
+      "learning_rate": 1.4853540117321095e-09,
+      "loss": 1.3618,
+      "step": 9295
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.08636922925763955,
+      "learning_rate": 1.0912806417961995e-09,
+      "loss": 1.3992,
+      "step": 9296
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.09516627466784042,
+      "learning_rate": 7.578338632741044e-10,
+      "loss": 1.423,
+      "step": 9297
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.10331085370003165,
+      "learning_rate": 4.850137165779422e-10,
+      "loss": 1.3025,
+      "step": 9298
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.08985317351843392,
+      "learning_rate": 2.728202349033815e-10,
+      "loss": 1.3703,
+      "step": 9299
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.09341887519280219,
+      "learning_rate": 1.2125344384106285e-10,
+      "loss": 1.4087,
+      "step": 9300
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.08905643836734109,
+      "learning_rate": 3.031336187619971e-11,
+      "loss": 1.4227,
+      "step": 9301
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.08366057997777612,
+      "learning_rate": 0.0,
+      "loss": 1.5995,
+      "step": 9302
+    },
+    {
+      "epoch": 1.0,
+      "step": 9302,
+      "total_flos": 1.5268155381671526e+18,
+      "train_loss": 1.4317849794912534,
+      "train_runtime": 52932.8416,
+      "train_samples_per_second": 11.248,
+      "train_steps_per_second": 0.176
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 9302,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 2000,
+  "total_flos": 1.5268155381671526e+18,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}