diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,5923 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 20.0,
+  "eval_steps": 25.0,
+  "global_step": 840,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.023880597014925373,
+      "grad_norm": Infinity,
+      "learning_rate": 1e-05,
+      "loss": 76.8913,
+      "step": 1
+    },
+    {
+      "epoch": 0.04776119402985075,
+      "grad_norm": 395.8459777832031,
+      "learning_rate": 1e-05,
+      "loss": 77.3235,
+      "step": 2
+    },
+    {
+      "epoch": 0.07164179104477612,
+      "grad_norm": 161.73968505859375,
+      "learning_rate": 9.988095238095239e-06,
+      "loss": 67.3668,
+      "step": 3
+    },
+    {
+      "epoch": 0.0955223880597015,
+      "grad_norm": 145.1616973876953,
+      "learning_rate": 9.976190476190477e-06,
+      "loss": 62.8542,
+      "step": 4
+    },
+    {
+      "epoch": 0.11940298507462686,
+      "grad_norm": 52.10374069213867,
+      "learning_rate": 9.964285714285714e-06,
+      "loss": 59.8627,
+      "step": 5
+    },
+    {
+      "epoch": 0.14328358208955225,
+      "grad_norm": 31.430763244628906,
+      "learning_rate": 9.952380952380954e-06,
+      "loss": 59.2378,
+      "step": 6
+    },
+    {
+      "epoch": 0.16716417910447762,
+      "grad_norm": 37.42692947387695,
+      "learning_rate": 9.940476190476192e-06,
+      "loss": 58.3114,
+      "step": 7
+    },
+    {
+      "epoch": 0.191044776119403,
+      "grad_norm": 32.284332275390625,
+      "learning_rate": 9.92857142857143e-06,
+      "loss": 57.3408,
+      "step": 8
+    },
+    {
+      "epoch": 0.21492537313432836,
+      "grad_norm": 28.520849227905273,
+      "learning_rate": 9.916666666666668e-06,
+      "loss": 56.9256,
+      "step": 9
+    },
+    {
+      "epoch": 0.23880597014925373,
+      "grad_norm": 40.480167388916016,
+      "learning_rate": 9.904761904761906e-06,
+      "loss": 57.4842,
+      "step": 10
+    },
+    {
+      "epoch": 0.2626865671641791,
+      "grad_norm": 20.3512020111084,
+      "learning_rate": 9.892857142857143e-06,
+      "loss": 56.7597,
+      "step": 11
+    },
+    {
+      "epoch": 0.2865671641791045,
+      "grad_norm": 22.308382034301758,
+      "learning_rate": 9.880952380952381e-06,
+      "loss": 56.594,
+      "step": 12
+    },
+    {
+      "epoch": 0.31044776119402984,
+      "grad_norm": 14.086284637451172,
+      "learning_rate": 9.869047619047621e-06,
+      "loss": 56.6327,
+      "step": 13
+    },
+    {
+      "epoch": 0.33432835820895523,
+      "grad_norm": NaN,
+      "learning_rate": 9.857142857142859e-06,
+      "loss": 63.0791,
+      "step": 14
+    },
+    {
+      "epoch": 0.3582089552238806,
+      "grad_norm": 17.055604934692383,
+      "learning_rate": 9.857142857142859e-06,
+      "loss": 56.1551,
+      "step": 15
+    },
+    {
+      "epoch": 0.382089552238806,
+      "grad_norm": 18.52287483215332,
+      "learning_rate": 9.845238095238097e-06,
+      "loss": 54.9502,
+      "step": 16
+    },
+    {
+      "epoch": 0.4059701492537313,
+      "grad_norm": 32.66905212402344,
+      "learning_rate": 9.833333333333333e-06,
+      "loss": 55.6494,
+      "step": 17
+    },
+    {
+      "epoch": 0.4298507462686567,
+      "grad_norm": 24.075742721557617,
+      "learning_rate": 9.821428571428573e-06,
+      "loss": 55.4766,
+      "step": 18
+    },
+    {
+      "epoch": 0.4537313432835821,
+      "grad_norm": 31.505783081054688,
+      "learning_rate": 9.80952380952381e-06,
+      "loss": 55.1481,
+      "step": 19
+    },
+    {
+      "epoch": 0.47761194029850745,
+      "grad_norm": 30.53020477294922,
+      "learning_rate": 9.797619047619048e-06,
+      "loss": 55.158,
+      "step": 20
+    },
+    {
+      "epoch": 0.5014925373134328,
+      "grad_norm": 14.44444751739502,
+      "learning_rate": 9.785714285714286e-06,
+      "loss": 54.8286,
+      "step": 21
+    },
+    {
+      "epoch": 0.5253731343283582,
+      "grad_norm": 53.601078033447266,
+      "learning_rate": 9.773809523809524e-06,
+      "loss": 55.1466,
+      "step": 22
+    },
+    {
+      "epoch": 0.5492537313432836,
+      "grad_norm": 37.156028747558594,
+      "learning_rate": 9.761904761904762e-06,
+      "loss": 54.629,
+      "step": 23
+    },
+    {
+      "epoch": 0.573134328358209,
+      "grad_norm": 41.84994125366211,
+      "learning_rate": 9.75e-06,
+      "loss": 55.2766,
+      "step": 24
+    },
+    {
+      "epoch": 0.5970149253731343,
+      "grad_norm": 50.61705017089844,
+      "learning_rate": 9.73809523809524e-06,
+      "loss": 54.3497,
+      "step": 25
+    },
+    {
+      "epoch": 0.6208955223880597,
+      "grad_norm": 16.964982986450195,
+      "learning_rate": 9.726190476190477e-06,
+      "loss": 55.1673,
+      "step": 26
+    },
+    {
+      "epoch": 0.6447761194029851,
+      "grad_norm": 22.71157455444336,
+      "learning_rate": 9.714285714285715e-06,
+      "loss": 54.8001,
+      "step": 27
+    },
+    {
+      "epoch": 0.6686567164179105,
+      "grad_norm": 16.372802734375,
+      "learning_rate": 9.702380952380953e-06,
+      "loss": 54.4506,
+      "step": 28
+    },
+    {
+      "epoch": 0.6925373134328359,
+      "grad_norm": 13.21664047241211,
+      "learning_rate": 9.690476190476191e-06,
+      "loss": 53.0488,
+      "step": 29
+    },
+    {
+      "epoch": 0.7164179104477612,
+      "grad_norm": NaN,
+      "learning_rate": 9.678571428571429e-06,
+      "loss": 74.7707,
+      "step": 30
+    },
+    {
+      "epoch": 0.7402985074626866,
+      "grad_norm": 23.045652389526367,
+      "learning_rate": 9.678571428571429e-06,
+      "loss": 54.1909,
+      "step": 31
+    },
+    {
+      "epoch": 0.764179104477612,
+      "grad_norm": 15.034178733825684,
+      "learning_rate": 9.666666666666667e-06,
+      "loss": 53.3253,
+      "step": 32
+    },
+    {
+      "epoch": 0.7880597014925373,
+      "grad_norm": 14.148232460021973,
+      "learning_rate": 9.654761904761906e-06,
+      "loss": 53.693,
+      "step": 33
+    },
+    {
+      "epoch": 0.8119402985074626,
+      "grad_norm": NaN,
+      "learning_rate": 9.642857142857144e-06,
+      "loss": 61.3517,
+      "step": 34
+    },
+    {
+      "epoch": 0.835820895522388,
+      "grad_norm": 14.757994651794434,
+      "learning_rate": 9.642857142857144e-06,
+      "loss": 53.3175,
+      "step": 35
+    },
+    {
+      "epoch": 0.8597014925373134,
+      "grad_norm": 10.875706672668457,
+      "learning_rate": 9.630952380952382e-06,
+      "loss": 54.2592,
+      "step": 36
+    },
+    {
+      "epoch": 0.8835820895522388,
+      "grad_norm": 9.926539421081543,
+      "learning_rate": 9.61904761904762e-06,
+      "loss": 53.8721,
+      "step": 37
+    },
+    {
+      "epoch": 0.9074626865671642,
+      "grad_norm": 17.697235107421875,
+      "learning_rate": 9.607142857142858e-06,
+      "loss": 54.2901,
+      "step": 38
+    },
+    {
+      "epoch": 0.9313432835820895,
+      "grad_norm": 29.19430160522461,
+      "learning_rate": 9.595238095238096e-06,
+      "loss": 53.1261,
+      "step": 39
+    },
+    {
+      "epoch": 0.9552238805970149,
+      "grad_norm": 12.3985595703125,
+      "learning_rate": 9.583333333333335e-06,
+      "loss": 53.9815,
+      "step": 40
+    },
+    {
+      "epoch": 0.9791044776119403,
+      "grad_norm": 17.109691619873047,
+      "learning_rate": 9.571428571428573e-06,
+      "loss": 52.5838,
+      "step": 41
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 14.529239654541016,
+      "learning_rate": 9.559523809523811e-06,
+      "loss": 46.1888,
+      "step": 42
+    },
+    {
+      "epoch": 1.0238805970149254,
+      "grad_norm": 15.683514595031738,
+      "learning_rate": 9.547619047619049e-06,
+      "loss": 52.2043,
+      "step": 43
+    },
+    {
+      "epoch": 1.0477611940298508,
+      "grad_norm": 26.219507217407227,
+      "learning_rate": 9.535714285714287e-06,
+      "loss": 53.09,
+      "step": 44
+    },
+    {
+      "epoch": 1.0716417910447762,
+      "grad_norm": 19.859697341918945,
+      "learning_rate": 9.523809523809525e-06,
+      "loss": 52.5858,
+      "step": 45
+    },
+    {
+      "epoch": 1.0955223880597016,
+      "grad_norm": 11.090332984924316,
+      "learning_rate": 9.511904761904763e-06,
+      "loss": 53.7674,
+      "step": 46
+    },
+    {
+      "epoch": 1.1194029850746268,
+      "grad_norm": 15.586993217468262,
+      "learning_rate": 9.5e-06,
+      "loss": 53.4901,
+      "step": 47
+    },
+    {
+      "epoch": 1.1432835820895522,
+      "grad_norm": 22.734928131103516,
+      "learning_rate": 9.488095238095238e-06,
+      "loss": 53.6667,
+      "step": 48
+    },
+    {
+      "epoch": 1.1671641791044776,
+      "grad_norm": 16.382047653198242,
+      "learning_rate": 9.476190476190476e-06,
+      "loss": 53.2914,
+      "step": 49
+    },
+    {
+      "epoch": 1.191044776119403,
+      "grad_norm": 15.916092872619629,
+      "learning_rate": 9.464285714285714e-06,
+      "loss": 51.729,
+      "step": 50
+    },
+    {
+      "epoch": 1.2149253731343284,
+      "grad_norm": NaN,
+      "learning_rate": 9.452380952380952e-06,
+      "loss": 66.3151,
+      "step": 51
+    },
+    {
+      "epoch": 1.2388059701492538,
+      "grad_norm": 20.121395111083984,
+      "learning_rate": 9.452380952380952e-06,
+      "loss": 53.866,
+      "step": 52
+    },
+    {
+      "epoch": 1.2626865671641792,
+      "grad_norm": NaN,
+      "learning_rate": 9.440476190476192e-06,
+      "loss": 67.3538,
+      "step": 53
+    },
+    {
+      "epoch": 1.2865671641791045,
+      "grad_norm": 13.869222640991211,
+      "learning_rate": 9.440476190476192e-06,
+      "loss": 53.5431,
+      "step": 54
+    },
+    {
+      "epoch": 1.31044776119403,
+      "grad_norm": 25.478107452392578,
+      "learning_rate": 9.42857142857143e-06,
+      "loss": 52.7126,
+      "step": 55
+    },
+    {
+      "epoch": 1.3343283582089551,
+      "grad_norm": 35.76942825317383,
+      "learning_rate": 9.416666666666667e-06,
+      "loss": 53.8082,
+      "step": 56
+    },
+    {
+      "epoch": 1.3582089552238805,
+      "grad_norm": 17.95109748840332,
+      "learning_rate": 9.404761904761905e-06,
+      "loss": 53.0275,
+      "step": 57
+    },
+    {
+      "epoch": 1.382089552238806,
+      "grad_norm": 26.80129623413086,
+      "learning_rate": 9.392857142857143e-06,
+      "loss": 53.3295,
+      "step": 58
+    },
+    {
+      "epoch": 1.4059701492537313,
+      "grad_norm": 25.803054809570312,
+      "learning_rate": 9.380952380952381e-06,
+      "loss": 52.3771,
+      "step": 59
+    },
+    {
+      "epoch": 1.4298507462686567,
+      "grad_norm": 35.39850616455078,
+      "learning_rate": 9.36904761904762e-06,
+      "loss": 52.9467,
+      "step": 60
+    },
+    {
+      "epoch": 1.4537313432835821,
+      "grad_norm": 27.43315887451172,
+      "learning_rate": 9.357142857142859e-06,
+      "loss": 53.1359,
+      "step": 61
+    },
+    {
+      "epoch": 1.4776119402985075,
+      "grad_norm": 33.431400299072266,
+      "learning_rate": 9.345238095238096e-06,
+      "loss": 52.2006,
+      "step": 62
+    },
+    {
+      "epoch": 1.5014925373134327,
+      "grad_norm": 33.08237075805664,
+      "learning_rate": 9.333333333333334e-06,
+      "loss": 52.6816,
+      "step": 63
+    },
+    {
+      "epoch": 1.5253731343283583,
+      "grad_norm": 25.711997985839844,
+      "learning_rate": 9.321428571428572e-06,
+      "loss": 52.6113,
+      "step": 64
+    },
+    {
+      "epoch": 1.5492537313432835,
+      "grad_norm": 32.688297271728516,
+      "learning_rate": 9.30952380952381e-06,
+      "loss": 51.9086,
+      "step": 65
+    },
+    {
+      "epoch": 1.573134328358209,
+      "grad_norm": 31.856857299804688,
+      "learning_rate": 9.297619047619048e-06,
+      "loss": 53.1913,
+      "step": 66
+    },
+    {
+      "epoch": 1.5970149253731343,
+      "grad_norm": 26.231773376464844,
+      "learning_rate": 9.285714285714288e-06,
+      "loss": 51.464,
+      "step": 67
+    },
+    {
+      "epoch": 1.6208955223880597,
+      "grad_norm": 29.39109230041504,
+      "learning_rate": 9.273809523809525e-06,
+      "loss": 52.0572,
+      "step": 68
+    },
+    {
+      "epoch": 1.644776119402985,
+      "grad_norm": 28.86277198791504,
+      "learning_rate": 9.261904761904763e-06,
+      "loss": 52.7115,
+      "step": 69
+    },
+    {
+      "epoch": 1.6686567164179105,
+      "grad_norm": 28.82640266418457,
+      "learning_rate": 9.250000000000001e-06,
+      "loss": 53.1756,
+      "step": 70
+    },
+    {
+      "epoch": 1.6925373134328359,
+      "grad_norm": 31.32577133178711,
+      "learning_rate": 9.238095238095239e-06,
+      "loss": 52.369,
+      "step": 71
+    },
+    {
+      "epoch": 1.716417910447761,
+      "grad_norm": 18.739269256591797,
+      "learning_rate": 9.226190476190477e-06,
+      "loss": 52.6631,
+      "step": 72
+    },
+    {
+      "epoch": 1.7402985074626867,
+      "grad_norm": 22.889320373535156,
+      "learning_rate": 9.214285714285715e-06,
+      "loss": 52.4786,
+      "step": 73
+    },
+    {
+      "epoch": 1.7641791044776118,
+      "grad_norm": 24.175626754760742,
+      "learning_rate": 9.202380952380953e-06,
+      "loss": 52.1792,
+      "step": 74
+    },
+    {
+      "epoch": 1.7880597014925375,
+      "grad_norm": 25.257095336914062,
+      "learning_rate": 9.19047619047619e-06,
+      "loss": 51.39,
+      "step": 75
+    },
+    {
+      "epoch": 1.8119402985074626,
+      "grad_norm": 35.745208740234375,
+      "learning_rate": 9.178571428571429e-06,
+      "loss": 53.331,
+      "step": 76
+    },
+    {
+      "epoch": 1.835820895522388,
+      "grad_norm": 23.815813064575195,
+      "learning_rate": 9.166666666666666e-06,
+      "loss": 52.7632,
+      "step": 77
+    },
+    {
+      "epoch": 1.8597014925373134,
+      "grad_norm": 32.405757904052734,
+      "learning_rate": 9.154761904761906e-06,
+      "loss": 52.3165,
+      "step": 78
+    },
+    {
+      "epoch": 1.8835820895522388,
+      "grad_norm": 38.95046615600586,
+      "learning_rate": 9.142857142857144e-06,
+      "loss": 52.0931,
+      "step": 79
+    },
+    {
+      "epoch": 1.9074626865671642,
+      "grad_norm": 22.412342071533203,
+      "learning_rate": 9.130952380952382e-06,
+      "loss": 51.8732,
+      "step": 80
+    },
+    {
+      "epoch": 1.9313432835820894,
+      "grad_norm": 35.088253021240234,
+      "learning_rate": 9.11904761904762e-06,
+      "loss": 52.0182,
+      "step": 81
+    },
+    {
+      "epoch": 1.955223880597015,
+      "grad_norm": 20.136964797973633,
+      "learning_rate": 9.107142857142858e-06,
+      "loss": 52.7956,
+      "step": 82
+    },
+    {
+      "epoch": 1.9791044776119402,
+      "grad_norm": 20.860034942626953,
+      "learning_rate": 9.095238095238095e-06,
+      "loss": 50.339,
+      "step": 83
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 20.38931655883789,
+      "learning_rate": 9.083333333333333e-06,
+      "loss": 45.3651,
+      "step": 84
+    },
+    {
+      "epoch": 2.023880597014925,
+      "grad_norm": 21.519498825073242,
+      "learning_rate": 9.071428571428573e-06,
+      "loss": 52.1228,
+      "step": 85
+    },
+    {
+      "epoch": 2.047761194029851,
+      "grad_norm": 30.47000503540039,
+      "learning_rate": 9.05952380952381e-06,
+      "loss": 51.1968,
+      "step": 86
+    },
+    {
+      "epoch": 2.071641791044776,
+      "grad_norm": 19.728044509887695,
+      "learning_rate": 9.047619047619049e-06,
+      "loss": 52.1514,
+      "step": 87
+    },
+    {
+      "epoch": 2.0955223880597016,
+      "grad_norm": 42.647281646728516,
+      "learning_rate": 9.035714285714287e-06,
+      "loss": 52.5597,
+      "step": 88
+    },
+    {
+      "epoch": 2.1194029850746268,
+      "grad_norm": 55.45186233520508,
+      "learning_rate": 9.023809523809524e-06,
+      "loss": 53.7621,
+      "step": 89
+    },
+    {
+      "epoch": 2.1432835820895524,
+      "grad_norm": 14.879026412963867,
+      "learning_rate": 9.011904761904762e-06,
+      "loss": 52.0743,
+      "step": 90
+    },
+    {
+      "epoch": 2.1671641791044776,
+      "grad_norm": 48.55704116821289,
+      "learning_rate": 9e-06,
+      "loss": 52.0543,
+      "step": 91
+    },
+    {
+      "epoch": 2.191044776119403,
+      "grad_norm": 37.96782302856445,
+      "learning_rate": 8.98809523809524e-06,
+      "loss": 52.4684,
+      "step": 92
+    },
+    {
+      "epoch": 2.2149253731343284,
+      "grad_norm": 27.714475631713867,
+      "learning_rate": 8.976190476190478e-06,
+      "loss": 52.725,
+      "step": 93
+    },
+    {
+      "epoch": 2.2388059701492535,
+      "grad_norm": 32.12433624267578,
+      "learning_rate": 8.964285714285716e-06,
+      "loss": 52.7161,
+      "step": 94
+    },
+    {
+      "epoch": 2.262686567164179,
+      "grad_norm": 23.8153018951416,
+      "learning_rate": 8.952380952380953e-06,
+      "loss": 51.3124,
+      "step": 95
+    },
+    {
+      "epoch": 2.2865671641791043,
+      "grad_norm": 31.269794464111328,
+      "learning_rate": 8.940476190476191e-06,
+      "loss": 51.9646,
+      "step": 96
+    },
+    {
+      "epoch": 2.31044776119403,
+      "grad_norm": 16.611865997314453,
+      "learning_rate": 8.92857142857143e-06,
+      "loss": 51.8503,
+      "step": 97
+    },
+    {
+      "epoch": 2.334328358208955,
+      "grad_norm": 26.69631004333496,
+      "learning_rate": 8.916666666666667e-06,
+      "loss": 52.4857,
+      "step": 98
+    },
+    {
+      "epoch": 2.3582089552238807,
+      "grad_norm": 21.10638999938965,
+      "learning_rate": 8.904761904761905e-06,
+      "loss": 52.0022,
+      "step": 99
+    },
+    {
+      "epoch": 2.382089552238806,
+      "grad_norm": 16.273351669311523,
+      "learning_rate": 8.892857142857143e-06,
+      "loss": 50.4367,
+      "step": 100
+    },
+    {
+      "epoch": 2.405970149253731,
+      "grad_norm": 16.407167434692383,
+      "learning_rate": 8.88095238095238e-06,
+      "loss": 51.2079,
+      "step": 101
+    },
+    {
+      "epoch": 2.4298507462686567,
+      "grad_norm": 16.22024154663086,
+      "learning_rate": 8.869047619047619e-06,
+      "loss": 50.4939,
+      "step": 102
+    },
+    {
+      "epoch": 2.4537313432835823,
+      "grad_norm": 27.11235809326172,
+      "learning_rate": 8.857142857142858e-06,
+      "loss": 50.0872,
+      "step": 103
+    },
+    {
+      "epoch": 2.4776119402985075,
+      "grad_norm": 18.912181854248047,
+      "learning_rate": 8.845238095238096e-06,
+      "loss": 51.8135,
+      "step": 104
+    },
+    {
+      "epoch": 2.5014925373134327,
+      "grad_norm": 29.597028732299805,
+      "learning_rate": 8.833333333333334e-06,
+      "loss": 49.4789,
+      "step": 105
+    },
+    {
+      "epoch": 2.5253731343283583,
+      "grad_norm": 30.51687240600586,
+      "learning_rate": 8.821428571428572e-06,
+      "loss": 52.5555,
+      "step": 106
+    },
+    {
+      "epoch": 2.5492537313432835,
+      "grad_norm": 31.4583797454834,
+      "learning_rate": 8.80952380952381e-06,
+      "loss": 51.0073,
+      "step": 107
+    },
+    {
+      "epoch": 2.573134328358209,
+      "grad_norm": 30.35653305053711,
+      "learning_rate": 8.797619047619048e-06,
+      "loss": 50.9501,
+      "step": 108
+    },
+    {
+      "epoch": 2.5970149253731343,
+      "grad_norm": 24.041545867919922,
+      "learning_rate": 8.785714285714286e-06,
+      "loss": 49.5162,
+      "step": 109
+    },
+    {
+      "epoch": 2.62089552238806,
+      "grad_norm": 23.52166175842285,
+      "learning_rate": 8.773809523809525e-06,
+      "loss": 52.9747,
+      "step": 110
+    },
+    {
+      "epoch": 2.644776119402985,
+      "grad_norm": 28.871065139770508,
+      "learning_rate": 8.761904761904763e-06,
+      "loss": 50.2273,
+      "step": 111
+    },
+    {
+      "epoch": 2.6686567164179102,
+      "grad_norm": 26.484140396118164,
+      "learning_rate": 8.750000000000001e-06,
+      "loss": 51.2286,
+      "step": 112
+    },
+    {
+      "epoch": 2.692537313432836,
+      "grad_norm": 37.570743560791016,
+      "learning_rate": 8.738095238095239e-06,
+      "loss": 49.7131,
+      "step": 113
+    },
+    {
+      "epoch": 2.716417910447761,
+      "grad_norm": 23.827178955078125,
+      "learning_rate": 8.726190476190477e-06,
+      "loss": 51.913,
+      "step": 114
+    },
+    {
+      "epoch": 2.7402985074626867,
+      "grad_norm": 33.89924621582031,
+      "learning_rate": 8.714285714285715e-06,
+      "loss": 52.2382,
+      "step": 115
+    },
+    {
+      "epoch": 2.764179104477612,
+      "grad_norm": 29.397851943969727,
+      "learning_rate": 8.702380952380952e-06,
+      "loss": 52.1548,
+      "step": 116
+    },
+    {
+      "epoch": 2.7880597014925375,
+      "grad_norm": 28.73517417907715,
+      "learning_rate": 8.690476190476192e-06,
+      "loss": 51.2892,
+      "step": 117
+    },
+    {
+      "epoch": 2.8119402985074626,
+      "grad_norm": 32.068138122558594,
+      "learning_rate": 8.67857142857143e-06,
+      "loss": 51.542,
+      "step": 118
+    },
+    {
+      "epoch": 2.835820895522388,
+      "grad_norm": 22.76898956298828,
+      "learning_rate": 8.666666666666668e-06,
+      "loss": 50.373,
+      "step": 119
+    },
+    {
+      "epoch": 2.8597014925373134,
+      "grad_norm": 33.528263092041016,
+      "learning_rate": 8.654761904761906e-06,
+      "loss": 51.3075,
+      "step": 120
+    },
+    {
+      "epoch": 2.883582089552239,
+      "grad_norm": 21.655696868896484,
+      "learning_rate": 8.642857142857144e-06,
+      "loss": 51.1397,
+      "step": 121
+    },
+    {
+      "epoch": 2.9074626865671642,
+      "grad_norm": 25.94880485534668,
+      "learning_rate": 8.630952380952381e-06,
+      "loss": 51.4326,
+      "step": 122
+    },
+    {
+      "epoch": 2.9313432835820894,
+      "grad_norm": 36.14421844482422,
+      "learning_rate": 8.61904761904762e-06,
+      "loss": 50.9524,
+      "step": 123
+    },
+    {
+      "epoch": 2.955223880597015,
+      "grad_norm": 28.361459732055664,
+      "learning_rate": 8.607142857142859e-06,
+      "loss": 51.3171,
+      "step": 124
+    },
+    {
+      "epoch": 2.97910447761194,
+      "grad_norm": 30.784954071044922,
+      "learning_rate": 8.595238095238097e-06,
+      "loss": 49.9797,
+      "step": 125
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 22.98565101623535,
+      "learning_rate": 8.583333333333333e-06,
+      "loss": 44.3471,
+      "step": 126
+    },
+    {
+      "epoch": 3.023880597014925,
+      "grad_norm": 25.601985931396484,
+      "learning_rate": 8.571428571428571e-06,
+      "loss": 51.6574,
+      "step": 127
+    },
+    {
+      "epoch": 3.047761194029851,
+      "grad_norm": 27.648792266845703,
+      "learning_rate": 8.55952380952381e-06,
+      "loss": 51.3385,
+      "step": 128
+    },
+    {
+      "epoch": 3.071641791044776,
+      "grad_norm": 18.773529052734375,
+      "learning_rate": 8.547619047619048e-06,
+      "loss": 51.1451,
+      "step": 129
+    },
+    {
+      "epoch": 3.0955223880597016,
+      "grad_norm": 29.439353942871094,
+      "learning_rate": 8.535714285714286e-06,
+      "loss": 51.6092,
+      "step": 130
+    },
+    {
+      "epoch": 3.1194029850746268,
+      "grad_norm": 32.41486740112305,
+      "learning_rate": 8.523809523809524e-06,
+      "loss": 50.9068,
+      "step": 131
+    },
+    {
+      "epoch": 3.1432835820895524,
+      "grad_norm": 23.441896438598633,
+      "learning_rate": 8.511904761904762e-06,
+      "loss": 51.7453,
+      "step": 132
+    },
+    {
+      "epoch": 3.1671641791044776,
+      "grad_norm": 29.218734741210938,
+      "learning_rate": 8.5e-06,
+      "loss": 49.9124,
+      "step": 133
+    },
+    {
+      "epoch": 3.191044776119403,
+      "grad_norm": 20.988981246948242,
+      "learning_rate": 8.488095238095238e-06,
+      "loss": 50.9788,
+      "step": 134
+    },
+    {
+      "epoch": 3.2149253731343284,
+      "grad_norm": 22.57052993774414,
+      "learning_rate": 8.476190476190477e-06,
+      "loss": 51.4228,
+      "step": 135
+    },
+    {
+      "epoch": 3.2388059701492535,
+      "grad_norm": 26.112573623657227,
+      "learning_rate": 8.464285714285715e-06,
+      "loss": 50.3332,
+      "step": 136
+    },
+    {
+      "epoch": 3.262686567164179,
+      "grad_norm": 23.8747615814209,
+      "learning_rate": 8.452380952380953e-06,
+      "loss": 51.1763,
+      "step": 137
+    },
+    {
+      "epoch": 3.2865671641791043,
+      "grad_norm": 24.12811851501465,
+      "learning_rate": 8.440476190476191e-06,
+      "loss": 49.8539,
+      "step": 138
+    },
+    {
+      "epoch": 3.31044776119403,
+      "grad_norm": 27.462984085083008,
+      "learning_rate": 8.428571428571429e-06,
+      "loss": 50.7766,
+      "step": 139
+    },
+    {
+      "epoch": 3.334328358208955,
+      "grad_norm": 31.261472702026367,
+      "learning_rate": 8.416666666666667e-06,
+      "loss": 49.7599,
+      "step": 140
+    },
+    {
+      "epoch": 3.3582089552238807,
+      "grad_norm": 21.049545288085938,
+      "learning_rate": 8.404761904761905e-06,
+      "loss": 49.6827,
+      "step": 141
+    },
+    {
+      "epoch": 3.382089552238806,
+      "grad_norm": 30.103389739990234,
+      "learning_rate": 8.392857142857144e-06,
+      "loss": 49.3866,
+      "step": 142
+    },
+    {
+      "epoch": 3.405970149253731,
+      "grad_norm": 31.348888397216797,
+      "learning_rate": 8.380952380952382e-06,
+      "loss": 51.4607,
+      "step": 143
+    },
+    {
+      "epoch": 3.4298507462686567,
+      "grad_norm": 28.910200119018555,
+      "learning_rate": 8.36904761904762e-06,
+      "loss": 51.2337,
+      "step": 144
+    },
+    {
+      "epoch": 3.4537313432835823,
+      "grad_norm": 21.00281524658203,
+      "learning_rate": 8.357142857142858e-06,
+      "loss": 50.9557,
+      "step": 145
+    },
+    {
+      "epoch": 3.4776119402985075,
+      "grad_norm": 45.842002868652344,
+      "learning_rate": 8.345238095238096e-06,
+      "loss": 49.7377,
+      "step": 146
+    },
+    {
+      "epoch": 3.5014925373134327,
+      "grad_norm": 30.77996253967285,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 51.1234,
+      "step": 147
+    },
+    {
+      "epoch": 3.5253731343283583,
+      "grad_norm": 31.492767333984375,
+      "learning_rate": 8.321428571428573e-06,
+      "loss": 50.5733,
+      "step": 148
+    },
+    {
+      "epoch": 3.5492537313432835,
+      "grad_norm": 36.57206344604492,
+      "learning_rate": 8.309523809523811e-06,
+      "loss": 50.6762,
+      "step": 149
+    },
+    {
+      "epoch": 3.573134328358209,
+      "grad_norm": 33.86347198486328,
+      "learning_rate": 8.297619047619049e-06,
+      "loss": 50.8281,
+      "step": 150
+    },
+    {
+      "epoch": 3.5970149253731343,
+      "grad_norm": 30.812152862548828,
+      "learning_rate": 8.285714285714287e-06,
+      "loss": 50.6509,
+      "step": 151
+    },
+    {
+      "epoch": 3.62089552238806,
+      "grad_norm": 24.536882400512695,
+      "learning_rate": 8.273809523809523e-06,
+      "loss": 50.1112,
+      "step": 152
+    },
+    {
+      "epoch": 3.644776119402985,
+      "grad_norm": 29.8430233001709,
+      "learning_rate": 8.261904761904763e-06,
+      "loss": 50.846,
+      "step": 153
+    },
+    {
+      "epoch": 3.6686567164179102,
+      "grad_norm": 26.18596076965332,
+      "learning_rate": 8.25e-06,
+      "loss": 50.3806,
+      "step": 154
+    },
+    {
+      "epoch": 3.692537313432836,
+      "grad_norm": 38.75019836425781,
+      "learning_rate": 8.238095238095239e-06,
+      "loss": 49.8915,
+      "step": 155
+    },
+    {
+      "epoch": 3.716417910447761,
+      "grad_norm": 34.30149841308594,
+      "learning_rate": 8.226190476190476e-06,
+      "loss": 50.7886,
+      "step": 156
+    },
+    {
+      "epoch": 3.7402985074626867,
+      "grad_norm": 33.179298400878906,
+      "learning_rate": 8.214285714285714e-06,
+      "loss": 50.8175,
+      "step": 157
+    },
+    {
+      "epoch": 3.764179104477612,
+      "grad_norm": 34.90909957885742,
+      "learning_rate": 8.202380952380952e-06,
+      "loss": 50.3521,
+      "step": 158
+    },
+    {
+      "epoch": 3.7880597014925375,
+      "grad_norm": 33.2717399597168,
+      "learning_rate": 8.190476190476192e-06,
+      "loss": 51.2006,
+      "step": 159
+    },
+    {
+      "epoch": 3.8119402985074626,
+      "grad_norm": 33.082672119140625,
+      "learning_rate": 8.17857142857143e-06,
+      "loss": 49.5627,
+      "step": 160
+    },
+    {
+      "epoch": 3.835820895522388,
+      "grad_norm": 23.65228843688965,
+      "learning_rate": 8.166666666666668e-06,
+      "loss": 49.9631,
+      "step": 161
+    },
+    {
+      "epoch": 3.8597014925373134,
+      "grad_norm": 37.3172492980957,
+      "learning_rate": 8.154761904761905e-06,
+      "loss": 50.7175,
+      "step": 162
+    },
+    {
+      "epoch": 3.883582089552239,
+      "grad_norm": 29.369930267333984,
+      "learning_rate": 8.142857142857143e-06,
+      "loss": 51.1435,
+      "step": 163
+    },
+    {
+      "epoch": 3.9074626865671642,
+      "grad_norm": 28.807470321655273,
+      "learning_rate": 8.130952380952381e-06,
+      "loss": 50.3349,
+      "step": 164
+    },
+    {
+      "epoch": 3.9313432835820894,
+      "grad_norm": 33.90628433227539,
+      "learning_rate": 8.119047619047619e-06,
+      "loss": 50.7241,
+      "step": 165
+    },
+    {
+      "epoch": 3.955223880597015,
+      "grad_norm": 21.72952651977539,
+      "learning_rate": 8.107142857142859e-06,
+      "loss": 49.9013,
+      "step": 166
+    },
+    {
+      "epoch": 3.97910447761194,
+      "grad_norm": 26.831520080566406,
+      "learning_rate": 8.095238095238097e-06,
+      "loss": 51.0161,
+      "step": 167
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 24.49069595336914,
+      "learning_rate": 8.083333333333334e-06,
+      "loss": 44.6758,
+      "step": 168
+    },
+    {
+      "epoch": 4.023880597014926,
+      "grad_norm": 36.32711410522461,
+      "learning_rate": 8.071428571428572e-06,
+      "loss": 49.8601,
+      "step": 169
+    },
+    {
+      "epoch": 4.04776119402985,
+      "grad_norm": 29.862812042236328,
+      "learning_rate": 8.05952380952381e-06,
+      "loss": 51.712,
+      "step": 170
+    },
+    {
+      "epoch": 4.071641791044776,
+      "grad_norm": 40.245887756347656,
+      "learning_rate": 8.047619047619048e-06,
+      "loss": 50.3353,
+      "step": 171
+    },
+    {
+      "epoch": 4.095522388059702,
+      "grad_norm": 34.22684097290039,
+      "learning_rate": 8.035714285714286e-06,
+      "loss": 50.6474,
+      "step": 172
+    },
+    {
+      "epoch": 4.119402985074627,
+      "grad_norm": 36.754669189453125,
+      "learning_rate": 8.023809523809526e-06,
+      "loss": 50.1623,
+      "step": 173
+    },
+    {
+      "epoch": 4.143283582089552,
+      "grad_norm": 35.76541519165039,
+      "learning_rate": 8.011904761904763e-06,
+      "loss": 50.2426,
+      "step": 174
+    },
+    {
+      "epoch": 4.167164179104478,
+      "grad_norm": 25.851362228393555,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 49.9525,
+      "step": 175
+    },
+    {
+      "epoch": 4.191044776119403,
+      "grad_norm": 24.48278045654297,
+      "learning_rate": 7.98809523809524e-06,
+      "loss": 49.1466,
+      "step": 176
+    },
+    {
+      "epoch": 4.214925373134328,
+      "grad_norm": 28.79146385192871,
+      "learning_rate": 7.976190476190477e-06,
+      "loss": 49.9365,
+      "step": 177
+    },
+    {
+      "epoch": 4.2388059701492535,
+      "grad_norm": 29.29482650756836,
+      "learning_rate": 7.964285714285715e-06,
+      "loss": 50.7427,
+      "step": 178
+    },
+    {
+      "epoch": 4.262686567164179,
+      "grad_norm": 23.50571060180664,
+      "learning_rate": 7.952380952380953e-06,
+      "loss": 49.7287,
+      "step": 179
+    },
+    {
+      "epoch": 4.286567164179105,
+      "grad_norm": 27.805828094482422,
+      "learning_rate": 7.94047619047619e-06,
+      "loss": 50.4316,
+      "step": 180
+    },
+    {
+      "epoch": 4.3104477611940295,
+      "grad_norm": 28.323888778686523,
+      "learning_rate": 7.928571428571429e-06,
+      "loss": 50.0263,
+      "step": 181
+    },
+    {
+      "epoch": 4.334328358208955,
+      "grad_norm": 25.43438148498535,
+      "learning_rate": 7.916666666666667e-06,
+      "loss": 49.426,
+      "step": 182
+    },
+    {
+      "epoch": 4.358208955223881,
+      "grad_norm": 22.169496536254883,
+      "learning_rate": 7.904761904761904e-06,
+      "loss": 51.1048,
+      "step": 183
+    },
+    {
+      "epoch": 4.382089552238806,
+      "grad_norm": 33.660545349121094,
+      "learning_rate": 7.892857142857144e-06,
+      "loss": 49.7654,
+      "step": 184
+    },
+    {
+      "epoch": 4.405970149253731,
+      "grad_norm": 24.276273727416992,
+      "learning_rate": 7.880952380952382e-06,
+      "loss": 50.4976,
+      "step": 185
+    },
+    {
+      "epoch": 4.429850746268657,
+      "grad_norm": 41.48741149902344,
+      "learning_rate": 7.86904761904762e-06,
+      "loss": 52.0386,
+      "step": 186
+    },
+    {
+      "epoch": 4.453731343283582,
+      "grad_norm": 25.86789894104004,
+      "learning_rate": 7.857142857142858e-06,
+      "loss": 49.129,
+      "step": 187
+    },
+    {
+      "epoch": 4.477611940298507,
+      "grad_norm": 26.607038497924805,
+      "learning_rate": 7.845238095238096e-06,
+      "loss": 49.3561,
+      "step": 188
+    },
+    {
+      "epoch": 4.501492537313433,
+      "grad_norm": 43.54303741455078,
+      "learning_rate": 7.833333333333333e-06,
+      "loss": 50.1143,
+      "step": 189
+    },
+    {
+      "epoch": 4.525373134328358,
+      "grad_norm": 45.6146354675293,
+      "learning_rate": 7.821428571428571e-06,
+      "loss": 49.3217,
+      "step": 190
+    },
+    {
+      "epoch": 4.549253731343284,
+      "grad_norm": 24.00080680847168,
+      "learning_rate": 7.809523809523811e-06,
+      "loss": 50.484,
+      "step": 191
+    },
+    {
+      "epoch": 4.573134328358209,
+      "grad_norm": 29.736740112304688,
+      "learning_rate": 7.797619047619049e-06,
+      "loss": 49.748,
+      "step": 192
+    },
+    {
+      "epoch": 4.597014925373134,
+      "grad_norm": 33.08702850341797,
+      "learning_rate": 7.785714285714287e-06,
+      "loss": 50.2142,
+      "step": 193
+    },
+    {
+      "epoch": 4.62089552238806,
+      "grad_norm": 19.16411018371582,
+      "learning_rate": 7.773809523809525e-06,
+      "loss": 50.1073,
+      "step": 194
+    },
+    {
+      "epoch": 4.6447761194029855,
+      "grad_norm": 32.145721435546875,
+      "learning_rate": 7.761904761904762e-06,
+      "loss": 48.5769,
+      "step": 195
+    },
+    {
+      "epoch": 4.66865671641791,
+      "grad_norm": 38.768341064453125,
+      "learning_rate": 7.75e-06,
+      "loss": 49.681,
+      "step": 196
+    },
+    {
+      "epoch": 4.692537313432836,
+      "grad_norm": 26.108245849609375,
+      "learning_rate": 7.738095238095238e-06,
+      "loss": 49.9193,
+      "step": 197
+    },
+    {
+      "epoch": 4.7164179104477615,
+      "grad_norm": 28.86294174194336,
+      "learning_rate": 7.726190476190478e-06,
+      "loss": 50.4584,
+      "step": 198
+    },
+    {
+      "epoch": 4.740298507462686,
+      "grad_norm": 31.089380264282227,
+      "learning_rate": 7.714285714285716e-06,
+      "loss": 50.7873,
+      "step": 199
+    },
+    {
+      "epoch": 4.764179104477612,
+      "grad_norm": 22.934032440185547,
+      "learning_rate": 7.702380952380954e-06,
+      "loss": 50.611,
+      "step": 200
+    },
+    {
+      "epoch": 4.7880597014925375,
+      "grad_norm": 27.986371994018555,
+      "learning_rate": 7.690476190476191e-06,
+      "loss": 49.275,
+      "step": 201
+    },
+    {
+      "epoch": 4.811940298507462,
+      "grad_norm": 23.44196319580078,
+      "learning_rate": 7.67857142857143e-06,
+      "loss": 50.2035,
+      "step": 202
+    },
+    {
+      "epoch": 4.835820895522388,
+      "grad_norm": 22.05059242248535,
+      "learning_rate": 7.666666666666667e-06,
+      "loss": 48.9595,
+      "step": 203
+    },
+    {
+      "epoch": 4.859701492537313,
+      "grad_norm": 29.709396362304688,
+      "learning_rate": 7.654761904761905e-06,
+      "loss": 50.5343,
+      "step": 204
+    },
+    {
+      "epoch": 4.883582089552239,
+      "grad_norm": 23.702781677246094,
+      "learning_rate": 7.642857142857143e-06,
+      "loss": 50.2627,
+      "step": 205
+    },
+    {
+      "epoch": 4.907462686567165,
+      "grad_norm": 20.144807815551758,
+      "learning_rate": 7.630952380952381e-06,
+      "loss": 51.0125,
+      "step": 206
+    },
+    {
+      "epoch": 4.931343283582089,
+      "grad_norm": 28.83676528930664,
+      "learning_rate": 7.61904761904762e-06,
+      "loss": 50.5985,
+      "step": 207
+    },
+    {
+      "epoch": 4.955223880597015,
+      "grad_norm": 34.40160369873047,
+      "learning_rate": 7.6071428571428575e-06,
+      "loss": 49.6469,
+      "step": 208
+    },
+    {
+      "epoch": 4.979104477611941,
+      "grad_norm": 26.982925415039062,
+      "learning_rate": 7.595238095238095e-06,
+      "loss": 50.1666,
+      "step": 209
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 19.569746017456055,
+      "learning_rate": 7.583333333333333e-06,
+      "loss": 43.6715,
+      "step": 210
+    },
+    {
+      "epoch": 5.023880597014926,
+      "grad_norm": 23.753328323364258,
+      "learning_rate": 7.571428571428572e-06,
+      "loss": 49.9273,
+      "step": 211
+    },
+    {
+      "epoch": 5.04776119402985,
+      "grad_norm": 22.463659286499023,
+      "learning_rate": 7.55952380952381e-06,
+      "loss": 48.8499,
+      "step": 212
+    },
+    {
+      "epoch": 5.071641791044776,
+      "grad_norm": 24.507875442504883,
+      "learning_rate": 7.547619047619048e-06,
+      "loss": 49.3275,
+      "step": 213
+    },
+    {
+      "epoch": 5.095522388059702,
+      "grad_norm": 21.727603912353516,
+      "learning_rate": 7.5357142857142865e-06,
+      "loss": 49.1879,
+      "step": 214
+    },
+    {
+      "epoch": 5.119402985074627,
+      "grad_norm": 26.122251510620117,
+      "learning_rate": 7.523809523809524e-06,
+      "loss": 50.1094,
+      "step": 215
+    },
+    {
+      "epoch": 5.143283582089552,
+      "grad_norm": 24.142263412475586,
+      "learning_rate": 7.511904761904762e-06,
+      "loss": 50.2708,
+      "step": 216
+    },
+    {
+      "epoch": 5.167164179104478,
+      "grad_norm": 22.762237548828125,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 50.441,
+      "step": 217
+    },
+    {
+      "epoch": 5.191044776119403,
+      "grad_norm": 35.74570846557617,
+      "learning_rate": 7.488095238095239e-06,
+      "loss": 48.5121,
+      "step": 218
+    },
+    {
+      "epoch": 5.214925373134328,
+      "grad_norm": 30.92180824279785,
+      "learning_rate": 7.476190476190477e-06,
+      "loss": 49.4257,
+      "step": 219
+    },
+    {
+      "epoch": 5.2388059701492535,
+      "grad_norm": 26.90997314453125,
+      "learning_rate": 7.464285714285715e-06,
+      "loss": 50.9712,
+      "step": 220
+    },
+    {
+      "epoch": 5.262686567164179,
+      "grad_norm": 35.544700622558594,
+      "learning_rate": 7.4523809523809534e-06,
+      "loss": 49.7908,
+      "step": 221
+    },
+    {
+      "epoch": 5.286567164179105,
+      "grad_norm": 33.78145217895508,
+      "learning_rate": 7.440476190476191e-06,
+      "loss": 49.2105,
+      "step": 222
+    },
+    {
+      "epoch": 5.3104477611940295,
+      "grad_norm": 32.16508102416992,
+      "learning_rate": 7.428571428571429e-06,
+      "loss": 49.8545,
+      "step": 223
+    },
+    {
+      "epoch": 5.334328358208955,
+      "grad_norm": 30.4263973236084,
+      "learning_rate": 7.416666666666668e-06,
+      "loss": 50.0994,
+      "step": 224
+    },
+    {
+      "epoch": 5.358208955223881,
+      "grad_norm": 25.801084518432617,
+      "learning_rate": 7.404761904761906e-06,
+      "loss": 49.6227,
+      "step": 225
+    },
+    {
+      "epoch": 5.382089552238806,
+      "grad_norm": 27.16851234436035,
+      "learning_rate": 7.392857142857144e-06,
+      "loss": 50.0005,
+      "step": 226
+    },
+    {
+      "epoch": 5.405970149253731,
+      "grad_norm": 30.102867126464844,
+      "learning_rate": 7.380952380952382e-06,
+      "loss": 50.7114,
+      "step": 227
+    },
+    {
+      "epoch": 5.429850746268657,
+      "grad_norm": 26.032968521118164,
+      "learning_rate": 7.36904761904762e-06,
+      "loss": 48.99,
+      "step": 228
+    },
+    {
+      "epoch": 5.453731343283582,
+      "grad_norm": 24.074424743652344,
+      "learning_rate": 7.357142857142858e-06,
+      "loss": 49.6965,
+      "step": 229
+    },
+    {
+      "epoch": 5.477611940298507,
+      "grad_norm": 24.5870361328125,
+      "learning_rate": 7.345238095238096e-06,
+      "loss": 48.8593,
+      "step": 230
+    },
+    {
+      "epoch": 5.501492537313433,
+      "grad_norm": 22.831932067871094,
+      "learning_rate": 7.333333333333333e-06,
+      "loss": 49.9975,
+      "step": 231
+    },
+    {
+      "epoch": 5.525373134328358,
+      "grad_norm": 26.88197135925293,
+      "learning_rate": 7.321428571428572e-06,
+      "loss": 49.5131,
+      "step": 232
+    },
+    {
+      "epoch": 5.549253731343284,
+      "grad_norm": 28.986154556274414,
+      "learning_rate": 7.30952380952381e-06,
+      "loss": 48.9042,
+      "step": 233
+    },
+    {
+      "epoch": 5.573134328358209,
+      "grad_norm": 17.220605850219727,
+      "learning_rate": 7.297619047619048e-06,
+      "loss": 49.453,
+      "step": 234
+    },
+    {
+      "epoch": 5.597014925373134,
+      "grad_norm": 25.110107421875,
+      "learning_rate": 7.285714285714286e-06,
+      "loss": 49.8196,
+      "step": 235
+    },
+    {
+      "epoch": 5.62089552238806,
+      "grad_norm": 34.680870056152344,
+      "learning_rate": 7.273809523809524e-06,
+      "loss": 49.9709,
+      "step": 236
+    },
+    {
+      "epoch": 5.6447761194029855,
+      "grad_norm": 24.10121726989746,
+      "learning_rate": 7.261904761904762e-06,
+      "loss": 49.4685,
+      "step": 237
+    },
+    {
+      "epoch": 5.66865671641791,
+      "grad_norm": 28.65550422668457,
+      "learning_rate": 7.25e-06,
+      "loss": 50.3439,
+      "step": 238
+    },
+    {
+      "epoch": 5.692537313432836,
+      "grad_norm": 27.49604606628418,
+      "learning_rate": 7.238095238095239e-06,
+      "loss": 50.0419,
+      "step": 239
+    },
+    {
+      "epoch": 5.7164179104477615,
+      "grad_norm": 22.843509674072266,
+      "learning_rate": 7.226190476190477e-06,
+      "loss": 49.0357,
+      "step": 240
+    },
+    {
+      "epoch": 5.740298507462686,
+      "grad_norm": 36.56801223754883,
+      "learning_rate": 7.2142857142857145e-06,
+      "loss": 49.4478,
+      "step": 241
+    },
+    {
+      "epoch": 5.764179104477612,
+      "grad_norm": 38.233734130859375,
+      "learning_rate": 7.202380952380953e-06,
+      "loss": 50.4473,
+      "step": 242
+    },
+    {
+      "epoch": 5.7880597014925375,
+      "grad_norm": 29.198333740234375,
+      "learning_rate": 7.190476190476191e-06,
+      "loss": 49.8598,
+      "step": 243
+    },
+    {
+      "epoch": 5.811940298507462,
+      "grad_norm": 34.49404525756836,
+      "learning_rate": 7.178571428571429e-06,
+      "loss": 49.1441,
+      "step": 244
+    },
+    {
+      "epoch": 5.835820895522388,
+      "grad_norm": 35.568359375,
+      "learning_rate": 7.166666666666667e-06,
+      "loss": 49.9402,
+      "step": 245
+    },
+    {
+      "epoch": 5.859701492537313,
+      "grad_norm": 31.041446685791016,
+      "learning_rate": 7.154761904761906e-06,
+      "loss": 50.1265,
+      "step": 246
+    },
+    {
+      "epoch": 5.883582089552239,
+      "grad_norm": 48.34186935424805,
+      "learning_rate": 7.1428571428571436e-06,
+      "loss": 50.7649,
+      "step": 247
+    },
+    {
+      "epoch": 5.907462686567165,
+      "grad_norm": 39.171661376953125,
+      "learning_rate": 7.1309523809523814e-06,
+      "loss": 48.943,
+      "step": 248
+    },
+    {
+      "epoch": 5.931343283582089,
+      "grad_norm": 28.724523544311523,
+      "learning_rate": 7.11904761904762e-06,
+      "loss": 50.8039,
+      "step": 249
+    },
+    {
+      "epoch": 5.955223880597015,
+      "grad_norm": 36.57830810546875,
+      "learning_rate": 7.107142857142858e-06,
+      "loss": 50.2311,
+      "step": 250
+    },
+    {
+      "epoch": 5.979104477611941,
+      "grad_norm": 39.91551971435547,
+      "learning_rate": 7.095238095238096e-06,
+      "loss": 49.1617,
+      "step": 251
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 17.104145050048828,
+      "learning_rate": 7.083333333333335e-06,
+      "loss": 42.8003,
+      "step": 252
+    },
+    {
+      "epoch": 6.023880597014926,
+      "grad_norm": 33.03441619873047,
+      "learning_rate": 7.0714285714285726e-06,
+      "loss": 48.2896,
+      "step": 253
+    },
+    {
+      "epoch": 6.04776119402985,
+      "grad_norm": 26.487470626831055,
+      "learning_rate": 7.0595238095238105e-06,
+      "loss": 49.205,
+      "step": 254
+    },
+    {
+      "epoch": 6.071641791044776,
+      "grad_norm": 26.752981185913086,
+      "learning_rate": 7.047619047619048e-06,
+      "loss": 50.3943,
+      "step": 255
+    },
+    {
+      "epoch": 6.095522388059702,
+      "grad_norm": 22.44376564025879,
+      "learning_rate": 7.035714285714287e-06,
+      "loss": 49.285,
+      "step": 256
+    },
+    {
+      "epoch": 6.119402985074627,
+      "grad_norm": 31.066368103027344,
+      "learning_rate": 7.023809523809524e-06,
+      "loss": 49.3131,
+      "step": 257
+    },
+    {
+      "epoch": 6.143283582089552,
+      "grad_norm": 28.67262840270996,
+      "learning_rate": 7.011904761904762e-06,
+      "loss": 50.6188,
+      "step": 258
+    },
+    {
+      "epoch": 6.167164179104478,
+      "grad_norm": 24.013134002685547,
+      "learning_rate": 7e-06,
+      "loss": 50.4382,
+      "step": 259
+    },
+    {
+      "epoch": 6.191044776119403,
+      "grad_norm": 26.5673828125,
+      "learning_rate": 6.988095238095239e-06,
+      "loss": 49.7058,
+      "step": 260
+    },
+    {
+      "epoch": 6.214925373134328,
+      "grad_norm": 20.803695678710938,
+      "learning_rate": 6.9761904761904765e-06,
+      "loss": 48.9389,
+      "step": 261
+    },
+    {
+      "epoch": 6.2388059701492535,
+      "grad_norm": 23.450183868408203,
+      "learning_rate": 6.964285714285714e-06,
+      "loss": 49.0091,
+      "step": 262
+    },
+    {
+      "epoch": 6.262686567164179,
+      "grad_norm": 36.94446563720703,
+      "learning_rate": 6.952380952380952e-06,
+      "loss": 50.2589,
+      "step": 263
+    },
+    {
+      "epoch": 6.286567164179105,
+      "grad_norm": 39.548095703125,
+      "learning_rate": 6.940476190476191e-06,
+      "loss": 49.3129,
+      "step": 264
+    },
+    {
+      "epoch": 6.3104477611940295,
+      "grad_norm": 30.536083221435547,
+      "learning_rate": 6.928571428571429e-06,
+      "loss": 49.1838,
+      "step": 265
+    },
+    {
+      "epoch": 6.334328358208955,
+      "grad_norm": 27.97296714782715,
+      "learning_rate": 6.916666666666667e-06,
+      "loss": 50.3184,
+      "step": 266
+    },
+    {
+      "epoch": 6.358208955223881,
+      "grad_norm": 25.69655418395996,
+      "learning_rate": 6.9047619047619055e-06,
+      "loss": 49.2226,
+      "step": 267
+    },
+    {
+      "epoch": 6.382089552238806,
+      "grad_norm": 22.114097595214844,
+      "learning_rate": 6.892857142857143e-06,
+      "loss": 49.4455,
+      "step": 268
+    },
+    {
+      "epoch": 6.405970149253731,
+      "grad_norm": 30.47511100769043,
+      "learning_rate": 6.880952380952381e-06,
+      "loss": 49.7409,
+      "step": 269
+    },
+    {
+      "epoch": 6.429850746268657,
+      "grad_norm": 26.32929039001465,
+      "learning_rate": 6.86904761904762e-06,
+      "loss": 50.3336,
+      "step": 270
+    },
+    {
+      "epoch": 6.453731343283582,
+      "grad_norm": 28.09309196472168,
+      "learning_rate": 6.857142857142858e-06,
+      "loss": 49.6044,
+      "step": 271
+    },
+    {
+      "epoch": 6.477611940298507,
+      "grad_norm": 25.840974807739258,
+      "learning_rate": 6.845238095238096e-06,
+      "loss": 49.9185,
+      "step": 272
+    },
+    {
+      "epoch": 6.501492537313433,
+      "grad_norm": 31.89126205444336,
+      "learning_rate": 6.833333333333334e-06,
+      "loss": 48.2732,
+      "step": 273
+    },
+    {
+      "epoch": 6.525373134328358,
+      "grad_norm": 24.013029098510742,
+      "learning_rate": 6.8214285714285724e-06,
+      "loss": 49.9752,
+      "step": 274
+    },
+    {
+      "epoch": 6.549253731343284,
+      "grad_norm": 25.509836196899414,
+      "learning_rate": 6.80952380952381e-06,
+      "loss": 50.5493,
+      "step": 275
+    },
+    {
+      "epoch": 6.573134328358209,
+      "grad_norm": 35.25442886352539,
+      "learning_rate": 6.797619047619048e-06,
+      "loss": 49.2553,
+      "step": 276
+    },
+    {
+      "epoch": 6.597014925373134,
+      "grad_norm": 29.42585563659668,
+      "learning_rate": 6.785714285714287e-06,
+      "loss": 48.776,
+      "step": 277
+    },
+    {
+      "epoch": 6.62089552238806,
+      "grad_norm": 25.90894889831543,
+      "learning_rate": 6.773809523809525e-06,
+      "loss": 49.1964,
+      "step": 278
+    },
+    {
+      "epoch": 6.6447761194029855,
+      "grad_norm": 25.63600730895996,
+      "learning_rate": 6.761904761904763e-06,
+      "loss": 48.4528,
+      "step": 279
+    },
+    {
+      "epoch": 6.66865671641791,
+      "grad_norm": 29.943740844726562,
+      "learning_rate": 6.750000000000001e-06,
+      "loss": 49.9026,
+      "step": 280
+    },
+    {
+      "epoch": 6.692537313432836,
+      "grad_norm": 33.253910064697266,
+      "learning_rate": 6.738095238095239e-06,
+      "loss": 49.2364,
+      "step": 281
+    },
+    {
+      "epoch": 6.7164179104477615,
+      "grad_norm": 23.465354919433594,
+      "learning_rate": 6.726190476190477e-06,
+      "loss": 49.2759,
+      "step": 282
+    },
+    {
+      "epoch": 6.740298507462686,
+      "grad_norm": 31.023218154907227,
+      "learning_rate": 6.714285714285714e-06,
+      "loss": 49.3256,
+      "step": 283
+    },
+    {
+      "epoch": 6.764179104477612,
+      "grad_norm": 32.376991271972656,
+      "learning_rate": 6.702380952380952e-06,
+      "loss": 47.7239,
+      "step": 284
+    },
+    {
+      "epoch": 6.7880597014925375,
+      "grad_norm": 18.388896942138672,
+      "learning_rate": 6.690476190476191e-06,
+      "loss": 49.5751,
+      "step": 285
+    },
+    {
+      "epoch": 6.811940298507462,
+      "grad_norm": 22.1639404296875,
+      "learning_rate": 6.678571428571429e-06,
+      "loss": 48.8153,
+      "step": 286
+    },
+    {
+      "epoch": 6.835820895522388,
+      "grad_norm": 31.39455223083496,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 49.7872,
+      "step": 287
+    },
+    {
+      "epoch": 6.859701492537313,
+      "grad_norm": 36.480533599853516,
+      "learning_rate": 6.654761904761905e-06,
+      "loss": 48.9679,
+      "step": 288
+    },
+    {
+      "epoch": 6.883582089552239,
+      "grad_norm": 23.432872772216797,
+      "learning_rate": 6.642857142857143e-06,
+      "loss": 48.4627,
+      "step": 289
+    },
+    {
+      "epoch": 6.907462686567165,
+      "grad_norm": 33.31097412109375,
+      "learning_rate": 6.630952380952381e-06,
+      "loss": 49.617,
+      "step": 290
+    },
+    {
+      "epoch": 6.931343283582089,
+      "grad_norm": 34.07685852050781,
+      "learning_rate": 6.619047619047619e-06,
+      "loss": 49.3365,
+      "step": 291
+    },
+    {
+      "epoch": 6.955223880597015,
+      "grad_norm": 25.68811798095703,
+      "learning_rate": 6.607142857142858e-06,
+      "loss": 49.7821,
+      "step": 292
+    },
+    {
+      "epoch": 6.979104477611941,
+      "grad_norm": 24.179588317871094,
+      "learning_rate": 6.595238095238096e-06,
+      "loss": 48.5466,
+      "step": 293
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 26.158781051635742,
+      "learning_rate": 6.5833333333333335e-06,
+      "loss": 43.2838,
+      "step": 294
+    },
+    {
+      "epoch": 7.023880597014926,
+      "grad_norm": 22.84689712524414,
+      "learning_rate": 6.571428571428572e-06,
+      "loss": 49.3253,
+      "step": 295
+    },
+    {
+      "epoch": 7.04776119402985,
+      "grad_norm": NaN,
+      "learning_rate": 6.55952380952381e-06,
+      "loss": 75.2786,
+      "step": 296
+    },
+    {
+      "epoch": 7.071641791044776,
+      "grad_norm": 31.13886070251465,
+      "learning_rate": 6.55952380952381e-06,
+      "loss": 49.984,
+      "step": 297
+    },
+    {
+      "epoch": 7.095522388059702,
+      "grad_norm": 32.37982940673828,
+      "learning_rate": 6.547619047619048e-06,
+      "loss": 49.6632,
+      "step": 298
+    },
+    {
+      "epoch": 7.119402985074627,
+      "grad_norm": 22.977916717529297,
+      "learning_rate": 6.535714285714286e-06,
+      "loss": 48.7802,
+      "step": 299
+    },
+    {
+      "epoch": 7.143283582089552,
+      "grad_norm": NaN,
+      "learning_rate": 6.523809523809525e-06,
+      "loss": 60.3381,
+      "step": 300
+    },
+    {
+      "epoch": 7.167164179104478,
+      "grad_norm": 32.18650817871094,
+      "learning_rate": 6.523809523809525e-06,
+      "loss": 49.2689,
+      "step": 301
+    },
+    {
+      "epoch": 7.191044776119403,
+      "grad_norm": 30.0800724029541,
+      "learning_rate": 6.5119047619047626e-06,
+      "loss": 49.3891,
+      "step": 302
+    },
+    {
+      "epoch": 7.214925373134328,
+      "grad_norm": 32.35110855102539,
+      "learning_rate": 6.5000000000000004e-06,
+      "loss": 48.4497,
+      "step": 303
+    },
+    {
+      "epoch": 7.2388059701492535,
+      "grad_norm": 34.08786392211914,
+      "learning_rate": 6.488095238095239e-06,
+      "loss": 49.1321,
+      "step": 304
+    },
+    {
+      "epoch": 7.262686567164179,
+      "grad_norm": 25.25969696044922,
+      "learning_rate": 6.476190476190477e-06,
+      "loss": 49.0524,
+      "step": 305
+    },
+    {
+      "epoch": 7.286567164179105,
+      "grad_norm": 25.843929290771484,
+      "learning_rate": 6.464285714285715e-06,
+      "loss": 49.8077,
+      "step": 306
+    },
+    {
+      "epoch": 7.3104477611940295,
+      "grad_norm": 34.57284927368164,
+      "learning_rate": 6.452380952380954e-06,
+      "loss": 49.5393,
+      "step": 307
+    },
+    {
+      "epoch": 7.334328358208955,
+      "grad_norm": 33.44814682006836,
+      "learning_rate": 6.4404761904761916e-06,
+      "loss": 49.0375,
+      "step": 308
+    },
+    {
+      "epoch": 7.358208955223881,
+      "grad_norm": 25.127429962158203,
+      "learning_rate": 6.4285714285714295e-06,
+      "loss": 48.8145,
+      "step": 309
+    },
+    {
+      "epoch": 7.382089552238806,
+      "grad_norm": 31.81999969482422,
+      "learning_rate": 6.416666666666667e-06,
+      "loss": 49.6432,
+      "step": 310
+    },
+    {
+      "epoch": 7.405970149253731,
+      "grad_norm": 22.428335189819336,
+      "learning_rate": 6.404761904761904e-06,
+      "loss": 47.6232,
+      "step": 311
+    },
+    {
+      "epoch": 7.429850746268657,
+      "grad_norm": 45.87803268432617,
+      "learning_rate": 6.392857142857143e-06,
+      "loss": 48.3479,
+      "step": 312
+    },
+    {
+      "epoch": 7.453731343283582,
+      "grad_norm": 37.441253662109375,
+      "learning_rate": 6.380952380952381e-06,
+      "loss": 48.593,
+      "step": 313
+    },
+    {
+      "epoch": 7.477611940298507,
+      "grad_norm": 23.15785789489746,
+      "learning_rate": 6.369047619047619e-06,
+      "loss": 49.1204,
+      "step": 314
+    },
+    {
+      "epoch": 7.501492537313433,
+      "grad_norm": 35.8905029296875,
+      "learning_rate": 6.357142857142858e-06,
+      "loss": 49.2918,
+      "step": 315
+    },
+    {
+      "epoch": 7.525373134328358,
+      "grad_norm": 37.41954040527344,
+      "learning_rate": 6.3452380952380955e-06,
+      "loss": 47.7495,
+      "step": 316
+    },
+    {
+      "epoch": 7.549253731343284,
+      "grad_norm": 31.173114776611328,
+      "learning_rate": 6.333333333333333e-06,
+      "loss": 49.539,
+      "step": 317
+    },
+    {
+      "epoch": 7.573134328358209,
+      "grad_norm": 23.941965103149414,
+      "learning_rate": 6.321428571428571e-06,
+      "loss": 49.0958,
+      "step": 318
+    },
+    {
+      "epoch": 7.597014925373134,
+      "grad_norm": 31.949769973754883,
+      "learning_rate": 6.30952380952381e-06,
+      "loss": 49.1945,
+      "step": 319
+    },
+    {
+      "epoch": 7.62089552238806,
+      "grad_norm": 21.299409866333008,
+      "learning_rate": 6.297619047619048e-06,
+      "loss": 49.3823,
+      "step": 320
+    },
+    {
+      "epoch": 7.6447761194029855,
+      "grad_norm": 34.93647766113281,
+      "learning_rate": 6.285714285714286e-06,
+      "loss": 48.8867,
+      "step": 321
+    },
+    {
+      "epoch": 7.66865671641791,
+      "grad_norm": 30.189655303955078,
+      "learning_rate": 6.2738095238095245e-06,
+      "loss": 49.8644,
+      "step": 322
+    },
+    {
+      "epoch": 7.692537313432836,
+      "grad_norm": 19.964523315429688,
+      "learning_rate": 6.261904761904762e-06,
+      "loss": 49.6489,
+      "step": 323
+    },
+    {
+      "epoch": 7.7164179104477615,
+      "grad_norm": 22.253337860107422,
+      "learning_rate": 6.25e-06,
+      "loss": 48.0582,
+      "step": 324
+    },
+    {
+      "epoch": 7.740298507462686,
+      "grad_norm": 26.631391525268555,
+      "learning_rate": 6.238095238095239e-06,
+      "loss": 48.5585,
+      "step": 325
+    },
+    {
+      "epoch": 7.764179104477612,
+      "grad_norm": 26.0469913482666,
+      "learning_rate": 6.226190476190477e-06,
+      "loss": 49.4969,
+      "step": 326
+    },
+    {
+      "epoch": 7.7880597014925375,
+      "grad_norm": 30.000507354736328,
+      "learning_rate": 6.214285714285715e-06,
+      "loss": 49.6044,
+      "step": 327
+    },
+    {
+      "epoch": 7.811940298507462,
+      "grad_norm": 29.44800567626953,
+      "learning_rate": 6.202380952380953e-06,
+      "loss": 50.3622,
+      "step": 328
+    },
+    {
+      "epoch": 7.835820895522388,
+      "grad_norm": 24.83717918395996,
+      "learning_rate": 6.1904761904761914e-06,
+      "loss": 50.0974,
+      "step": 329
+    },
+    {
+      "epoch": 7.859701492537313,
+      "grad_norm": 30.0760555267334,
+      "learning_rate": 6.178571428571429e-06,
+      "loss": 48.9307,
+      "step": 330
+    },
+    {
+      "epoch": 7.883582089552239,
+      "grad_norm": 21.087966918945312,
+      "learning_rate": 6.166666666666667e-06,
+      "loss": 49.3432,
+      "step": 331
+    },
+    {
+      "epoch": 7.907462686567165,
+      "grad_norm": 23.193716049194336,
+      "learning_rate": 6.154761904761906e-06,
+      "loss": 48.6664,
+      "step": 332
+    },
+    {
+      "epoch": 7.931343283582089,
+      "grad_norm": 22.764123916625977,
+      "learning_rate": 6.142857142857144e-06,
+      "loss": 49.3497,
+      "step": 333
+    },
+    {
+      "epoch": 7.955223880597015,
+      "grad_norm": 22.411897659301758,
+      "learning_rate": 6.130952380952382e-06,
+      "loss": 49.4106,
+      "step": 334
+    },
+    {
+      "epoch": 7.979104477611941,
+      "grad_norm": 29.535375595092773,
+      "learning_rate": 6.11904761904762e-06,
+      "loss": 49.0695,
+      "step": 335
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 21.094457626342773,
+      "learning_rate": 6.107142857142858e-06,
+      "loss": 42.1367,
+      "step": 336
+    },
+    {
+      "epoch": 8.023880597014925,
+      "grad_norm": 33.74859619140625,
+      "learning_rate": 6.095238095238096e-06,
+      "loss": 47.0065,
+      "step": 337
+    },
+    {
+      "epoch": 8.047761194029851,
+      "grad_norm": 32.539127349853516,
+      "learning_rate": 6.083333333333333e-06,
+      "loss": 47.9697,
+      "step": 338
+    },
+    {
+      "epoch": 8.071641791044776,
+      "grad_norm": 19.168655395507812,
+      "learning_rate": 6.071428571428571e-06,
+      "loss": 49.4919,
+      "step": 339
+    },
+    {
+      "epoch": 8.0955223880597,
+      "grad_norm": 30.041269302368164,
+      "learning_rate": 6.05952380952381e-06,
+      "loss": 48.7887,
+      "step": 340
+    },
+    {
+      "epoch": 8.119402985074627,
+      "grad_norm": 21.070598602294922,
+      "learning_rate": 6.047619047619048e-06,
+      "loss": 48.5064,
+      "step": 341
+    },
+    {
+      "epoch": 8.143283582089552,
+      "grad_norm": 29.560287475585938,
+      "learning_rate": 6.035714285714286e-06,
+      "loss": 47.7472,
+      "step": 342
+    },
+    {
+      "epoch": 8.167164179104478,
+      "grad_norm": 24.256393432617188,
+      "learning_rate": 6.023809523809524e-06,
+      "loss": 48.8917,
+      "step": 343
+    },
+    {
+      "epoch": 8.191044776119403,
+      "grad_norm": 29.970674514770508,
+      "learning_rate": 6.011904761904762e-06,
+      "loss": 48.3464,
+      "step": 344
+    },
+    {
+      "epoch": 8.214925373134328,
+      "grad_norm": 25.274595260620117,
+      "learning_rate": 6e-06,
+      "loss": 49.1565,
+      "step": 345
+    },
+    {
+      "epoch": 8.238805970149254,
+      "grad_norm": 24.000280380249023,
+      "learning_rate": 5.988095238095238e-06,
+      "loss": 49.3396,
+      "step": 346
+    },
+    {
+      "epoch": 8.26268656716418,
+      "grad_norm": 25.110261917114258,
+      "learning_rate": 5.976190476190477e-06,
+      "loss": 49.786,
+      "step": 347
+    },
+    {
+      "epoch": 8.286567164179104,
+      "grad_norm": 26.188514709472656,
+      "learning_rate": 5.964285714285715e-06,
+      "loss": 50.1652,
+      "step": 348
+    },
+    {
+      "epoch": 8.31044776119403,
+      "grad_norm": 18.536714553833008,
+      "learning_rate": 5.9523809523809525e-06,
+      "loss": 49.7224,
+      "step": 349
+    },
+    {
+      "epoch": 8.334328358208955,
+      "grad_norm": 33.79502868652344,
+      "learning_rate": 5.940476190476191e-06,
+      "loss": 48.2923,
+      "step": 350
+    },
+    {
+      "epoch": 8.35820895522388,
+      "grad_norm": 33.03609085083008,
+      "learning_rate": 5.928571428571429e-06,
+      "loss": 49.128,
+      "step": 351
+    },
+    {
+      "epoch": 8.382089552238806,
+      "grad_norm": 23.88555145263672,
+      "learning_rate": 5.916666666666667e-06,
+      "loss": 49.6072,
+      "step": 352
+    },
+    {
+      "epoch": 8.405970149253731,
+      "grad_norm": 29.688135147094727,
+      "learning_rate": 5.904761904761905e-06,
+      "loss": 49.0984,
+      "step": 353
+    },
+    {
+      "epoch": 8.429850746268656,
+      "grad_norm": 23.166162490844727,
+      "learning_rate": 5.892857142857144e-06,
+      "loss": 48.8104,
+      "step": 354
+    },
+    {
+      "epoch": 8.453731343283582,
+      "grad_norm": 27.68876838684082,
+      "learning_rate": 5.8809523809523816e-06,
+      "loss": 48.7745,
+      "step": 355
+    },
+    {
+      "epoch": 8.477611940298507,
+      "grad_norm": 26.520286560058594,
+      "learning_rate": 5.8690476190476194e-06,
+      "loss": 47.883,
+      "step": 356
+    },
+    {
+      "epoch": 8.501492537313434,
+      "grad_norm": 28.830135345458984,
+      "learning_rate": 5.857142857142858e-06,
+      "loss": 49.1347,
+      "step": 357
+    },
+    {
+      "epoch": 8.525373134328358,
+      "grad_norm": 27.387250900268555,
+      "learning_rate": 5.845238095238096e-06,
+      "loss": 48.2092,
+      "step": 358
+    },
+    {
+      "epoch": 8.549253731343283,
+      "grad_norm": 23.53616714477539,
+      "learning_rate": 5.833333333333334e-06,
+      "loss": 48.437,
+      "step": 359
+    },
+    {
+      "epoch": 8.57313432835821,
+      "grad_norm": 25.665664672851562,
+      "learning_rate": 5.821428571428573e-06,
+      "loss": 49.3006,
+      "step": 360
+    },
+    {
+      "epoch": 8.597014925373134,
+      "grad_norm": 24.35331153869629,
+      "learning_rate": 5.8095238095238106e-06,
+      "loss": 49.5249,
+      "step": 361
+    },
+    {
+      "epoch": 8.620895522388059,
+      "grad_norm": 28.612688064575195,
+      "learning_rate": 5.7976190476190485e-06,
+      "loss": 50.1344,
+      "step": 362
+    },
+    {
+      "epoch": 8.644776119402986,
+      "grad_norm": 25.055545806884766,
+      "learning_rate": 5.785714285714286e-06,
+      "loss": 48.6014,
+      "step": 363
+    },
+    {
+      "epoch": 8.66865671641791,
+      "grad_norm": 27.645490646362305,
+      "learning_rate": 5.773809523809523e-06,
+      "loss": 48.953,
+      "step": 364
+    },
+    {
+      "epoch": 8.692537313432837,
+      "grad_norm": 26.791471481323242,
+      "learning_rate": 5.761904761904762e-06,
+      "loss": 49.5912,
+      "step": 365
+    },
+    {
+      "epoch": 8.716417910447761,
+      "grad_norm": 27.57213592529297,
+      "learning_rate": 5.75e-06,
+      "loss": 48.9958,
+      "step": 366
+    },
+    {
+      "epoch": 8.740298507462686,
+      "grad_norm": 20.936344146728516,
+      "learning_rate": 5.738095238095238e-06,
+      "loss": 48.3449,
+      "step": 367
+    },
+    {
+      "epoch": 8.764179104477613,
+      "grad_norm": 31.695810317993164,
+      "learning_rate": 5.726190476190477e-06,
+      "loss": 49.1015,
+      "step": 368
+    },
+    {
+      "epoch": 8.788059701492537,
+      "grad_norm": 31.584064483642578,
+      "learning_rate": 5.7142857142857145e-06,
+      "loss": 48.8249,
+      "step": 369
+    },
+    {
+      "epoch": 8.811940298507462,
+      "grad_norm": 30.70412826538086,
+      "learning_rate": 5.702380952380952e-06,
+      "loss": 49.2984,
+      "step": 370
+    },
+    {
+      "epoch": 8.835820895522389,
+      "grad_norm": 36.31315231323242,
+      "learning_rate": 5.690476190476191e-06,
+      "loss": 48.6769,
+      "step": 371
+    },
+    {
+      "epoch": 8.859701492537313,
+      "grad_norm": 28.98838996887207,
+      "learning_rate": 5.678571428571429e-06,
+      "loss": 50.2101,
+      "step": 372
+    },
+    {
+      "epoch": 8.883582089552238,
+      "grad_norm": 29.07052230834961,
+      "learning_rate": 5.666666666666667e-06,
+      "loss": 49.9206,
+      "step": 373
+    },
+    {
+      "epoch": 8.907462686567165,
+      "grad_norm": 31.653087615966797,
+      "learning_rate": 5.654761904761905e-06,
+      "loss": 48.3035,
+      "step": 374
+    },
+    {
+      "epoch": 8.93134328358209,
+      "grad_norm": 27.019704818725586,
+      "learning_rate": 5.6428571428571435e-06,
+      "loss": 48.6833,
+      "step": 375
+    },
+    {
+      "epoch": 8.955223880597014,
+      "grad_norm": 30.919578552246094,
+      "learning_rate": 5.630952380952381e-06,
+      "loss": 47.7973,
+      "step": 376
+    },
+    {
+      "epoch": 8.97910447761194,
+      "grad_norm": 28.002975463867188,
+      "learning_rate": 5.619047619047619e-06,
+      "loss": 49.5539,
+      "step": 377
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 27.587263107299805,
+      "learning_rate": 5.607142857142858e-06,
+      "loss": 42.9343,
+      "step": 378
+    },
+    {
+      "epoch": 9.023880597014925,
+      "grad_norm": 31.024024963378906,
+      "learning_rate": 5.595238095238096e-06,
+      "loss": 48.6774,
+      "step": 379
+    },
+    {
+      "epoch": 9.047761194029851,
+      "grad_norm": 27.262426376342773,
+      "learning_rate": 5.583333333333334e-06,
+      "loss": 47.8833,
+      "step": 380
+    },
+    {
+      "epoch": 9.071641791044776,
+      "grad_norm": 29.223133087158203,
+      "learning_rate": 5.571428571428572e-06,
+      "loss": 47.1563,
+      "step": 381
+    },
+    {
+      "epoch": 9.0955223880597,
+      "grad_norm": 21.004749298095703,
+      "learning_rate": 5.5595238095238104e-06,
+      "loss": 49.0407,
+      "step": 382
+    },
+    {
+      "epoch": 9.119402985074627,
+      "grad_norm": 25.157907485961914,
+      "learning_rate": 5.547619047619048e-06,
+      "loss": 48.5233,
+      "step": 383
+    },
+    {
+      "epoch": 9.143283582089552,
+      "grad_norm": 17.611478805541992,
+      "learning_rate": 5.535714285714286e-06,
+      "loss": 47.4846,
+      "step": 384
+    },
+    {
+      "epoch": 9.167164179104478,
+      "grad_norm": 21.00395965576172,
+      "learning_rate": 5.523809523809525e-06,
+      "loss": 49.7533,
+      "step": 385
+    },
+    {
+      "epoch": 9.191044776119403,
+      "grad_norm": 22.07697296142578,
+      "learning_rate": 5.511904761904763e-06,
+      "loss": 48.5003,
+      "step": 386
+    },
+    {
+      "epoch": 9.214925373134328,
+      "grad_norm": 21.743778228759766,
+      "learning_rate": 5.500000000000001e-06,
+      "loss": 48.149,
+      "step": 387
+    },
+    {
+      "epoch": 9.238805970149254,
+      "grad_norm": 23.499980926513672,
+      "learning_rate": 5.4880952380952394e-06,
+      "loss": 48.2213,
+      "step": 388
+    },
+    {
+      "epoch": 9.26268656716418,
+      "grad_norm": 22.22580337524414,
+      "learning_rate": 5.476190476190477e-06,
+      "loss": 48.4671,
+      "step": 389
+    },
+    {
+      "epoch": 9.286567164179104,
+      "grad_norm": 26.5915470123291,
+      "learning_rate": 5.464285714285714e-06,
+      "loss": 49.2343,
+      "step": 390
+    },
+    {
+      "epoch": 9.31044776119403,
+      "grad_norm": 22.510892868041992,
+      "learning_rate": 5.452380952380952e-06,
+      "loss": 48.9363,
+      "step": 391
+    },
+    {
+      "epoch": 9.334328358208955,
+      "grad_norm": 27.17405128479004,
+      "learning_rate": 5.44047619047619e-06,
+      "loss": 49.1814,
+      "step": 392
+    },
+    {
+      "epoch": 9.35820895522388,
+      "grad_norm": 29.143529891967773,
+      "learning_rate": 5.428571428571429e-06,
+      "loss": 48.4786,
+      "step": 393
+    },
+    {
+      "epoch": 9.382089552238806,
+      "grad_norm": 20.24784278869629,
+      "learning_rate": 5.416666666666667e-06,
+      "loss": 49.2987,
+      "step": 394
+    },
+    {
+      "epoch": 9.405970149253731,
+      "grad_norm": 31.44426155090332,
+      "learning_rate": 5.404761904761905e-06,
+      "loss": 49.9466,
+      "step": 395
+    },
+    {
+      "epoch": 9.429850746268656,
+      "grad_norm": 23.775951385498047,
+      "learning_rate": 5.392857142857143e-06,
+      "loss": 49.1681,
+      "step": 396
+    },
+    {
+      "epoch": 9.453731343283582,
+      "grad_norm": 22.168636322021484,
+      "learning_rate": 5.380952380952381e-06,
+      "loss": 48.8523,
+      "step": 397
+    },
+    {
+      "epoch": 9.477611940298507,
+      "grad_norm": 20.944936752319336,
+      "learning_rate": 5.369047619047619e-06,
+      "loss": 48.7369,
+      "step": 398
+    },
+    {
+      "epoch": 9.501492537313434,
+      "grad_norm": 23.880292892456055,
+      "learning_rate": 5.357142857142857e-06,
+      "loss": 48.4703,
+      "step": 399
+    },
+    {
+      "epoch": 9.525373134328358,
+      "grad_norm": 25.316978454589844,
+      "learning_rate": 5.345238095238096e-06,
+      "loss": 48.3752,
+      "step": 400
+    },
+    {
+      "epoch": 9.549253731343283,
+      "grad_norm": 24.398311614990234,
+      "learning_rate": 5.333333333333334e-06,
+      "loss": 47.532,
+      "step": 401
+    },
+    {
+      "epoch": 9.57313432835821,
+      "grad_norm": 23.157140731811523,
+      "learning_rate": 5.3214285714285715e-06,
+      "loss": 49.1824,
+      "step": 402
+    },
+    {
+      "epoch": 9.597014925373134,
+      "grad_norm": 21.641061782836914,
+      "learning_rate": 5.30952380952381e-06,
+      "loss": 49.6601,
+      "step": 403
+    },
+    {
+      "epoch": 9.620895522388059,
+      "grad_norm": 23.863712310791016,
+      "learning_rate": 5.297619047619048e-06,
+      "loss": 49.2146,
+      "step": 404
+    },
+    {
+      "epoch": 9.644776119402986,
+      "grad_norm": 21.876007080078125,
+      "learning_rate": 5.285714285714286e-06,
+      "loss": 48.0027,
+      "step": 405
+    },
+    {
+      "epoch": 9.66865671641791,
+      "grad_norm": 25.783042907714844,
+      "learning_rate": 5.273809523809525e-06,
+      "loss": 48.2702,
+      "step": 406
+    },
+    {
+      "epoch": 9.692537313432837,
+      "grad_norm": 18.782087326049805,
+      "learning_rate": 5.261904761904763e-06,
+      "loss": 48.9365,
+      "step": 407
+    },
+    {
+      "epoch": 9.716417910447761,
+      "grad_norm": 20.206588745117188,
+      "learning_rate": 5.2500000000000006e-06,
+      "loss": 49.4144,
+      "step": 408
+    },
+    {
+      "epoch": 9.740298507462686,
+      "grad_norm": 20.98710823059082,
+      "learning_rate": 5.2380952380952384e-06,
+      "loss": 49.7442,
+      "step": 409
+    },
+    {
+      "epoch": 9.764179104477613,
+      "grad_norm": 19.24452018737793,
+      "learning_rate": 5.226190476190477e-06,
+      "loss": 49.0249,
+      "step": 410
+    },
+    {
+      "epoch": 9.788059701492537,
+      "grad_norm": 23.18075180053711,
+      "learning_rate": 5.214285714285715e-06,
+      "loss": 48.8795,
+      "step": 411
+    },
+    {
+      "epoch": 9.811940298507462,
+      "grad_norm": 17.233261108398438,
+      "learning_rate": 5.202380952380953e-06,
+      "loss": 49.2985,
+      "step": 412
+    },
+    {
+      "epoch": 9.835820895522389,
+      "grad_norm": 24.74007797241211,
+      "learning_rate": 5.190476190476192e-06,
+      "loss": 48.8793,
+      "step": 413
+    },
+    {
+      "epoch": 9.859701492537313,
+      "grad_norm": 20.26863670349121,
+      "learning_rate": 5.1785714285714296e-06,
+      "loss": 49.6989,
+      "step": 414
+    },
+    {
+      "epoch": 9.883582089552238,
+      "grad_norm": 26.168167114257812,
+      "learning_rate": 5.1666666666666675e-06,
+      "loss": 48.7413,
+      "step": 415
+    },
+    {
+      "epoch": 9.907462686567165,
+      "grad_norm": 29.008501052856445,
+      "learning_rate": 5.1547619047619045e-06,
+      "loss": 48.7414,
+      "step": 416
+    },
+    {
+      "epoch": 9.93134328358209,
+      "grad_norm": 18.459829330444336,
+      "learning_rate": 5.142857142857142e-06,
+      "loss": 47.7865,
+      "step": 417
+    },
+    {
+      "epoch": 9.955223880597014,
+      "grad_norm": 20.898181915283203,
+      "learning_rate": 5.130952380952381e-06,
+      "loss": 47.7274,
+      "step": 418
+    },
+    {
+      "epoch": 9.97910447761194,
+      "grad_norm": 23.5065860748291,
+      "learning_rate": 5.119047619047619e-06,
+      "loss": 48.471,
+      "step": 419
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 23.147043228149414,
+      "learning_rate": 5.107142857142857e-06,
+      "loss": 42.3971,
+      "step": 420
+    },
+    {
+      "epoch": 10.023880597014925,
+      "grad_norm": 28.423707962036133,
+      "learning_rate": 5.095238095238096e-06,
+      "loss": 49.4977,
+      "step": 421
+    },
+    {
+      "epoch": 10.047761194029851,
+      "grad_norm": 22.017820358276367,
+      "learning_rate": 5.0833333333333335e-06,
+      "loss": 47.0638,
+      "step": 422
+    },
+    {
+      "epoch": 10.071641791044776,
+      "grad_norm": 18.173845291137695,
+      "learning_rate": 5.071428571428571e-06,
+      "loss": 48.338,
+      "step": 423
+    },
+    {
+      "epoch": 10.0955223880597,
+      "grad_norm": 17.628551483154297,
+      "learning_rate": 5.05952380952381e-06,
+      "loss": 48.2847,
+      "step": 424
+    },
+    {
+      "epoch": 10.119402985074627,
+      "grad_norm": 19.974040985107422,
+      "learning_rate": 5.047619047619048e-06,
+      "loss": 49.2284,
+      "step": 425
+    },
+    {
+      "epoch": 10.143283582089552,
+      "grad_norm": 22.45549774169922,
+      "learning_rate": 5.035714285714286e-06,
+      "loss": 49.6345,
+      "step": 426
+    },
+    {
+      "epoch": 10.167164179104478,
+      "grad_norm": 21.609479904174805,
+      "learning_rate": 5.023809523809524e-06,
+      "loss": 48.2098,
+      "step": 427
+    },
+    {
+      "epoch": 10.191044776119403,
+      "grad_norm": 24.7137451171875,
+      "learning_rate": 5.0119047619047625e-06,
+      "loss": 47.9527,
+      "step": 428
+    },
+    {
+      "epoch": 10.214925373134328,
+      "grad_norm": 22.888975143432617,
+      "learning_rate": 5e-06,
+      "loss": 49.781,
+      "step": 429
+    },
+    {
+      "epoch": 10.238805970149254,
+      "grad_norm": 25.53217124938965,
+      "learning_rate": 4.988095238095238e-06,
+      "loss": 48.9902,
+      "step": 430
+    },
+    {
+      "epoch": 10.26268656716418,
+      "grad_norm": 27.80384063720703,
+      "learning_rate": 4.976190476190477e-06,
+      "loss": 48.2545,
+      "step": 431
+    },
+    {
+      "epoch": 10.286567164179104,
+      "grad_norm": 21.421342849731445,
+      "learning_rate": 4.964285714285715e-06,
+      "loss": 49.1483,
+      "step": 432
+    },
+    {
+      "epoch": 10.31044776119403,
+      "grad_norm": 26.178152084350586,
+      "learning_rate": 4.952380952380953e-06,
+      "loss": 49.1129,
+      "step": 433
+    },
+    {
+      "epoch": 10.334328358208955,
+      "grad_norm": 27.993371963500977,
+      "learning_rate": 4.940476190476191e-06,
+      "loss": 48.1783,
+      "step": 434
+    },
+    {
+      "epoch": 10.35820895522388,
+      "grad_norm": 26.75821876525879,
+      "learning_rate": 4.928571428571429e-06,
+      "loss": 48.1773,
+      "step": 435
+    },
+    {
+      "epoch": 10.382089552238806,
+      "grad_norm": 25.641353607177734,
+      "learning_rate": 4.9166666666666665e-06,
+      "loss": 48.9295,
+      "step": 436
+    },
+    {
+      "epoch": 10.405970149253731,
+      "grad_norm": 23.26271629333496,
+      "learning_rate": 4.904761904761905e-06,
+      "loss": 49.5486,
+      "step": 437
+    },
+    {
+      "epoch": 10.429850746268656,
+      "grad_norm": 23.637466430664062,
+      "learning_rate": 4.892857142857143e-06,
+      "loss": 48.1263,
+      "step": 438
+    },
+    {
+      "epoch": 10.453731343283582,
+      "grad_norm": 29.285432815551758,
+      "learning_rate": 4.880952380952381e-06,
+      "loss": 48.2424,
+      "step": 439
+    },
+    {
+      "epoch": 10.477611940298507,
+      "grad_norm": 29.91914939880371,
+      "learning_rate": 4.86904761904762e-06,
+      "loss": 48.3695,
+      "step": 440
+    },
+    {
+      "epoch": 10.501492537313434,
+      "grad_norm": 25.249099731445312,
+      "learning_rate": 4.857142857142858e-06,
+      "loss": 48.3644,
+      "step": 441
+    },
+    {
+      "epoch": 10.525373134328358,
+      "grad_norm": 22.37591552734375,
+      "learning_rate": 4.8452380952380955e-06,
+      "loss": 49.397,
+      "step": 442
+    },
+    {
+      "epoch": 10.549253731343283,
+      "grad_norm": 22.805437088012695,
+      "learning_rate": 4.833333333333333e-06,
+      "loss": 48.6522,
+      "step": 443
+    },
+    {
+      "epoch": 10.57313432835821,
+      "grad_norm": 21.229095458984375,
+      "learning_rate": 4.821428571428572e-06,
+      "loss": 47.6681,
+      "step": 444
+    },
+    {
+      "epoch": 10.597014925373134,
+      "grad_norm": 23.359468460083008,
+      "learning_rate": 4.80952380952381e-06,
+      "loss": 48.602,
+      "step": 445
+    },
+    {
+      "epoch": 10.620895522388059,
+      "grad_norm": 20.953310012817383,
+      "learning_rate": 4.797619047619048e-06,
+      "loss": 49.3366,
+      "step": 446
+    },
+    {
+      "epoch": 10.644776119402986,
+      "grad_norm": 21.970388412475586,
+      "learning_rate": 4.785714285714287e-06,
+      "loss": 46.964,
+      "step": 447
+    },
+    {
+      "epoch": 10.66865671641791,
+      "grad_norm": 24.282426834106445,
+      "learning_rate": 4.7738095238095245e-06,
+      "loss": 48.2676,
+      "step": 448
+    },
+    {
+      "epoch": 10.692537313432837,
+      "grad_norm": 15.47967529296875,
+      "learning_rate": 4.761904761904762e-06,
+      "loss": 48.1993,
+      "step": 449
+    },
+    {
+      "epoch": 10.716417910447761,
+      "grad_norm": 23.230947494506836,
+      "learning_rate": 4.75e-06,
+      "loss": 48.5229,
+      "step": 450
+    },
+    {
+      "epoch": 10.740298507462686,
+      "grad_norm": 20.514225006103516,
+      "learning_rate": 4.738095238095238e-06,
+      "loss": 48.062,
+      "step": 451
+    },
+    {
+      "epoch": 10.764179104477613,
+      "grad_norm": 19.060667037963867,
+      "learning_rate": 4.726190476190476e-06,
+      "loss": 48.3893,
+      "step": 452
+    },
+    {
+      "epoch": 10.788059701492537,
+      "grad_norm": 29.78558349609375,
+      "learning_rate": 4.714285714285715e-06,
+      "loss": 48.9921,
+      "step": 453
+    },
+    {
+      "epoch": 10.811940298507462,
+      "grad_norm": 23.262001037597656,
+      "learning_rate": 4.702380952380953e-06,
+      "loss": 48.5597,
+      "step": 454
+    },
+    {
+      "epoch": 10.835820895522389,
+      "grad_norm": 25.83403778076172,
+      "learning_rate": 4.6904761904761905e-06,
+      "loss": 49.2911,
+      "step": 455
+    },
+    {
+      "epoch": 10.859701492537313,
+      "grad_norm": 21.846391677856445,
+      "learning_rate": 4.678571428571429e-06,
+      "loss": 47.3256,
+      "step": 456
+    },
+    {
+      "epoch": 10.883582089552238,
+      "grad_norm": 17.09532356262207,
+      "learning_rate": 4.666666666666667e-06,
+      "loss": 48.3647,
+      "step": 457
+    },
+    {
+      "epoch": 10.907462686567165,
+      "grad_norm": 31.050525665283203,
+      "learning_rate": 4.654761904761905e-06,
+      "loss": 48.3605,
+      "step": 458
+    },
+    {
+      "epoch": 10.93134328358209,
+      "grad_norm": 22.532379150390625,
+      "learning_rate": 4.642857142857144e-06,
+      "loss": 49.0826,
+      "step": 459
+    },
+    {
+      "epoch": 10.955223880597014,
+      "grad_norm": 23.585033416748047,
+      "learning_rate": 4.630952380952382e-06,
+      "loss": 48.5111,
+      "step": 460
+    },
+    {
+      "epoch": 10.97910447761194,
+      "grad_norm": NaN,
+      "learning_rate": 4.6190476190476196e-06,
+      "loss": 66.9717,
+      "step": 461
+    },
+    {
+      "epoch": 11.0,
+      "grad_norm": 24.73590087890625,
+      "learning_rate": 4.6190476190476196e-06,
+      "loss": 41.9122,
+      "step": 462
+    },
+    {
+      "epoch": 11.023880597014925,
+      "grad_norm": 27.4709415435791,
+      "learning_rate": 4.6071428571428574e-06,
+      "loss": 48.4682,
+      "step": 463
+    },
+    {
+      "epoch": 11.047761194029851,
+      "grad_norm": 26.158245086669922,
+      "learning_rate": 4.595238095238095e-06,
+      "loss": 48.1845,
+      "step": 464
+    },
+    {
+      "epoch": 11.071641791044776,
+      "grad_norm": 25.14693260192871,
+      "learning_rate": 4.583333333333333e-06,
+      "loss": 48.4229,
+      "step": 465
+    },
+    {
+      "epoch": 11.0955223880597,
+      "grad_norm": 22.229764938354492,
+      "learning_rate": 4.571428571428572e-06,
+      "loss": 47.8876,
+      "step": 466
+    },
+    {
+      "epoch": 11.119402985074627,
+      "grad_norm": 24.202686309814453,
+      "learning_rate": 4.55952380952381e-06,
+      "loss": 48.4304,
+      "step": 467
+    },
+    {
+      "epoch": 11.143283582089552,
+      "grad_norm": 21.449726104736328,
+      "learning_rate": 4.547619047619048e-06,
+      "loss": 47.6457,
+      "step": 468
+    },
+    {
+      "epoch": 11.167164179104478,
+      "grad_norm": 23.769763946533203,
+      "learning_rate": 4.5357142857142865e-06,
+      "loss": 49.1031,
+      "step": 469
+    },
+    {
+      "epoch": 11.191044776119403,
+      "grad_norm": 21.20684814453125,
+      "learning_rate": 4.523809523809524e-06,
+      "loss": 47.6488,
+      "step": 470
+    },
+    {
+      "epoch": 11.214925373134328,
+      "grad_norm": 17.992631912231445,
+      "learning_rate": 4.511904761904762e-06,
+      "loss": 47.9435,
+      "step": 471
+    },
+    {
+      "epoch": 11.238805970149254,
+      "grad_norm": 22.017776489257812,
+      "learning_rate": 4.5e-06,
+      "loss": 48.5224,
+      "step": 472
+    },
+    {
+      "epoch": 11.26268656716418,
+      "grad_norm": 22.98673439025879,
+      "learning_rate": 4.488095238095239e-06,
+      "loss": 47.9258,
+      "step": 473
+    },
+    {
+      "epoch": 11.286567164179104,
+      "grad_norm": 16.146743774414062,
+      "learning_rate": 4.476190476190477e-06,
+      "loss": 48.3957,
+      "step": 474
+    },
+    {
+      "epoch": 11.31044776119403,
+      "grad_norm": 23.30071258544922,
+      "learning_rate": 4.464285714285715e-06,
+      "loss": 48.5472,
+      "step": 475
+    },
+    {
+      "epoch": 11.334328358208955,
+      "grad_norm": 24.949913024902344,
+      "learning_rate": 4.4523809523809525e-06,
+      "loss": 48.2387,
+      "step": 476
+    },
+    {
+      "epoch": 11.35820895522388,
+      "grad_norm": 23.10662841796875,
+      "learning_rate": 4.44047619047619e-06,
+      "loss": 49.0681,
+      "step": 477
+    },
+    {
+      "epoch": 11.382089552238806,
+      "grad_norm": 19.024614334106445,
+      "learning_rate": 4.428571428571429e-06,
+      "loss": 49.3255,
+      "step": 478
+    },
+    {
+      "epoch": 11.405970149253731,
+      "grad_norm": 22.34437370300293,
+      "learning_rate": 4.416666666666667e-06,
+      "loss": 47.0069,
+      "step": 479
+    },
+    {
+      "epoch": 11.429850746268656,
+      "grad_norm": 23.563596725463867,
+      "learning_rate": 4.404761904761905e-06,
+      "loss": 46.8188,
+      "step": 480
+    },
+    {
+      "epoch": 11.453731343283582,
+      "grad_norm": 20.5488338470459,
+      "learning_rate": 4.392857142857143e-06,
+      "loss": 47.8277,
+      "step": 481
+    },
+    {
+      "epoch": 11.477611940298507,
+      "grad_norm": 18.416519165039062,
+      "learning_rate": 4.3809523809523815e-06,
+      "loss": 48.2203,
+      "step": 482
+    },
+    {
+      "epoch": 11.501492537313434,
+      "grad_norm": 28.21132469177246,
+      "learning_rate": 4.369047619047619e-06,
+      "loss": 48.0691,
+      "step": 483
+    },
+    {
+      "epoch": 11.525373134328358,
+      "grad_norm": 21.36182975769043,
+      "learning_rate": 4.357142857142857e-06,
+      "loss": 48.273,
+      "step": 484
+    },
+    {
+      "epoch": 11.549253731343283,
+      "grad_norm": 25.726530075073242,
+      "learning_rate": 4.345238095238096e-06,
+      "loss": 48.7529,
+      "step": 485
+    },
+    {
+      "epoch": 11.57313432835821,
+      "grad_norm": 21.686412811279297,
+      "learning_rate": 4.333333333333334e-06,
+      "loss": 48.3005,
+      "step": 486
+    },
+    {
+      "epoch": 11.597014925373134,
+      "grad_norm": 20.56638526916504,
+      "learning_rate": 4.321428571428572e-06,
+      "loss": 50.1248,
+      "step": 487
+    },
+    {
+      "epoch": 11.620895522388059,
+      "grad_norm": 24.193323135375977,
+      "learning_rate": 4.30952380952381e-06,
+      "loss": 48.6031,
+      "step": 488
+    },
+    {
+      "epoch": 11.644776119402986,
+      "grad_norm": 17.18548583984375,
+      "learning_rate": 4.297619047619048e-06,
+      "loss": 49.2039,
+      "step": 489
+    },
+    {
+      "epoch": 11.66865671641791,
+      "grad_norm": 19.07050895690918,
+      "learning_rate": 4.2857142857142855e-06,
+      "loss": 48.0961,
+      "step": 490
+    },
+    {
+      "epoch": 11.692537313432837,
+      "grad_norm": 19.831188201904297,
+      "learning_rate": 4.273809523809524e-06,
+      "loss": 48.5481,
+      "step": 491
+    },
+    {
+      "epoch": 11.716417910447761,
+      "grad_norm": 23.408592224121094,
+      "learning_rate": 4.261904761904762e-06,
+      "loss": 48.583,
+      "step": 492
+    },
+    {
+      "epoch": 11.740298507462686,
+      "grad_norm": 22.152788162231445,
+      "learning_rate": 4.25e-06,
+      "loss": 48.6684,
+      "step": 493
+    },
+    {
+      "epoch": 11.764179104477613,
+      "grad_norm": NaN,
+      "learning_rate": 4.238095238095239e-06,
+      "loss": 54.7097,
+      "step": 494
+    },
+    {
+      "epoch": 11.788059701492537,
+      "grad_norm": 23.1225528717041,
+      "learning_rate": 4.238095238095239e-06,
+      "loss": 48.5439,
+      "step": 495
+    },
+    {
+      "epoch": 11.811940298507462,
+      "grad_norm": 24.673904418945312,
+      "learning_rate": 4.226190476190477e-06,
+      "loss": 48.2645,
+      "step": 496
+    },
+    {
+      "epoch": 11.835820895522389,
+      "grad_norm": 23.318784713745117,
+      "learning_rate": 4.2142857142857145e-06,
+      "loss": 47.9159,
+      "step": 497
+    },
+    {
+      "epoch": 11.859701492537313,
+      "grad_norm": 24.62889289855957,
+      "learning_rate": 4.202380952380952e-06,
+      "loss": 48.1392,
+      "step": 498
+    },
+    {
+      "epoch": 11.883582089552238,
+      "grad_norm": 17.315168380737305,
+      "learning_rate": 4.190476190476191e-06,
+      "loss": 49.399,
+      "step": 499
+    },
+    {
+      "epoch": 11.907462686567165,
+      "grad_norm": 24.458532333374023,
+      "learning_rate": 4.178571428571429e-06,
+      "loss": 49.2189,
+      "step": 500
+    },
+    {
+      "epoch": 11.93134328358209,
+      "grad_norm": 28.294036865234375,
+      "learning_rate": 4.166666666666667e-06,
+      "loss": 48.4759,
+      "step": 501
+    },
+    {
+      "epoch": 11.955223880597014,
+      "grad_norm": 22.393577575683594,
+      "learning_rate": 4.154761904761906e-06,
+      "loss": 48.9718,
+      "step": 502
+    },
+    {
+      "epoch": 11.97910447761194,
+      "grad_norm": 20.199522018432617,
+      "learning_rate": 4.1428571428571435e-06,
+      "loss": 47.5364,
+      "step": 503
+    },
+    {
+      "epoch": 12.0,
+      "grad_norm": 22.080204010009766,
+      "learning_rate": 4.130952380952381e-06,
+      "loss": 42.5308,
+      "step": 504
+    },
+    {
+      "epoch": 12.023880597014925,
+      "grad_norm": 28.897024154663086,
+      "learning_rate": 4.119047619047619e-06,
+      "loss": 48.9022,
+      "step": 505
+    },
+    {
+      "epoch": 12.047761194029851,
+      "grad_norm": 28.31342887878418,
+      "learning_rate": 4.107142857142857e-06,
+      "loss": 47.6489,
+      "step": 506
+    },
+    {
+      "epoch": 12.071641791044776,
+      "grad_norm": 22.62079620361328,
+      "learning_rate": 4.095238095238096e-06,
+      "loss": 48.1606,
+      "step": 507
+    },
+    {
+      "epoch": 12.0955223880597,
+      "grad_norm": 33.49858474731445,
+      "learning_rate": 4.083333333333334e-06,
+      "loss": 47.8462,
+      "step": 508
+    },
+    {
+      "epoch": 12.119402985074627,
+      "grad_norm": 22.20858383178711,
+      "learning_rate": 4.071428571428572e-06,
+      "loss": 47.2505,
+      "step": 509
+    },
+    {
+      "epoch": 12.143283582089552,
+      "grad_norm": 25.425495147705078,
+      "learning_rate": 4.0595238095238095e-06,
+      "loss": 48.6289,
+      "step": 510
+    },
+    {
+      "epoch": 12.167164179104478,
+      "grad_norm": 29.32784652709961,
+      "learning_rate": 4.047619047619048e-06,
+      "loss": 47.7772,
+      "step": 511
+    },
+    {
+      "epoch": 12.191044776119403,
+      "grad_norm": 20.661781311035156,
+      "learning_rate": 4.035714285714286e-06,
+      "loss": 47.1414,
+      "step": 512
+    },
+    {
+      "epoch": 12.214925373134328,
+      "grad_norm": 31.4210205078125,
+      "learning_rate": 4.023809523809524e-06,
+      "loss": 47.4312,
+      "step": 513
+    },
+    {
+      "epoch": 12.238805970149254,
+      "grad_norm": 32.390071868896484,
+      "learning_rate": 4.011904761904763e-06,
+      "loss": 49.9899,
+      "step": 514
+    },
+    {
+      "epoch": 12.26268656716418,
+      "grad_norm": 17.431835174560547,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 48.8975,
+      "step": 515
+    },
+    {
+      "epoch": 12.286567164179104,
+      "grad_norm": 29.32766342163086,
+      "learning_rate": 3.9880952380952386e-06,
+      "loss": 48.8764,
+      "step": 516
+    },
+    {
+      "epoch": 12.31044776119403,
+      "grad_norm": 29.523069381713867,
+      "learning_rate": 3.9761904761904764e-06,
+      "loss": 48.2602,
+      "step": 517
+    },
+    {
+      "epoch": 12.334328358208955,
+      "grad_norm": 23.866840362548828,
+      "learning_rate": 3.964285714285714e-06,
+      "loss": 47.4016,
+      "step": 518
+    },
+    {
+      "epoch": 12.35820895522388,
+      "grad_norm": 27.464962005615234,
+      "learning_rate": 3.952380952380952e-06,
+      "loss": 48.2559,
+      "step": 519
+    },
+    {
+      "epoch": 12.382089552238806,
+      "grad_norm": 19.796552658081055,
+      "learning_rate": 3.940476190476191e-06,
+      "loss": 48.7665,
+      "step": 520
+    },
+    {
+      "epoch": 12.405970149253731,
+      "grad_norm": 18.637983322143555,
+      "learning_rate": 3.928571428571429e-06,
+      "loss": 48.1456,
+      "step": 521
+    },
+    {
+      "epoch": 12.429850746268656,
+      "grad_norm": 22.065799713134766,
+      "learning_rate": 3.916666666666667e-06,
+      "loss": 48.7803,
+      "step": 522
+    },
+    {
+      "epoch": 12.453731343283582,
+      "grad_norm": 22.648218154907227,
+      "learning_rate": 3.9047619047619055e-06,
+      "loss": 47.3376,
+      "step": 523
+    },
+    {
+      "epoch": 12.477611940298507,
+      "grad_norm": 17.55946922302246,
+      "learning_rate": 3.892857142857143e-06,
+      "loss": 47.6002,
+      "step": 524
+    },
+    {
+      "epoch": 12.501492537313434,
+      "grad_norm": 19.173139572143555,
+      "learning_rate": 3.880952380952381e-06,
+      "loss": 48.8976,
+      "step": 525
+    },
+    {
+      "epoch": 12.525373134328358,
+      "grad_norm": 24.052696228027344,
+      "learning_rate": 3.869047619047619e-06,
+      "loss": 48.1851,
+      "step": 526
+    },
+    {
+      "epoch": 12.549253731343283,
+      "grad_norm": 19.28683090209961,
+      "learning_rate": 3.857142857142858e-06,
+      "loss": 48.0342,
+      "step": 527
+    },
+    {
+      "epoch": 12.57313432835821,
+      "grad_norm": 21.528470993041992,
+      "learning_rate": 3.845238095238096e-06,
+      "loss": 49.3597,
+      "step": 528
+    },
+    {
+      "epoch": 12.597014925373134,
+      "grad_norm": 22.880159378051758,
+      "learning_rate": 3.833333333333334e-06,
+      "loss": 47.9594,
+      "step": 529
+    },
+    {
+      "epoch": 12.620895522388059,
+      "grad_norm": 19.00438117980957,
+      "learning_rate": 3.8214285714285715e-06,
+      "loss": 47.2837,
+      "step": 530
+    },
+    {
+      "epoch": 12.644776119402986,
+      "grad_norm": 22.21845054626465,
+      "learning_rate": 3.80952380952381e-06,
+      "loss": 47.1453,
+      "step": 531
+    },
+    {
+      "epoch": 12.66865671641791,
+      "grad_norm": 18.551712036132812,
+      "learning_rate": 3.7976190476190477e-06,
+      "loss": 47.9594,
+      "step": 532
+    },
+    {
+      "epoch": 12.692537313432837,
+      "grad_norm": 17.805360794067383,
+      "learning_rate": 3.785714285714286e-06,
+      "loss": 49.1036,
+      "step": 533
+    },
+    {
+      "epoch": 12.716417910447761,
+      "grad_norm": 14.508918762207031,
+      "learning_rate": 3.773809523809524e-06,
+      "loss": 48.1203,
+      "step": 534
+    },
+    {
+      "epoch": 12.740298507462686,
+      "grad_norm": 19.395994186401367,
+      "learning_rate": 3.761904761904762e-06,
+      "loss": 47.7891,
+      "step": 535
+    },
+    {
+      "epoch": 12.764179104477613,
+      "grad_norm": 27.492908477783203,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 48.9027,
+      "step": 536
+    },
+    {
+      "epoch": 12.788059701492537,
+      "grad_norm": 21.751968383789062,
+      "learning_rate": 3.7380952380952384e-06,
+      "loss": 48.0929,
+      "step": 537
+    },
+    {
+      "epoch": 12.811940298507462,
+      "grad_norm": 24.78274917602539,
+      "learning_rate": 3.7261904761904767e-06,
+      "loss": 48.1678,
+      "step": 538
+    },
+    {
+      "epoch": 12.835820895522389,
+      "grad_norm": 26.319196701049805,
+      "learning_rate": 3.7142857142857146e-06,
+      "loss": 49.1874,
+      "step": 539
+    },
+    {
+      "epoch": 12.859701492537313,
+      "grad_norm": 20.670148849487305,
+      "learning_rate": 3.702380952380953e-06,
+      "loss": 48.8441,
+      "step": 540
+    },
+    {
+      "epoch": 12.883582089552238,
+      "grad_norm": 23.578706741333008,
+      "learning_rate": 3.690476190476191e-06,
+      "loss": 47.1627,
+      "step": 541
+    },
+    {
+      "epoch": 12.907462686567165,
+      "grad_norm": 23.807973861694336,
+      "learning_rate": 3.678571428571429e-06,
+      "loss": 47.493,
+      "step": 542
+    },
+    {
+      "epoch": 12.93134328358209,
+      "grad_norm": 20.977373123168945,
+      "learning_rate": 3.6666666666666666e-06,
+      "loss": 49.3489,
+      "step": 543
+    },
+    {
+      "epoch": 12.955223880597014,
+      "grad_norm": 21.219995498657227,
+      "learning_rate": 3.654761904761905e-06,
+      "loss": 49.8562,
+      "step": 544
+    },
+    {
+      "epoch": 12.97910447761194,
+      "grad_norm": 17.777210235595703,
+      "learning_rate": 3.642857142857143e-06,
+      "loss": 48.4018,
+      "step": 545
+    },
+    {
+      "epoch": 13.0,
+      "grad_norm": 17.52475929260254,
+      "learning_rate": 3.630952380952381e-06,
+      "loss": 42.3621,
+      "step": 546
+    },
+    {
+      "epoch": 13.023880597014925,
+      "grad_norm": 23.431884765625,
+      "learning_rate": 3.6190476190476194e-06,
+      "loss": 49.0982,
+      "step": 547
+    },
+    {
+      "epoch": 13.047761194029851,
+      "grad_norm": 25.512338638305664,
+      "learning_rate": 3.6071428571428573e-06,
+      "loss": 47.5758,
+      "step": 548
+    },
+    {
+      "epoch": 13.071641791044776,
+      "grad_norm": 25.41205406188965,
+      "learning_rate": 3.5952380952380956e-06,
+      "loss": 49.0519,
+      "step": 549
+    },
+    {
+      "epoch": 13.0955223880597,
+      "grad_norm": 20.511945724487305,
+      "learning_rate": 3.5833333333333335e-06,
+      "loss": 48.9739,
+      "step": 550
+    },
+    {
+      "epoch": 13.119402985074627,
+      "grad_norm": 18.88302993774414,
+      "learning_rate": 3.5714285714285718e-06,
+      "loss": 47.0551,
+      "step": 551
+    },
+    {
+      "epoch": 13.143283582089552,
+      "grad_norm": 17.176782608032227,
+      "learning_rate": 3.55952380952381e-06,
+      "loss": 48.0771,
+      "step": 552
+    },
+    {
+      "epoch": 13.167164179104478,
+      "grad_norm": 19.72154426574707,
+      "learning_rate": 3.547619047619048e-06,
+      "loss": 49.5084,
+      "step": 553
+    },
+    {
+      "epoch": 13.191044776119403,
+      "grad_norm": 24.780994415283203,
+      "learning_rate": 3.5357142857142863e-06,
+      "loss": 46.5557,
+      "step": 554
+    },
+    {
+      "epoch": 13.214925373134328,
+      "grad_norm": 20.380996704101562,
+      "learning_rate": 3.523809523809524e-06,
+      "loss": 48.841,
+      "step": 555
+    },
+    {
+      "epoch": 13.238805970149254,
+      "grad_norm": 26.90860939025879,
+      "learning_rate": 3.511904761904762e-06,
+      "loss": 47.6185,
+      "step": 556
+    },
+    {
+      "epoch": 13.26268656716418,
+      "grad_norm": 21.40388298034668,
+      "learning_rate": 3.5e-06,
+      "loss": 47.787,
+      "step": 557
+    },
+    {
+      "epoch": 13.286567164179104,
+      "grad_norm": 24.708845138549805,
+      "learning_rate": 3.4880952380952383e-06,
+      "loss": 47.1974,
+      "step": 558
+    },
+    {
+      "epoch": 13.31044776119403,
+      "grad_norm": 25.317148208618164,
+      "learning_rate": 3.476190476190476e-06,
+      "loss": 49.2282,
+      "step": 559
+    },
+    {
+      "epoch": 13.334328358208955,
+      "grad_norm": 22.903011322021484,
+      "learning_rate": 3.4642857142857145e-06,
+      "loss": 47.0762,
+      "step": 560
+    },
+    {
+      "epoch": 13.35820895522388,
+      "grad_norm": 23.626604080200195,
+      "learning_rate": 3.4523809523809528e-06,
+      "loss": 47.3622,
+      "step": 561
+    },
+    {
+      "epoch": 13.382089552238806,
+      "grad_norm": 16.69061279296875,
+      "learning_rate": 3.4404761904761907e-06,
+      "loss": 48.5621,
+      "step": 562
+    },
+    {
+      "epoch": 13.405970149253731,
+      "grad_norm": 20.52508544921875,
+      "learning_rate": 3.428571428571429e-06,
+      "loss": 47.6565,
+      "step": 563
+    },
+    {
+      "epoch": 13.429850746268656,
+      "grad_norm": 25.125743865966797,
+      "learning_rate": 3.416666666666667e-06,
+      "loss": 48.1353,
+      "step": 564
+    },
+    {
+      "epoch": 13.453731343283582,
+      "grad_norm": 20.697166442871094,
+      "learning_rate": 3.404761904761905e-06,
+      "loss": 47.9368,
+      "step": 565
+    },
+    {
+      "epoch": 13.477611940298507,
+      "grad_norm": 22.396892547607422,
+      "learning_rate": 3.3928571428571435e-06,
+      "loss": 48.2956,
+      "step": 566
+    },
+    {
+      "epoch": 13.501492537313434,
+      "grad_norm": 24.770437240600586,
+      "learning_rate": 3.3809523809523814e-06,
+      "loss": 48.4467,
+      "step": 567
+    },
+    {
+      "epoch": 13.525373134328358,
+      "grad_norm": 19.44706153869629,
+      "learning_rate": 3.3690476190476197e-06,
+      "loss": 48.3155,
+      "step": 568
+    },
+    {
+      "epoch": 13.549253731343283,
+      "grad_norm": 27.680660247802734,
+      "learning_rate": 3.357142857142857e-06,
+      "loss": 47.9039,
+      "step": 569
+    },
+    {
+      "epoch": 13.57313432835821,
+      "grad_norm": 21.89419174194336,
+      "learning_rate": 3.3452380952380954e-06,
+      "loss": 47.9416,
+      "step": 570
+    },
+    {
+      "epoch": 13.597014925373134,
+      "grad_norm": 19.10918426513672,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 48.45,
+      "step": 571
+    },
+    {
+      "epoch": 13.620895522388059,
+      "grad_norm": 29.83106231689453,
+      "learning_rate": 3.3214285714285716e-06,
+      "loss": 48.9583,
+      "step": 572
+    },
+    {
+      "epoch": 13.644776119402986,
+      "grad_norm": 28.05882453918457,
+      "learning_rate": 3.3095238095238095e-06,
+      "loss": 49.108,
+      "step": 573
+    },
+    {
+      "epoch": 13.66865671641791,
+      "grad_norm": 17.379384994506836,
+      "learning_rate": 3.297619047619048e-06,
+      "loss": 48.4707,
+      "step": 574
+    },
+    {
+      "epoch": 13.692537313432837,
+      "grad_norm": 19.15117645263672,
+      "learning_rate": 3.285714285714286e-06,
+      "loss": 48.476,
+      "step": 575
+    },
+    {
+      "epoch": 13.716417910447761,
+      "grad_norm": 23.892152786254883,
+      "learning_rate": 3.273809523809524e-06,
+      "loss": 48.0321,
+      "step": 576
+    },
+    {
+      "epoch": 13.740298507462686,
+      "grad_norm": 18.658008575439453,
+      "learning_rate": 3.2619047619047623e-06,
+      "loss": 47.2192,
+      "step": 577
+    },
+    {
+      "epoch": 13.764179104477613,
+      "grad_norm": 16.940099716186523,
+      "learning_rate": 3.2500000000000002e-06,
+      "loss": 49.2263,
+      "step": 578
+    },
+    {
+      "epoch": 13.788059701492537,
+      "grad_norm": 25.7972412109375,
+      "learning_rate": 3.2380952380952385e-06,
+      "loss": 47.5039,
+      "step": 579
+    },
+    {
+      "epoch": 13.811940298507462,
+      "grad_norm": 28.928129196166992,
+      "learning_rate": 3.226190476190477e-06,
+      "loss": 47.9264,
+      "step": 580
+    },
+    {
+      "epoch": 13.835820895522389,
+      "grad_norm": 23.67597007751465,
+      "learning_rate": 3.2142857142857147e-06,
+      "loss": 49.1464,
+      "step": 581
+    },
+    {
+      "epoch": 13.859701492537313,
+      "grad_norm": 18.345443725585938,
+      "learning_rate": 3.202380952380952e-06,
+      "loss": 47.888,
+      "step": 582
+    },
+    {
+      "epoch": 13.883582089552238,
+      "grad_norm": 19.80716896057129,
+      "learning_rate": 3.1904761904761905e-06,
+      "loss": 47.4324,
+      "step": 583
+    },
+    {
+      "epoch": 13.907462686567165,
+      "grad_norm": 20.488346099853516,
+      "learning_rate": 3.178571428571429e-06,
+      "loss": 48.3033,
+      "step": 584
+    },
+    {
+      "epoch": 13.93134328358209,
+      "grad_norm": 22.3657283782959,
+      "learning_rate": 3.1666666666666667e-06,
+      "loss": 48.1474,
+      "step": 585
+    },
+    {
+      "epoch": 13.955223880597014,
+      "grad_norm": 17.457408905029297,
+      "learning_rate": 3.154761904761905e-06,
+      "loss": 47.2418,
+      "step": 586
+    },
+    {
+      "epoch": 13.97910447761194,
+      "grad_norm": NaN,
+      "learning_rate": 3.142857142857143e-06,
+      "loss": 54.1812,
+      "step": 587
+    },
+    {
+      "epoch": 14.0,
+      "grad_norm": 17.137672424316406,
+      "learning_rate": 3.142857142857143e-06,
+      "loss": 42.3703,
+      "step": 588
+    },
+    {
+      "epoch": 14.023880597014925,
+      "grad_norm": 20.55642318725586,
+      "learning_rate": 3.130952380952381e-06,
+      "loss": 49.4628,
+      "step": 589
+    },
+    {
+      "epoch": 14.047761194029851,
+      "grad_norm": 19.925596237182617,
+      "learning_rate": 3.1190476190476195e-06,
+      "loss": 47.5266,
+      "step": 590
+    },
+    {
+      "epoch": 14.071641791044776,
+      "grad_norm": 12.49276065826416,
+      "learning_rate": 3.1071428571428574e-06,
+      "loss": 47.8654,
+      "step": 591
+    },
+    {
+      "epoch": 14.0955223880597,
+      "grad_norm": 17.266550064086914,
+      "learning_rate": 3.0952380952380957e-06,
+      "loss": 48.4362,
+      "step": 592
+    },
+    {
+      "epoch": 14.119402985074627,
+      "grad_norm": 18.234397888183594,
+      "learning_rate": 3.0833333333333336e-06,
+      "loss": 48.9532,
+      "step": 593
+    },
+    {
+      "epoch": 14.143283582089552,
+      "grad_norm": 19.880165100097656,
+      "learning_rate": 3.071428571428572e-06,
+      "loss": 48.0088,
+      "step": 594
+    },
+    {
+      "epoch": 14.167164179104478,
+      "grad_norm": 23.04216766357422,
+      "learning_rate": 3.05952380952381e-06,
+      "loss": 48.0934,
+      "step": 595
+    },
+    {
+      "epoch": 14.191044776119403,
+      "grad_norm": 19.199676513671875,
+      "learning_rate": 3.047619047619048e-06,
+      "loss": 48.3845,
+      "step": 596
+    },
+    {
+      "epoch": 14.214925373134328,
+      "grad_norm": 20.758337020874023,
+      "learning_rate": 3.0357142857142856e-06,
+      "loss": 47.4652,
+      "step": 597
+    },
+    {
+      "epoch": 14.238805970149254,
+      "grad_norm": 17.532787322998047,
+      "learning_rate": 3.023809523809524e-06,
+      "loss": 48.0212,
+      "step": 598
+    },
+    {
+      "epoch": 14.26268656716418,
+      "grad_norm": 16.547094345092773,
+      "learning_rate": 3.011904761904762e-06,
+      "loss": 48.6113,
+      "step": 599
+    },
+    {
+      "epoch": 14.286567164179104,
+      "grad_norm": 16.324464797973633,
+      "learning_rate": 3e-06,
+      "loss": 47.9735,
+      "step": 600
+    },
+    {
+      "epoch": 14.31044776119403,
+      "grad_norm": 16.54167938232422,
+      "learning_rate": 2.9880952380952384e-06,
+      "loss": 47.4436,
+      "step": 601
+    },
+    {
+      "epoch": 14.334328358208955,
+      "grad_norm": 23.455759048461914,
+      "learning_rate": 2.9761904761904763e-06,
+      "loss": 47.6631,
+      "step": 602
+    },
+    {
+      "epoch": 14.35820895522388,
+      "grad_norm": 19.159008026123047,
+      "learning_rate": 2.9642857142857146e-06,
+      "loss": 48.0291,
+      "step": 603
+    },
+    {
+      "epoch": 14.382089552238806,
+      "grad_norm": 18.66881561279297,
+      "learning_rate": 2.9523809523809525e-06,
+      "loss": 46.4582,
+      "step": 604
+    },
+    {
+      "epoch": 14.405970149253731,
+      "grad_norm": 19.129064559936523,
+      "learning_rate": 2.9404761904761908e-06,
+      "loss": 49.4455,
+      "step": 605
+    },
+    {
+      "epoch": 14.429850746268656,
+      "grad_norm": NaN,
+      "learning_rate": 2.928571428571429e-06,
+      "loss": 78.6564,
+      "step": 606
+    },
+    {
+      "epoch": 14.453731343283582,
+      "grad_norm": 18.47364044189453,
+      "learning_rate": 2.928571428571429e-06,
+      "loss": 48.1748,
+      "step": 607
+    },
+    {
+      "epoch": 14.477611940298507,
+      "grad_norm": 17.920883178710938,
+      "learning_rate": 2.916666666666667e-06,
+      "loss": 47.6447,
+      "step": 608
+    },
+    {
+      "epoch": 14.501492537313434,
+      "grad_norm": 18.263038635253906,
+      "learning_rate": 2.9047619047619053e-06,
+      "loss": 48.7324,
+      "step": 609
+    },
+    {
+      "epoch": 14.525373134328358,
+      "grad_norm": 24.323266983032227,
+      "learning_rate": 2.892857142857143e-06,
+      "loss": 48.6135,
+      "step": 610
+    },
+    {
+      "epoch": 14.549253731343283,
+      "grad_norm": 21.56492042541504,
+      "learning_rate": 2.880952380952381e-06,
+      "loss": 47.0007,
+      "step": 611
+    },
+    {
+      "epoch": 14.57313432835821,
+      "grad_norm": 17.741748809814453,
+      "learning_rate": 2.869047619047619e-06,
+      "loss": 46.3136,
+      "step": 612
+    },
+    {
+      "epoch": 14.597014925373134,
+      "grad_norm": 17.218914031982422,
+      "learning_rate": 2.8571428571428573e-06,
+      "loss": 47.6417,
+      "step": 613
+    },
+    {
+      "epoch": 14.620895522388059,
+      "grad_norm": 22.856996536254883,
+      "learning_rate": 2.8452380952380956e-06,
+      "loss": 47.6898,
+      "step": 614
+    },
+    {
+      "epoch": 14.644776119402986,
+      "grad_norm": NaN,
+      "learning_rate": 2.8333333333333335e-06,
+      "loss": 53.529,
+      "step": 615
+    },
+    {
+      "epoch": 14.66865671641791,
+      "grad_norm": 23.29751968383789,
+      "learning_rate": 2.8333333333333335e-06,
+      "loss": 48.164,
+      "step": 616
+    },
+    {
+      "epoch": 14.692537313432837,
+      "grad_norm": 15.633321762084961,
+      "learning_rate": 2.8214285714285718e-06,
+      "loss": 46.9866,
+      "step": 617
+    },
+    {
+      "epoch": 14.716417910447761,
+      "grad_norm": 21.713376998901367,
+      "learning_rate": 2.8095238095238096e-06,
+      "loss": 48.2856,
+      "step": 618
+    },
+    {
+      "epoch": 14.740298507462686,
+      "grad_norm": 17.07369613647461,
+      "learning_rate": 2.797619047619048e-06,
+      "loss": 46.4404,
+      "step": 619
+    },
+    {
+      "epoch": 14.764179104477613,
+      "grad_norm": 14.855449676513672,
+      "learning_rate": 2.785714285714286e-06,
+      "loss": 48.2668,
+      "step": 620
+    },
+    {
+      "epoch": 14.788059701492537,
+      "grad_norm": 16.479616165161133,
+      "learning_rate": 2.773809523809524e-06,
+      "loss": 49.2661,
+      "step": 621
+    },
+    {
+      "epoch": 14.811940298507462,
+      "grad_norm": 14.471490859985352,
+      "learning_rate": 2.7619047619047625e-06,
+      "loss": 47.0484,
+      "step": 622
+    },
+    {
+      "epoch": 14.835820895522389,
+      "grad_norm": 19.018714904785156,
+      "learning_rate": 2.7500000000000004e-06,
+      "loss": 49.2253,
+      "step": 623
+    },
+    {
+      "epoch": 14.859701492537313,
+      "grad_norm": 16.21799087524414,
+      "learning_rate": 2.7380952380952387e-06,
+      "loss": 49.0738,
+      "step": 624
+    },
+    {
+      "epoch": 14.883582089552238,
+      "grad_norm": 20.86383628845215,
+      "learning_rate": 2.726190476190476e-06,
+      "loss": 48.4231,
+      "step": 625
+    },
+    {
+      "epoch": 14.907462686567165,
+      "grad_norm": 20.60930633544922,
+      "learning_rate": 2.7142857142857144e-06,
+      "loss": 47.7464,
+      "step": 626
+    },
+    {
+      "epoch": 14.93134328358209,
+      "grad_norm": 20.909135818481445,
+      "learning_rate": 2.7023809523809523e-06,
+      "loss": 48.519,
+      "step": 627
+    },
+    {
+      "epoch": 14.955223880597014,
+      "grad_norm": 18.555694580078125,
+      "learning_rate": 2.6904761904761906e-06,
+      "loss": 48.1625,
+      "step": 628
+    },
+    {
+      "epoch": 14.97910447761194,
+      "grad_norm": 18.154813766479492,
+      "learning_rate": 2.6785714285714285e-06,
+      "loss": 48.9444,
+      "step": 629
+    },
+    {
+      "epoch": 15.0,
+      "grad_norm": 19.308523178100586,
+      "learning_rate": 2.666666666666667e-06,
+      "loss": 42.1936,
+      "step": 630
+    },
+    {
+      "epoch": 15.023880597014925,
+      "grad_norm": 22.725357055664062,
+      "learning_rate": 2.654761904761905e-06,
+      "loss": 49.5597,
+      "step": 631
+    },
+    {
+      "epoch": 15.047761194029851,
+      "grad_norm": 18.862451553344727,
+      "learning_rate": 2.642857142857143e-06,
+      "loss": 46.914,
+      "step": 632
+    },
+    {
+      "epoch": 15.071641791044776,
+      "grad_norm": 19.017065048217773,
+      "learning_rate": 2.6309523809523813e-06,
+      "loss": 47.7233,
+      "step": 633
+    },
+    {
+      "epoch": 15.0955223880597,
+      "grad_norm": 19.03627586364746,
+      "learning_rate": 2.6190476190476192e-06,
+      "loss": 48.0406,
+      "step": 634
+    },
+    {
+      "epoch": 15.119402985074627,
+      "grad_norm": 18.53116798400879,
+      "learning_rate": 2.6071428571428575e-06,
+      "loss": 47.3259,
+      "step": 635
+    },
+    {
+      "epoch": 15.143283582089552,
+      "grad_norm": 19.265275955200195,
+      "learning_rate": 2.595238095238096e-06,
+      "loss": 47.2465,
+      "step": 636
+    },
+    {
+      "epoch": 15.167164179104478,
+      "grad_norm": 19.497289657592773,
+      "learning_rate": 2.5833333333333337e-06,
+      "loss": 48.5984,
+      "step": 637
+    },
+    {
+      "epoch": 15.191044776119403,
+      "grad_norm": 20.183780670166016,
+      "learning_rate": 2.571428571428571e-06,
+      "loss": 46.6221,
+      "step": 638
+    },
+    {
+      "epoch": 15.214925373134328,
+      "grad_norm": 22.911672592163086,
+      "learning_rate": 2.5595238095238095e-06,
+      "loss": 48.0178,
+      "step": 639
+    },
+    {
+      "epoch": 15.238805970149254,
+      "grad_norm": 20.678709030151367,
+      "learning_rate": 2.547619047619048e-06,
+      "loss": 47.0322,
+      "step": 640
+    },
+    {
+      "epoch": 15.26268656716418,
+      "grad_norm": 18.579042434692383,
+      "learning_rate": 2.5357142857142857e-06,
+      "loss": 48.1428,
+      "step": 641
+    },
+    {
+      "epoch": 15.286567164179104,
+      "grad_norm": 23.61576271057129,
+      "learning_rate": 2.523809523809524e-06,
+      "loss": 48.444,
+      "step": 642
+    },
+    {
+      "epoch": 15.31044776119403,
+      "grad_norm": 19.602746963500977,
+      "learning_rate": 2.511904761904762e-06,
+      "loss": 48.2582,
+      "step": 643
+    },
+    {
+      "epoch": 15.334328358208955,
+      "grad_norm": 12.509607315063477,
+      "learning_rate": 2.5e-06,
+      "loss": 48.8531,
+      "step": 644
+    },
+    {
+      "epoch": 15.35820895522388,
+      "grad_norm": 18.749767303466797,
+      "learning_rate": 2.4880952380952385e-06,
+      "loss": 47.6453,
+      "step": 645
+    },
+    {
+      "epoch": 15.382089552238806,
+      "grad_norm": 20.612041473388672,
+      "learning_rate": 2.4761904761904764e-06,
+      "loss": 48.7038,
+      "step": 646
+    },
+    {
+      "epoch": 15.405970149253731,
+      "grad_norm": 18.65719985961914,
+      "learning_rate": 2.4642857142857147e-06,
+      "loss": 47.7954,
+      "step": 647
+    },
+    {
+      "epoch": 15.429850746268656,
+      "grad_norm": 22.636686325073242,
+      "learning_rate": 2.4523809523809526e-06,
+      "loss": 48.1164,
+      "step": 648
+    },
+    {
+      "epoch": 15.453731343283582,
+      "grad_norm": 20.93446922302246,
+      "learning_rate": 2.4404761904761905e-06,
+      "loss": 48.5955,
+      "step": 649
+    },
+    {
+      "epoch": 15.477611940298507,
+      "grad_norm": 20.77125358581543,
+      "learning_rate": 2.428571428571429e-06,
+      "loss": 48.4369,
+      "step": 650
+    },
+    {
+      "epoch": 15.501492537313434,
+      "grad_norm": 17.003498077392578,
+      "learning_rate": 2.4166666666666667e-06,
+      "loss": 49.0355,
+      "step": 651
+    },
+    {
+      "epoch": 15.525373134328358,
+      "grad_norm": 20.743436813354492,
+      "learning_rate": 2.404761904761905e-06,
+      "loss": 47.8368,
+      "step": 652
+    },
+    {
+      "epoch": 15.549253731343283,
+      "grad_norm": NaN,
+      "learning_rate": 2.3928571428571433e-06,
+      "loss": 41.6371,
+      "step": 653
+    },
+    {
+      "epoch": 15.57313432835821,
+      "grad_norm": 21.716781616210938,
+      "learning_rate": 2.3928571428571433e-06,
+      "loss": 48.5806,
+      "step": 654
+    },
+    {
+      "epoch": 15.597014925373134,
+      "grad_norm": 18.8812198638916,
+      "learning_rate": 2.380952380952381e-06,
+      "loss": 49.0707,
+      "step": 655
+    },
+    {
+      "epoch": 15.620895522388059,
+      "grad_norm": 22.305049896240234,
+      "learning_rate": 2.369047619047619e-06,
+      "loss": 47.7556,
+      "step": 656
+    },
+    {
+      "epoch": 15.644776119402986,
+      "grad_norm": 20.51401710510254,
+      "learning_rate": 2.3571428571428574e-06,
+      "loss": 48.1588,
+      "step": 657
+    },
+    {
+      "epoch": 15.66865671641791,
+      "grad_norm": 17.691770553588867,
+      "learning_rate": 2.3452380952380953e-06,
+      "loss": 47.5187,
+      "step": 658
+    },
+    {
+      "epoch": 15.692537313432837,
+      "grad_norm": 22.343585968017578,
+      "learning_rate": 2.3333333333333336e-06,
+      "loss": 47.5725,
+      "step": 659
+    },
+    {
+      "epoch": 15.716417910447761,
+      "grad_norm": 21.656587600708008,
+      "learning_rate": 2.321428571428572e-06,
+      "loss": 47.6903,
+      "step": 660
+    },
+    {
+      "epoch": 15.740298507462686,
+      "grad_norm": 20.632055282592773,
+      "learning_rate": 2.3095238095238098e-06,
+      "loss": 47.4526,
+      "step": 661
+    },
+    {
+      "epoch": 15.764179104477613,
+      "grad_norm": 22.324811935424805,
+      "learning_rate": 2.2976190476190477e-06,
+      "loss": 47.3316,
+      "step": 662
+    },
+    {
+      "epoch": 15.788059701492537,
+      "grad_norm": 19.320737838745117,
+      "learning_rate": 2.285714285714286e-06,
+      "loss": 48.2315,
+      "step": 663
+    },
+    {
+      "epoch": 15.811940298507462,
+      "grad_norm": 18.58050537109375,
+      "learning_rate": 2.273809523809524e-06,
+      "loss": 47.9147,
+      "step": 664
+    },
+    {
+      "epoch": 15.835820895522389,
+      "grad_norm": 20.37384796142578,
+      "learning_rate": 2.261904761904762e-06,
+      "loss": 47.5874,
+      "step": 665
+    },
+    {
+      "epoch": 15.859701492537313,
+      "grad_norm": 20.893856048583984,
+      "learning_rate": 2.25e-06,
+      "loss": 49.225,
+      "step": 666
+    },
+    {
+      "epoch": 15.883582089552238,
+      "grad_norm": 18.4589786529541,
+      "learning_rate": 2.2380952380952384e-06,
+      "loss": 47.3042,
+      "step": 667
+    },
+    {
+      "epoch": 15.907462686567165,
+      "grad_norm": 20.845996856689453,
+      "learning_rate": 2.2261904761904763e-06,
+      "loss": 47.3255,
+      "step": 668
+    },
+    {
+      "epoch": 15.93134328358209,
+      "grad_norm": 20.149137496948242,
+      "learning_rate": 2.2142857142857146e-06,
+      "loss": 48.6543,
+      "step": 669
+    },
+    {
+      "epoch": 15.955223880597014,
+      "grad_norm": 14.768882751464844,
+      "learning_rate": 2.2023809523809525e-06,
+      "loss": 46.8274,
+      "step": 670
+    },
+    {
+      "epoch": 15.97910447761194,
+      "grad_norm": 26.926074981689453,
+      "learning_rate": 2.1904761904761908e-06,
+      "loss": 48.2035,
+      "step": 671
+    },
+    {
+      "epoch": 16.0,
+      "grad_norm": 22.840618133544922,
+      "learning_rate": 2.1785714285714286e-06,
+      "loss": 42.8242,
+      "step": 672
+    },
+    {
+      "epoch": 16.023880597014927,
+      "grad_norm": 16.183008193969727,
+      "learning_rate": 2.166666666666667e-06,
+      "loss": 47.8309,
+      "step": 673
+    },
+    {
+      "epoch": 16.04776119402985,
+      "grad_norm": 20.603744506835938,
+      "learning_rate": 2.154761904761905e-06,
+      "loss": 48.5197,
+      "step": 674
+    },
+    {
+      "epoch": 16.071641791044776,
+      "grad_norm": 26.492107391357422,
+      "learning_rate": 2.1428571428571427e-06,
+      "loss": 47.2312,
+      "step": 675
+    },
+    {
+      "epoch": 16.095522388059702,
+      "grad_norm": 19.786901473999023,
+      "learning_rate": 2.130952380952381e-06,
+      "loss": 49.6201,
+      "step": 676
+    },
+    {
+      "epoch": 16.119402985074625,
+      "grad_norm": 18.150909423828125,
+      "learning_rate": 2.1190476190476194e-06,
+      "loss": 48.7407,
+      "step": 677
+    },
+    {
+      "epoch": 16.143283582089552,
+      "grad_norm": 18.797983169555664,
+      "learning_rate": 2.1071428571428572e-06,
+      "loss": 47.0801,
+      "step": 678
+    },
+    {
+      "epoch": 16.16716417910448,
+      "grad_norm": 16.088953018188477,
+      "learning_rate": 2.0952380952380955e-06,
+      "loss": 47.6509,
+      "step": 679
+    },
+    {
+      "epoch": 16.1910447761194,
+      "grad_norm": 20.359085083007812,
+      "learning_rate": 2.0833333333333334e-06,
+      "loss": 48.9226,
+      "step": 680
+    },
+    {
+      "epoch": 16.214925373134328,
+      "grad_norm": 21.99265480041504,
+      "learning_rate": 2.0714285714285717e-06,
+      "loss": 47.3775,
+      "step": 681
+    },
+    {
+      "epoch": 16.238805970149254,
+      "grad_norm": 18.616743087768555,
+      "learning_rate": 2.0595238095238096e-06,
+      "loss": 45.8448,
+      "step": 682
+    },
+    {
+      "epoch": 16.262686567164177,
+      "grad_norm": 19.6337947845459,
+      "learning_rate": 2.047619047619048e-06,
+      "loss": 48.2077,
+      "step": 683
+    },
+    {
+      "epoch": 16.286567164179104,
+      "grad_norm": 23.881439208984375,
+      "learning_rate": 2.035714285714286e-06,
+      "loss": 48.6796,
+      "step": 684
+    },
+    {
+      "epoch": 16.31044776119403,
+      "grad_norm": 19.665023803710938,
+      "learning_rate": 2.023809523809524e-06,
+      "loss": 48.7275,
+      "step": 685
+    },
+    {
+      "epoch": 16.334328358208957,
+      "grad_norm": 18.438793182373047,
+      "learning_rate": 2.011904761904762e-06,
+      "loss": 49.585,
+      "step": 686
+    },
+    {
+      "epoch": 16.35820895522388,
+      "grad_norm": 17.073816299438477,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 47.4548,
+      "step": 687
+    },
+    {
+      "epoch": 16.382089552238806,
+      "grad_norm": 20.504276275634766,
+      "learning_rate": 1.9880952380952382e-06,
+      "loss": 47.3555,
+      "step": 688
+    },
+    {
+      "epoch": 16.405970149253733,
+      "grad_norm": 21.564546585083008,
+      "learning_rate": 1.976190476190476e-06,
+      "loss": 47.6304,
+      "step": 689
+    },
+    {
+      "epoch": 16.429850746268656,
+      "grad_norm": 16.773197174072266,
+      "learning_rate": 1.9642857142857144e-06,
+      "loss": 49.078,
+      "step": 690
+    },
+    {
+      "epoch": 16.453731343283582,
+      "grad_norm": 22.77934455871582,
+      "learning_rate": 1.9523809523809527e-06,
+      "loss": 47.8289,
+      "step": 691
+    },
+    {
+      "epoch": 16.47761194029851,
+      "grad_norm": 17.375993728637695,
+      "learning_rate": 1.9404761904761906e-06,
+      "loss": 48.4812,
+      "step": 692
+    },
+    {
+      "epoch": 16.501492537313432,
+      "grad_norm": 21.407329559326172,
+      "learning_rate": 1.928571428571429e-06,
+      "loss": 48.2934,
+      "step": 693
+    },
+    {
+      "epoch": 16.52537313432836,
+      "grad_norm": 15.673316955566406,
+      "learning_rate": 1.916666666666667e-06,
+      "loss": 46.7304,
+      "step": 694
+    },
+    {
+      "epoch": 16.549253731343285,
+      "grad_norm": 24.577089309692383,
+      "learning_rate": 1.904761904761905e-06,
+      "loss": 47.9352,
+      "step": 695
+    },
+    {
+      "epoch": 16.573134328358208,
+      "grad_norm": 24.46076774597168,
+      "learning_rate": 1.892857142857143e-06,
+      "loss": 48.8173,
+      "step": 696
+    },
+    {
+      "epoch": 16.597014925373134,
+      "grad_norm": 14.248388290405273,
+      "learning_rate": 1.880952380952381e-06,
+      "loss": 48.5858,
+      "step": 697
+    },
+    {
+      "epoch": 16.62089552238806,
+      "grad_norm": 16.925329208374023,
+      "learning_rate": 1.8690476190476192e-06,
+      "loss": 47.8278,
+      "step": 698
+    },
+    {
+      "epoch": 16.644776119402984,
+      "grad_norm": 25.52614402770996,
+      "learning_rate": 1.8571428571428573e-06,
+      "loss": 48.1248,
+      "step": 699
+    },
+    {
+      "epoch": 16.66865671641791,
+      "grad_norm": 21.011341094970703,
+      "learning_rate": 1.8452380952380954e-06,
+      "loss": 47.8154,
+      "step": 700
+    },
+    {
+      "epoch": 16.692537313432837,
+      "grad_norm": 14.694896697998047,
+      "learning_rate": 1.8333333333333333e-06,
+      "loss": 47.9668,
+      "step": 701
+    },
+    {
+      "epoch": 16.71641791044776,
+      "grad_norm": 22.32903480529785,
+      "learning_rate": 1.8214285714285716e-06,
+      "loss": 48.6784,
+      "step": 702
+    },
+    {
+      "epoch": 16.740298507462686,
+      "grad_norm": 17.19482421875,
+      "learning_rate": 1.8095238095238097e-06,
+      "loss": 46.9973,
+      "step": 703
+    },
+    {
+      "epoch": 16.764179104477613,
+      "grad_norm": 14.590733528137207,
+      "learning_rate": 1.7976190476190478e-06,
+      "loss": 47.2393,
+      "step": 704
+    },
+    {
+      "epoch": 16.788059701492536,
+      "grad_norm": 17.131982803344727,
+      "learning_rate": 1.7857142857142859e-06,
+      "loss": 47.9412,
+      "step": 705
+    },
+    {
+      "epoch": 16.811940298507462,
+      "grad_norm": 18.513992309570312,
+      "learning_rate": 1.773809523809524e-06,
+      "loss": 48.8777,
+      "step": 706
+    },
+    {
+      "epoch": 16.83582089552239,
+      "grad_norm": 17.625539779663086,
+      "learning_rate": 1.761904761904762e-06,
+      "loss": 48.3885,
+      "step": 707
+    },
+    {
+      "epoch": 16.85970149253731,
+      "grad_norm": 16.540056228637695,
+      "learning_rate": 1.75e-06,
+      "loss": 47.8561,
+      "step": 708
+    },
+    {
+      "epoch": 16.883582089552238,
+      "grad_norm": 20.070533752441406,
+      "learning_rate": 1.738095238095238e-06,
+      "loss": 46.6418,
+      "step": 709
+    },
+    {
+      "epoch": 16.907462686567165,
+      "grad_norm": 18.742460250854492,
+      "learning_rate": 1.7261904761904764e-06,
+      "loss": 46.7471,
+      "step": 710
+    },
+    {
+      "epoch": 16.93134328358209,
+      "grad_norm": 17.491954803466797,
+      "learning_rate": 1.7142857142857145e-06,
+      "loss": 47.5558,
+      "step": 711
+    },
+    {
+      "epoch": 16.955223880597014,
+      "grad_norm": 17.457130432128906,
+      "learning_rate": 1.7023809523809526e-06,
+      "loss": 47.4441,
+      "step": 712
+    },
+    {
+      "epoch": 16.97910447761194,
+      "grad_norm": 21.053844451904297,
+      "learning_rate": 1.6904761904761907e-06,
+      "loss": 48.1931,
+      "step": 713
+    },
+    {
+      "epoch": 17.0,
+      "grad_norm": 16.943801879882812,
+      "learning_rate": 1.6785714285714286e-06,
+      "loss": 41.9934,
+      "step": 714
+    },
+    {
+      "epoch": 17.023880597014927,
+      "grad_norm": 21.56785011291504,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 47.1652,
+      "step": 715
+    },
+    {
+      "epoch": 17.04776119402985,
+      "grad_norm": 21.193382263183594,
+      "learning_rate": 1.6547619047619048e-06,
+      "loss": 47.6751,
+      "step": 716
+    },
+    {
+      "epoch": 17.071641791044776,
+      "grad_norm": 16.245115280151367,
+      "learning_rate": 1.642857142857143e-06,
+      "loss": 47.4133,
+      "step": 717
+    },
+    {
+      "epoch": 17.095522388059702,
+      "grad_norm": 18.834646224975586,
+      "learning_rate": 1.6309523809523812e-06,
+      "loss": 48.145,
+      "step": 718
+    },
+    {
+      "epoch": 17.119402985074625,
+      "grad_norm": 15.769698143005371,
+      "learning_rate": 1.6190476190476193e-06,
+      "loss": 48.1181,
+      "step": 719
+    },
+    {
+      "epoch": 17.143283582089552,
+      "grad_norm": 13.460511207580566,
+      "learning_rate": 1.6071428571428574e-06,
+      "loss": 49.1229,
+      "step": 720
+    },
+    {
+      "epoch": 17.16716417910448,
+      "grad_norm": 18.58087158203125,
+      "learning_rate": 1.5952380952380953e-06,
+      "loss": 47.5095,
+      "step": 721
+    },
+    {
+      "epoch": 17.1910447761194,
+      "grad_norm": 18.607332229614258,
+      "learning_rate": 1.5833333333333333e-06,
+      "loss": 49.1334,
+      "step": 722
+    },
+    {
+      "epoch": 17.214925373134328,
+      "grad_norm": 15.046488761901855,
+      "learning_rate": 1.5714285714285714e-06,
+      "loss": 47.6151,
+      "step": 723
+    },
+    {
+      "epoch": 17.238805970149254,
+      "grad_norm": 17.442358016967773,
+      "learning_rate": 1.5595238095238098e-06,
+      "loss": 47.3771,
+      "step": 724
+    },
+    {
+      "epoch": 17.262686567164177,
+      "grad_norm": 11.690101623535156,
+      "learning_rate": 1.5476190476190479e-06,
+      "loss": 48.1095,
+      "step": 725
+    },
+    {
+      "epoch": 17.286567164179104,
+      "grad_norm": 17.945192337036133,
+      "learning_rate": 1.535714285714286e-06,
+      "loss": 47.8941,
+      "step": 726
+    },
+    {
+      "epoch": 17.31044776119403,
+      "grad_norm": 13.878116607666016,
+      "learning_rate": 1.523809523809524e-06,
+      "loss": 47.6422,
+      "step": 727
+    },
+    {
+      "epoch": 17.334328358208957,
+      "grad_norm": 15.942928314208984,
+      "learning_rate": 1.511904761904762e-06,
+      "loss": 46.9964,
+      "step": 728
+    },
+    {
+      "epoch": 17.35820895522388,
+      "grad_norm": 13.57482624053955,
+      "learning_rate": 1.5e-06,
+      "loss": 47.1832,
+      "step": 729
+    },
+    {
+      "epoch": 17.382089552238806,
+      "grad_norm": 13.781617164611816,
+      "learning_rate": 1.4880952380952381e-06,
+      "loss": 48.9621,
+      "step": 730
+    },
+    {
+      "epoch": 17.405970149253733,
+      "grad_norm": 14.26857852935791,
+      "learning_rate": 1.4761904761904762e-06,
+      "loss": 48.6631,
+      "step": 731
+    },
+    {
+      "epoch": 17.429850746268656,
+      "grad_norm": 16.23444938659668,
+      "learning_rate": 1.4642857142857145e-06,
+      "loss": 46.84,
+      "step": 732
+    },
+    {
+      "epoch": 17.453731343283582,
+      "grad_norm": 17.442630767822266,
+      "learning_rate": 1.4523809523809526e-06,
+      "loss": 48.2996,
+      "step": 733
+    },
+    {
+      "epoch": 17.47761194029851,
+      "grad_norm": 14.329082489013672,
+      "learning_rate": 1.4404761904761905e-06,
+      "loss": 47.396,
+      "step": 734
+    },
+    {
+      "epoch": 17.501492537313432,
+      "grad_norm": 14.772257804870605,
+      "learning_rate": 1.4285714285714286e-06,
+      "loss": 48.5733,
+      "step": 735
+    },
+    {
+      "epoch": 17.52537313432836,
+      "grad_norm": 14.331324577331543,
+      "learning_rate": 1.4166666666666667e-06,
+      "loss": 48.2969,
+      "step": 736
+    },
+    {
+      "epoch": 17.549253731343285,
+      "grad_norm": 17.498600006103516,
+      "learning_rate": 1.4047619047619048e-06,
+      "loss": 48.0221,
+      "step": 737
+    },
+    {
+      "epoch": 17.573134328358208,
+      "grad_norm": 16.155025482177734,
+      "learning_rate": 1.392857142857143e-06,
+      "loss": 47.9848,
+      "step": 738
+    },
+    {
+      "epoch": 17.597014925373134,
+      "grad_norm": 15.552813529968262,
+      "learning_rate": 1.3809523809523812e-06,
+      "loss": 48.4413,
+      "step": 739
+    },
+    {
+      "epoch": 17.62089552238806,
+      "grad_norm": 15.887310981750488,
+      "learning_rate": 1.3690476190476193e-06,
+      "loss": 47.6463,
+      "step": 740
+    },
+    {
+      "epoch": 17.644776119402984,
+      "grad_norm": 17.783411026000977,
+      "learning_rate": 1.3571428571428572e-06,
+      "loss": 47.8009,
+      "step": 741
+    },
+    {
+      "epoch": 17.66865671641791,
+      "grad_norm": 17.108932495117188,
+      "learning_rate": 1.3452380952380953e-06,
+      "loss": 47.9888,
+      "step": 742
+    },
+    {
+      "epoch": 17.692537313432837,
+      "grad_norm": 19.79203224182129,
+      "learning_rate": 1.3333333333333334e-06,
+      "loss": 48.5732,
+      "step": 743
+    },
+    {
+      "epoch": 17.71641791044776,
+      "grad_norm": 17.06324005126953,
+      "learning_rate": 1.3214285714285715e-06,
+      "loss": 48.4815,
+      "step": 744
+    },
+    {
+      "epoch": 17.740298507462686,
+      "grad_norm": 17.399097442626953,
+      "learning_rate": 1.3095238095238096e-06,
+      "loss": 47.5591,
+      "step": 745
+    },
+    {
+      "epoch": 17.764179104477613,
+      "grad_norm": 15.836935997009277,
+      "learning_rate": 1.297619047619048e-06,
+      "loss": 47.994,
+      "step": 746
+    },
+    {
+      "epoch": 17.788059701492536,
+      "grad_norm": 18.20856475830078,
+      "learning_rate": 1.2857142857142856e-06,
+      "loss": 47.9979,
+      "step": 747
+    },
+    {
+      "epoch": 17.811940298507462,
+      "grad_norm": 19.10239601135254,
+      "learning_rate": 1.273809523809524e-06,
+      "loss": 48.1196,
+      "step": 748
+    },
+    {
+      "epoch": 17.83582089552239,
+      "grad_norm": 17.21087646484375,
+      "learning_rate": 1.261904761904762e-06,
+      "loss": 47.8816,
+      "step": 749
+    },
+    {
+      "epoch": 17.85970149253731,
+      "grad_norm": 14.792268753051758,
+      "learning_rate": 1.25e-06,
+      "loss": 47.8182,
+      "step": 750
+    },
+    {
+      "epoch": 17.883582089552238,
+      "grad_norm": 13.695488929748535,
+      "learning_rate": 1.2380952380952382e-06,
+      "loss": 47.4298,
+      "step": 751
+    },
+    {
+      "epoch": 17.907462686567165,
+      "grad_norm": 15.197646141052246,
+      "learning_rate": 1.2261904761904763e-06,
+      "loss": 47.7132,
+      "step": 752
+    },
+    {
+      "epoch": 17.93134328358209,
+      "grad_norm": 19.13431739807129,
+      "learning_rate": 1.2142857142857144e-06,
+      "loss": 48.431,
+      "step": 753
+    },
+    {
+      "epoch": 17.955223880597014,
+      "grad_norm": 15.690411567687988,
+      "learning_rate": 1.2023809523809525e-06,
+      "loss": 47.4529,
+      "step": 754
+    },
+    {
+      "epoch": 17.97910447761194,
+      "grad_norm": 14.75414752960205,
+      "learning_rate": 1.1904761904761906e-06,
+      "loss": 47.9668,
+      "step": 755
+    },
+    {
+      "epoch": 18.0,
+      "grad_norm": 11.497115135192871,
+      "learning_rate": 1.1785714285714287e-06,
+      "loss": 41.8653,
+      "step": 756
+    },
+    {
+      "epoch": 18.023880597014927,
+      "grad_norm": 16.20159339904785,
+      "learning_rate": 1.1666666666666668e-06,
+      "loss": 47.2871,
+      "step": 757
+    },
+    {
+      "epoch": 18.04776119402985,
+      "grad_norm": 15.400497436523438,
+      "learning_rate": 1.1547619047619049e-06,
+      "loss": 46.3673,
+      "step": 758
+    },
+    {
+      "epoch": 18.071641791044776,
+      "grad_norm": 12.16773509979248,
+      "learning_rate": 1.142857142857143e-06,
+      "loss": 47.7463,
+      "step": 759
+    },
+    {
+      "epoch": 18.095522388059702,
+      "grad_norm": 19.978351593017578,
+      "learning_rate": 1.130952380952381e-06,
+      "loss": 47.4632,
+      "step": 760
+    },
+    {
+      "epoch": 18.119402985074625,
+      "grad_norm": 14.090561866760254,
+      "learning_rate": 1.1190476190476192e-06,
+      "loss": 48.9356,
+      "step": 761
+    },
+    {
+      "epoch": 18.143283582089552,
+      "grad_norm": 13.143173217773438,
+      "learning_rate": 1.1071428571428573e-06,
+      "loss": 48.1129,
+      "step": 762
+    },
+    {
+      "epoch": 18.16716417910448,
+      "grad_norm": 15.609000205993652,
+      "learning_rate": 1.0952380952380954e-06,
+      "loss": 48.8554,
+      "step": 763
+    },
+    {
+      "epoch": 18.1910447761194,
+      "grad_norm": 14.012611389160156,
+      "learning_rate": 1.0833333333333335e-06,
+      "loss": 47.6785,
+      "step": 764
+    },
+    {
+      "epoch": 18.214925373134328,
+      "grad_norm": 13.417494773864746,
+      "learning_rate": 1.0714285714285714e-06,
+      "loss": 48.2733,
+      "step": 765
+    },
+    {
+      "epoch": 18.238805970149254,
+      "grad_norm": 15.562864303588867,
+      "learning_rate": 1.0595238095238097e-06,
+      "loss": 48.0488,
+      "step": 766
+    },
+    {
+      "epoch": 18.262686567164177,
+      "grad_norm": 17.083723068237305,
+      "learning_rate": 1.0476190476190478e-06,
+      "loss": 49.3136,
+      "step": 767
+    },
+    {
+      "epoch": 18.286567164179104,
+      "grad_norm": 16.564395904541016,
+      "learning_rate": 1.0357142857142859e-06,
+      "loss": 48.1835,
+      "step": 768
+    },
+    {
+      "epoch": 18.31044776119403,
+      "grad_norm": 14.694929122924805,
+      "learning_rate": 1.023809523809524e-06,
+      "loss": 48.8632,
+      "step": 769
+    },
+    {
+      "epoch": 18.334328358208957,
+      "grad_norm": 13.4928560256958,
+      "learning_rate": 1.011904761904762e-06,
+      "loss": 47.3333,
+      "step": 770
+    },
+    {
+      "epoch": 18.35820895522388,
+      "grad_norm": 12.980204582214355,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 47.7961,
+      "step": 771
+    },
+    {
+      "epoch": 18.382089552238806,
+      "grad_norm": 14.19666862487793,
+      "learning_rate": 9.88095238095238e-07,
+      "loss": 47.7806,
+      "step": 772
+    },
+    {
+      "epoch": 18.405970149253733,
+      "grad_norm": 13.2017183303833,
+      "learning_rate": 9.761904761904764e-07,
+      "loss": 46.1119,
+      "step": 773
+    },
+    {
+      "epoch": 18.429850746268656,
+      "grad_norm": 15.064650535583496,
+      "learning_rate": 9.642857142857145e-07,
+      "loss": 47.8105,
+      "step": 774
+    },
+    {
+      "epoch": 18.453731343283582,
+      "grad_norm": 14.713834762573242,
+      "learning_rate": 9.523809523809525e-07,
+      "loss": 47.5723,
+      "step": 775
+    },
+    {
+      "epoch": 18.47761194029851,
+      "grad_norm": 13.394201278686523,
+      "learning_rate": 9.404761904761906e-07,
+      "loss": 48.1632,
+      "step": 776
+    },
+    {
+      "epoch": 18.501492537313432,
+      "grad_norm": 10.742532730102539,
+      "learning_rate": 9.285714285714287e-07,
+      "loss": 48.6118,
+      "step": 777
+    },
+    {
+      "epoch": 18.52537313432836,
+      "grad_norm": 12.402650833129883,
+      "learning_rate": 9.166666666666666e-07,
+      "loss": 48.6597,
+      "step": 778
+    },
+    {
+      "epoch": 18.549253731343285,
+      "grad_norm": 15.73616886138916,
+      "learning_rate": 9.047619047619048e-07,
+      "loss": 47.9931,
+      "step": 779
+    },
+    {
+      "epoch": 18.573134328358208,
+      "grad_norm": 14.188780784606934,
+      "learning_rate": 8.928571428571429e-07,
+      "loss": 47.0081,
+      "step": 780
+    },
+    {
+      "epoch": 18.597014925373134,
+      "grad_norm": 12.516701698303223,
+      "learning_rate": 8.80952380952381e-07,
+      "loss": 49.0287,
+      "step": 781
+    },
+    {
+      "epoch": 18.62089552238806,
+      "grad_norm": 15.069429397583008,
+      "learning_rate": 8.69047619047619e-07,
+      "loss": 47.8721,
+      "step": 782
+    },
+    {
+      "epoch": 18.644776119402984,
+      "grad_norm": 13.091047286987305,
+      "learning_rate": 8.571428571428572e-07,
+      "loss": 48.1678,
+      "step": 783
+    },
+    {
+      "epoch": 18.66865671641791,
+      "grad_norm": 15.017065048217773,
+      "learning_rate": 8.452380952380953e-07,
+      "loss": 47.1277,
+      "step": 784
+    },
+    {
+      "epoch": 18.692537313432837,
+      "grad_norm": 12.091531753540039,
+      "learning_rate": 8.333333333333333e-07,
+      "loss": 47.4962,
+      "step": 785
+    },
+    {
+      "epoch": 18.71641791044776,
+      "grad_norm": 15.20182991027832,
+      "learning_rate": 8.214285714285715e-07,
+      "loss": 48.6224,
+      "step": 786
+    },
+    {
+      "epoch": 18.740298507462686,
+      "grad_norm": 11.17827320098877,
+      "learning_rate": 8.095238095238096e-07,
+      "loss": 47.9759,
+      "step": 787
+    },
+    {
+      "epoch": 18.764179104477613,
+      "grad_norm": 14.884525299072266,
+      "learning_rate": 7.976190476190476e-07,
+      "loss": 47.9749,
+      "step": 788
+    },
+    {
+      "epoch": 18.788059701492536,
+      "grad_norm": 14.360984802246094,
+      "learning_rate": 7.857142857142857e-07,
+      "loss": 48.9952,
+      "step": 789
+    },
+    {
+      "epoch": 18.811940298507462,
+      "grad_norm": 11.265621185302734,
+      "learning_rate": 7.738095238095239e-07,
+      "loss": 47.4274,
+      "step": 790
+    },
+    {
+      "epoch": 18.83582089552239,
+      "grad_norm": 18.072290420532227,
+      "learning_rate": 7.61904761904762e-07,
+      "loss": 47.8815,
+      "step": 791
+    },
+    {
+      "epoch": 18.85970149253731,
+      "grad_norm": 15.310029029846191,
+      "learning_rate": 7.5e-07,
+      "loss": 47.3907,
+      "step": 792
+    },
+    {
+      "epoch": 18.883582089552238,
+      "grad_norm": 14.032752990722656,
+      "learning_rate": 7.380952380952381e-07,
+      "loss": 48.0883,
+      "step": 793
+    },
+    {
+      "epoch": 18.907462686567165,
+      "grad_norm": 12.853668212890625,
+      "learning_rate": 7.261904761904763e-07,
+      "loss": 47.1452,
+      "step": 794
+    },
+    {
+      "epoch": 18.93134328358209,
+      "grad_norm": 15.6067476272583,
+      "learning_rate": 7.142857142857143e-07,
+      "loss": 46.6303,
+      "step": 795
+    },
+    {
+      "epoch": 18.955223880597014,
+      "grad_norm": 12.828201293945312,
+      "learning_rate": 7.023809523809524e-07,
+      "loss": 47.9885,
+      "step": 796
+    },
+    {
+      "epoch": 18.97910447761194,
+      "grad_norm": 13.336589813232422,
+      "learning_rate": 6.904761904761906e-07,
+      "loss": 48.2315,
+      "step": 797
+    },
+    {
+      "epoch": 19.0,
+      "grad_norm": 13.629434585571289,
+      "learning_rate": 6.785714285714286e-07,
+      "loss": 41.9374,
+      "step": 798
+    },
+    {
+      "epoch": 19.023880597014927,
+      "grad_norm": 13.237930297851562,
+      "learning_rate": 6.666666666666667e-07,
+      "loss": 46.6802,
+      "step": 799
+    },
+    {
+      "epoch": 19.04776119402985,
+      "grad_norm": 13.715863227844238,
+      "learning_rate": 6.547619047619048e-07,
+      "loss": 49.0494,
+      "step": 800
+    },
+    {
+      "epoch": 19.071641791044776,
+      "grad_norm": 13.439970016479492,
+      "learning_rate": 6.428571428571428e-07,
+      "loss": 46.3647,
+      "step": 801
+    },
+    {
+      "epoch": 19.095522388059702,
+      "grad_norm": 15.468942642211914,
+      "learning_rate": 6.30952380952381e-07,
+      "loss": 48.4725,
+      "step": 802
+    },
+    {
+      "epoch": 19.119402985074625,
+      "grad_norm": 14.160257339477539,
+      "learning_rate": 6.190476190476191e-07,
+      "loss": 47.4033,
+      "step": 803
+    },
+    {
+      "epoch": 19.143283582089552,
+      "grad_norm": 13.667155265808105,
+      "learning_rate": 6.071428571428572e-07,
+      "loss": 48.4729,
+      "step": 804
+    },
+    {
+      "epoch": 19.16716417910448,
+      "grad_norm": 12.428313255310059,
+      "learning_rate": 5.952380952380953e-07,
+      "loss": 48.8939,
+      "step": 805
+    },
+    {
+      "epoch": 19.1910447761194,
+      "grad_norm": 12.985882759094238,
+      "learning_rate": 5.833333333333334e-07,
+      "loss": 47.0663,
+      "step": 806
+    },
+    {
+      "epoch": 19.214925373134328,
+      "grad_norm": 12.827404975891113,
+      "learning_rate": 5.714285714285715e-07,
+      "loss": 47.5614,
+      "step": 807
+    },
+    {
+      "epoch": 19.238805970149254,
+      "grad_norm": 11.078653335571289,
+      "learning_rate": 5.595238095238096e-07,
+      "loss": 48.564,
+      "step": 808
+    },
+    {
+      "epoch": 19.262686567164177,
+      "grad_norm": 13.346016883850098,
+      "learning_rate": 5.476190476190477e-07,
+      "loss": 48.0823,
+      "step": 809
+    },
+    {
+      "epoch": 19.286567164179104,
+      "grad_norm": 14.523963928222656,
+      "learning_rate": 5.357142857142857e-07,
+      "loss": 48.4225,
+      "step": 810
+    },
+    {
+      "epoch": 19.31044776119403,
+      "grad_norm": 12.598445892333984,
+      "learning_rate": 5.238095238095239e-07,
+      "loss": 47.2514,
+      "step": 811
+    },
+    {
+      "epoch": 19.334328358208957,
+      "grad_norm": 12.203497886657715,
+      "learning_rate": 5.11904761904762e-07,
+      "loss": 47.217,
+      "step": 812
+    },
+    {
+      "epoch": 19.35820895522388,
+      "grad_norm": 12.144754409790039,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 47.191,
+      "step": 813
+    },
+    {
+      "epoch": 19.382089552238806,
+      "grad_norm": 12.585047721862793,
+      "learning_rate": 4.880952380952382e-07,
+      "loss": 48.4947,
+      "step": 814
+    },
+    {
+      "epoch": 19.405970149253733,
+      "grad_norm": 11.295561790466309,
+      "learning_rate": 4.7619047619047623e-07,
+      "loss": 46.9444,
+      "step": 815
+    },
+    {
+      "epoch": 19.429850746268656,
+      "grad_norm": 13.055256843566895,
+      "learning_rate": 4.642857142857143e-07,
+      "loss": 48.4469,
+      "step": 816
+    },
+    {
+      "epoch": 19.453731343283582,
+      "grad_norm": 12.051807403564453,
+      "learning_rate": 4.523809523809524e-07,
+      "loss": 48.0547,
+      "step": 817
+    },
+    {
+      "epoch": 19.47761194029851,
+      "grad_norm": 13.44185733795166,
+      "learning_rate": 4.404761904761905e-07,
+      "loss": 48.3155,
+      "step": 818
+    },
+    {
+      "epoch": 19.501492537313432,
+      "grad_norm": 12.405723571777344,
+      "learning_rate": 4.285714285714286e-07,
+      "loss": 48.3982,
+      "step": 819
+    },
+    {
+      "epoch": 19.52537313432836,
+      "grad_norm": 14.900402069091797,
+      "learning_rate": 4.1666666666666667e-07,
+      "loss": 48.2653,
+      "step": 820
+    },
+    {
+      "epoch": 19.549253731343285,
+      "grad_norm": 10.70801067352295,
+      "learning_rate": 4.047619047619048e-07,
+      "loss": 48.0384,
+      "step": 821
+    },
+    {
+      "epoch": 19.573134328358208,
+      "grad_norm": 12.318074226379395,
+      "learning_rate": 3.9285714285714286e-07,
+      "loss": 47.554,
+      "step": 822
+    },
+    {
+      "epoch": 19.597014925373134,
+      "grad_norm": 12.898431777954102,
+      "learning_rate": 3.80952380952381e-07,
+      "loss": 48.3586,
+      "step": 823
+    },
+    {
+      "epoch": 19.62089552238806,
+      "grad_norm": 15.45779800415039,
+      "learning_rate": 3.6904761904761906e-07,
+      "loss": 48.4193,
+      "step": 824
+    },
+    {
+      "epoch": 19.644776119402984,
+      "grad_norm": 11.230570793151855,
+      "learning_rate": 3.5714285714285716e-07,
+      "loss": 48.5294,
+      "step": 825
+    },
+    {
+      "epoch": 19.66865671641791,
+      "grad_norm": 13.647272109985352,
+      "learning_rate": 3.452380952380953e-07,
+      "loss": 47.2569,
+      "step": 826
+    },
+    {
+      "epoch": 19.692537313432837,
+      "grad_norm": 11.521178245544434,
+      "learning_rate": 3.3333333333333335e-07,
+      "loss": 47.2899,
+      "step": 827
+    },
+    {
+      "epoch": 19.71641791044776,
+      "grad_norm": 11.537907600402832,
+      "learning_rate": 3.214285714285714e-07,
+      "loss": 46.6462,
+      "step": 828
+    },
+    {
+      "epoch": 19.740298507462686,
+      "grad_norm": 11.670267105102539,
+      "learning_rate": 3.0952380952380955e-07,
+      "loss": 47.9797,
+      "step": 829
+    },
+    {
+      "epoch": 19.764179104477613,
+      "grad_norm": 11.660557746887207,
+      "learning_rate": 2.9761904761904765e-07,
+      "loss": 47.9744,
+      "step": 830
+    },
+    {
+      "epoch": 19.788059701492536,
+      "grad_norm": 12.332269668579102,
+      "learning_rate": 2.8571428571428575e-07,
+      "loss": 48.6015,
+      "step": 831
+    },
+    {
+      "epoch": 19.811940298507462,
+      "grad_norm": 12.228848457336426,
+      "learning_rate": 2.7380952380952385e-07,
+      "loss": 47.3215,
+      "step": 832
+    },
+    {
+      "epoch": 19.83582089552239,
+      "grad_norm": 13.780754089355469,
+      "learning_rate": 2.6190476190476194e-07,
+      "loss": 48.853,
+      "step": 833
+    },
+    {
+      "epoch": 19.85970149253731,
+      "grad_norm": 11.639240264892578,
+      "learning_rate": 2.5000000000000004e-07,
+      "loss": 48.8199,
+      "step": 834
+    },
+    {
+      "epoch": 19.883582089552238,
+      "grad_norm": 10.796862602233887,
+      "learning_rate": 2.3809523809523811e-07,
+      "loss": 47.5373,
+      "step": 835
+    },
+    {
+      "epoch": 19.907462686567165,
+      "grad_norm": 13.573180198669434,
+      "learning_rate": 2.261904761904762e-07,
+      "loss": 47.8368,
+      "step": 836
+    },
+    {
+      "epoch": 19.93134328358209,
+      "grad_norm": 11.497776985168457,
+      "learning_rate": 2.142857142857143e-07,
+      "loss": 47.8226,
+      "step": 837
+    },
+    {
+      "epoch": 19.955223880597014,
+      "grad_norm": 10.777889251708984,
+      "learning_rate": 2.023809523809524e-07,
+      "loss": 47.6424,
+      "step": 838
+    },
+    {
+      "epoch": 19.97910447761194,
+      "grad_norm": 10.77852725982666,
+      "learning_rate": 1.904761904761905e-07,
+      "loss": 46.379,
+      "step": 839
+    },
+    {
+      "epoch": 20.0,
+      "grad_norm": 13.582564353942871,
+      "learning_rate": 1.7857142857142858e-07,
+      "loss": 42.5239,
+      "step": 840
+    },
+    {
+      "epoch": 20.0,
+      "step": 840,
+      "total_flos": 4.130470305428237e+16,
+      "train_loss": 49.47331008002872,
+      "train_runtime": 26137.3223,
+      "train_samples_per_second": 4.095,
+      "train_steps_per_second": 0.032
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 840,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 20,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4.130470305428237e+16,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}