diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,58857 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.9998215029452013,
+  "eval_steps": 500,
+  "global_step": 8402,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00023799607306479442,
+      "grad_norm": 5.3633106806179,
+      "learning_rate": 0.0,
+      "loss": 0.7123,
+      "step": 1
+    },
+    {
+      "epoch": 0.00047599214612958885,
+      "grad_norm": 3.8186829684019856,
+      "learning_rate": 3.952569169960474e-08,
+      "loss": 0.6005,
+      "step": 2
+    },
+    {
+      "epoch": 0.0007139882191943833,
+      "grad_norm": 3.650677826595722,
+      "learning_rate": 7.905138339920948e-08,
+      "loss": 0.5807,
+      "step": 3
+    },
+    {
+      "epoch": 0.0009519842922591777,
+      "grad_norm": 3.7298913667173106,
+      "learning_rate": 1.1857707509881423e-07,
+      "loss": 0.7382,
+      "step": 4
+    },
+    {
+      "epoch": 0.0011899803653239722,
+      "grad_norm": 5.228820067839456,
+      "learning_rate": 1.5810276679841897e-07,
+      "loss": 0.6882,
+      "step": 5
+    },
+    {
+      "epoch": 0.0014279764383887665,
+      "grad_norm": 3.4388897772378844,
+      "learning_rate": 1.9762845849802374e-07,
+      "loss": 0.5336,
+      "step": 6
+    },
+    {
+      "epoch": 0.001665972511453561,
+      "grad_norm": 3.788288992632434,
+      "learning_rate": 2.3715415019762845e-07,
+      "loss": 0.6794,
+      "step": 7
+    },
+    {
+      "epoch": 0.0019039685845183554,
+      "grad_norm": 4.769198995741846,
+      "learning_rate": 2.766798418972332e-07,
+      "loss": 0.6865,
+      "step": 8
+    },
+    {
+      "epoch": 0.00214196465758315,
+      "grad_norm": 4.131645314597065,
+      "learning_rate": 3.1620553359683794e-07,
+      "loss": 0.6119,
+      "step": 9
+    },
+    {
+      "epoch": 0.0023799607306479445,
+      "grad_norm": 3.453158794881233,
+      "learning_rate": 3.5573122529644276e-07,
+      "loss": 0.5578,
+      "step": 10
+    },
+    {
+      "epoch": 0.0026179568037127386,
+      "grad_norm": 4.116364371771232,
+      "learning_rate": 3.9525691699604747e-07,
+      "loss": 0.6975,
+      "step": 11
+    },
+    {
+      "epoch": 0.002855952876777533,
+      "grad_norm": 4.918270216385396,
+      "learning_rate": 4.347826086956522e-07,
+      "loss": 0.6732,
+      "step": 12
+    },
+    {
+      "epoch": 0.0030939489498423276,
+      "grad_norm": 3.817650403686872,
+      "learning_rate": 4.743083003952569e-07,
+      "loss": 0.6094,
+      "step": 13
+    },
+    {
+      "epoch": 0.003331945022907122,
+      "grad_norm": 3.369346282433572,
+      "learning_rate": 5.138339920948617e-07,
+      "loss": 0.6053,
+      "step": 14
+    },
+    {
+      "epoch": 0.0035699410959719163,
+      "grad_norm": 3.988126056910898,
+      "learning_rate": 5.533596837944664e-07,
+      "loss": 0.7204,
+      "step": 15
+    },
+    {
+      "epoch": 0.003807937169036711,
+      "grad_norm": 3.8112595042647035,
+      "learning_rate": 5.928853754940712e-07,
+      "loss": 0.5855,
+      "step": 16
+    },
+    {
+      "epoch": 0.004045933242101505,
+      "grad_norm": 2.48224547819952,
+      "learning_rate": 6.324110671936759e-07,
+      "loss": 0.5257,
+      "step": 17
+    },
+    {
+      "epoch": 0.0042839293151663,
+      "grad_norm": 2.733961457300926,
+      "learning_rate": 6.719367588932807e-07,
+      "loss": 0.6612,
+      "step": 18
+    },
+    {
+      "epoch": 0.004521925388231094,
+      "grad_norm": 3.2790957516477235,
+      "learning_rate": 7.114624505928855e-07,
+      "loss": 0.6498,
+      "step": 19
+    },
+    {
+      "epoch": 0.004759921461295889,
+      "grad_norm": 2.5124910550047352,
+      "learning_rate": 7.509881422924902e-07,
+      "loss": 0.5692,
+      "step": 20
+    },
+    {
+      "epoch": 0.0049979175343606835,
+      "grad_norm": 2.659608784971935,
+      "learning_rate": 7.905138339920949e-07,
+      "loss": 0.5649,
+      "step": 21
+    },
+    {
+      "epoch": 0.005235913607425477,
+      "grad_norm": 2.198967450226314,
+      "learning_rate": 8.300395256916997e-07,
+      "loss": 0.6861,
+      "step": 22
+    },
+    {
+      "epoch": 0.005473909680490272,
+      "grad_norm": 1.8298179674027764,
+      "learning_rate": 8.695652173913044e-07,
+      "loss": 0.5675,
+      "step": 23
+    },
+    {
+      "epoch": 0.005711905753555066,
+      "grad_norm": 1.696370720789889,
+      "learning_rate": 9.090909090909091e-07,
+      "loss": 0.5139,
+      "step": 24
+    },
+    {
+      "epoch": 0.005949901826619861,
+      "grad_norm": 1.8785219368535397,
+      "learning_rate": 9.486166007905138e-07,
+      "loss": 0.6436,
+      "step": 25
+    },
+    {
+      "epoch": 0.006187897899684655,
+      "grad_norm": 2.08357255427808,
+      "learning_rate": 9.881422924901187e-07,
+      "loss": 0.6934,
+      "step": 26
+    },
+    {
+      "epoch": 0.00642589397274945,
+      "grad_norm": 1.6150355998820158,
+      "learning_rate": 1.0276679841897233e-06,
+      "loss": 0.5594,
+      "step": 27
+    },
+    {
+      "epoch": 0.006663890045814244,
+      "grad_norm": 1.5788592814751126,
+      "learning_rate": 1.067193675889328e-06,
+      "loss": 0.5212,
+      "step": 28
+    },
+    {
+      "epoch": 0.006901886118879039,
+      "grad_norm": 1.7395751013863352,
+      "learning_rate": 1.1067193675889329e-06,
+      "loss": 0.6369,
+      "step": 29
+    },
+    {
+      "epoch": 0.0071398821919438325,
+      "grad_norm": 1.7730246621108292,
+      "learning_rate": 1.1462450592885378e-06,
+      "loss": 0.636,
+      "step": 30
+    },
+    {
+      "epoch": 0.007377878265008627,
+      "grad_norm": 1.245992545689414,
+      "learning_rate": 1.1857707509881424e-06,
+      "loss": 0.5341,
+      "step": 31
+    },
+    {
+      "epoch": 0.007615874338073422,
+      "grad_norm": 1.259904616724805,
+      "learning_rate": 1.225296442687747e-06,
+      "loss": 0.5601,
+      "step": 32
+    },
+    {
+      "epoch": 0.007853870411138216,
+      "grad_norm": 1.427704060064298,
+      "learning_rate": 1.2648221343873517e-06,
+      "loss": 0.6261,
+      "step": 33
+    },
+    {
+      "epoch": 0.00809186648420301,
+      "grad_norm": 1.4266849280066096,
+      "learning_rate": 1.3043478260869566e-06,
+      "loss": 0.5241,
+      "step": 34
+    },
+    {
+      "epoch": 0.008329862557267805,
+      "grad_norm": 1.1492899123050067,
+      "learning_rate": 1.3438735177865615e-06,
+      "loss": 0.5351,
+      "step": 35
+    },
+    {
+      "epoch": 0.0085678586303326,
+      "grad_norm": 1.3426107399500602,
+      "learning_rate": 1.3833992094861662e-06,
+      "loss": 0.6201,
+      "step": 36
+    },
+    {
+      "epoch": 0.008805854703397394,
+      "grad_norm": 1.3943538319404358,
+      "learning_rate": 1.422924901185771e-06,
+      "loss": 0.5752,
+      "step": 37
+    },
+    {
+      "epoch": 0.009043850776462189,
+      "grad_norm": 1.2295476918388797,
+      "learning_rate": 1.4624505928853755e-06,
+      "loss": 0.5134,
+      "step": 38
+    },
+    {
+      "epoch": 0.009281846849526983,
+      "grad_norm": 1.0233852069436877,
+      "learning_rate": 1.5019762845849804e-06,
+      "loss": 0.5413,
+      "step": 39
+    },
+    {
+      "epoch": 0.009519842922591778,
+      "grad_norm": 1.095723970715187,
+      "learning_rate": 1.541501976284585e-06,
+      "loss": 0.6027,
+      "step": 40
+    },
+    {
+      "epoch": 0.009757838995656572,
+      "grad_norm": 0.9021710359927738,
+      "learning_rate": 1.5810276679841899e-06,
+      "loss": 0.5096,
+      "step": 41
+    },
+    {
+      "epoch": 0.009995835068721367,
+      "grad_norm": 0.8271645037826616,
+      "learning_rate": 1.6205533596837948e-06,
+      "loss": 0.4744,
+      "step": 42
+    },
+    {
+      "epoch": 0.01023383114178616,
+      "grad_norm": 0.9178185602286086,
+      "learning_rate": 1.6600790513833994e-06,
+      "loss": 0.5669,
+      "step": 43
+    },
+    {
+      "epoch": 0.010471827214850954,
+      "grad_norm": 0.7642553443300366,
+      "learning_rate": 1.699604743083004e-06,
+      "loss": 0.5444,
+      "step": 44
+    },
+    {
+      "epoch": 0.010709823287915749,
+      "grad_norm": 0.7527668145998719,
+      "learning_rate": 1.7391304347826088e-06,
+      "loss": 0.48,
+      "step": 45
+    },
+    {
+      "epoch": 0.010947819360980543,
+      "grad_norm": 0.7022856285565289,
+      "learning_rate": 1.7786561264822136e-06,
+      "loss": 0.5277,
+      "step": 46
+    },
+    {
+      "epoch": 0.011185815434045338,
+      "grad_norm": 0.7455577433162014,
+      "learning_rate": 1.8181818181818183e-06,
+      "loss": 0.6125,
+      "step": 47
+    },
+    {
+      "epoch": 0.011423811507110132,
+      "grad_norm": 0.7225816964027125,
+      "learning_rate": 1.8577075098814232e-06,
+      "loss": 0.4642,
+      "step": 48
+    },
+    {
+      "epoch": 0.011661807580174927,
+      "grad_norm": 0.6856376423837439,
+      "learning_rate": 1.8972332015810276e-06,
+      "loss": 0.4817,
+      "step": 49
+    },
+    {
+      "epoch": 0.011899803653239721,
+      "grad_norm": 0.7776257134551714,
+      "learning_rate": 1.9367588932806323e-06,
+      "loss": 0.5714,
+      "step": 50
+    },
+    {
+      "epoch": 0.012137799726304516,
+      "grad_norm": 0.6598276781972773,
+      "learning_rate": 1.9762845849802374e-06,
+      "loss": 0.5405,
+      "step": 51
+    },
+    {
+      "epoch": 0.01237579579936931,
+      "grad_norm": 0.7519528179240484,
+      "learning_rate": 2.015810276679842e-06,
+      "loss": 0.4796,
+      "step": 52
+    },
+    {
+      "epoch": 0.012613791872434105,
+      "grad_norm": 0.6641024419350181,
+      "learning_rate": 2.0553359683794467e-06,
+      "loss": 0.4447,
+      "step": 53
+    },
+    {
+      "epoch": 0.0128517879454989,
+      "grad_norm": 0.7974261768047928,
+      "learning_rate": 2.0948616600790518e-06,
+      "loss": 0.5726,
+      "step": 54
+    },
+    {
+      "epoch": 0.013089784018563694,
+      "grad_norm": 0.701806597200162,
+      "learning_rate": 2.134387351778656e-06,
+      "loss": 0.5425,
+      "step": 55
+    },
+    {
+      "epoch": 0.013327780091628489,
+      "grad_norm": 0.6533475855597715,
+      "learning_rate": 2.173913043478261e-06,
+      "loss": 0.4617,
+      "step": 56
+    },
+    {
+      "epoch": 0.013565776164693283,
+      "grad_norm": 0.6176303785724955,
+      "learning_rate": 2.2134387351778658e-06,
+      "loss": 0.525,
+      "step": 57
+    },
+    {
+      "epoch": 0.013803772237758078,
+      "grad_norm": 0.6092588831474826,
+      "learning_rate": 2.2529644268774704e-06,
+      "loss": 0.5486,
+      "step": 58
+    },
+    {
+      "epoch": 0.014041768310822872,
+      "grad_norm": 0.6409287405924915,
+      "learning_rate": 2.2924901185770755e-06,
+      "loss": 0.4525,
+      "step": 59
+    },
+    {
+      "epoch": 0.014279764383887665,
+      "grad_norm": 0.5448182821305082,
+      "learning_rate": 2.33201581027668e-06,
+      "loss": 0.4333,
+      "step": 60
+    },
+    {
+      "epoch": 0.01451776045695246,
+      "grad_norm": 0.5985899326268923,
+      "learning_rate": 2.371541501976285e-06,
+      "loss": 0.5382,
+      "step": 61
+    },
+    {
+      "epoch": 0.014755756530017254,
+      "grad_norm": 0.5562460345435642,
+      "learning_rate": 2.4110671936758895e-06,
+      "loss": 0.5455,
+      "step": 62
+    },
+    {
+      "epoch": 0.014993752603082049,
+      "grad_norm": 0.5762099087323772,
+      "learning_rate": 2.450592885375494e-06,
+      "loss": 0.4017,
+      "step": 63
+    },
+    {
+      "epoch": 0.015231748676146843,
+      "grad_norm": 0.5743174512558051,
+      "learning_rate": 2.4901185770750993e-06,
+      "loss": 0.4655,
+      "step": 64
+    },
+    {
+      "epoch": 0.015469744749211638,
+      "grad_norm": 0.5478944257204411,
+      "learning_rate": 2.5296442687747035e-06,
+      "loss": 0.5516,
+      "step": 65
+    },
+    {
+      "epoch": 0.015707740822276432,
+      "grad_norm": 0.5348787537296499,
+      "learning_rate": 2.5691699604743086e-06,
+      "loss": 0.4859,
+      "step": 66
+    },
+    {
+      "epoch": 0.015945736895341227,
+      "grad_norm": 0.513662426262249,
+      "learning_rate": 2.6086956521739132e-06,
+      "loss": 0.4405,
+      "step": 67
+    },
+    {
+      "epoch": 0.01618373296840602,
+      "grad_norm": 0.5477522117902747,
+      "learning_rate": 2.6482213438735183e-06,
+      "loss": 0.5469,
+      "step": 68
+    },
+    {
+      "epoch": 0.016421729041470816,
+      "grad_norm": 0.5353133875838187,
+      "learning_rate": 2.687747035573123e-06,
+      "loss": 0.5546,
+      "step": 69
+    },
+    {
+      "epoch": 0.01665972511453561,
+      "grad_norm": 0.5247379742690176,
+      "learning_rate": 2.7272727272727272e-06,
+      "loss": 0.45,
+      "step": 70
+    },
+    {
+      "epoch": 0.016897721187600405,
+      "grad_norm": 0.48632866471872,
+      "learning_rate": 2.7667984189723323e-06,
+      "loss": 0.4319,
+      "step": 71
+    },
+    {
+      "epoch": 0.0171357172606652,
+      "grad_norm": 0.5339643875178105,
+      "learning_rate": 2.806324110671937e-06,
+      "loss": 0.5189,
+      "step": 72
+    },
+    {
+      "epoch": 0.017373713333729994,
+      "grad_norm": 0.5148248669651565,
+      "learning_rate": 2.845849802371542e-06,
+      "loss": 0.5148,
+      "step": 73
+    },
+    {
+      "epoch": 0.01761170940679479,
+      "grad_norm": 0.4906003639934845,
+      "learning_rate": 2.8853754940711463e-06,
+      "loss": 0.4181,
+      "step": 74
+    },
+    {
+      "epoch": 0.017849705479859583,
+      "grad_norm": 0.5090667895636138,
+      "learning_rate": 2.924901185770751e-06,
+      "loss": 0.4989,
+      "step": 75
+    },
+    {
+      "epoch": 0.018087701552924378,
+      "grad_norm": 0.47929822302895847,
+      "learning_rate": 2.964426877470356e-06,
+      "loss": 0.51,
+      "step": 76
+    },
+    {
+      "epoch": 0.018325697625989172,
+      "grad_norm": 0.49337262495522655,
+      "learning_rate": 3.0039525691699607e-06,
+      "loss": 0.416,
+      "step": 77
+    },
+    {
+      "epoch": 0.018563693699053967,
+      "grad_norm": 0.46721565809919324,
+      "learning_rate": 3.043478260869566e-06,
+      "loss": 0.4243,
+      "step": 78
+    },
+    {
+      "epoch": 0.01880168977211876,
+      "grad_norm": 0.47787380347630104,
+      "learning_rate": 3.08300395256917e-06,
+      "loss": 0.5134,
+      "step": 79
+    },
+    {
+      "epoch": 0.019039685845183556,
+      "grad_norm": 0.4888904145322131,
+      "learning_rate": 3.1225296442687747e-06,
+      "loss": 0.4965,
+      "step": 80
+    },
+    {
+      "epoch": 0.01927768191824835,
+      "grad_norm": 0.474440273062018,
+      "learning_rate": 3.1620553359683798e-06,
+      "loss": 0.4231,
+      "step": 81
+    },
+    {
+      "epoch": 0.019515677991313145,
+      "grad_norm": 0.4917931480050054,
+      "learning_rate": 3.2015810276679844e-06,
+      "loss": 0.5044,
+      "step": 82
+    },
+    {
+      "epoch": 0.01975367406437794,
+      "grad_norm": 0.4850752880482312,
+      "learning_rate": 3.2411067193675895e-06,
+      "loss": 0.4966,
+      "step": 83
+    },
+    {
+      "epoch": 0.019991670137442734,
+      "grad_norm": 0.504493166936379,
+      "learning_rate": 3.2806324110671938e-06,
+      "loss": 0.416,
+      "step": 84
+    },
+    {
+      "epoch": 0.020229666210507525,
+      "grad_norm": 0.5145462573417079,
+      "learning_rate": 3.320158102766799e-06,
+      "loss": 0.405,
+      "step": 85
+    },
+    {
+      "epoch": 0.02046766228357232,
+      "grad_norm": 0.4950180063312114,
+      "learning_rate": 3.3596837944664035e-06,
+      "loss": 0.5179,
+      "step": 86
+    },
+    {
+      "epoch": 0.020705658356637114,
+      "grad_norm": 0.4499055758325075,
+      "learning_rate": 3.399209486166008e-06,
+      "loss": 0.5045,
+      "step": 87
+    },
+    {
+      "epoch": 0.02094365442970191,
+      "grad_norm": 0.4886696541409708,
+      "learning_rate": 3.4387351778656133e-06,
+      "loss": 0.4223,
+      "step": 88
+    },
+    {
+      "epoch": 0.021181650502766703,
+      "grad_norm": 0.47670535706005207,
+      "learning_rate": 3.4782608695652175e-06,
+      "loss": 0.4569,
+      "step": 89
+    },
+    {
+      "epoch": 0.021419646575831498,
+      "grad_norm": 0.4382725012174603,
+      "learning_rate": 3.5177865612648226e-06,
+      "loss": 0.5505,
+      "step": 90
+    },
+    {
+      "epoch": 0.021657642648896292,
+      "grad_norm": 0.47014527706939824,
+      "learning_rate": 3.5573122529644273e-06,
+      "loss": 0.4423,
+      "step": 91
+    },
+    {
+      "epoch": 0.021895638721961087,
+      "grad_norm": 0.4776076310968795,
+      "learning_rate": 3.5968379446640315e-06,
+      "loss": 0.4014,
+      "step": 92
+    },
+    {
+      "epoch": 0.02213363479502588,
+      "grad_norm": 0.48030073876619656,
+      "learning_rate": 3.6363636363636366e-06,
+      "loss": 0.483,
+      "step": 93
+    },
+    {
+      "epoch": 0.022371630868090676,
+      "grad_norm": 0.4298351648691775,
+      "learning_rate": 3.6758893280632412e-06,
+      "loss": 0.5171,
+      "step": 94
+    },
+    {
+      "epoch": 0.02260962694115547,
+      "grad_norm": 0.4368816040607989,
+      "learning_rate": 3.7154150197628463e-06,
+      "loss": 0.4107,
+      "step": 95
+    },
+    {
+      "epoch": 0.022847623014220265,
+      "grad_norm": 0.47265191527306455,
+      "learning_rate": 3.754940711462451e-06,
+      "loss": 0.4379,
+      "step": 96
+    },
+    {
+      "epoch": 0.02308561908728506,
+      "grad_norm": 0.44472115691541,
+      "learning_rate": 3.7944664031620552e-06,
+      "loss": 0.5029,
+      "step": 97
+    },
+    {
+      "epoch": 0.023323615160349854,
+      "grad_norm": 0.47516731747233143,
+      "learning_rate": 3.833992094861661e-06,
+      "loss": 0.4948,
+      "step": 98
+    },
+    {
+      "epoch": 0.02356161123341465,
+      "grad_norm": 0.4756183678548411,
+      "learning_rate": 3.8735177865612646e-06,
+      "loss": 0.4284,
+      "step": 99
+    },
+    {
+      "epoch": 0.023799607306479443,
+      "grad_norm": 0.4648909223572498,
+      "learning_rate": 3.91304347826087e-06,
+      "loss": 0.4613,
+      "step": 100
+    },
+    {
+      "epoch": 0.024037603379544237,
+      "grad_norm": 0.45494168373864374,
+      "learning_rate": 3.952569169960475e-06,
+      "loss": 0.5071,
+      "step": 101
+    },
+    {
+      "epoch": 0.024275599452609032,
+      "grad_norm": 0.47165993720085936,
+      "learning_rate": 3.992094861660079e-06,
+      "loss": 0.4297,
+      "step": 102
+    },
+    {
+      "epoch": 0.024513595525673827,
+      "grad_norm": 0.46600664566797045,
+      "learning_rate": 4.031620553359684e-06,
+      "loss": 0.4006,
+      "step": 103
+    },
+    {
+      "epoch": 0.02475159159873862,
+      "grad_norm": 0.440136381866,
+      "learning_rate": 4.071146245059289e-06,
+      "loss": 0.5229,
+      "step": 104
+    },
+    {
+      "epoch": 0.024989587671803416,
+      "grad_norm": 0.4874483108078875,
+      "learning_rate": 4.110671936758893e-06,
+      "loss": 0.4793,
+      "step": 105
+    },
+    {
+      "epoch": 0.02522758374486821,
+      "grad_norm": 0.4412999202522836,
+      "learning_rate": 4.150197628458498e-06,
+      "loss": 0.3896,
+      "step": 106
+    },
+    {
+      "epoch": 0.025465579817933005,
+      "grad_norm": 0.4843482283829763,
+      "learning_rate": 4.1897233201581036e-06,
+      "loss": 0.5004,
+      "step": 107
+    },
+    {
+      "epoch": 0.0257035758909978,
+      "grad_norm": 0.5192505494175473,
+      "learning_rate": 4.229249011857708e-06,
+      "loss": 0.4848,
+      "step": 108
+    },
+    {
+      "epoch": 0.025941571964062594,
+      "grad_norm": 0.4697882778878796,
+      "learning_rate": 4.268774703557312e-06,
+      "loss": 0.4183,
+      "step": 109
+    },
+    {
+      "epoch": 0.026179568037127388,
+      "grad_norm": 0.47243383971128333,
+      "learning_rate": 4.3083003952569175e-06,
+      "loss": 0.4165,
+      "step": 110
+    },
+    {
+      "epoch": 0.026417564110192183,
+      "grad_norm": 0.46058732243836664,
+      "learning_rate": 4.347826086956522e-06,
+      "loss": 0.4516,
+      "step": 111
+    },
+    {
+      "epoch": 0.026655560183256977,
+      "grad_norm": 0.439182011456468,
+      "learning_rate": 4.387351778656127e-06,
+      "loss": 0.4817,
+      "step": 112
+    },
+    {
+      "epoch": 0.026893556256321772,
+      "grad_norm": 0.43801194497466545,
+      "learning_rate": 4.4268774703557315e-06,
+      "loss": 0.4033,
+      "step": 113
+    },
+    {
+      "epoch": 0.027131552329386566,
+      "grad_norm": 0.47578753276213764,
+      "learning_rate": 4.466403162055336e-06,
+      "loss": 0.4751,
+      "step": 114
+    },
+    {
+      "epoch": 0.02736954840245136,
+      "grad_norm": 0.44222504076871677,
+      "learning_rate": 4.505928853754941e-06,
+      "loss": 0.5385,
+      "step": 115
+    },
+    {
+      "epoch": 0.027607544475516155,
+      "grad_norm": 0.4619468918214023,
+      "learning_rate": 4.5454545454545455e-06,
+      "loss": 0.4023,
+      "step": 116
+    },
+    {
+      "epoch": 0.02784554054858095,
+      "grad_norm": 0.4554362965127849,
+      "learning_rate": 4.584980237154151e-06,
+      "loss": 0.3981,
+      "step": 117
+    },
+    {
+      "epoch": 0.028083536621645745,
+      "grad_norm": 0.42154073719188745,
+      "learning_rate": 4.624505928853755e-06,
+      "loss": 0.4805,
+      "step": 118
+    },
+    {
+      "epoch": 0.028321532694710536,
+      "grad_norm": 0.4407938952654761,
+      "learning_rate": 4.66403162055336e-06,
+      "loss": 0.4883,
+      "step": 119
+    },
+    {
+      "epoch": 0.02855952876777533,
+      "grad_norm": 0.4533865445704307,
+      "learning_rate": 4.703557312252965e-06,
+      "loss": 0.4043,
+      "step": 120
+    },
+    {
+      "epoch": 0.028797524840840125,
+      "grad_norm": 0.4501769062330425,
+      "learning_rate": 4.74308300395257e-06,
+      "loss": 0.424,
+      "step": 121
+    },
+    {
+      "epoch": 0.02903552091390492,
+      "grad_norm": 0.4964204618463292,
+      "learning_rate": 4.782608695652174e-06,
+      "loss": 0.5057,
+      "step": 122
+    },
+    {
+      "epoch": 0.029273516986969714,
+      "grad_norm": 0.43608831339404674,
+      "learning_rate": 4.822134387351779e-06,
+      "loss": 0.4411,
+      "step": 123
+    },
+    {
+      "epoch": 0.029511513060034508,
+      "grad_norm": 0.5134189283501325,
+      "learning_rate": 4.861660079051384e-06,
+      "loss": 0.3686,
+      "step": 124
+    },
+    {
+      "epoch": 0.029749509133099303,
+      "grad_norm": 0.4454413745824674,
+      "learning_rate": 4.901185770750988e-06,
+      "loss": 0.4342,
+      "step": 125
+    },
+    {
+      "epoch": 0.029987505206164097,
+      "grad_norm": 0.4206745271958203,
+      "learning_rate": 4.940711462450593e-06,
+      "loss": 0.5246,
+      "step": 126
+    },
+    {
+      "epoch": 0.030225501279228892,
+      "grad_norm": 0.45531729242028907,
+      "learning_rate": 4.9802371541501985e-06,
+      "loss": 0.4269,
+      "step": 127
+    },
+    {
+      "epoch": 0.030463497352293686,
+      "grad_norm": 0.43018413490903146,
+      "learning_rate": 5.019762845849802e-06,
+      "loss": 0.4158,
+      "step": 128
+    },
+    {
+      "epoch": 0.03070149342535848,
+      "grad_norm": 0.4452267415803146,
+      "learning_rate": 5.059288537549407e-06,
+      "loss": 0.498,
+      "step": 129
+    },
+    {
+      "epoch": 0.030939489498423275,
+      "grad_norm": 0.44751910045276433,
+      "learning_rate": 5.0988142292490125e-06,
+      "loss": 0.4583,
+      "step": 130
+    },
+    {
+      "epoch": 0.03117748557148807,
+      "grad_norm": 0.4474828998483563,
+      "learning_rate": 5.138339920948617e-06,
+      "loss": 0.3873,
+      "step": 131
+    },
+    {
+      "epoch": 0.031415481644552865,
+      "grad_norm": 0.46141860455017775,
+      "learning_rate": 5.177865612648222e-06,
+      "loss": 0.4635,
+      "step": 132
+    },
+    {
+      "epoch": 0.03165347771761766,
+      "grad_norm": 0.4406433909527717,
+      "learning_rate": 5.2173913043478265e-06,
+      "loss": 0.5175,
+      "step": 133
+    },
+    {
+      "epoch": 0.031891473790682454,
+      "grad_norm": 0.46817970001375386,
+      "learning_rate": 5.256916996047431e-06,
+      "loss": 0.4269,
+      "step": 134
+    },
+    {
+      "epoch": 0.03212946986374725,
+      "grad_norm": 0.4582029027978222,
+      "learning_rate": 5.296442687747037e-06,
+      "loss": 0.3855,
+      "step": 135
+    },
+    {
+      "epoch": 0.03236746593681204,
+      "grad_norm": 0.4483408124347049,
+      "learning_rate": 5.335968379446641e-06,
+      "loss": 0.5035,
+      "step": 136
+    },
+    {
+      "epoch": 0.03260546200987684,
+      "grad_norm": 0.4521188947407404,
+      "learning_rate": 5.375494071146246e-06,
+      "loss": 0.4773,
+      "step": 137
+    },
+    {
+      "epoch": 0.03284345808294163,
+      "grad_norm": 0.4584775312885165,
+      "learning_rate": 5.41501976284585e-06,
+      "loss": 0.4263,
+      "step": 138
+    },
+    {
+      "epoch": 0.033081454156006426,
+      "grad_norm": 0.44388523673544455,
+      "learning_rate": 5.4545454545454545e-06,
+      "loss": 0.4206,
+      "step": 139
+    },
+    {
+      "epoch": 0.03331945022907122,
+      "grad_norm": 0.4443963263357755,
+      "learning_rate": 5.49407114624506e-06,
+      "loss": 0.5354,
+      "step": 140
+    },
+    {
+      "epoch": 0.033557446302136015,
+      "grad_norm": 0.4624952425984824,
+      "learning_rate": 5.533596837944665e-06,
+      "loss": 0.4237,
+      "step": 141
+    },
+    {
+      "epoch": 0.03379544237520081,
+      "grad_norm": 0.4372307380751775,
+      "learning_rate": 5.573122529644269e-06,
+      "loss": 0.4082,
+      "step": 142
+    },
+    {
+      "epoch": 0.034033438448265604,
+      "grad_norm": 0.4494964774973684,
+      "learning_rate": 5.612648221343874e-06,
+      "loss": 0.5033,
+      "step": 143
+    },
+    {
+      "epoch": 0.0342714345213304,
+      "grad_norm": 0.4308495087586641,
+      "learning_rate": 5.652173913043479e-06,
+      "loss": 0.4855,
+      "step": 144
+    },
+    {
+      "epoch": 0.03450943059439519,
+      "grad_norm": 0.45898397608033087,
+      "learning_rate": 5.691699604743084e-06,
+      "loss": 0.3943,
+      "step": 145
+    },
+    {
+      "epoch": 0.03474742666745999,
+      "grad_norm": 0.444667683614168,
+      "learning_rate": 5.731225296442689e-06,
+      "loss": 0.3907,
+      "step": 146
+    },
+    {
+      "epoch": 0.03498542274052478,
+      "grad_norm": 0.43190369082425845,
+      "learning_rate": 5.770750988142293e-06,
+      "loss": 0.5306,
+      "step": 147
+    },
+    {
+      "epoch": 0.03522341881358958,
+      "grad_norm": 0.47772442543782057,
+      "learning_rate": 5.810276679841897e-06,
+      "loss": 0.4328,
+      "step": 148
+    },
+    {
+      "epoch": 0.03546141488665437,
+      "grad_norm": 0.5149054927207557,
+      "learning_rate": 5.849802371541502e-06,
+      "loss": 0.4038,
+      "step": 149
+    },
+    {
+      "epoch": 0.035699410959719166,
+      "grad_norm": 0.46489298468292706,
+      "learning_rate": 5.8893280632411074e-06,
+      "loss": 0.4641,
+      "step": 150
+    },
+    {
+      "epoch": 0.03593740703278396,
+      "grad_norm": 0.43822016358847443,
+      "learning_rate": 5.928853754940712e-06,
+      "loss": 0.511,
+      "step": 151
+    },
+    {
+      "epoch": 0.036175403105848755,
+      "grad_norm": 0.4780094956775413,
+      "learning_rate": 5.968379446640317e-06,
+      "loss": 0.4091,
+      "step": 152
+    },
+    {
+      "epoch": 0.03641339917891355,
+      "grad_norm": 0.5193809604213568,
+      "learning_rate": 6.007905138339921e-06,
+      "loss": 0.3704,
+      "step": 153
+    },
+    {
+      "epoch": 0.036651395251978344,
+      "grad_norm": 0.47203611226409087,
+      "learning_rate": 6.047430830039526e-06,
+      "loss": 0.5057,
+      "step": 154
+    },
+    {
+      "epoch": 0.03688939132504314,
+      "grad_norm": 0.4255995467291753,
+      "learning_rate": 6.086956521739132e-06,
+      "loss": 0.4863,
+      "step": 155
+    },
+    {
+      "epoch": 0.03712738739810793,
+      "grad_norm": 0.4317511844695257,
+      "learning_rate": 6.126482213438736e-06,
+      "loss": 0.4005,
+      "step": 156
+    },
+    {
+      "epoch": 0.03736538347117273,
+      "grad_norm": 0.4463988044172272,
+      "learning_rate": 6.16600790513834e-06,
+      "loss": 0.4169,
+      "step": 157
+    },
+    {
+      "epoch": 0.03760337954423752,
+      "grad_norm": 0.4301302395556984,
+      "learning_rate": 6.205533596837945e-06,
+      "loss": 0.4749,
+      "step": 158
+    },
+    {
+      "epoch": 0.03784137561730232,
+      "grad_norm": 0.4885006908777435,
+      "learning_rate": 6.245059288537549e-06,
+      "loss": 0.4176,
+      "step": 159
+    },
+    {
+      "epoch": 0.03807937169036711,
+      "grad_norm": 0.490807059909575,
+      "learning_rate": 6.284584980237155e-06,
+      "loss": 0.3878,
+      "step": 160
+    },
+    {
+      "epoch": 0.038317367763431906,
+      "grad_norm": 0.43132409383451087,
+      "learning_rate": 6.3241106719367596e-06,
+      "loss": 0.4834,
+      "step": 161
+    },
+    {
+      "epoch": 0.0385553638364967,
+      "grad_norm": 0.44865911014332444,
+      "learning_rate": 6.363636363636364e-06,
+      "loss": 0.4821,
+      "step": 162
+    },
+    {
+      "epoch": 0.038793359909561495,
+      "grad_norm": 0.4260431090618812,
+      "learning_rate": 6.403162055335969e-06,
+      "loss": 0.4174,
+      "step": 163
+    },
+    {
+      "epoch": 0.03903135598262629,
+      "grad_norm": 0.49901659393595554,
+      "learning_rate": 6.442687747035574e-06,
+      "loss": 0.4296,
+      "step": 164
+    },
+    {
+      "epoch": 0.039269352055691084,
+      "grad_norm": 0.4897547343200327,
+      "learning_rate": 6.482213438735179e-06,
+      "loss": 0.512,
+      "step": 165
+    },
+    {
+      "epoch": 0.03950734812875588,
+      "grad_norm": 0.4813236527289336,
+      "learning_rate": 6.521739130434783e-06,
+      "loss": 0.431,
+      "step": 166
+    },
+    {
+      "epoch": 0.03974534420182067,
+      "grad_norm": 0.45927226014349076,
+      "learning_rate": 6.5612648221343875e-06,
+      "loss": 0.3718,
+      "step": 167
+    },
+    {
+      "epoch": 0.03998334027488547,
+      "grad_norm": 0.4965136015728435,
+      "learning_rate": 6.600790513833992e-06,
+      "loss": 0.4695,
+      "step": 168
+    },
+    {
+      "epoch": 0.04022133634795026,
+      "grad_norm": 0.4683557415256707,
+      "learning_rate": 6.640316205533598e-06,
+      "loss": 0.4778,
+      "step": 169
+    },
+    {
+      "epoch": 0.04045933242101505,
+      "grad_norm": 0.4634555253102173,
+      "learning_rate": 6.679841897233202e-06,
+      "loss": 0.3995,
+      "step": 170
+    },
+    {
+      "epoch": 0.040697328494079844,
+      "grad_norm": 0.48973289383270246,
+      "learning_rate": 6.719367588932807e-06,
+      "loss": 0.3976,
+      "step": 171
+    },
+    {
+      "epoch": 0.04093532456714464,
+      "grad_norm": 0.42723303350236147,
+      "learning_rate": 6.758893280632412e-06,
+      "loss": 0.494,
+      "step": 172
+    },
+    {
+      "epoch": 0.04117332064020943,
+      "grad_norm": 0.7587112192005009,
+      "learning_rate": 6.798418972332016e-06,
+      "loss": 0.3958,
+      "step": 173
+    },
+    {
+      "epoch": 0.04141131671327423,
+      "grad_norm": 0.4915235015843637,
+      "learning_rate": 6.837944664031622e-06,
+      "loss": 0.3719,
+      "step": 174
+    },
+    {
+      "epoch": 0.04164931278633902,
+      "grad_norm": 0.48028138363352874,
+      "learning_rate": 6.8774703557312265e-06,
+      "loss": 0.4144,
+      "step": 175
+    },
+    {
+      "epoch": 0.04188730885940382,
+      "grad_norm": 0.43339371237253094,
+      "learning_rate": 6.91699604743083e-06,
+      "loss": 0.523,
+      "step": 176
+    },
+    {
+      "epoch": 0.04212530493246861,
+      "grad_norm": 0.45447194682552483,
+      "learning_rate": 6.956521739130435e-06,
+      "loss": 0.3567,
+      "step": 177
+    },
+    {
+      "epoch": 0.042363301005533406,
+      "grad_norm": 0.4552581773897187,
+      "learning_rate": 6.99604743083004e-06,
+      "loss": 0.3722,
+      "step": 178
+    },
+    {
+      "epoch": 0.0426012970785982,
+      "grad_norm": 0.4796620850546896,
+      "learning_rate": 7.035573122529645e-06,
+      "loss": 0.4874,
+      "step": 179
+    },
+    {
+      "epoch": 0.042839293151662995,
+      "grad_norm": 0.43749845312110847,
+      "learning_rate": 7.07509881422925e-06,
+      "loss": 0.4922,
+      "step": 180
+    },
+    {
+      "epoch": 0.04307728922472779,
+      "grad_norm": 0.46789960493230326,
+      "learning_rate": 7.1146245059288545e-06,
+      "loss": 0.383,
+      "step": 181
+    },
+    {
+      "epoch": 0.043315285297792584,
+      "grad_norm": 0.5224298682432643,
+      "learning_rate": 7.154150197628459e-06,
+      "loss": 0.4526,
+      "step": 182
+    },
+    {
+      "epoch": 0.04355328137085738,
+      "grad_norm": 0.4746636793562093,
+      "learning_rate": 7.193675889328063e-06,
+      "loss": 0.4934,
+      "step": 183
+    },
+    {
+      "epoch": 0.04379127744392217,
+      "grad_norm": 0.4701682318036117,
+      "learning_rate": 7.233201581027669e-06,
+      "loss": 0.4034,
+      "step": 184
+    },
+    {
+      "epoch": 0.04402927351698697,
+      "grad_norm": 0.5015499630814017,
+      "learning_rate": 7.272727272727273e-06,
+      "loss": 0.3577,
+      "step": 185
+    },
+    {
+      "epoch": 0.04426726959005176,
+      "grad_norm": 0.447979905112201,
+      "learning_rate": 7.312252964426878e-06,
+      "loss": 0.4454,
+      "step": 186
+    },
+    {
+      "epoch": 0.04450526566311656,
+      "grad_norm": 0.4360535244392955,
+      "learning_rate": 7.3517786561264825e-06,
+      "loss": 0.4708,
+      "step": 187
+    },
+    {
+      "epoch": 0.04474326173618135,
+      "grad_norm": 0.4906094221266247,
+      "learning_rate": 7.391304347826087e-06,
+      "loss": 0.3734,
+      "step": 188
+    },
+    {
+      "epoch": 0.044981257809246146,
+      "grad_norm": 0.47012438815001395,
+      "learning_rate": 7.430830039525693e-06,
+      "loss": 0.4194,
+      "step": 189
+    },
+    {
+      "epoch": 0.04521925388231094,
+      "grad_norm": 0.4668496482235658,
+      "learning_rate": 7.470355731225297e-06,
+      "loss": 0.4804,
+      "step": 190
+    },
+    {
+      "epoch": 0.045457249955375735,
+      "grad_norm": 0.43523540977913394,
+      "learning_rate": 7.509881422924902e-06,
+      "loss": 0.4067,
+      "step": 191
+    },
+    {
+      "epoch": 0.04569524602844053,
+      "grad_norm": 0.4388184201414626,
+      "learning_rate": 7.549407114624507e-06,
+      "loss": 0.405,
+      "step": 192
+    },
+    {
+      "epoch": 0.045933242101505324,
+      "grad_norm": 0.4710177222648498,
+      "learning_rate": 7.5889328063241105e-06,
+      "loss": 0.4461,
+      "step": 193
+    },
+    {
+      "epoch": 0.04617123817457012,
+      "grad_norm": 0.45506913551818734,
+      "learning_rate": 7.628458498023717e-06,
+      "loss": 0.5013,
+      "step": 194
+    },
+    {
+      "epoch": 0.04640923424763491,
+      "grad_norm": 0.4503230582686603,
+      "learning_rate": 7.667984189723321e-06,
+      "loss": 0.3812,
+      "step": 195
+    },
+    {
+      "epoch": 0.04664723032069971,
+      "grad_norm": 0.4900687410351956,
+      "learning_rate": 7.707509881422925e-06,
+      "loss": 0.3929,
+      "step": 196
+    },
+    {
+      "epoch": 0.0468852263937645,
+      "grad_norm": 0.4422937878413219,
+      "learning_rate": 7.747035573122529e-06,
+      "loss": 0.4826,
+      "step": 197
+    },
+    {
+      "epoch": 0.0471232224668293,
+      "grad_norm": 0.4696233806991265,
+      "learning_rate": 7.786561264822135e-06,
+      "loss": 0.4321,
+      "step": 198
+    },
+    {
+      "epoch": 0.04736121853989409,
+      "grad_norm": 0.5036764448381432,
+      "learning_rate": 7.82608695652174e-06,
+      "loss": 0.3737,
+      "step": 199
+    },
+    {
+      "epoch": 0.047599214612958886,
+      "grad_norm": 0.4460954367231951,
+      "learning_rate": 7.865612648221344e-06,
+      "loss": 0.4502,
+      "step": 200
+    },
+    {
+      "epoch": 0.04783721068602368,
+      "grad_norm": 0.4121996924774416,
+      "learning_rate": 7.90513833992095e-06,
+      "loss": 0.4872,
+      "step": 201
+    },
+    {
+      "epoch": 0.048075206759088475,
+      "grad_norm": 0.5377437642053337,
+      "learning_rate": 7.944664031620553e-06,
+      "loss": 0.3756,
+      "step": 202
+    },
+    {
+      "epoch": 0.04831320283215327,
+      "grad_norm": 0.4869568221691017,
+      "learning_rate": 7.984189723320159e-06,
+      "loss": 0.3994,
+      "step": 203
+    },
+    {
+      "epoch": 0.048551198905218064,
+      "grad_norm": 0.45051544052407894,
+      "learning_rate": 8.023715415019764e-06,
+      "loss": 0.4758,
+      "step": 204
+    },
+    {
+      "epoch": 0.04878919497828286,
+      "grad_norm": 0.46693597119879243,
+      "learning_rate": 8.063241106719368e-06,
+      "loss": 0.4724,
+      "step": 205
+    },
+    {
+      "epoch": 0.04902719105134765,
+      "grad_norm": 0.4434729096123986,
+      "learning_rate": 8.102766798418974e-06,
+      "loss": 0.3654,
+      "step": 206
+    },
+    {
+      "epoch": 0.04926518712441245,
+      "grad_norm": 0.4516410800019458,
+      "learning_rate": 8.142292490118577e-06,
+      "loss": 0.4328,
+      "step": 207
+    },
+    {
+      "epoch": 0.04950318319747724,
+      "grad_norm": 0.4092851533761814,
+      "learning_rate": 8.181818181818183e-06,
+      "loss": 0.4973,
+      "step": 208
+    },
+    {
+      "epoch": 0.04974117927054204,
+      "grad_norm": 0.4458979348243189,
+      "learning_rate": 8.221343873517787e-06,
+      "loss": 0.3875,
+      "step": 209
+    },
+    {
+      "epoch": 0.04997917534360683,
+      "grad_norm": 0.4575088617794585,
+      "learning_rate": 8.260869565217392e-06,
+      "loss": 0.3812,
+      "step": 210
+    },
+    {
+      "epoch": 0.050217171416671626,
+      "grad_norm": 0.42998279360778746,
+      "learning_rate": 8.300395256916996e-06,
+      "loss": 0.4689,
+      "step": 211
+    },
+    {
+      "epoch": 0.05045516748973642,
+      "grad_norm": 0.4664154763779984,
+      "learning_rate": 8.339920948616602e-06,
+      "loss": 0.4628,
+      "step": 212
+    },
+    {
+      "epoch": 0.050693163562801215,
+      "grad_norm": 0.48372522149295644,
+      "learning_rate": 8.379446640316207e-06,
+      "loss": 0.3713,
+      "step": 213
+    },
+    {
+      "epoch": 0.05093115963586601,
+      "grad_norm": 0.45614467932799935,
+      "learning_rate": 8.418972332015811e-06,
+      "loss": 0.4565,
+      "step": 214
+    },
+    {
+      "epoch": 0.051169155708930804,
+      "grad_norm": 0.4521533500555388,
+      "learning_rate": 8.458498023715416e-06,
+      "loss": 0.4939,
+      "step": 215
+    },
+    {
+      "epoch": 0.0514071517819956,
+      "grad_norm": 0.44133282573284666,
+      "learning_rate": 8.49802371541502e-06,
+      "loss": 0.4448,
+      "step": 216
+    },
+    {
+      "epoch": 0.05164514785506039,
+      "grad_norm": 0.4461748694701973,
+      "learning_rate": 8.537549407114624e-06,
+      "loss": 0.3685,
+      "step": 217
+    },
+    {
+      "epoch": 0.05188314392812519,
+      "grad_norm": 0.43361066733503273,
+      "learning_rate": 8.57707509881423e-06,
+      "loss": 0.4443,
+      "step": 218
+    },
+    {
+      "epoch": 0.05212114000118998,
+      "grad_norm": 0.44192635288128085,
+      "learning_rate": 8.616600790513835e-06,
+      "loss": 0.4858,
+      "step": 219
+    },
+    {
+      "epoch": 0.052359136074254777,
+      "grad_norm": 0.4765904637426461,
+      "learning_rate": 8.656126482213439e-06,
+      "loss": 0.37,
+      "step": 220
+    },
+    {
+      "epoch": 0.05259713214731957,
+      "grad_norm": 0.42950301730198376,
+      "learning_rate": 8.695652173913044e-06,
+      "loss": 0.4087,
+      "step": 221
+    },
+    {
+      "epoch": 0.052835128220384366,
+      "grad_norm": 0.4493355585707809,
+      "learning_rate": 8.735177865612648e-06,
+      "loss": 0.4808,
+      "step": 222
+    },
+    {
+      "epoch": 0.05307312429344916,
+      "grad_norm": 0.4473843798566356,
+      "learning_rate": 8.774703557312254e-06,
+      "loss": 0.4421,
+      "step": 223
+    },
+    {
+      "epoch": 0.053311120366513955,
+      "grad_norm": 0.4662717413273855,
+      "learning_rate": 8.81422924901186e-06,
+      "loss": 0.3826,
+      "step": 224
+    },
+    {
+      "epoch": 0.05354911643957875,
+      "grad_norm": 0.4753294327238269,
+      "learning_rate": 8.853754940711463e-06,
+      "loss": 0.435,
+      "step": 225
+    },
+    {
+      "epoch": 0.053787112512643544,
+      "grad_norm": 0.43479966227192995,
+      "learning_rate": 8.893280632411067e-06,
+      "loss": 0.4577,
+      "step": 226
+    },
+    {
+      "epoch": 0.05402510858570834,
+      "grad_norm": 0.4540970891602704,
+      "learning_rate": 8.932806324110672e-06,
+      "loss": 0.3734,
+      "step": 227
+    },
+    {
+      "epoch": 0.05426310465877313,
+      "grad_norm": 0.47014652353944864,
+      "learning_rate": 8.972332015810278e-06,
+      "loss": 0.4018,
+      "step": 228
+    },
+    {
+      "epoch": 0.05450110073183793,
+      "grad_norm": 0.4588196915897715,
+      "learning_rate": 9.011857707509882e-06,
+      "loss": 0.4821,
+      "step": 229
+    },
+    {
+      "epoch": 0.05473909680490272,
+      "grad_norm": 0.4290397451443549,
+      "learning_rate": 9.051383399209487e-06,
+      "loss": 0.4443,
+      "step": 230
+    },
+    {
+      "epoch": 0.054977092877967516,
+      "grad_norm": 0.4634653667823266,
+      "learning_rate": 9.090909090909091e-06,
+      "loss": 0.3622,
+      "step": 231
+    },
+    {
+      "epoch": 0.05521508895103231,
+      "grad_norm": 0.4799790289966676,
+      "learning_rate": 9.130434782608697e-06,
+      "loss": 0.4462,
+      "step": 232
+    },
+    {
+      "epoch": 0.055453085024097105,
+      "grad_norm": 0.4480981798171716,
+      "learning_rate": 9.169960474308302e-06,
+      "loss": 0.4328,
+      "step": 233
+    },
+    {
+      "epoch": 0.0556910810971619,
+      "grad_norm": 0.46767604607383606,
+      "learning_rate": 9.209486166007906e-06,
+      "loss": 0.412,
+      "step": 234
+    },
+    {
+      "epoch": 0.055929077170226695,
+      "grad_norm": 0.5333947256403828,
+      "learning_rate": 9.24901185770751e-06,
+      "loss": 0.3851,
+      "step": 235
+    },
+    {
+      "epoch": 0.05616707324329149,
+      "grad_norm": 0.44207186013928046,
+      "learning_rate": 9.288537549407115e-06,
+      "loss": 0.4775,
+      "step": 236
+    },
+    {
+      "epoch": 0.056405069316356284,
+      "grad_norm": 0.4579089731862815,
+      "learning_rate": 9.32806324110672e-06,
+      "loss": 0.4796,
+      "step": 237
+    },
+    {
+      "epoch": 0.05664306538942107,
+      "grad_norm": 0.4916045682404132,
+      "learning_rate": 9.367588932806325e-06,
+      "loss": 0.3495,
+      "step": 238
+    },
+    {
+      "epoch": 0.056881061462485866,
+      "grad_norm": 0.48592356270000286,
+      "learning_rate": 9.40711462450593e-06,
+      "loss": 0.3762,
+      "step": 239
+    },
+    {
+      "epoch": 0.05711905753555066,
+      "grad_norm": 0.40760679787076876,
+      "learning_rate": 9.446640316205534e-06,
+      "loss": 0.4946,
+      "step": 240
+    },
+    {
+      "epoch": 0.057357053608615455,
+      "grad_norm": 0.5404446805301054,
+      "learning_rate": 9.48616600790514e-06,
+      "loss": 0.3878,
+      "step": 241
+    },
+    {
+      "epoch": 0.05759504968168025,
+      "grad_norm": 0.5023864741183144,
+      "learning_rate": 9.525691699604745e-06,
+      "loss": 0.3769,
+      "step": 242
+    },
+    {
+      "epoch": 0.057833045754745044,
+      "grad_norm": 0.4291166201410665,
+      "learning_rate": 9.565217391304349e-06,
+      "loss": 0.4405,
+      "step": 243
+    },
+    {
+      "epoch": 0.05807104182780984,
+      "grad_norm": 0.4370205569354994,
+      "learning_rate": 9.604743083003954e-06,
+      "loss": 0.4543,
+      "step": 244
+    },
+    {
+      "epoch": 0.05830903790087463,
+      "grad_norm": 0.4726343079593376,
+      "learning_rate": 9.644268774703558e-06,
+      "loss": 0.3977,
+      "step": 245
+    },
+    {
+      "epoch": 0.05854703397393943,
+      "grad_norm": 0.48561129532487857,
+      "learning_rate": 9.683794466403162e-06,
+      "loss": 0.3915,
+      "step": 246
+    },
+    {
+      "epoch": 0.05878503004700422,
+      "grad_norm": 0.4757606568781063,
+      "learning_rate": 9.723320158102767e-06,
+      "loss": 0.4472,
+      "step": 247
+    },
+    {
+      "epoch": 0.059023026120069016,
+      "grad_norm": 0.515327431255371,
+      "learning_rate": 9.762845849802373e-06,
+      "loss": 0.4027,
+      "step": 248
+    },
+    {
+      "epoch": 0.05926102219313381,
+      "grad_norm": 0.4762224245393142,
+      "learning_rate": 9.802371541501977e-06,
+      "loss": 0.3695,
+      "step": 249
+    },
+    {
+      "epoch": 0.059499018266198606,
+      "grad_norm": 0.4459966492940902,
+      "learning_rate": 9.841897233201582e-06,
+      "loss": 0.4651,
+      "step": 250
+    },
+    {
+      "epoch": 0.0597370143392634,
+      "grad_norm": 0.4460885254332536,
+      "learning_rate": 9.881422924901186e-06,
+      "loss": 0.4789,
+      "step": 251
+    },
+    {
+      "epoch": 0.059975010412328195,
+      "grad_norm": 0.5301749599674928,
+      "learning_rate": 9.920948616600791e-06,
+      "loss": 0.3894,
+      "step": 252
+    },
+    {
+      "epoch": 0.06021300648539299,
+      "grad_norm": 0.5074601139708311,
+      "learning_rate": 9.960474308300397e-06,
+      "loss": 0.379,
+      "step": 253
+    },
+    {
+      "epoch": 0.060451002558457784,
+      "grad_norm": 0.46943388848534723,
+      "learning_rate": 1e-05,
+      "loss": 0.4968,
+      "step": 254
+    },
+    {
+      "epoch": 0.06068899863152258,
+      "grad_norm": 0.4849058663631589,
+      "learning_rate": 9.999999628438155e-06,
+      "loss": 0.4239,
+      "step": 255
+    },
+    {
+      "epoch": 0.06092699470458737,
+      "grad_norm": 0.5159279619948409,
+      "learning_rate": 9.999998513752668e-06,
+      "loss": 0.3897,
+      "step": 256
+    },
+    {
+      "epoch": 0.06116499077765217,
+      "grad_norm": 0.4913799031325082,
+      "learning_rate": 9.999996655943708e-06,
+      "loss": 0.4329,
+      "step": 257
+    },
+    {
+      "epoch": 0.06140298685071696,
+      "grad_norm": 0.4609134555633486,
+      "learning_rate": 9.999994055011552e-06,
+      "loss": 0.4635,
+      "step": 258
+    },
+    {
+      "epoch": 0.061640982923781756,
+      "grad_norm": 0.4600665037690504,
+      "learning_rate": 9.999990710956586e-06,
+      "loss": 0.361,
+      "step": 259
+    },
+    {
+      "epoch": 0.06187897899684655,
+      "grad_norm": 0.4893892245206209,
+      "learning_rate": 9.999986623779307e-06,
+      "loss": 0.3929,
+      "step": 260
+    },
+    {
+      "epoch": 0.062116975069911345,
+      "grad_norm": 0.47317754883053104,
+      "learning_rate": 9.99998179348032e-06,
+      "loss": 0.4411,
+      "step": 261
+    },
+    {
+      "epoch": 0.06235497114297614,
+      "grad_norm": 0.46027884774340916,
+      "learning_rate": 9.999976220060347e-06,
+      "loss": 0.4357,
+      "step": 262
+    },
+    {
+      "epoch": 0.06259296721604093,
+      "grad_norm": 0.5110985377219065,
+      "learning_rate": 9.999969903520212e-06,
+      "loss": 0.368,
+      "step": 263
+    },
+    {
+      "epoch": 0.06283096328910573,
+      "grad_norm": 0.5424172208623367,
+      "learning_rate": 9.999962843860858e-06,
+      "loss": 0.3945,
+      "step": 264
+    },
+    {
+      "epoch": 0.06306895936217052,
+      "grad_norm": 0.4455661724192156,
+      "learning_rate": 9.999955041083332e-06,
+      "loss": 0.4574,
+      "step": 265
+    },
+    {
+      "epoch": 0.06330695543523532,
+      "grad_norm": 0.4967774114157658,
+      "learning_rate": 9.999946495188793e-06,
+      "loss": 0.4139,
+      "step": 266
+    },
+    {
+      "epoch": 0.06354495150830011,
+      "grad_norm": 0.48548245385801053,
+      "learning_rate": 9.999937206178512e-06,
+      "loss": 0.369,
+      "step": 267
+    },
+    {
+      "epoch": 0.06378294758136491,
+      "grad_norm": 0.44764818752534175,
+      "learning_rate": 9.999927174053872e-06,
+      "loss": 0.4449,
+      "step": 268
+    },
+    {
+      "epoch": 0.0640209436544297,
+      "grad_norm": 0.47571255336778956,
+      "learning_rate": 9.999916398816359e-06,
+      "loss": 0.4252,
+      "step": 269
+    },
+    {
+      "epoch": 0.0642589397274945,
+      "grad_norm": 0.5176454688899305,
+      "learning_rate": 9.999904880467579e-06,
+      "loss": 0.4172,
+      "step": 270
+    },
+    {
+      "epoch": 0.06449693580055929,
+      "grad_norm": 0.47797256428291063,
+      "learning_rate": 9.99989261900924e-06,
+      "loss": 0.4234,
+      "step": 271
+    },
+    {
+      "epoch": 0.06473493187362409,
+      "grad_norm": 0.45140623928149853,
+      "learning_rate": 9.999879614443168e-06,
+      "loss": 0.4936,
+      "step": 272
+    },
+    {
+      "epoch": 0.06497292794668888,
+      "grad_norm": 0.4619192022175719,
+      "learning_rate": 9.999865866771295e-06,
+      "loss": 0.472,
+      "step": 273
+    },
+    {
+      "epoch": 0.06521092401975367,
+      "grad_norm": 0.5262244410688721,
+      "learning_rate": 9.999851375995662e-06,
+      "loss": 0.3445,
+      "step": 274
+    },
+    {
+      "epoch": 0.06544892009281847,
+      "grad_norm": 0.4520208956154366,
+      "learning_rate": 9.999836142118424e-06,
+      "loss": 0.4155,
+      "step": 275
+    },
+    {
+      "epoch": 0.06568691616588326,
+      "grad_norm": 0.46440432038472335,
+      "learning_rate": 9.999820165141845e-06,
+      "loss": 0.4586,
+      "step": 276
+    },
+    {
+      "epoch": 0.06592491223894806,
+      "grad_norm": 0.4573062726758063,
+      "learning_rate": 9.9998034450683e-06,
+      "loss": 0.3835,
+      "step": 277
+    },
+    {
+      "epoch": 0.06616290831201285,
+      "grad_norm": 0.4475402421191899,
+      "learning_rate": 9.999785981900277e-06,
+      "loss": 0.3686,
+      "step": 278
+    },
+    {
+      "epoch": 0.06640090438507765,
+      "grad_norm": 0.4696018693236434,
+      "learning_rate": 9.999767775640364e-06,
+      "loss": 0.4623,
+      "step": 279
+    },
+    {
+      "epoch": 0.06663890045814244,
+      "grad_norm": 0.4618737104284479,
+      "learning_rate": 9.999748826291273e-06,
+      "loss": 0.4177,
+      "step": 280
+    },
+    {
+      "epoch": 0.06687689653120724,
+      "grad_norm": 0.4799653750371601,
+      "learning_rate": 9.99972913385582e-06,
+      "loss": 0.3504,
+      "step": 281
+    },
+    {
+      "epoch": 0.06711489260427203,
+      "grad_norm": 0.4497108981102174,
+      "learning_rate": 9.999708698336929e-06,
+      "loss": 0.4563,
+      "step": 282
+    },
+    {
+      "epoch": 0.06735288867733683,
+      "grad_norm": 0.44395236286884626,
+      "learning_rate": 9.999687519737639e-06,
+      "loss": 0.4782,
+      "step": 283
+    },
+    {
+      "epoch": 0.06759088475040162,
+      "grad_norm": 0.6091422041147246,
+      "learning_rate": 9.999665598061097e-06,
+      "loss": 0.3892,
+      "step": 284
+    },
+    {
+      "epoch": 0.06782888082346641,
+      "grad_norm": 0.49705585154406684,
+      "learning_rate": 9.999642933310561e-06,
+      "loss": 0.3826,
+      "step": 285
+    },
+    {
+      "epoch": 0.06806687689653121,
+      "grad_norm": 0.4568270789179068,
+      "learning_rate": 9.9996195254894e-06,
+      "loss": 0.4423,
+      "step": 286
+    },
+    {
+      "epoch": 0.068304872969596,
+      "grad_norm": 0.45526888232059454,
+      "learning_rate": 9.999595374601093e-06,
+      "loss": 0.4749,
+      "step": 287
+    },
+    {
+      "epoch": 0.0685428690426608,
+      "grad_norm": 0.46039335829876826,
+      "learning_rate": 9.99957048064923e-06,
+      "loss": 0.3574,
+      "step": 288
+    },
+    {
+      "epoch": 0.06878086511572559,
+      "grad_norm": 0.4484215237280295,
+      "learning_rate": 9.999544843637509e-06,
+      "loss": 0.4244,
+      "step": 289
+    },
+    {
+      "epoch": 0.06901886118879039,
+      "grad_norm": 0.4156896188113592,
+      "learning_rate": 9.999518463569742e-06,
+      "loss": 0.4771,
+      "step": 290
+    },
+    {
+      "epoch": 0.06925685726185518,
+      "grad_norm": 0.4207781838747178,
+      "learning_rate": 9.99949134044985e-06,
+      "loss": 0.3879,
+      "step": 291
+    },
+    {
+      "epoch": 0.06949485333491998,
+      "grad_norm": 0.45221148207375844,
+      "learning_rate": 9.999463474281862e-06,
+      "loss": 0.403,
+      "step": 292
+    },
+    {
+      "epoch": 0.06973284940798477,
+      "grad_norm": 0.4377876809771397,
+      "learning_rate": 9.999434865069922e-06,
+      "loss": 0.4193,
+      "step": 293
+    },
+    {
+      "epoch": 0.06997084548104957,
+      "grad_norm": 0.4024216767935093,
+      "learning_rate": 9.99940551281828e-06,
+      "loss": 0.4875,
+      "step": 294
+    },
+    {
+      "epoch": 0.07020884155411436,
+      "grad_norm": 0.43322473693066027,
+      "learning_rate": 9.999375417531301e-06,
+      "loss": 0.3583,
+      "step": 295
+    },
+    {
+      "epoch": 0.07044683762717915,
+      "grad_norm": 0.471599149075093,
+      "learning_rate": 9.999344579213455e-06,
+      "loss": 0.42,
+      "step": 296
+    },
+    {
+      "epoch": 0.07068483370024395,
+      "grad_norm": 0.44747663908303237,
+      "learning_rate": 9.999312997869326e-06,
+      "loss": 0.503,
+      "step": 297
+    },
+    {
+      "epoch": 0.07092282977330874,
+      "grad_norm": 0.5069193101593006,
+      "learning_rate": 9.99928067350361e-06,
+      "loss": 0.4093,
+      "step": 298
+    },
+    {
+      "epoch": 0.07116082584637354,
+      "grad_norm": 0.4859190395369954,
+      "learning_rate": 9.99924760612111e-06,
+      "loss": 0.3736,
+      "step": 299
+    },
+    {
+      "epoch": 0.07139882191943833,
+      "grad_norm": 0.4497898536509786,
+      "learning_rate": 9.999213795726738e-06,
+      "loss": 0.4518,
+      "step": 300
+    },
+    {
+      "epoch": 0.07163681799250313,
+      "grad_norm": 0.4399132919727615,
+      "learning_rate": 9.999179242325523e-06,
+      "loss": 0.5143,
+      "step": 301
+    },
+    {
+      "epoch": 0.07187481406556792,
+      "grad_norm": 0.40980962676569144,
+      "learning_rate": 9.999143945922599e-06,
+      "loss": 0.3701,
+      "step": 302
+    },
+    {
+      "epoch": 0.07211281013863272,
+      "grad_norm": 0.4266818493401362,
+      "learning_rate": 9.999107906523212e-06,
+      "loss": 0.3847,
+      "step": 303
+    },
+    {
+      "epoch": 0.07235080621169751,
+      "grad_norm": 0.4120201125570514,
+      "learning_rate": 9.999071124132717e-06,
+      "loss": 0.4753,
+      "step": 304
+    },
+    {
+      "epoch": 0.0725888022847623,
+      "grad_norm": 0.5036108776564726,
+      "learning_rate": 9.999033598756583e-06,
+      "loss": 0.4513,
+      "step": 305
+    },
+    {
+      "epoch": 0.0728267983578271,
+      "grad_norm": 0.45433796497230666,
+      "learning_rate": 9.998995330400385e-06,
+      "loss": 0.38,
+      "step": 306
+    },
+    {
+      "epoch": 0.0730647944308919,
+      "grad_norm": 0.4604572613372604,
+      "learning_rate": 9.998956319069813e-06,
+      "loss": 0.4294,
+      "step": 307
+    },
+    {
+      "epoch": 0.07330279050395669,
+      "grad_norm": 0.4239655605570626,
+      "learning_rate": 9.998916564770662e-06,
+      "loss": 0.5023,
+      "step": 308
+    },
+    {
+      "epoch": 0.07354078657702148,
+      "grad_norm": 0.49639603018946726,
+      "learning_rate": 9.998876067508846e-06,
+      "loss": 0.3821,
+      "step": 309
+    },
+    {
+      "epoch": 0.07377878265008628,
+      "grad_norm": 0.4498046706633747,
+      "learning_rate": 9.998834827290376e-06,
+      "loss": 0.3897,
+      "step": 310
+    },
+    {
+      "epoch": 0.07401677872315107,
+      "grad_norm": 0.4847626369166094,
+      "learning_rate": 9.998792844121386e-06,
+      "loss": 0.4587,
+      "step": 311
+    },
+    {
+      "epoch": 0.07425477479621587,
+      "grad_norm": 0.4754430506699406,
+      "learning_rate": 9.998750118008117e-06,
+      "loss": 0.4725,
+      "step": 312
+    },
+    {
+      "epoch": 0.07449277086928066,
+      "grad_norm": 0.48598157686223126,
+      "learning_rate": 9.998706648956916e-06,
+      "loss": 0.3919,
+      "step": 313
+    },
+    {
+      "epoch": 0.07473076694234546,
+      "grad_norm": 0.4660284468209422,
+      "learning_rate": 9.998662436974246e-06,
+      "loss": 0.422,
+      "step": 314
+    },
+    {
+      "epoch": 0.07496876301541025,
+      "grad_norm": 0.4755957103806313,
+      "learning_rate": 9.998617482066677e-06,
+      "loss": 0.4716,
+      "step": 315
+    },
+    {
+      "epoch": 0.07520675908847504,
+      "grad_norm": 0.49198072627395434,
+      "learning_rate": 9.998571784240889e-06,
+      "loss": 0.3805,
+      "step": 316
+    },
+    {
+      "epoch": 0.07544475516153984,
+      "grad_norm": 0.45039434089900343,
+      "learning_rate": 9.998525343503676e-06,
+      "loss": 0.3578,
+      "step": 317
+    },
+    {
+      "epoch": 0.07568275123460463,
+      "grad_norm": 0.4567179079014157,
+      "learning_rate": 9.998478159861938e-06,
+      "loss": 0.4384,
+      "step": 318
+    },
+    {
+      "epoch": 0.07592074730766943,
+      "grad_norm": 0.44281871668390654,
+      "learning_rate": 9.99843023332269e-06,
+      "loss": 0.4717,
+      "step": 319
+    },
+    {
+      "epoch": 0.07615874338073422,
+      "grad_norm": 0.5238647769899825,
+      "learning_rate": 9.998381563893056e-06,
+      "loss": 0.4097,
+      "step": 320
+    },
+    {
+      "epoch": 0.07639673945379902,
+      "grad_norm": 0.4477956066826517,
+      "learning_rate": 9.998332151580266e-06,
+      "loss": 0.3769,
+      "step": 321
+    },
+    {
+      "epoch": 0.07663473552686381,
+      "grad_norm": 0.44274206034195474,
+      "learning_rate": 9.998281996391665e-06,
+      "loss": 0.4546,
+      "step": 322
+    },
+    {
+      "epoch": 0.0768727315999286,
+      "grad_norm": 0.43869730085122627,
+      "learning_rate": 9.998231098334708e-06,
+      "loss": 0.4046,
+      "step": 323
+    },
+    {
+      "epoch": 0.0771107276729934,
+      "grad_norm": 0.5146875829687024,
+      "learning_rate": 9.99817945741696e-06,
+      "loss": 0.3706,
+      "step": 324
+    },
+    {
+      "epoch": 0.0773487237460582,
+      "grad_norm": 0.4036470374218696,
+      "learning_rate": 9.998127073646095e-06,
+      "loss": 0.4352,
+      "step": 325
+    },
+    {
+      "epoch": 0.07758671981912299,
+      "grad_norm": 0.4630902945017874,
+      "learning_rate": 9.998073947029899e-06,
+      "loss": 0.4768,
+      "step": 326
+    },
+    {
+      "epoch": 0.07782471589218778,
+      "grad_norm": 0.5458520408430351,
+      "learning_rate": 9.99802007757627e-06,
+      "loss": 0.3639,
+      "step": 327
+    },
+    {
+      "epoch": 0.07806271196525258,
+      "grad_norm": 0.42241640395474855,
+      "learning_rate": 9.997965465293208e-06,
+      "loss": 0.3646,
+      "step": 328
+    },
+    {
+      "epoch": 0.07830070803831737,
+      "grad_norm": 0.4473593312203704,
+      "learning_rate": 9.99791011018884e-06,
+      "loss": 0.4494,
+      "step": 329
+    },
+    {
+      "epoch": 0.07853870411138217,
+      "grad_norm": 0.4617723155081359,
+      "learning_rate": 9.997854012271383e-06,
+      "loss": 0.4168,
+      "step": 330
+    },
+    {
+      "epoch": 0.07877670018444696,
+      "grad_norm": 0.4796048928242152,
+      "learning_rate": 9.99779717154918e-06,
+      "loss": 0.3781,
+      "step": 331
+    },
+    {
+      "epoch": 0.07901469625751176,
+      "grad_norm": 0.4409776472086194,
+      "learning_rate": 9.99773958803068e-06,
+      "loss": 0.4194,
+      "step": 332
+    },
+    {
+      "epoch": 0.07925269233057655,
+      "grad_norm": 0.49694630682051916,
+      "learning_rate": 9.997681261724436e-06,
+      "loss": 0.4955,
+      "step": 333
+    },
+    {
+      "epoch": 0.07949068840364135,
+      "grad_norm": 0.5521495917382526,
+      "learning_rate": 9.99762219263912e-06,
+      "loss": 0.3829,
+      "step": 334
+    },
+    {
+      "epoch": 0.07972868447670614,
+      "grad_norm": 0.4448039189655745,
+      "learning_rate": 9.997562380783512e-06,
+      "loss": 0.3625,
+      "step": 335
+    },
+    {
+      "epoch": 0.07996668054977094,
+      "grad_norm": 0.4233213810496793,
+      "learning_rate": 9.997501826166502e-06,
+      "loss": 0.4431,
+      "step": 336
+    },
+    {
+      "epoch": 0.08020467662283573,
+      "grad_norm": 0.4511409424625309,
+      "learning_rate": 9.997440528797087e-06,
+      "loss": 0.4574,
+      "step": 337
+    },
+    {
+      "epoch": 0.08044267269590052,
+      "grad_norm": 0.5076129237328914,
+      "learning_rate": 9.997378488684376e-06,
+      "loss": 0.3615,
+      "step": 338
+    },
+    {
+      "epoch": 0.0806806687689653,
+      "grad_norm": 0.41838318534134256,
+      "learning_rate": 9.997315705837596e-06,
+      "loss": 0.3779,
+      "step": 339
+    },
+    {
+      "epoch": 0.0809186648420301,
+      "grad_norm": 0.44373668851344966,
+      "learning_rate": 9.997252180266074e-06,
+      "loss": 0.4479,
+      "step": 340
+    },
+    {
+      "epoch": 0.0811566609150949,
+      "grad_norm": 0.5204118151673891,
+      "learning_rate": 9.997187911979252e-06,
+      "loss": 0.4187,
+      "step": 341
+    },
+    {
+      "epoch": 0.08139465698815969,
+      "grad_norm": 0.5043252971282625,
+      "learning_rate": 9.99712290098668e-06,
+      "loss": 0.3434,
+      "step": 342
+    },
+    {
+      "epoch": 0.08163265306122448,
+      "grad_norm": 0.44526891013124525,
+      "learning_rate": 9.997057147298024e-06,
+      "loss": 0.4474,
+      "step": 343
+    },
+    {
+      "epoch": 0.08187064913428928,
+      "grad_norm": 0.477769396408338,
+      "learning_rate": 9.996990650923053e-06,
+      "loss": 0.4611,
+      "step": 344
+    },
+    {
+      "epoch": 0.08210864520735407,
+      "grad_norm": 0.5121141417084246,
+      "learning_rate": 9.996923411871653e-06,
+      "loss": 0.3596,
+      "step": 345
+    },
+    {
+      "epoch": 0.08234664128041887,
+      "grad_norm": 0.4424669784191923,
+      "learning_rate": 9.996855430153816e-06,
+      "loss": 0.3949,
+      "step": 346
+    },
+    {
+      "epoch": 0.08258463735348366,
+      "grad_norm": 0.4508136073799909,
+      "learning_rate": 9.996786705779645e-06,
+      "loss": 0.4567,
+      "step": 347
+    },
+    {
+      "epoch": 0.08282263342654846,
+      "grad_norm": 0.5106951134749175,
+      "learning_rate": 9.996717238759355e-06,
+      "loss": 0.3982,
+      "step": 348
+    },
+    {
+      "epoch": 0.08306062949961325,
+      "grad_norm": 0.4710547751602934,
+      "learning_rate": 9.996647029103271e-06,
+      "loss": 0.334,
+      "step": 349
+    },
+    {
+      "epoch": 0.08329862557267805,
+      "grad_norm": 0.45527398227968663,
+      "learning_rate": 9.996576076821827e-06,
+      "loss": 0.4494,
+      "step": 350
+    },
+    {
+      "epoch": 0.08353662164574284,
+      "grad_norm": 0.43961786486243276,
+      "learning_rate": 9.99650438192557e-06,
+      "loss": 0.4702,
+      "step": 351
+    },
+    {
+      "epoch": 0.08377461771880763,
+      "grad_norm": 0.4920860251140083,
+      "learning_rate": 9.996431944425154e-06,
+      "loss": 0.3757,
+      "step": 352
+    },
+    {
+      "epoch": 0.08401261379187243,
+      "grad_norm": 0.4544455547851751,
+      "learning_rate": 9.996358764331344e-06,
+      "loss": 0.3715,
+      "step": 353
+    },
+    {
+      "epoch": 0.08425060986493722,
+      "grad_norm": 0.4383129033433832,
+      "learning_rate": 9.996284841655017e-06,
+      "loss": 0.4829,
+      "step": 354
+    },
+    {
+      "epoch": 0.08448860593800202,
+      "grad_norm": 0.4911519487114972,
+      "learning_rate": 9.996210176407163e-06,
+      "loss": 0.4288,
+      "step": 355
+    },
+    {
+      "epoch": 0.08472660201106681,
+      "grad_norm": 0.4165648910362421,
+      "learning_rate": 9.996134768598874e-06,
+      "loss": 0.3398,
+      "step": 356
+    },
+    {
+      "epoch": 0.0849645980841316,
+      "grad_norm": 0.4583575292505661,
+      "learning_rate": 9.996058618241362e-06,
+      "loss": 0.4005,
+      "step": 357
+    },
+    {
+      "epoch": 0.0852025941571964,
+      "grad_norm": 0.41965836884164454,
+      "learning_rate": 9.995981725345941e-06,
+      "loss": 0.4569,
+      "step": 358
+    },
+    {
+      "epoch": 0.0854405902302612,
+      "grad_norm": 0.45798270095341687,
+      "learning_rate": 9.995904089924043e-06,
+      "loss": 0.4057,
+      "step": 359
+    },
+    {
+      "epoch": 0.08567858630332599,
+      "grad_norm": 0.46062392413472286,
+      "learning_rate": 9.995825711987202e-06,
+      "loss": 0.3556,
+      "step": 360
+    },
+    {
+      "epoch": 0.08591658237639078,
+      "grad_norm": 0.4533838682476381,
+      "learning_rate": 9.99574659154707e-06,
+      "loss": 0.4538,
+      "step": 361
+    },
+    {
+      "epoch": 0.08615457844945558,
+      "grad_norm": 0.46749855571344134,
+      "learning_rate": 9.995666728615407e-06,
+      "loss": 0.4573,
+      "step": 362
+    },
+    {
+      "epoch": 0.08639257452252037,
+      "grad_norm": 0.41387520189963245,
+      "learning_rate": 9.995586123204083e-06,
+      "loss": 0.3657,
+      "step": 363
+    },
+    {
+      "epoch": 0.08663057059558517,
+      "grad_norm": 0.4516587880902474,
+      "learning_rate": 9.995504775325073e-06,
+      "loss": 0.4202,
+      "step": 364
+    },
+    {
+      "epoch": 0.08686856666864996,
+      "grad_norm": 0.42225242395315815,
+      "learning_rate": 9.99542268499047e-06,
+      "loss": 0.4667,
+      "step": 365
+    },
+    {
+      "epoch": 0.08710656274171476,
+      "grad_norm": 0.44765554289425097,
+      "learning_rate": 9.995339852212478e-06,
+      "loss": 0.4053,
+      "step": 366
+    },
+    {
+      "epoch": 0.08734455881477955,
+      "grad_norm": 0.40737919321911104,
+      "learning_rate": 9.995256277003403e-06,
+      "loss": 0.3616,
+      "step": 367
+    },
+    {
+      "epoch": 0.08758255488784435,
+      "grad_norm": 0.47373973714542855,
+      "learning_rate": 9.99517195937567e-06,
+      "loss": 0.4084,
+      "step": 368
+    },
+    {
+      "epoch": 0.08782055096090914,
+      "grad_norm": 0.4530682065428116,
+      "learning_rate": 9.995086899341808e-06,
+      "loss": 0.4571,
+      "step": 369
+    },
+    {
+      "epoch": 0.08805854703397394,
+      "grad_norm": 1.3913274762621268,
+      "learning_rate": 9.995001096914462e-06,
+      "loss": 0.3454,
+      "step": 370
+    },
+    {
+      "epoch": 0.08829654310703873,
+      "grad_norm": 0.47478901107913046,
+      "learning_rate": 9.99491455210638e-06,
+      "loss": 0.3953,
+      "step": 371
+    },
+    {
+      "epoch": 0.08853453918010352,
+      "grad_norm": 0.5018434335802849,
+      "learning_rate": 9.994827264930432e-06,
+      "loss": 0.4965,
+      "step": 372
+    },
+    {
+      "epoch": 0.08877253525316832,
+      "grad_norm": 0.4450551192390846,
+      "learning_rate": 9.994739235399581e-06,
+      "loss": 0.4243,
+      "step": 373
+    },
+    {
+      "epoch": 0.08901053132623311,
+      "grad_norm": 0.4515333750014817,
+      "learning_rate": 9.994650463526918e-06,
+      "loss": 0.3676,
+      "step": 374
+    },
+    {
+      "epoch": 0.08924852739929791,
+      "grad_norm": 0.48851016231525185,
+      "learning_rate": 9.994560949325636e-06,
+      "loss": 0.467,
+      "step": 375
+    },
+    {
+      "epoch": 0.0894865234723627,
+      "grad_norm": 0.4195656189924732,
+      "learning_rate": 9.994470692809036e-06,
+      "loss": 0.4756,
+      "step": 376
+    },
+    {
+      "epoch": 0.0897245195454275,
+      "grad_norm": 0.43584449035060197,
+      "learning_rate": 9.994379693990533e-06,
+      "loss": 0.3715,
+      "step": 377
+    },
+    {
+      "epoch": 0.08996251561849229,
+      "grad_norm": 0.4385497336895332,
+      "learning_rate": 9.994287952883652e-06,
+      "loss": 0.3578,
+      "step": 378
+    },
+    {
+      "epoch": 0.09020051169155709,
+      "grad_norm": 0.47979109705837614,
+      "learning_rate": 9.994195469502031e-06,
+      "loss": 0.4422,
+      "step": 379
+    },
+    {
+      "epoch": 0.09043850776462188,
+      "grad_norm": 0.4035725189989237,
+      "learning_rate": 9.994102243859412e-06,
+      "loss": 0.4285,
+      "step": 380
+    },
+    {
+      "epoch": 0.09067650383768668,
+      "grad_norm": 0.40727590335667724,
+      "learning_rate": 9.99400827596965e-06,
+      "loss": 0.3412,
+      "step": 381
+    },
+    {
+      "epoch": 0.09091449991075147,
+      "grad_norm": 0.4480428671921397,
+      "learning_rate": 9.993913565846713e-06,
+      "loss": 0.4233,
+      "step": 382
+    },
+    {
+      "epoch": 0.09115249598381626,
+      "grad_norm": 0.45603055794812214,
+      "learning_rate": 9.993818113504678e-06,
+      "loss": 0.4748,
+      "step": 383
+    },
+    {
+      "epoch": 0.09139049205688106,
+      "grad_norm": 0.4662947179384981,
+      "learning_rate": 9.993721918957728e-06,
+      "loss": 0.3761,
+      "step": 384
+    },
+    {
+      "epoch": 0.09162848812994585,
+      "grad_norm": 0.4974648803891763,
+      "learning_rate": 9.993624982220164e-06,
+      "loss": 0.3555,
+      "step": 385
+    },
+    {
+      "epoch": 0.09186648420301065,
+      "grad_norm": 0.4643066974713487,
+      "learning_rate": 9.99352730330639e-06,
+      "loss": 0.4187,
+      "step": 386
+    },
+    {
+      "epoch": 0.09210448027607544,
+      "grad_norm": 0.43172365645329547,
+      "learning_rate": 9.993428882230925e-06,
+      "loss": 0.4664,
+      "step": 387
+    },
+    {
+      "epoch": 0.09234247634914024,
+      "grad_norm": 0.4714894865640499,
+      "learning_rate": 9.993329719008397e-06,
+      "loss": 0.3685,
+      "step": 388
+    },
+    {
+      "epoch": 0.09258047242220503,
+      "grad_norm": 0.41913422983030557,
+      "learning_rate": 9.993229813653544e-06,
+      "loss": 0.3675,
+      "step": 389
+    },
+    {
+      "epoch": 0.09281846849526983,
+      "grad_norm": 0.43975663296852435,
+      "learning_rate": 9.993129166181215e-06,
+      "loss": 0.4556,
+      "step": 390
+    },
+    {
+      "epoch": 0.09305646456833462,
+      "grad_norm": 0.4494552356684672,
+      "learning_rate": 9.993027776606365e-06,
+      "loss": 0.4227,
+      "step": 391
+    },
+    {
+      "epoch": 0.09329446064139942,
+      "grad_norm": 0.4288268452457883,
+      "learning_rate": 9.992925644944068e-06,
+      "loss": 0.3461,
+      "step": 392
+    },
+    {
+      "epoch": 0.09353245671446421,
+      "grad_norm": 0.4489798380818497,
+      "learning_rate": 9.992822771209501e-06,
+      "loss": 0.4179,
+      "step": 393
+    },
+    {
+      "epoch": 0.093770452787529,
+      "grad_norm": 0.4509441961780933,
+      "learning_rate": 9.992719155417954e-06,
+      "loss": 0.4886,
+      "step": 394
+    },
+    {
+      "epoch": 0.0940084488605938,
+      "grad_norm": 0.4449441936420067,
+      "learning_rate": 9.992614797584825e-06,
+      "loss": 0.3609,
+      "step": 395
+    },
+    {
+      "epoch": 0.0942464449336586,
+      "grad_norm": 0.4517902195455245,
+      "learning_rate": 9.992509697725627e-06,
+      "loss": 0.3913,
+      "step": 396
+    },
+    {
+      "epoch": 0.09448444100672339,
+      "grad_norm": 0.4272371117143219,
+      "learning_rate": 9.992403855855979e-06,
+      "loss": 0.4825,
+      "step": 397
+    },
+    {
+      "epoch": 0.09472243707978818,
+      "grad_norm": 0.49237207377468334,
+      "learning_rate": 9.992297271991611e-06,
+      "loss": 0.4321,
+      "step": 398
+    },
+    {
+      "epoch": 0.09496043315285298,
+      "grad_norm": 0.4742011509459983,
+      "learning_rate": 9.992189946148366e-06,
+      "loss": 0.3706,
+      "step": 399
+    },
+    {
+      "epoch": 0.09519842922591777,
+      "grad_norm": 0.481623368264721,
+      "learning_rate": 9.992081878342196e-06,
+      "loss": 0.433,
+      "step": 400
+    },
+    {
+      "epoch": 0.09543642529898257,
+      "grad_norm": 0.43969308523488704,
+      "learning_rate": 9.991973068589157e-06,
+      "loss": 0.497,
+      "step": 401
+    },
+    {
+      "epoch": 0.09567442137204736,
+      "grad_norm": 0.45002640361694246,
+      "learning_rate": 9.991863516905428e-06,
+      "loss": 0.3763,
+      "step": 402
+    },
+    {
+      "epoch": 0.09591241744511216,
+      "grad_norm": 0.4140916324473366,
+      "learning_rate": 9.991753223307283e-06,
+      "loss": 0.4141,
+      "step": 403
+    },
+    {
+      "epoch": 0.09615041351817695,
+      "grad_norm": 0.4987232508457122,
+      "learning_rate": 9.991642187811122e-06,
+      "loss": 0.4597,
+      "step": 404
+    },
+    {
+      "epoch": 0.09638840959124174,
+      "grad_norm": 0.502335430018636,
+      "learning_rate": 9.991530410433446e-06,
+      "loss": 0.4302,
+      "step": 405
+    },
+    {
+      "epoch": 0.09662640566430654,
+      "grad_norm": 0.4984534349500167,
+      "learning_rate": 9.991417891190864e-06,
+      "loss": 0.3477,
+      "step": 406
+    },
+    {
+      "epoch": 0.09686440173737133,
+      "grad_norm": 0.43148480208491835,
+      "learning_rate": 9.991304630100103e-06,
+      "loss": 0.4022,
+      "step": 407
+    },
+    {
+      "epoch": 0.09710239781043613,
+      "grad_norm": 0.42523411463325067,
+      "learning_rate": 9.991190627177993e-06,
+      "loss": 0.4572,
+      "step": 408
+    },
+    {
+      "epoch": 0.09734039388350092,
+      "grad_norm": 0.5251992385161286,
+      "learning_rate": 9.991075882441482e-06,
+      "loss": 0.3742,
+      "step": 409
+    },
+    {
+      "epoch": 0.09757838995656572,
+      "grad_norm": 0.5071319884695096,
+      "learning_rate": 9.990960395907621e-06,
+      "loss": 0.3707,
+      "step": 410
+    },
+    {
+      "epoch": 0.09781638602963051,
+      "grad_norm": 0.449266770757162,
+      "learning_rate": 9.990844167593574e-06,
+      "loss": 0.4296,
+      "step": 411
+    },
+    {
+      "epoch": 0.0980543821026953,
+      "grad_norm": 0.48923465192436794,
+      "learning_rate": 9.990727197516617e-06,
+      "loss": 0.4722,
+      "step": 412
+    },
+    {
+      "epoch": 0.0982923781757601,
+      "grad_norm": 0.525715202859848,
+      "learning_rate": 9.990609485694133e-06,
+      "loss": 0.3591,
+      "step": 413
+    },
+    {
+      "epoch": 0.0985303742488249,
+      "grad_norm": 0.4460753119588959,
+      "learning_rate": 9.990491032143619e-06,
+      "loss": 0.3947,
+      "step": 414
+    },
+    {
+      "epoch": 0.09876837032188969,
+      "grad_norm": 0.4540712349304704,
+      "learning_rate": 9.990371836882678e-06,
+      "loss": 0.4902,
+      "step": 415
+    },
+    {
+      "epoch": 0.09900636639495448,
+      "grad_norm": 0.45967087446771393,
+      "learning_rate": 9.990251899929026e-06,
+      "loss": 0.3735,
+      "step": 416
+    },
+    {
+      "epoch": 0.09924436246801928,
+      "grad_norm": 0.5344550556485966,
+      "learning_rate": 9.990131221300489e-06,
+      "loss": 0.3667,
+      "step": 417
+    },
+    {
+      "epoch": 0.09948235854108407,
+      "grad_norm": 0.44602952588771577,
+      "learning_rate": 9.990009801015003e-06,
+      "loss": 0.4267,
+      "step": 418
+    },
+    {
+      "epoch": 0.09972035461414887,
+      "grad_norm": 0.5195582033175876,
+      "learning_rate": 9.989887639090614e-06,
+      "loss": 0.4233,
+      "step": 419
+    },
+    {
+      "epoch": 0.09995835068721366,
+      "grad_norm": 0.5142299290549335,
+      "learning_rate": 9.989764735545477e-06,
+      "loss": 0.378,
+      "step": 420
+    },
+    {
+      "epoch": 0.10019634676027846,
+      "grad_norm": 0.4854121376426571,
+      "learning_rate": 9.98964109039786e-06,
+      "loss": 0.361,
+      "step": 421
+    },
+    {
+      "epoch": 0.10043434283334325,
+      "grad_norm": 0.4550651737012537,
+      "learning_rate": 9.98951670366614e-06,
+      "loss": 0.4618,
+      "step": 422
+    },
+    {
+      "epoch": 0.10067233890640805,
+      "grad_norm": 0.5234336286250443,
+      "learning_rate": 9.989391575368802e-06,
+      "loss": 0.3958,
+      "step": 423
+    },
+    {
+      "epoch": 0.10091033497947284,
+      "grad_norm": 0.47063073761951685,
+      "learning_rate": 9.989265705524444e-06,
+      "loss": 0.3747,
+      "step": 424
+    },
+    {
+      "epoch": 0.10114833105253764,
+      "grad_norm": 0.46952341342697657,
+      "learning_rate": 9.989139094151773e-06,
+      "loss": 0.4112,
+      "step": 425
+    },
+    {
+      "epoch": 0.10138632712560243,
+      "grad_norm": 0.42526909585554723,
+      "learning_rate": 9.98901174126961e-06,
+      "loss": 0.4603,
+      "step": 426
+    },
+    {
+      "epoch": 0.10162432319866722,
+      "grad_norm": 0.43907211057538165,
+      "learning_rate": 9.988883646896877e-06,
+      "loss": 0.3813,
+      "step": 427
+    },
+    {
+      "epoch": 0.10186231927173202,
+      "grad_norm": 0.4940877209700441,
+      "learning_rate": 9.988754811052616e-06,
+      "loss": 0.355,
+      "step": 428
+    },
+    {
+      "epoch": 0.10210031534479681,
+      "grad_norm": 0.4302285563808865,
+      "learning_rate": 9.988625233755975e-06,
+      "loss": 0.4783,
+      "step": 429
+    },
+    {
+      "epoch": 0.10233831141786161,
+      "grad_norm": 0.4399773493860619,
+      "learning_rate": 9.988494915026213e-06,
+      "loss": 0.4771,
+      "step": 430
+    },
+    {
+      "epoch": 0.1025763074909264,
+      "grad_norm": 0.5749592576382808,
+      "learning_rate": 9.988363854882694e-06,
+      "loss": 0.3388,
+      "step": 431
+    },
+    {
+      "epoch": 0.1028143035639912,
+      "grad_norm": 0.4171737508333019,
+      "learning_rate": 9.988232053344901e-06,
+      "loss": 0.4032,
+      "step": 432
+    },
+    {
+      "epoch": 0.10305229963705599,
+      "grad_norm": 0.44690528937324775,
+      "learning_rate": 9.988099510432422e-06,
+      "loss": 0.4354,
+      "step": 433
+    },
+    {
+      "epoch": 0.10329029571012079,
+      "grad_norm": 0.7413219284083276,
+      "learning_rate": 9.987966226164958e-06,
+      "loss": 0.3871,
+      "step": 434
+    },
+    {
+      "epoch": 0.10352829178318558,
+      "grad_norm": 0.572534433584837,
+      "learning_rate": 9.987832200562315e-06,
+      "loss": 0.3343,
+      "step": 435
+    },
+    {
+      "epoch": 0.10376628785625037,
+      "grad_norm": 0.4024474823593315,
+      "learning_rate": 9.987697433644414e-06,
+      "loss": 0.4353,
+      "step": 436
+    },
+    {
+      "epoch": 0.10400428392931517,
+      "grad_norm": 0.5053428492479292,
+      "learning_rate": 9.987561925431283e-06,
+      "loss": 0.456,
+      "step": 437
+    },
+    {
+      "epoch": 0.10424228000237996,
+      "grad_norm": 0.5684801045695161,
+      "learning_rate": 9.987425675943067e-06,
+      "loss": 0.3426,
+      "step": 438
+    },
+    {
+      "epoch": 0.10448027607544476,
+      "grad_norm": 0.5703263329939863,
+      "learning_rate": 9.98728868520001e-06,
+      "loss": 0.417,
+      "step": 439
+    },
+    {
+      "epoch": 0.10471827214850955,
+      "grad_norm": 0.5736516226532222,
+      "learning_rate": 9.987150953222476e-06,
+      "loss": 0.4352,
+      "step": 440
+    },
+    {
+      "epoch": 0.10495626822157435,
+      "grad_norm": 0.4287015110638932,
+      "learning_rate": 9.987012480030934e-06,
+      "loss": 0.4033,
+      "step": 441
+    },
+    {
+      "epoch": 0.10519426429463914,
+      "grad_norm": 0.46545342863477973,
+      "learning_rate": 9.986873265645965e-06,
+      "loss": 0.3712,
+      "step": 442
+    },
+    {
+      "epoch": 0.10543226036770394,
+      "grad_norm": 0.4709132222065424,
+      "learning_rate": 9.98673331008826e-06,
+      "loss": 0.4198,
+      "step": 443
+    },
+    {
+      "epoch": 0.10567025644076873,
+      "grad_norm": 0.5082536813625272,
+      "learning_rate": 9.986592613378616e-06,
+      "loss": 0.4497,
+      "step": 444
+    },
+    {
+      "epoch": 0.10590825251383353,
+      "grad_norm": 0.4722946115830572,
+      "learning_rate": 9.98645117553795e-06,
+      "loss": 0.3466,
+      "step": 445
+    },
+    {
+      "epoch": 0.10614624858689832,
+      "grad_norm": 1.2081855072445622,
+      "learning_rate": 9.98630899658728e-06,
+      "loss": 0.3711,
+      "step": 446
+    },
+    {
+      "epoch": 0.10638424465996311,
+      "grad_norm": 0.49995351880997874,
+      "learning_rate": 9.98616607654774e-06,
+      "loss": 0.4477,
+      "step": 447
+    },
+    {
+      "epoch": 0.10662224073302791,
+      "grad_norm": 0.5032188297810777,
+      "learning_rate": 9.986022415440564e-06,
+      "loss": 0.4162,
+      "step": 448
+    },
+    {
+      "epoch": 0.1068602368060927,
+      "grad_norm": 0.7647181678703971,
+      "learning_rate": 9.985878013287113e-06,
+      "loss": 0.3694,
+      "step": 449
+    },
+    {
+      "epoch": 0.1070982328791575,
+      "grad_norm": 0.6789559238085723,
+      "learning_rate": 9.985732870108843e-06,
+      "loss": 0.4079,
+      "step": 450
+    },
+    {
+      "epoch": 0.10733622895222229,
+      "grad_norm": 0.6758534206958869,
+      "learning_rate": 9.985586985927328e-06,
+      "loss": 0.4949,
+      "step": 451
+    },
+    {
+      "epoch": 0.10757422502528709,
+      "grad_norm": 0.63305874198617,
+      "learning_rate": 9.98544036076425e-06,
+      "loss": 0.3443,
+      "step": 452
+    },
+    {
+      "epoch": 0.10781222109835188,
+      "grad_norm": 0.5712028770173395,
+      "learning_rate": 9.985292994641398e-06,
+      "loss": 0.3686,
+      "step": 453
+    },
+    {
+      "epoch": 0.10805021717141668,
+      "grad_norm": 0.42711184550985515,
+      "learning_rate": 9.98514488758068e-06,
+      "loss": 0.4819,
+      "step": 454
+    },
+    {
+      "epoch": 0.10828821324448147,
+      "grad_norm": 0.45616716468152596,
+      "learning_rate": 9.984996039604102e-06,
+      "loss": 0.4337,
+      "step": 455
+    },
+    {
+      "epoch": 0.10852620931754627,
+      "grad_norm": 0.44388087258594616,
+      "learning_rate": 9.98484645073379e-06,
+      "loss": 0.3341,
+      "step": 456
+    },
+    {
+      "epoch": 0.10876420539061106,
+      "grad_norm": 0.5001916577790845,
+      "learning_rate": 9.984696120991979e-06,
+      "loss": 0.431,
+      "step": 457
+    },
+    {
+      "epoch": 0.10900220146367585,
+      "grad_norm": 0.44705716190669925,
+      "learning_rate": 9.984545050401007e-06,
+      "loss": 0.4467,
+      "step": 458
+    },
+    {
+      "epoch": 0.10924019753674065,
+      "grad_norm": 0.45150670551113553,
+      "learning_rate": 9.98439323898333e-06,
+      "loss": 0.3919,
+      "step": 459
+    },
+    {
+      "epoch": 0.10947819360980544,
+      "grad_norm": 0.520796434754494,
+      "learning_rate": 9.98424068676151e-06,
+      "loss": 0.3459,
+      "step": 460
+    },
+    {
+      "epoch": 0.10971618968287024,
+      "grad_norm": 0.4208531093058114,
+      "learning_rate": 9.984087393758218e-06,
+      "loss": 0.444,
+      "step": 461
+    },
+    {
+      "epoch": 0.10995418575593503,
+      "grad_norm": 0.45915739817950607,
+      "learning_rate": 9.983933359996241e-06,
+      "loss": 0.4736,
+      "step": 462
+    },
+    {
+      "epoch": 0.11019218182899983,
+      "grad_norm": 0.4461278219351183,
+      "learning_rate": 9.983778585498468e-06,
+      "loss": 0.3765,
+      "step": 463
+    },
+    {
+      "epoch": 0.11043017790206462,
+      "grad_norm": 0.40738699092689734,
+      "learning_rate": 9.983623070287905e-06,
+      "loss": 0.3933,
+      "step": 464
+    },
+    {
+      "epoch": 0.11066817397512942,
+      "grad_norm": 0.4913832741443862,
+      "learning_rate": 9.983466814387666e-06,
+      "loss": 0.4552,
+      "step": 465
+    },
+    {
+      "epoch": 0.11090617004819421,
+      "grad_norm": 0.530231130629648,
+      "learning_rate": 9.983309817820972e-06,
+      "loss": 0.3586,
+      "step": 466
+    },
+    {
+      "epoch": 0.111144166121259,
+      "grad_norm": 0.4241816209307345,
+      "learning_rate": 9.983152080611158e-06,
+      "loss": 0.326,
+      "step": 467
+    },
+    {
+      "epoch": 0.1113821621943238,
+      "grad_norm": 0.43804873868170846,
+      "learning_rate": 9.98299360278167e-06,
+      "loss": 0.4251,
+      "step": 468
+    },
+    {
+      "epoch": 0.1116201582673886,
+      "grad_norm": 0.4393810388350614,
+      "learning_rate": 9.982834384356057e-06,
+      "loss": 0.4675,
+      "step": 469
+    },
+    {
+      "epoch": 0.11185815434045339,
+      "grad_norm": 0.46372765200679533,
+      "learning_rate": 9.982674425357985e-06,
+      "loss": 0.3493,
+      "step": 470
+    },
+    {
+      "epoch": 0.11209615041351818,
+      "grad_norm": 0.4436475358645403,
+      "learning_rate": 9.982513725811228e-06,
+      "loss": 0.4243,
+      "step": 471
+    },
+    {
+      "epoch": 0.11233414648658298,
+      "grad_norm": 0.4173010199129987,
+      "learning_rate": 9.98235228573967e-06,
+      "loss": 0.472,
+      "step": 472
+    },
+    {
+      "epoch": 0.11257214255964777,
+      "grad_norm": 0.535729642414528,
+      "learning_rate": 9.982190105167306e-06,
+      "loss": 0.4164,
+      "step": 473
+    },
+    {
+      "epoch": 0.11281013863271257,
+      "grad_norm": 0.4772334936227505,
+      "learning_rate": 9.982027184118236e-06,
+      "loss": 0.349,
+      "step": 474
+    },
+    {
+      "epoch": 0.11304813470577735,
+      "grad_norm": 0.47313830652761185,
+      "learning_rate": 9.981863522616681e-06,
+      "loss": 0.4435,
+      "step": 475
+    },
+    {
+      "epoch": 0.11328613077884214,
+      "grad_norm": 0.47425353304897183,
+      "learning_rate": 9.981699120686959e-06,
+      "loss": 0.4695,
+      "step": 476
+    },
+    {
+      "epoch": 0.11352412685190694,
+      "grad_norm": 0.44292841465417715,
+      "learning_rate": 9.981533978353508e-06,
+      "loss": 0.3857,
+      "step": 477
+    },
+    {
+      "epoch": 0.11376212292497173,
+      "grad_norm": 0.5239164407002358,
+      "learning_rate": 9.981368095640868e-06,
+      "loss": 0.3726,
+      "step": 478
+    },
+    {
+      "epoch": 0.11400011899803653,
+      "grad_norm": 0.44488449687995085,
+      "learning_rate": 9.981201472573698e-06,
+      "loss": 0.4432,
+      "step": 479
+    },
+    {
+      "epoch": 0.11423811507110132,
+      "grad_norm": 0.43967538016218216,
+      "learning_rate": 9.98103410917676e-06,
+      "loss": 0.4076,
+      "step": 480
+    },
+    {
+      "epoch": 0.11447611114416611,
+      "grad_norm": 0.739424564589288,
+      "learning_rate": 9.980866005474928e-06,
+      "loss": 0.3675,
+      "step": 481
+    },
+    {
+      "epoch": 0.11471410721723091,
+      "grad_norm": 0.4993736830245461,
+      "learning_rate": 9.980697161493185e-06,
+      "loss": 0.4131,
+      "step": 482
+    },
+    {
+      "epoch": 0.1149521032902957,
+      "grad_norm": 1.3276469711665722,
+      "learning_rate": 9.980527577256629e-06,
+      "loss": 0.4678,
+      "step": 483
+    },
+    {
+      "epoch": 0.1151900993633605,
+      "grad_norm": 0.46128218449640923,
+      "learning_rate": 9.980357252790464e-06,
+      "loss": 0.3607,
+      "step": 484
+    },
+    {
+      "epoch": 0.1154280954364253,
+      "grad_norm": 0.41316023220047166,
+      "learning_rate": 9.980186188120002e-06,
+      "loss": 0.351,
+      "step": 485
+    },
+    {
+      "epoch": 0.11566609150949009,
+      "grad_norm": 0.482093707108989,
+      "learning_rate": 9.980014383270668e-06,
+      "loss": 0.3955,
+      "step": 486
+    },
+    {
+      "epoch": 0.11590408758255488,
+      "grad_norm": 0.5011032501000284,
+      "learning_rate": 9.979841838267999e-06,
+      "loss": 0.458,
+      "step": 487
+    },
+    {
+      "epoch": 0.11614208365561968,
+      "grad_norm": 0.4135106798022244,
+      "learning_rate": 9.979668553137635e-06,
+      "loss": 0.3878,
+      "step": 488
+    },
+    {
+      "epoch": 0.11638007972868447,
+      "grad_norm": 0.44904272180640614,
+      "learning_rate": 9.979494527905334e-06,
+      "loss": 0.3706,
+      "step": 489
+    },
+    {
+      "epoch": 0.11661807580174927,
+      "grad_norm": 0.41254283183456614,
+      "learning_rate": 9.979319762596959e-06,
+      "loss": 0.4731,
+      "step": 490
+    },
+    {
+      "epoch": 0.11685607187481406,
+      "grad_norm": 0.4326477250584368,
+      "learning_rate": 9.979144257238484e-06,
+      "loss": 0.3882,
+      "step": 491
+    },
+    {
+      "epoch": 0.11709406794787885,
+      "grad_norm": 0.47385412746747546,
+      "learning_rate": 9.978968011855996e-06,
+      "loss": 0.3596,
+      "step": 492
+    },
+    {
+      "epoch": 0.11733206402094365,
+      "grad_norm": 0.48119232908966797,
+      "learning_rate": 9.978791026475689e-06,
+      "loss": 0.4342,
+      "step": 493
+    },
+    {
+      "epoch": 0.11757006009400844,
+      "grad_norm": 0.4441552118826538,
+      "learning_rate": 9.978613301123864e-06,
+      "loss": 0.4489,
+      "step": 494
+    },
+    {
+      "epoch": 0.11780805616707324,
+      "grad_norm": 0.46690414644784983,
+      "learning_rate": 9.978434835826937e-06,
+      "loss": 0.3777,
+      "step": 495
+    },
+    {
+      "epoch": 0.11804605224013803,
+      "grad_norm": 0.44970400252197734,
+      "learning_rate": 9.978255630611432e-06,
+      "loss": 0.3701,
+      "step": 496
+    },
+    {
+      "epoch": 0.11828404831320283,
+      "grad_norm": 0.42146124042697336,
+      "learning_rate": 9.978075685503988e-06,
+      "loss": 0.4581,
+      "step": 497
+    },
+    {
+      "epoch": 0.11852204438626762,
+      "grad_norm": 0.4287576458047421,
+      "learning_rate": 9.977895000531343e-06,
+      "loss": 0.4068,
+      "step": 498
+    },
+    {
+      "epoch": 0.11876004045933242,
+      "grad_norm": 0.4661011135993105,
+      "learning_rate": 9.977713575720354e-06,
+      "loss": 0.3648,
+      "step": 499
+    },
+    {
+      "epoch": 0.11899803653239721,
+      "grad_norm": 0.44336193791681305,
+      "learning_rate": 9.977531411097985e-06,
+      "loss": 0.3956,
+      "step": 500
+    },
+    {
+      "epoch": 0.119236032605462,
+      "grad_norm": 0.40135410660420456,
+      "learning_rate": 9.97734850669131e-06,
+      "loss": 0.4393,
+      "step": 501
+    },
+    {
+      "epoch": 0.1194740286785268,
+      "grad_norm": 0.4559807876897745,
+      "learning_rate": 9.977164862527512e-06,
+      "loss": 0.3646,
+      "step": 502
+    },
+    {
+      "epoch": 0.1197120247515916,
+      "grad_norm": 0.4347794693076568,
+      "learning_rate": 9.976980478633888e-06,
+      "loss": 0.3785,
+      "step": 503
+    },
+    {
+      "epoch": 0.11995002082465639,
+      "grad_norm": 0.46153771301045676,
+      "learning_rate": 9.97679535503784e-06,
+      "loss": 0.4069,
+      "step": 504
+    },
+    {
+      "epoch": 0.12018801689772118,
+      "grad_norm": 0.424717397560516,
+      "learning_rate": 9.976609491766883e-06,
+      "loss": 0.4598,
+      "step": 505
+    },
+    {
+      "epoch": 0.12042601297078598,
+      "grad_norm": 0.4434747636911381,
+      "learning_rate": 9.97642288884864e-06,
+      "loss": 0.356,
+      "step": 506
+    },
+    {
+      "epoch": 0.12066400904385077,
+      "grad_norm": 0.4452496263517166,
+      "learning_rate": 9.976235546310844e-06,
+      "loss": 0.4129,
+      "step": 507
+    },
+    {
+      "epoch": 0.12090200511691557,
+      "grad_norm": 0.4578774709736023,
+      "learning_rate": 9.97604746418134e-06,
+      "loss": 0.4812,
+      "step": 508
+    },
+    {
+      "epoch": 0.12114000118998036,
+      "grad_norm": 0.41675783377655806,
+      "learning_rate": 9.975858642488081e-06,
+      "loss": 0.3771,
+      "step": 509
+    },
+    {
+      "epoch": 0.12137799726304516,
+      "grad_norm": 0.40069287579572055,
+      "learning_rate": 9.975669081259132e-06,
+      "loss": 0.3734,
+      "step": 510
+    },
+    {
+      "epoch": 0.12161599333610995,
+      "grad_norm": 0.4309476794229487,
+      "learning_rate": 9.975478780522664e-06,
+      "loss": 0.4206,
+      "step": 511
+    },
+    {
+      "epoch": 0.12185398940917475,
+      "grad_norm": 0.46800262518826174,
+      "learning_rate": 9.975287740306962e-06,
+      "loss": 0.4462,
+      "step": 512
+    },
+    {
+      "epoch": 0.12209198548223954,
+      "grad_norm": 0.42699693782693693,
+      "learning_rate": 9.97509596064042e-06,
+      "loss": 0.3419,
+      "step": 513
+    },
+    {
+      "epoch": 0.12232998155530433,
+      "grad_norm": 0.4386357295282374,
+      "learning_rate": 9.97490344155154e-06,
+      "loss": 0.398,
+      "step": 514
+    },
+    {
+      "epoch": 0.12256797762836913,
+      "grad_norm": 0.5979447179868536,
+      "learning_rate": 9.974710183068935e-06,
+      "loss": 0.4266,
+      "step": 515
+    },
+    {
+      "epoch": 0.12280597370143392,
+      "grad_norm": 0.4391789380414681,
+      "learning_rate": 9.97451618522133e-06,
+      "loss": 0.3549,
+      "step": 516
+    },
+    {
+      "epoch": 0.12304396977449872,
+      "grad_norm": 0.4254442498856795,
+      "learning_rate": 9.974321448037553e-06,
+      "loss": 0.372,
+      "step": 517
+    },
+    {
+      "epoch": 0.12328196584756351,
+      "grad_norm": 0.4203710488596454,
+      "learning_rate": 9.974125971546553e-06,
+      "loss": 0.4029,
+      "step": 518
+    },
+    {
+      "epoch": 0.12351996192062831,
+      "grad_norm": 0.4235704728114593,
+      "learning_rate": 9.973929755777379e-06,
+      "loss": 0.4569,
+      "step": 519
+    },
+    {
+      "epoch": 0.1237579579936931,
+      "grad_norm": 0.43253576502031255,
+      "learning_rate": 9.973732800759193e-06,
+      "loss": 0.3386,
+      "step": 520
+    },
+    {
+      "epoch": 0.1239959540667579,
+      "grad_norm": 0.4130102595578799,
+      "learning_rate": 9.97353510652127e-06,
+      "loss": 0.3849,
+      "step": 521
+    },
+    {
+      "epoch": 0.12423395013982269,
+      "grad_norm": 0.43914912023331804,
+      "learning_rate": 9.97333667309299e-06,
+      "loss": 0.4497,
+      "step": 522
+    },
+    {
+      "epoch": 0.12447194621288749,
+      "grad_norm": 0.3972463467100798,
+      "learning_rate": 9.973137500503846e-06,
+      "loss": 0.3995,
+      "step": 523
+    },
+    {
+      "epoch": 0.12470994228595228,
+      "grad_norm": 0.4452000977643881,
+      "learning_rate": 9.97293758878344e-06,
+      "loss": 0.3296,
+      "step": 524
+    },
+    {
+      "epoch": 0.12494793835901707,
+      "grad_norm": 0.4194230322974543,
+      "learning_rate": 9.972736937961484e-06,
+      "loss": 0.4134,
+      "step": 525
+    },
+    {
+      "epoch": 0.12518593443208187,
+      "grad_norm": 0.4146732557797122,
+      "learning_rate": 9.9725355480678e-06,
+      "loss": 0.474,
+      "step": 526
+    },
+    {
+      "epoch": 0.12542393050514666,
+      "grad_norm": 0.40700998263113747,
+      "learning_rate": 9.972333419132319e-06,
+      "loss": 0.383,
+      "step": 527
+    },
+    {
+      "epoch": 0.12566192657821146,
+      "grad_norm": 0.4023415477618789,
+      "learning_rate": 9.97213055118508e-06,
+      "loss": 0.3658,
+      "step": 528
+    },
+    {
+      "epoch": 0.12589992265127625,
+      "grad_norm": 0.44336944898631714,
+      "learning_rate": 9.971926944256239e-06,
+      "loss": 0.476,
+      "step": 529
+    },
+    {
+      "epoch": 0.12613791872434105,
+      "grad_norm": 0.40291923826893716,
+      "learning_rate": 9.971722598376054e-06,
+      "loss": 0.4248,
+      "step": 530
+    },
+    {
+      "epoch": 0.12637591479740584,
+      "grad_norm": 0.3986265875109195,
+      "learning_rate": 9.971517513574896e-06,
+      "loss": 0.3415,
+      "step": 531
+    },
+    {
+      "epoch": 0.12661391087047064,
+      "grad_norm": 0.43032291172443293,
+      "learning_rate": 9.971311689883247e-06,
+      "loss": 0.4259,
+      "step": 532
+    },
+    {
+      "epoch": 0.12685190694353543,
+      "grad_norm": 0.39866454624485,
+      "learning_rate": 9.971105127331695e-06,
+      "loss": 0.4444,
+      "step": 533
+    },
+    {
+      "epoch": 0.12708990301660023,
+      "grad_norm": 0.43444250324758876,
+      "learning_rate": 9.970897825950942e-06,
+      "loss": 0.3788,
+      "step": 534
+    },
+    {
+      "epoch": 0.12732789908966502,
+      "grad_norm": 0.419118191593301,
+      "learning_rate": 9.970689785771798e-06,
+      "loss": 0.3715,
+      "step": 535
+    },
+    {
+      "epoch": 0.12756589516272981,
+      "grad_norm": 0.4333465547745647,
+      "learning_rate": 9.970481006825185e-06,
+      "loss": 0.4261,
+      "step": 536
+    },
+    {
+      "epoch": 0.1278038912357946,
+      "grad_norm": 0.4353660970004805,
+      "learning_rate": 9.970271489142127e-06,
+      "loss": 0.4315,
+      "step": 537
+    },
+    {
+      "epoch": 0.1280418873088594,
+      "grad_norm": 0.45342539645281116,
+      "learning_rate": 9.97006123275377e-06,
+      "loss": 0.4001,
+      "step": 538
+    },
+    {
+      "epoch": 0.1282798833819242,
+      "grad_norm": 0.44126076267178,
+      "learning_rate": 9.96985023769136e-06,
+      "loss": 0.3999,
+      "step": 539
+    },
+    {
+      "epoch": 0.128517879454989,
+      "grad_norm": 0.42194551508228195,
+      "learning_rate": 9.969638503986256e-06,
+      "loss": 0.4385,
+      "step": 540
+    },
+    {
+      "epoch": 0.1287558755280538,
+      "grad_norm": 0.4069592336480195,
+      "learning_rate": 9.969426031669928e-06,
+      "loss": 0.3883,
+      "step": 541
+    },
+    {
+      "epoch": 0.12899387160111858,
+      "grad_norm": 0.4178468875829833,
+      "learning_rate": 9.969212820773952e-06,
+      "loss": 0.3655,
+      "step": 542
+    },
+    {
+      "epoch": 0.12923186767418338,
+      "grad_norm": 0.4570434100827094,
+      "learning_rate": 9.968998871330021e-06,
+      "loss": 0.4334,
+      "step": 543
+    },
+    {
+      "epoch": 0.12946986374724817,
+      "grad_norm": 0.4166848553496299,
+      "learning_rate": 9.968784183369929e-06,
+      "loss": 0.4484,
+      "step": 544
+    },
+    {
+      "epoch": 0.12970785982031297,
+      "grad_norm": 0.42519002468336536,
+      "learning_rate": 9.968568756925588e-06,
+      "loss": 0.3469,
+      "step": 545
+    },
+    {
+      "epoch": 0.12994585589337776,
+      "grad_norm": 0.44962945270540455,
+      "learning_rate": 9.968352592029011e-06,
+      "loss": 0.3744,
+      "step": 546
+    },
+    {
+      "epoch": 0.13018385196644255,
+      "grad_norm": 0.40876516260944795,
+      "learning_rate": 9.968135688712328e-06,
+      "loss": 0.4431,
+      "step": 547
+    },
+    {
+      "epoch": 0.13042184803950735,
+      "grad_norm": 0.47209127285023517,
+      "learning_rate": 9.967918047007775e-06,
+      "loss": 0.3748,
+      "step": 548
+    },
+    {
+      "epoch": 0.13065984411257214,
+      "grad_norm": 0.4089317804173208,
+      "learning_rate": 9.967699666947702e-06,
+      "loss": 0.3127,
+      "step": 549
+    },
+    {
+      "epoch": 0.13089784018563694,
+      "grad_norm": 0.4235984982251744,
+      "learning_rate": 9.96748054856456e-06,
+      "loss": 0.4103,
+      "step": 550
+    },
+    {
+      "epoch": 0.13113583625870173,
+      "grad_norm": 0.47319011240693865,
+      "learning_rate": 9.967260691890924e-06,
+      "loss": 0.4523,
+      "step": 551
+    },
+    {
+      "epoch": 0.13137383233176653,
+      "grad_norm": 0.5145055531378172,
+      "learning_rate": 9.967040096959462e-06,
+      "loss": 0.3992,
+      "step": 552
+    },
+    {
+      "epoch": 0.13161182840483132,
+      "grad_norm": 0.4195472052055503,
+      "learning_rate": 9.966818763802963e-06,
+      "loss": 0.344,
+      "step": 553
+    },
+    {
+      "epoch": 0.13184982447789612,
+      "grad_norm": 0.43253895948369897,
+      "learning_rate": 9.966596692454323e-06,
+      "loss": 0.4628,
+      "step": 554
+    },
+    {
+      "epoch": 0.1320878205509609,
+      "grad_norm": 0.418918991234706,
+      "learning_rate": 9.966373882946546e-06,
+      "loss": 0.3967,
+      "step": 555
+    },
+    {
+      "epoch": 0.1323258166240257,
+      "grad_norm": 0.45069506384756325,
+      "learning_rate": 9.966150335312747e-06,
+      "loss": 0.3592,
+      "step": 556
+    },
+    {
+      "epoch": 0.1325638126970905,
+      "grad_norm": 0.4112688079328426,
+      "learning_rate": 9.965926049586154e-06,
+      "loss": 0.4051,
+      "step": 557
+    },
+    {
+      "epoch": 0.1328018087701553,
+      "grad_norm": 0.4306546680664368,
+      "learning_rate": 9.965701025800098e-06,
+      "loss": 0.4501,
+      "step": 558
+    },
+    {
+      "epoch": 0.1330398048432201,
+      "grad_norm": 0.47742995855119286,
+      "learning_rate": 9.965475263988024e-06,
+      "loss": 0.3709,
+      "step": 559
+    },
+    {
+      "epoch": 0.13327780091628488,
+      "grad_norm": 0.43252541349342494,
+      "learning_rate": 9.965248764183486e-06,
+      "loss": 0.3474,
+      "step": 560
+    },
+    {
+      "epoch": 0.13351579698934968,
+      "grad_norm": 0.4244516806580105,
+      "learning_rate": 9.965021526420146e-06,
+      "loss": 0.4157,
+      "step": 561
+    },
+    {
+      "epoch": 0.13375379306241447,
+      "grad_norm": 0.4501211384384804,
+      "learning_rate": 9.96479355073178e-06,
+      "loss": 0.4489,
+      "step": 562
+    },
+    {
+      "epoch": 0.13399178913547927,
+      "grad_norm": 0.433684978151453,
+      "learning_rate": 9.964564837152268e-06,
+      "loss": 0.3325,
+      "step": 563
+    },
+    {
+      "epoch": 0.13422978520854406,
+      "grad_norm": 0.4143350623973339,
+      "learning_rate": 9.964335385715607e-06,
+      "loss": 0.4103,
+      "step": 564
+    },
+    {
+      "epoch": 0.13446778128160886,
+      "grad_norm": 0.4356764057938835,
+      "learning_rate": 9.964105196455892e-06,
+      "loss": 0.4426,
+      "step": 565
+    },
+    {
+      "epoch": 0.13470577735467365,
+      "grad_norm": 0.43904123994773603,
+      "learning_rate": 9.963874269407342e-06,
+      "loss": 0.3823,
+      "step": 566
+    },
+    {
+      "epoch": 0.13494377342773844,
+      "grad_norm": 0.4195562006737919,
+      "learning_rate": 9.963642604604273e-06,
+      "loss": 0.358,
+      "step": 567
+    },
+    {
+      "epoch": 0.13518176950080324,
+      "grad_norm": 0.4134567373776786,
+      "learning_rate": 9.963410202081118e-06,
+      "loss": 0.3699,
+      "step": 568
+    },
+    {
+      "epoch": 0.13541976557386803,
+      "grad_norm": 0.4677127657760346,
+      "learning_rate": 9.96317706187242e-06,
+      "loss": 0.4375,
+      "step": 569
+    },
+    {
+      "epoch": 0.13565776164693283,
+      "grad_norm": 0.459487545289008,
+      "learning_rate": 9.962943184012826e-06,
+      "loss": 0.3227,
+      "step": 570
+    },
+    {
+      "epoch": 0.13589575771999762,
+      "grad_norm": 0.46349644278250035,
+      "learning_rate": 9.962708568537099e-06,
+      "loss": 0.382,
+      "step": 571
+    },
+    {
+      "epoch": 0.13613375379306242,
+      "grad_norm": 0.42793811916759666,
+      "learning_rate": 9.962473215480106e-06,
+      "loss": 0.4328,
+      "step": 572
+    },
+    {
+      "epoch": 0.1363717498661272,
+      "grad_norm": 0.4132709478039074,
+      "learning_rate": 9.962237124876828e-06,
+      "loss": 0.4126,
+      "step": 573
+    },
+    {
+      "epoch": 0.136609745939192,
+      "grad_norm": 0.4876228484575214,
+      "learning_rate": 9.962000296762352e-06,
+      "loss": 0.3429,
+      "step": 574
+    },
+    {
+      "epoch": 0.1368477420122568,
+      "grad_norm": 0.3977015514378766,
+      "learning_rate": 9.96176273117188e-06,
+      "loss": 0.4365,
+      "step": 575
+    },
+    {
+      "epoch": 0.1370857380853216,
+      "grad_norm": 0.4078608665053635,
+      "learning_rate": 9.961524428140716e-06,
+      "loss": 0.4907,
+      "step": 576
+    },
+    {
+      "epoch": 0.1373237341583864,
+      "grad_norm": 0.43371567468393196,
+      "learning_rate": 9.961285387704283e-06,
+      "loss": 0.3612,
+      "step": 577
+    },
+    {
+      "epoch": 0.13756173023145118,
+      "grad_norm": 0.4879930467640949,
+      "learning_rate": 9.961045609898103e-06,
+      "loss": 0.3276,
+      "step": 578
+    },
+    {
+      "epoch": 0.13779972630451598,
+      "grad_norm": 0.40451489231885396,
+      "learning_rate": 9.960805094757815e-06,
+      "loss": 0.4356,
+      "step": 579
+    },
+    {
+      "epoch": 0.13803772237758077,
+      "grad_norm": 0.44097468379761057,
+      "learning_rate": 9.960563842319164e-06,
+      "loss": 0.4218,
+      "step": 580
+    },
+    {
+      "epoch": 0.13827571845064557,
+      "grad_norm": 0.5834242401504829,
+      "learning_rate": 9.96032185261801e-06,
+      "loss": 0.3877,
+      "step": 581
+    },
+    {
+      "epoch": 0.13851371452371036,
+      "grad_norm": 0.45653361970962053,
+      "learning_rate": 9.960079125690317e-06,
+      "loss": 0.3879,
+      "step": 582
+    },
+    {
+      "epoch": 0.13875171059677516,
+      "grad_norm": 0.4476140474436517,
+      "learning_rate": 9.959835661572158e-06,
+      "loss": 0.4525,
+      "step": 583
+    },
+    {
+      "epoch": 0.13898970666983995,
+      "grad_norm": 0.4543506355962775,
+      "learning_rate": 9.959591460299719e-06,
+      "loss": 0.4012,
+      "step": 584
+    },
+    {
+      "epoch": 0.13922770274290475,
+      "grad_norm": 0.4800146363435457,
+      "learning_rate": 9.959346521909295e-06,
+      "loss": 0.3524,
+      "step": 585
+    },
+    {
+      "epoch": 0.13946569881596954,
+      "grad_norm": 0.4366124496205618,
+      "learning_rate": 9.95910084643729e-06,
+      "loss": 0.4189,
+      "step": 586
+    },
+    {
+      "epoch": 0.13970369488903434,
+      "grad_norm": 0.44571391749959416,
+      "learning_rate": 9.958854433920215e-06,
+      "loss": 0.4703,
+      "step": 587
+    },
+    {
+      "epoch": 0.13994169096209913,
+      "grad_norm": 0.49239121417665377,
+      "learning_rate": 9.958607284394696e-06,
+      "loss": 0.343,
+      "step": 588
+    },
+    {
+      "epoch": 0.14017968703516392,
+      "grad_norm": 0.4317901276773715,
+      "learning_rate": 9.958359397897465e-06,
+      "loss": 0.3861,
+      "step": 589
+    },
+    {
+      "epoch": 0.14041768310822872,
+      "grad_norm": 0.4147928486152337,
+      "learning_rate": 9.958110774465364e-06,
+      "loss": 0.4516,
+      "step": 590
+    },
+    {
+      "epoch": 0.1406556791812935,
+      "grad_norm": 0.4611062418346612,
+      "learning_rate": 9.957861414135343e-06,
+      "loss": 0.3849,
+      "step": 591
+    },
+    {
+      "epoch": 0.1408936752543583,
+      "grad_norm": 0.5052327412000163,
+      "learning_rate": 9.957611316944465e-06,
+      "loss": 0.4016,
+      "step": 592
+    },
+    {
+      "epoch": 0.1411316713274231,
+      "grad_norm": 0.4732644464070684,
+      "learning_rate": 9.957360482929898e-06,
+      "loss": 0.3911,
+      "step": 593
+    },
+    {
+      "epoch": 0.1413696674004879,
+      "grad_norm": 0.45055177739669783,
+      "learning_rate": 9.957108912128927e-06,
+      "loss": 0.4721,
+      "step": 594
+    },
+    {
+      "epoch": 0.1416076634735527,
+      "grad_norm": 0.5207702754193928,
+      "learning_rate": 9.956856604578937e-06,
+      "loss": 0.3663,
+      "step": 595
+    },
+    {
+      "epoch": 0.1418456595466175,
+      "grad_norm": 0.4532039740900448,
+      "learning_rate": 9.95660356031743e-06,
+      "loss": 0.3646,
+      "step": 596
+    },
+    {
+      "epoch": 0.14208365561968228,
+      "grad_norm": 0.45255299216112244,
+      "learning_rate": 9.956349779382014e-06,
+      "loss": 0.4761,
+      "step": 597
+    },
+    {
+      "epoch": 0.14232165169274708,
+      "grad_norm": 0.4294348039394876,
+      "learning_rate": 9.956095261810404e-06,
+      "loss": 0.4312,
+      "step": 598
+    },
+    {
+      "epoch": 0.14255964776581187,
+      "grad_norm": 0.4145607881038628,
+      "learning_rate": 9.955840007640432e-06,
+      "loss": 0.3167,
+      "step": 599
+    },
+    {
+      "epoch": 0.14279764383887666,
+      "grad_norm": 0.44934929004475394,
+      "learning_rate": 9.955584016910033e-06,
+      "loss": 0.3974,
+      "step": 600
+    },
+    {
+      "epoch": 0.14303563991194146,
+      "grad_norm": 0.39995571799722107,
+      "learning_rate": 9.955327289657253e-06,
+      "loss": 0.4574,
+      "step": 601
+    },
+    {
+      "epoch": 0.14327363598500625,
+      "grad_norm": 0.4365868901329556,
+      "learning_rate": 9.955069825920249e-06,
+      "loss": 0.3627,
+      "step": 602
+    },
+    {
+      "epoch": 0.14351163205807105,
+      "grad_norm": 0.4542607300367256,
+      "learning_rate": 9.954811625737289e-06,
+      "loss": 0.3208,
+      "step": 603
+    },
+    {
+      "epoch": 0.14374962813113584,
+      "grad_norm": 0.42854644501077194,
+      "learning_rate": 9.954552689146743e-06,
+      "loss": 0.4154,
+      "step": 604
+    },
+    {
+      "epoch": 0.14398762420420064,
+      "grad_norm": 0.43253450331881993,
+      "learning_rate": 9.954293016187098e-06,
+      "loss": 0.4209,
+      "step": 605
+    },
+    {
+      "epoch": 0.14422562027726543,
+      "grad_norm": 0.4283765316302236,
+      "learning_rate": 9.954032606896946e-06,
+      "loss": 0.3312,
+      "step": 606
+    },
+    {
+      "epoch": 0.14446361635033023,
+      "grad_norm": 0.4341091606907487,
+      "learning_rate": 9.953771461314994e-06,
+      "loss": 0.4196,
+      "step": 607
+    },
+    {
+      "epoch": 0.14470161242339502,
+      "grad_norm": 0.41550595013594926,
+      "learning_rate": 9.953509579480052e-06,
+      "loss": 0.4728,
+      "step": 608
+    },
+    {
+      "epoch": 0.14493960849645982,
+      "grad_norm": 0.41104915967555633,
+      "learning_rate": 9.953246961431043e-06,
+      "loss": 0.3532,
+      "step": 609
+    },
+    {
+      "epoch": 0.1451776045695246,
+      "grad_norm": 0.44491387505754176,
+      "learning_rate": 9.952983607206996e-06,
+      "loss": 0.3567,
+      "step": 610
+    },
+    {
+      "epoch": 0.1454156006425894,
+      "grad_norm": 0.41859179197947394,
+      "learning_rate": 9.952719516847055e-06,
+      "loss": 0.4096,
+      "step": 611
+    },
+    {
+      "epoch": 0.1456535967156542,
+      "grad_norm": 0.422145361038327,
+      "learning_rate": 9.95245469039047e-06,
+      "loss": 0.4246,
+      "step": 612
+    },
+    {
+      "epoch": 0.145891592788719,
+      "grad_norm": 0.4500980509847598,
+      "learning_rate": 9.9521891278766e-06,
+      "loss": 0.3426,
+      "step": 613
+    },
+    {
+      "epoch": 0.1461295888617838,
+      "grad_norm": 0.43000982194705367,
+      "learning_rate": 9.951922829344914e-06,
+      "loss": 0.3946,
+      "step": 614
+    },
+    {
+      "epoch": 0.14636758493484858,
+      "grad_norm": 0.41053172531397664,
+      "learning_rate": 9.951655794834991e-06,
+      "loss": 0.4663,
+      "step": 615
+    },
+    {
+      "epoch": 0.14660558100791338,
+      "grad_norm": 0.3923350216414785,
+      "learning_rate": 9.951388024386519e-06,
+      "loss": 0.3757,
+      "step": 616
+    },
+    {
+      "epoch": 0.14684357708097817,
+      "grad_norm": 0.4599433113284599,
+      "learning_rate": 9.951119518039297e-06,
+      "loss": 0.3385,
+      "step": 617
+    },
+    {
+      "epoch": 0.14708157315404297,
+      "grad_norm": 0.39908104730389854,
+      "learning_rate": 9.950850275833226e-06,
+      "loss": 0.4074,
+      "step": 618
+    },
+    {
+      "epoch": 0.14731956922710776,
+      "grad_norm": 0.4511403028324386,
+      "learning_rate": 9.950580297808329e-06,
+      "loss": 0.4194,
+      "step": 619
+    },
+    {
+      "epoch": 0.14755756530017256,
+      "grad_norm": 0.4617542670919041,
+      "learning_rate": 9.950309584004728e-06,
+      "loss": 0.3355,
+      "step": 620
+    },
+    {
+      "epoch": 0.14779556137323735,
+      "grad_norm": 0.4226130485474674,
+      "learning_rate": 9.950038134462655e-06,
+      "loss": 0.3681,
+      "step": 621
+    },
+    {
+      "epoch": 0.14803355744630214,
+      "grad_norm": 0.4247694573529077,
+      "learning_rate": 9.949765949222461e-06,
+      "loss": 0.4761,
+      "step": 622
+    },
+    {
+      "epoch": 0.14827155351936694,
+      "grad_norm": 0.4516914981541404,
+      "learning_rate": 9.949493028324593e-06,
+      "loss": 0.417,
+      "step": 623
+    },
+    {
+      "epoch": 0.14850954959243173,
+      "grad_norm": 0.45995529369267013,
+      "learning_rate": 9.949219371809618e-06,
+      "loss": 0.355,
+      "step": 624
+    },
+    {
+      "epoch": 0.14874754566549653,
+      "grad_norm": 0.4610870216065006,
+      "learning_rate": 9.948944979718206e-06,
+      "loss": 0.4184,
+      "step": 625
+    },
+    {
+      "epoch": 0.14898554173856132,
+      "grad_norm": 0.46221460180343904,
+      "learning_rate": 9.94866985209114e-06,
+      "loss": 0.4689,
+      "step": 626
+    },
+    {
+      "epoch": 0.14922353781162612,
+      "grad_norm": 0.4241292231021412,
+      "learning_rate": 9.948393988969307e-06,
+      "loss": 0.3911,
+      "step": 627
+    },
+    {
+      "epoch": 0.1494615338846909,
+      "grad_norm": 0.40337434368955516,
+      "learning_rate": 9.948117390393713e-06,
+      "loss": 0.3635,
+      "step": 628
+    },
+    {
+      "epoch": 0.1496995299577557,
+      "grad_norm": 0.40666940123628886,
+      "learning_rate": 9.947840056405461e-06,
+      "loss": 0.4265,
+      "step": 629
+    },
+    {
+      "epoch": 0.1499375260308205,
+      "grad_norm": 0.408720859695881,
+      "learning_rate": 9.947561987045777e-06,
+      "loss": 0.4675,
+      "step": 630
+    },
+    {
+      "epoch": 0.1501755221038853,
+      "grad_norm": 0.46059437009326065,
+      "learning_rate": 9.947283182355982e-06,
+      "loss": 0.3846,
+      "step": 631
+    },
+    {
+      "epoch": 0.1504135181769501,
+      "grad_norm": 0.40219876977724317,
+      "learning_rate": 9.947003642377517e-06,
+      "loss": 0.3583,
+      "step": 632
+    },
+    {
+      "epoch": 0.15065151425001488,
+      "grad_norm": 0.40741994114516356,
+      "learning_rate": 9.946723367151929e-06,
+      "loss": 0.4467,
+      "step": 633
+    },
+    {
+      "epoch": 0.15088951032307968,
+      "grad_norm": 0.4170974461418337,
+      "learning_rate": 9.94644235672087e-06,
+      "loss": 0.3964,
+      "step": 634
+    },
+    {
+      "epoch": 0.15112750639614447,
+      "grad_norm": 0.3959538443481256,
+      "learning_rate": 9.94616061112611e-06,
+      "loss": 0.3624,
+      "step": 635
+    },
+    {
+      "epoch": 0.15136550246920927,
+      "grad_norm": 0.41399916071040127,
+      "learning_rate": 9.94587813040952e-06,
+      "loss": 0.3988,
+      "step": 636
+    },
+    {
+      "epoch": 0.15160349854227406,
+      "grad_norm": 0.42565733949269147,
+      "learning_rate": 9.945594914613085e-06,
+      "loss": 0.4244,
+      "step": 637
+    },
+    {
+      "epoch": 0.15184149461533886,
+      "grad_norm": 0.4176195376045644,
+      "learning_rate": 9.945310963778897e-06,
+      "loss": 0.3422,
+      "step": 638
+    },
+    {
+      "epoch": 0.15207949068840365,
+      "grad_norm": 0.4091797435607871,
+      "learning_rate": 9.945026277949159e-06,
+      "loss": 0.391,
+      "step": 639
+    },
+    {
+      "epoch": 0.15231748676146845,
+      "grad_norm": 0.3752152693514474,
+      "learning_rate": 9.944740857166181e-06,
+      "loss": 0.4218,
+      "step": 640
+    },
+    {
+      "epoch": 0.15255548283453324,
+      "grad_norm": 0.7050883051614025,
+      "learning_rate": 9.944454701472387e-06,
+      "loss": 0.3827,
+      "step": 641
+    },
+    {
+      "epoch": 0.15279347890759803,
+      "grad_norm": 0.4433283675728129,
+      "learning_rate": 9.944167810910304e-06,
+      "loss": 0.3533,
+      "step": 642
+    },
+    {
+      "epoch": 0.15303147498066283,
+      "grad_norm": 0.42554987379842735,
+      "learning_rate": 9.94388018552257e-06,
+      "loss": 0.4089,
+      "step": 643
+    },
+    {
+      "epoch": 0.15326947105372762,
+      "grad_norm": 0.45679111789324556,
+      "learning_rate": 9.943591825351934e-06,
+      "loss": 0.4456,
+      "step": 644
+    },
+    {
+      "epoch": 0.15350746712679242,
+      "grad_norm": 0.42999137088818135,
+      "learning_rate": 9.943302730441258e-06,
+      "loss": 0.3426,
+      "step": 645
+    },
+    {
+      "epoch": 0.1537454631998572,
+      "grad_norm": 0.43170003868689516,
+      "learning_rate": 9.943012900833503e-06,
+      "loss": 0.3613,
+      "step": 646
+    },
+    {
+      "epoch": 0.153983459272922,
+      "grad_norm": 0.40514208501535537,
+      "learning_rate": 9.942722336571746e-06,
+      "loss": 0.459,
+      "step": 647
+    },
+    {
+      "epoch": 0.1542214553459868,
+      "grad_norm": 0.42503224905946857,
+      "learning_rate": 9.942431037699171e-06,
+      "loss": 0.4123,
+      "step": 648
+    },
+    {
+      "epoch": 0.1544594514190516,
+      "grad_norm": 0.43534422977552745,
+      "learning_rate": 9.942139004259077e-06,
+      "loss": 0.3489,
+      "step": 649
+    },
+    {
+      "epoch": 0.1546974474921164,
+      "grad_norm": 0.4042874898448957,
+      "learning_rate": 9.941846236294863e-06,
+      "loss": 0.417,
+      "step": 650
+    },
+    {
+      "epoch": 0.15493544356518119,
+      "grad_norm": 0.41685565190135837,
+      "learning_rate": 9.941552733850044e-06,
+      "loss": 0.4741,
+      "step": 651
+    },
+    {
+      "epoch": 0.15517343963824598,
+      "grad_norm": 0.41165103596561803,
+      "learning_rate": 9.941258496968238e-06,
+      "loss": 0.3744,
+      "step": 652
+    },
+    {
+      "epoch": 0.15541143571131077,
+      "grad_norm": 0.4215966209056301,
+      "learning_rate": 9.940963525693181e-06,
+      "loss": 0.3437,
+      "step": 653
+    },
+    {
+      "epoch": 0.15564943178437557,
+      "grad_norm": 0.4524386189484754,
+      "learning_rate": 9.94066782006871e-06,
+      "loss": 0.4291,
+      "step": 654
+    },
+    {
+      "epoch": 0.15588742785744036,
+      "grad_norm": 0.43482975599739715,
+      "learning_rate": 9.940371380138774e-06,
+      "loss": 0.4164,
+      "step": 655
+    },
+    {
+      "epoch": 0.15612542393050516,
+      "grad_norm": 0.4120821183177027,
+      "learning_rate": 9.940074205947432e-06,
+      "loss": 0.3363,
+      "step": 656
+    },
+    {
+      "epoch": 0.15636342000356995,
+      "grad_norm": 0.4309229287404229,
+      "learning_rate": 9.939776297538853e-06,
+      "loss": 0.389,
+      "step": 657
+    },
+    {
+      "epoch": 0.15660141607663475,
+      "grad_norm": 0.416688460486632,
+      "learning_rate": 9.93947765495731e-06,
+      "loss": 0.4333,
+      "step": 658
+    },
+    {
+      "epoch": 0.15683941214969954,
+      "grad_norm": 0.3965370666176588,
+      "learning_rate": 9.939178278247192e-06,
+      "loss": 0.3478,
+      "step": 659
+    },
+    {
+      "epoch": 0.15707740822276434,
+      "grad_norm": 0.439104289476318,
+      "learning_rate": 9.938878167452991e-06,
+      "loss": 0.3427,
+      "step": 660
+    },
+    {
+      "epoch": 0.15731540429582913,
+      "grad_norm": 0.398185404048493,
+      "learning_rate": 9.938577322619315e-06,
+      "loss": 0.3966,
+      "step": 661
+    },
+    {
+      "epoch": 0.15755340036889393,
+      "grad_norm": 0.4488481583859823,
+      "learning_rate": 9.938275743790872e-06,
+      "loss": 0.4251,
+      "step": 662
+    },
+    {
+      "epoch": 0.15779139644195872,
+      "grad_norm": 0.4417275681933899,
+      "learning_rate": 9.937973431012488e-06,
+      "loss": 0.3735,
+      "step": 663
+    },
+    {
+      "epoch": 0.15802939251502351,
+      "grad_norm": 0.4693818789749731,
+      "learning_rate": 9.937670384329092e-06,
+      "loss": 0.3718,
+      "step": 664
+    },
+    {
+      "epoch": 0.1582673885880883,
+      "grad_norm": 0.44060498282158367,
+      "learning_rate": 9.937366603785725e-06,
+      "loss": 0.4519,
+      "step": 665
+    },
+    {
+      "epoch": 0.1585053846611531,
+      "grad_norm": 0.4132724605397321,
+      "learning_rate": 9.937062089427534e-06,
+      "loss": 0.3665,
+      "step": 666
+    },
+    {
+      "epoch": 0.1587433807342179,
+      "grad_norm": 0.43487153487168945,
+      "learning_rate": 9.936756841299782e-06,
+      "loss": 0.3501,
+      "step": 667
+    },
+    {
+      "epoch": 0.1589813768072827,
+      "grad_norm": 0.4454465202025332,
+      "learning_rate": 9.936450859447833e-06,
+      "loss": 0.4349,
+      "step": 668
+    },
+    {
+      "epoch": 0.1592193728803475,
+      "grad_norm": 0.43517274964693736,
+      "learning_rate": 9.936144143917164e-06,
+      "loss": 0.4639,
+      "step": 669
+    },
+    {
+      "epoch": 0.15945736895341228,
+      "grad_norm": 0.4325185752961809,
+      "learning_rate": 9.935836694753363e-06,
+      "loss": 0.355,
+      "step": 670
+    },
+    {
+      "epoch": 0.15969536502647708,
+      "grad_norm": 0.41185000783549897,
+      "learning_rate": 9.93552851200212e-06,
+      "loss": 0.3643,
+      "step": 671
+    },
+    {
+      "epoch": 0.15993336109954187,
+      "grad_norm": 0.44265593925954544,
+      "learning_rate": 9.935219595709242e-06,
+      "loss": 0.4427,
+      "step": 672
+    },
+    {
+      "epoch": 0.16017135717260667,
+      "grad_norm": 0.47046598610709517,
+      "learning_rate": 9.93490994592064e-06,
+      "loss": 0.3921,
+      "step": 673
+    },
+    {
+      "epoch": 0.16040935324567146,
+      "grad_norm": 0.427596260881768,
+      "learning_rate": 9.934599562682337e-06,
+      "loss": 0.3758,
+      "step": 674
+    },
+    {
+      "epoch": 0.16064734931873625,
+      "grad_norm": 0.42891689676165107,
+      "learning_rate": 9.934288446040462e-06,
+      "loss": 0.3723,
+      "step": 675
+    },
+    {
+      "epoch": 0.16088534539180105,
+      "grad_norm": 0.40195671977273745,
+      "learning_rate": 9.933976596041257e-06,
+      "loss": 0.4701,
+      "step": 676
+    },
+    {
+      "epoch": 0.16112334146486584,
+      "grad_norm": 0.43868503957164495,
+      "learning_rate": 9.933664012731067e-06,
+      "loss": 0.3516,
+      "step": 677
+    },
+    {
+      "epoch": 0.1613613375379306,
+      "grad_norm": 0.45694726273935127,
+      "learning_rate": 9.933350696156354e-06,
+      "loss": 0.3707,
+      "step": 678
+    },
+    {
+      "epoch": 0.1615993336109954,
+      "grad_norm": 0.42986712701679414,
+      "learning_rate": 9.933036646363681e-06,
+      "loss": 0.4296,
+      "step": 679
+    },
+    {
+      "epoch": 0.1618373296840602,
+      "grad_norm": 0.38103639863281863,
+      "learning_rate": 9.932721863399726e-06,
+      "loss": 0.4153,
+      "step": 680
+    },
+    {
+      "epoch": 0.162075325757125,
+      "grad_norm": 0.4320260369164701,
+      "learning_rate": 9.93240634731127e-06,
+      "loss": 0.3337,
+      "step": 681
+    },
+    {
+      "epoch": 0.1623133218301898,
+      "grad_norm": 0.42202425560257006,
+      "learning_rate": 9.93209009814521e-06,
+      "loss": 0.3749,
+      "step": 682
+    },
+    {
+      "epoch": 0.16255131790325458,
+      "grad_norm": 0.39918790424517825,
+      "learning_rate": 9.93177311594855e-06,
+      "loss": 0.4151,
+      "step": 683
+    },
+    {
+      "epoch": 0.16278931397631938,
+      "grad_norm": 0.41085741596254716,
+      "learning_rate": 9.931455400768396e-06,
+      "loss": 0.3713,
+      "step": 684
+    },
+    {
+      "epoch": 0.16302731004938417,
+      "grad_norm": 0.4935289189499038,
+      "learning_rate": 9.931136952651971e-06,
+      "loss": 0.3496,
+      "step": 685
+    },
+    {
+      "epoch": 0.16326530612244897,
+      "grad_norm": 0.45454318553894524,
+      "learning_rate": 9.930817771646606e-06,
+      "loss": 0.4423,
+      "step": 686
+    },
+    {
+      "epoch": 0.16350330219551376,
+      "grad_norm": 0.48363104373210164,
+      "learning_rate": 9.930497857799737e-06,
+      "loss": 0.4216,
+      "step": 687
+    },
+    {
+      "epoch": 0.16374129826857856,
+      "grad_norm": 0.4024082102755574,
+      "learning_rate": 9.93017721115891e-06,
+      "loss": 0.3559,
+      "step": 688
+    },
+    {
+      "epoch": 0.16397929434164335,
+      "grad_norm": 0.4423776261521702,
+      "learning_rate": 9.929855831771787e-06,
+      "loss": 0.3834,
+      "step": 689
+    },
+    {
+      "epoch": 0.16421729041470814,
+      "grad_norm": 0.44126311420707715,
+      "learning_rate": 9.929533719686123e-06,
+      "loss": 0.4315,
+      "step": 690
+    },
+    {
+      "epoch": 0.16445528648777294,
+      "grad_norm": 0.4387429322912089,
+      "learning_rate": 9.929210874949802e-06,
+      "loss": 0.3487,
+      "step": 691
+    },
+    {
+      "epoch": 0.16469328256083773,
+      "grad_norm": 0.48730072451609946,
+      "learning_rate": 9.928887297610801e-06,
+      "loss": 0.3254,
+      "step": 692
+    },
+    {
+      "epoch": 0.16493127863390253,
+      "grad_norm": 0.4229571083939119,
+      "learning_rate": 9.928562987717211e-06,
+      "loss": 0.3936,
+      "step": 693
+    },
+    {
+      "epoch": 0.16516927470696732,
+      "grad_norm": 0.3987563067271361,
+      "learning_rate": 9.928237945317237e-06,
+      "loss": 0.4242,
+      "step": 694
+    },
+    {
+      "epoch": 0.16540727078003212,
+      "grad_norm": 0.4566722968242409,
+      "learning_rate": 9.927912170459183e-06,
+      "loss": 0.346,
+      "step": 695
+    },
+    {
+      "epoch": 0.1656452668530969,
+      "grad_norm": 0.46544117565173826,
+      "learning_rate": 9.927585663191472e-06,
+      "loss": 0.377,
+      "step": 696
+    },
+    {
+      "epoch": 0.1658832629261617,
+      "grad_norm": 0.42065889884419416,
+      "learning_rate": 9.927258423562628e-06,
+      "loss": 0.4574,
+      "step": 697
+    },
+    {
+      "epoch": 0.1661212589992265,
+      "grad_norm": 0.41117147375225527,
+      "learning_rate": 9.926930451621286e-06,
+      "loss": 0.411,
+      "step": 698
+    },
+    {
+      "epoch": 0.1663592550722913,
+      "grad_norm": 0.41917410595753857,
+      "learning_rate": 9.926601747416194e-06,
+      "loss": 0.3413,
+      "step": 699
+    },
+    {
+      "epoch": 0.1665972511453561,
+      "grad_norm": 0.513199643744508,
+      "learning_rate": 9.926272310996205e-06,
+      "loss": 0.4063,
+      "step": 700
+    },
+    {
+      "epoch": 0.16683524721842088,
+      "grad_norm": 0.42672332270069796,
+      "learning_rate": 9.92594214241028e-06,
+      "loss": 0.4861,
+      "step": 701
+    },
+    {
+      "epoch": 0.16707324329148568,
+      "grad_norm": 0.42431192157186515,
+      "learning_rate": 9.925611241707491e-06,
+      "loss": 0.3939,
+      "step": 702
+    },
+    {
+      "epoch": 0.16731123936455047,
+      "grad_norm": 0.44467532229285445,
+      "learning_rate": 9.925279608937014e-06,
+      "loss": 0.3696,
+      "step": 703
+    },
+    {
+      "epoch": 0.16754923543761527,
+      "grad_norm": 0.4515098096528452,
+      "learning_rate": 9.924947244148147e-06,
+      "loss": 0.4285,
+      "step": 704
+    },
+    {
+      "epoch": 0.16778723151068006,
+      "grad_norm": 0.5043978783638837,
+      "learning_rate": 9.924614147390278e-06,
+      "loss": 0.407,
+      "step": 705
+    },
+    {
+      "epoch": 0.16802522758374486,
+      "grad_norm": 0.44962332273904415,
+      "learning_rate": 9.92428031871292e-06,
+      "loss": 0.3473,
+      "step": 706
+    },
+    {
+      "epoch": 0.16826322365680965,
+      "grad_norm": 0.4212768922385841,
+      "learning_rate": 9.923945758165686e-06,
+      "loss": 0.3899,
+      "step": 707
+    },
+    {
+      "epoch": 0.16850121972987445,
+      "grad_norm": 0.42396104130969997,
+      "learning_rate": 9.923610465798298e-06,
+      "loss": 0.4148,
+      "step": 708
+    },
+    {
+      "epoch": 0.16873921580293924,
+      "grad_norm": 0.43780183253810606,
+      "learning_rate": 9.923274441660593e-06,
+      "loss": 0.3524,
+      "step": 709
+    },
+    {
+      "epoch": 0.16897721187600404,
+      "grad_norm": 0.4610507977132175,
+      "learning_rate": 9.922937685802508e-06,
+      "loss": 0.3738,
+      "step": 710
+    },
+    {
+      "epoch": 0.16921520794906883,
+      "grad_norm": 0.41170477402274724,
+      "learning_rate": 9.922600198274096e-06,
+      "loss": 0.4127,
+      "step": 711
+    },
+    {
+      "epoch": 0.16945320402213362,
+      "grad_norm": 0.42904482538288863,
+      "learning_rate": 9.922261979125516e-06,
+      "loss": 0.4375,
+      "step": 712
+    },
+    {
+      "epoch": 0.16969120009519842,
+      "grad_norm": 0.4288980919174409,
+      "learning_rate": 9.921923028407034e-06,
+      "loss": 0.3382,
+      "step": 713
+    },
+    {
+      "epoch": 0.1699291961682632,
+      "grad_norm": 0.400850498314156,
+      "learning_rate": 9.921583346169027e-06,
+      "loss": 0.3488,
+      "step": 714
+    },
+    {
+      "epoch": 0.170167192241328,
+      "grad_norm": 0.46108960454734477,
+      "learning_rate": 9.92124293246198e-06,
+      "loss": 0.4419,
+      "step": 715
+    },
+    {
+      "epoch": 0.1704051883143928,
+      "grad_norm": 0.41992077014825,
+      "learning_rate": 9.92090178733649e-06,
+      "loss": 0.4189,
+      "step": 716
+    },
+    {
+      "epoch": 0.1706431843874576,
+      "grad_norm": 0.48252193952318556,
+      "learning_rate": 9.920559910843255e-06,
+      "loss": 0.3313,
+      "step": 717
+    },
+    {
+      "epoch": 0.1708811804605224,
+      "grad_norm": 0.40035812359955575,
+      "learning_rate": 9.920217303033091e-06,
+      "loss": 0.4049,
+      "step": 718
+    },
+    {
+      "epoch": 0.1711191765335872,
+      "grad_norm": 0.4431295892233984,
+      "learning_rate": 9.919873963956914e-06,
+      "loss": 0.4627,
+      "step": 719
+    },
+    {
+      "epoch": 0.17135717260665198,
+      "grad_norm": 0.4546851162600903,
+      "learning_rate": 9.919529893665753e-06,
+      "loss": 0.3459,
+      "step": 720
+    },
+    {
+      "epoch": 0.17159516867971678,
+      "grad_norm": 0.41399727681843856,
+      "learning_rate": 9.919185092210748e-06,
+      "loss": 0.3504,
+      "step": 721
+    },
+    {
+      "epoch": 0.17183316475278157,
+      "grad_norm": 0.39447751341186205,
+      "learning_rate": 9.918839559643143e-06,
+      "loss": 0.4322,
+      "step": 722
+    },
+    {
+      "epoch": 0.17207116082584636,
+      "grad_norm": 0.4316965191862936,
+      "learning_rate": 9.918493296014294e-06,
+      "loss": 0.3914,
+      "step": 723
+    },
+    {
+      "epoch": 0.17230915689891116,
+      "grad_norm": 0.4190761857592741,
+      "learning_rate": 9.918146301375663e-06,
+      "loss": 0.3526,
+      "step": 724
+    },
+    {
+      "epoch": 0.17254715297197595,
+      "grad_norm": 0.4405259759469165,
+      "learning_rate": 9.917798575778821e-06,
+      "loss": 0.405,
+      "step": 725
+    },
+    {
+      "epoch": 0.17278514904504075,
+      "grad_norm": 0.39930877974901136,
+      "learning_rate": 9.917450119275452e-06,
+      "loss": 0.4576,
+      "step": 726
+    },
+    {
+      "epoch": 0.17302314511810554,
+      "grad_norm": 0.4312658018371688,
+      "learning_rate": 9.917100931917343e-06,
+      "loss": 0.3494,
+      "step": 727
+    },
+    {
+      "epoch": 0.17326114119117034,
+      "grad_norm": 0.3952105575090353,
+      "learning_rate": 9.916751013756393e-06,
+      "loss": 0.3349,
+      "step": 728
+    },
+    {
+      "epoch": 0.17349913726423513,
+      "grad_norm": 0.4114863899271299,
+      "learning_rate": 9.916400364844608e-06,
+      "loss": 0.4229,
+      "step": 729
+    },
+    {
+      "epoch": 0.17373713333729993,
+      "grad_norm": 0.42634304050454525,
+      "learning_rate": 9.916048985234102e-06,
+      "loss": 0.4014,
+      "step": 730
+    },
+    {
+      "epoch": 0.17397512941036472,
+      "grad_norm": 0.46609894460053153,
+      "learning_rate": 9.9156968749771e-06,
+      "loss": 0.3438,
+      "step": 731
+    },
+    {
+      "epoch": 0.17421312548342952,
+      "grad_norm": 0.4204390869113515,
+      "learning_rate": 9.915344034125931e-06,
+      "loss": 0.4,
+      "step": 732
+    },
+    {
+      "epoch": 0.1744511215564943,
+      "grad_norm": 0.411331032799109,
+      "learning_rate": 9.914990462733042e-06,
+      "loss": 0.4429,
+      "step": 733
+    },
+    {
+      "epoch": 0.1746891176295591,
+      "grad_norm": 0.4311784563828495,
+      "learning_rate": 9.914636160850979e-06,
+      "loss": 0.3685,
+      "step": 734
+    },
+    {
+      "epoch": 0.1749271137026239,
+      "grad_norm": 0.43611688575958385,
+      "learning_rate": 9.914281128532399e-06,
+      "loss": 0.3606,
+      "step": 735
+    },
+    {
+      "epoch": 0.1751651097756887,
+      "grad_norm": 0.4083035624121461,
+      "learning_rate": 9.91392536583007e-06,
+      "loss": 0.4403,
+      "step": 736
+    },
+    {
+      "epoch": 0.1754031058487535,
+      "grad_norm": 0.4515206561127267,
+      "learning_rate": 9.913568872796867e-06,
+      "loss": 0.4095,
+      "step": 737
+    },
+    {
+      "epoch": 0.17564110192181828,
+      "grad_norm": 0.39490410731530035,
+      "learning_rate": 9.913211649485776e-06,
+      "loss": 0.3601,
+      "step": 738
+    },
+    {
+      "epoch": 0.17587909799488308,
+      "grad_norm": 0.4383233004999705,
+      "learning_rate": 9.912853695949884e-06,
+      "loss": 0.3656,
+      "step": 739
+    },
+    {
+      "epoch": 0.17611709406794787,
+      "grad_norm": 0.4395334470031347,
+      "learning_rate": 9.912495012242396e-06,
+      "loss": 0.4211,
+      "step": 740
+    },
+    {
+      "epoch": 0.17635509014101267,
+      "grad_norm": 0.4397061820033599,
+      "learning_rate": 9.91213559841662e-06,
+      "loss": 0.3788,
+      "step": 741
+    },
+    {
+      "epoch": 0.17659308621407746,
+      "grad_norm": 0.4378799122528206,
+      "learning_rate": 9.911775454525974e-06,
+      "loss": 0.3375,
+      "step": 742
+    },
+    {
+      "epoch": 0.17683108228714225,
+      "grad_norm": 0.40859435165523883,
+      "learning_rate": 9.911414580623983e-06,
+      "loss": 0.3983,
+      "step": 743
+    },
+    {
+      "epoch": 0.17706907836020705,
+      "grad_norm": 0.4344842442523242,
+      "learning_rate": 9.911052976764282e-06,
+      "loss": 0.4477,
+      "step": 744
+    },
+    {
+      "epoch": 0.17730707443327184,
+      "grad_norm": 0.42409343119659304,
+      "learning_rate": 9.910690643000617e-06,
+      "loss": 0.331,
+      "step": 745
+    },
+    {
+      "epoch": 0.17754507050633664,
+      "grad_norm": 0.44353774653745326,
+      "learning_rate": 9.910327579386836e-06,
+      "loss": 0.3685,
+      "step": 746
+    },
+    {
+      "epoch": 0.17778306657940143,
+      "grad_norm": 0.39112060910154156,
+      "learning_rate": 9.909963785976902e-06,
+      "loss": 0.4222,
+      "step": 747
+    },
+    {
+      "epoch": 0.17802106265246623,
+      "grad_norm": 0.41719770556324215,
+      "learning_rate": 9.909599262824882e-06,
+      "loss": 0.4003,
+      "step": 748
+    },
+    {
+      "epoch": 0.17825905872553102,
+      "grad_norm": 0.43782801010917444,
+      "learning_rate": 9.909234009984956e-06,
+      "loss": 0.3264,
+      "step": 749
+    },
+    {
+      "epoch": 0.17849705479859582,
+      "grad_norm": 0.43932280484024017,
+      "learning_rate": 9.908868027511407e-06,
+      "loss": 0.3948,
+      "step": 750
+    },
+    {
+      "epoch": 0.1787350508716606,
+      "grad_norm": 0.40931591350618013,
+      "learning_rate": 9.908501315458628e-06,
+      "loss": 0.4369,
+      "step": 751
+    },
+    {
+      "epoch": 0.1789730469447254,
+      "grad_norm": 0.440394461817193,
+      "learning_rate": 9.908133873881125e-06,
+      "loss": 0.3308,
+      "step": 752
+    },
+    {
+      "epoch": 0.1792110430177902,
+      "grad_norm": 0.45585993680493553,
+      "learning_rate": 9.907765702833506e-06,
+      "loss": 0.3327,
+      "step": 753
+    },
+    {
+      "epoch": 0.179449039090855,
+      "grad_norm": 0.39891546180528054,
+      "learning_rate": 9.90739680237049e-06,
+      "loss": 0.4372,
+      "step": 754
+    },
+    {
+      "epoch": 0.1796870351639198,
+      "grad_norm": 0.4449856003881457,
+      "learning_rate": 9.907027172546907e-06,
+      "loss": 0.3646,
+      "step": 755
+    },
+    {
+      "epoch": 0.17992503123698458,
+      "grad_norm": 0.43483545394451123,
+      "learning_rate": 9.906656813417692e-06,
+      "loss": 0.366,
+      "step": 756
+    },
+    {
+      "epoch": 0.18016302731004938,
+      "grad_norm": 0.422673166996465,
+      "learning_rate": 9.90628572503789e-06,
+      "loss": 0.3759,
+      "step": 757
+    },
+    {
+      "epoch": 0.18040102338311417,
+      "grad_norm": 0.39567142752341666,
+      "learning_rate": 9.905913907462655e-06,
+      "loss": 0.4366,
+      "step": 758
+    },
+    {
+      "epoch": 0.18063901945617897,
+      "grad_norm": 0.4683314806260047,
+      "learning_rate": 9.905541360747244e-06,
+      "loss": 0.362,
+      "step": 759
+    },
+    {
+      "epoch": 0.18087701552924376,
+      "grad_norm": 0.41522141951305347,
+      "learning_rate": 9.90516808494703e-06,
+      "loss": 0.3395,
+      "step": 760
+    },
+    {
+      "epoch": 0.18111501160230856,
+      "grad_norm": 0.4303510055280143,
+      "learning_rate": 9.904794080117493e-06,
+      "loss": 0.4321,
+      "step": 761
+    },
+    {
+      "epoch": 0.18135300767537335,
+      "grad_norm": 0.4228437966345767,
+      "learning_rate": 9.904419346314216e-06,
+      "loss": 0.441,
+      "step": 762
+    },
+    {
+      "epoch": 0.18159100374843815,
+      "grad_norm": 0.4453386428802394,
+      "learning_rate": 9.904043883592893e-06,
+      "loss": 0.3773,
+      "step": 763
+    },
+    {
+      "epoch": 0.18182899982150294,
+      "grad_norm": 1.180486578415128,
+      "learning_rate": 9.90366769200933e-06,
+      "loss": 0.3609,
+      "step": 764
+    },
+    {
+      "epoch": 0.18206699589456773,
+      "grad_norm": 0.42781342072889006,
+      "learning_rate": 9.903290771619437e-06,
+      "loss": 0.4432,
+      "step": 765
+    },
+    {
+      "epoch": 0.18230499196763253,
+      "grad_norm": 0.38960119770687374,
+      "learning_rate": 9.902913122479235e-06,
+      "loss": 0.3728,
+      "step": 766
+    },
+    {
+      "epoch": 0.18254298804069732,
+      "grad_norm": 0.4321322538249792,
+      "learning_rate": 9.902534744644848e-06,
+      "loss": 0.3431,
+      "step": 767
+    },
+    {
+      "epoch": 0.18278098411376212,
+      "grad_norm": 0.41307757787063704,
+      "learning_rate": 9.902155638172517e-06,
+      "loss": 0.395,
+      "step": 768
+    },
+    {
+      "epoch": 0.1830189801868269,
+      "grad_norm": 0.43556469923967867,
+      "learning_rate": 9.901775803118584e-06,
+      "loss": 0.4441,
+      "step": 769
+    },
+    {
+      "epoch": 0.1832569762598917,
+      "grad_norm": 0.43167762486626765,
+      "learning_rate": 9.901395239539502e-06,
+      "loss": 0.3372,
+      "step": 770
+    },
+    {
+      "epoch": 0.1834949723329565,
+      "grad_norm": 0.416467157274216,
+      "learning_rate": 9.901013947491834e-06,
+      "loss": 0.355,
+      "step": 771
+    },
+    {
+      "epoch": 0.1837329684060213,
+      "grad_norm": 0.4141392350996442,
+      "learning_rate": 9.900631927032247e-06,
+      "loss": 0.4116,
+      "step": 772
+    },
+    {
+      "epoch": 0.1839709644790861,
+      "grad_norm": 0.42131938024420973,
+      "learning_rate": 9.90024917821752e-06,
+      "loss": 0.3775,
+      "step": 773
+    },
+    {
+      "epoch": 0.18420896055215089,
+      "grad_norm": 0.38656999194996833,
+      "learning_rate": 9.899865701104542e-06,
+      "loss": 0.3404,
+      "step": 774
+    },
+    {
+      "epoch": 0.18444695662521568,
+      "grad_norm": 0.4002326366629186,
+      "learning_rate": 9.8994814957503e-06,
+      "loss": 0.4006,
+      "step": 775
+    },
+    {
+      "epoch": 0.18468495269828047,
+      "grad_norm": 0.389329793747449,
+      "learning_rate": 9.899096562211902e-06,
+      "loss": 0.4635,
+      "step": 776
+    },
+    {
+      "epoch": 0.18492294877134527,
+      "grad_norm": 0.412679750416995,
+      "learning_rate": 9.898710900546557e-06,
+      "loss": 0.3451,
+      "step": 777
+    },
+    {
+      "epoch": 0.18516094484441006,
+      "grad_norm": 0.42487703444957303,
+      "learning_rate": 9.898324510811583e-06,
+      "loss": 0.3428,
+      "step": 778
+    },
+    {
+      "epoch": 0.18539894091747486,
+      "grad_norm": 0.3907619553000587,
+      "learning_rate": 9.89793739306441e-06,
+      "loss": 0.4392,
+      "step": 779
+    },
+    {
+      "epoch": 0.18563693699053965,
+      "grad_norm": 0.4193572680245468,
+      "learning_rate": 9.897549547362569e-06,
+      "loss": 0.4373,
+      "step": 780
+    },
+    {
+      "epoch": 0.18587493306360445,
+      "grad_norm": 0.39493589356753456,
+      "learning_rate": 9.897160973763706e-06,
+      "loss": 0.3539,
+      "step": 781
+    },
+    {
+      "epoch": 0.18611292913666924,
+      "grad_norm": 0.4440516227501279,
+      "learning_rate": 9.896771672325574e-06,
+      "loss": 0.3769,
+      "step": 782
+    },
+    {
+      "epoch": 0.18635092520973404,
+      "grad_norm": 0.44352023507130256,
+      "learning_rate": 9.89638164310603e-06,
+      "loss": 0.4338,
+      "step": 783
+    },
+    {
+      "epoch": 0.18658892128279883,
+      "grad_norm": 0.7671181173054992,
+      "learning_rate": 9.895990886163043e-06,
+      "loss": 0.3319,
+      "step": 784
+    },
+    {
+      "epoch": 0.18682691735586363,
+      "grad_norm": 0.4685322094472268,
+      "learning_rate": 9.89559940155469e-06,
+      "loss": 0.3454,
+      "step": 785
+    },
+    {
+      "epoch": 0.18706491342892842,
+      "grad_norm": 0.3828784304357339,
+      "learning_rate": 9.895207189339154e-06,
+      "loss": 0.4445,
+      "step": 786
+    },
+    {
+      "epoch": 0.18730290950199321,
+      "grad_norm": 0.42324641037940064,
+      "learning_rate": 9.89481424957473e-06,
+      "loss": 0.451,
+      "step": 787
+    },
+    {
+      "epoch": 0.187540905575058,
+      "grad_norm": 0.4365750038109837,
+      "learning_rate": 9.894420582319814e-06,
+      "loss": 0.3191,
+      "step": 788
+    },
+    {
+      "epoch": 0.1877789016481228,
+      "grad_norm": 0.40443174222338457,
+      "learning_rate": 9.894026187632917e-06,
+      "loss": 0.3748,
+      "step": 789
+    },
+    {
+      "epoch": 0.1880168977211876,
+      "grad_norm": 0.43152637845700487,
+      "learning_rate": 9.893631065572659e-06,
+      "loss": 0.4441,
+      "step": 790
+    },
+    {
+      "epoch": 0.1882548937942524,
+      "grad_norm": 0.43199571990589086,
+      "learning_rate": 9.893235216197761e-06,
+      "loss": 0.3711,
+      "step": 791
+    },
+    {
+      "epoch": 0.1884928898673172,
+      "grad_norm": 0.46404959339010077,
+      "learning_rate": 9.892838639567057e-06,
+      "loss": 0.359,
+      "step": 792
+    },
+    {
+      "epoch": 0.18873088594038198,
+      "grad_norm": 0.4441492172576122,
+      "learning_rate": 9.892441335739487e-06,
+      "loss": 0.4042,
+      "step": 793
+    },
+    {
+      "epoch": 0.18896888201344678,
+      "grad_norm": 0.42927081504797776,
+      "learning_rate": 9.892043304774102e-06,
+      "loss": 0.4278,
+      "step": 794
+    },
+    {
+      "epoch": 0.18920687808651157,
+      "grad_norm": 0.402576668783229,
+      "learning_rate": 9.89164454673006e-06,
+      "loss": 0.38,
+      "step": 795
+    },
+    {
+      "epoch": 0.18944487415957637,
+      "grad_norm": 0.37582462659523325,
+      "learning_rate": 9.891245061666622e-06,
+      "loss": 0.3614,
+      "step": 796
+    },
+    {
+      "epoch": 0.18968287023264116,
+      "grad_norm": 0.3988214342258959,
+      "learning_rate": 9.890844849643166e-06,
+      "loss": 0.4387,
+      "step": 797
+    },
+    {
+      "epoch": 0.18992086630570595,
+      "grad_norm": 0.4910334602397654,
+      "learning_rate": 9.890443910719171e-06,
+      "loss": 0.3962,
+      "step": 798
+    },
+    {
+      "epoch": 0.19015886237877075,
+      "grad_norm": 0.44461874782652483,
+      "learning_rate": 9.890042244954229e-06,
+      "loss": 0.3385,
+      "step": 799
+    },
+    {
+      "epoch": 0.19039685845183554,
+      "grad_norm": 0.4129631941986688,
+      "learning_rate": 9.889639852408035e-06,
+      "loss": 0.3928,
+      "step": 800
+    },
+    {
+      "epoch": 0.19063485452490034,
+      "grad_norm": 0.41914251475794334,
+      "learning_rate": 9.889236733140393e-06,
+      "loss": 0.4463,
+      "step": 801
+    },
+    {
+      "epoch": 0.19087285059796513,
+      "grad_norm": 0.41664422509463567,
+      "learning_rate": 9.88883288721122e-06,
+      "loss": 0.3765,
+      "step": 802
+    },
+    {
+      "epoch": 0.19111084667102993,
+      "grad_norm": 0.4045420912022183,
+      "learning_rate": 9.888428314680536e-06,
+      "loss": 0.3567,
+      "step": 803
+    },
+    {
+      "epoch": 0.19134884274409472,
+      "grad_norm": 0.4044005074831901,
+      "learning_rate": 9.888023015608471e-06,
+      "loss": 0.3931,
+      "step": 804
+    },
+    {
+      "epoch": 0.19158683881715952,
+      "grad_norm": 0.4368135710203652,
+      "learning_rate": 9.887616990055262e-06,
+      "loss": 0.4112,
+      "step": 805
+    },
+    {
+      "epoch": 0.1918248348902243,
+      "grad_norm": 0.4882695812113364,
+      "learning_rate": 9.887210238081253e-06,
+      "loss": 0.3695,
+      "step": 806
+    },
+    {
+      "epoch": 0.1920628309632891,
+      "grad_norm": 0.4246207428107935,
+      "learning_rate": 9.8868027597469e-06,
+      "loss": 0.3797,
+      "step": 807
+    },
+    {
+      "epoch": 0.1923008270363539,
+      "grad_norm": 0.46283957801455083,
+      "learning_rate": 9.886394555112764e-06,
+      "loss": 0.4513,
+      "step": 808
+    },
+    {
+      "epoch": 0.1925388231094187,
+      "grad_norm": 0.4047912499967522,
+      "learning_rate": 9.885985624239513e-06,
+      "loss": 0.3578,
+      "step": 809
+    },
+    {
+      "epoch": 0.1927768191824835,
+      "grad_norm": 0.4174808369463013,
+      "learning_rate": 9.885575967187924e-06,
+      "loss": 0.3537,
+      "step": 810
+    },
+    {
+      "epoch": 0.19301481525554828,
+      "grad_norm": 0.4347320503909233,
+      "learning_rate": 9.885165584018882e-06,
+      "loss": 0.3847,
+      "step": 811
+    },
+    {
+      "epoch": 0.19325281132861308,
+      "grad_norm": 0.44226534184677574,
+      "learning_rate": 9.884754474793383e-06,
+      "loss": 0.4511,
+      "step": 812
+    },
+    {
+      "epoch": 0.19349080740167787,
+      "grad_norm": 0.4019420309353,
+      "learning_rate": 9.884342639572526e-06,
+      "loss": 0.3352,
+      "step": 813
+    },
+    {
+      "epoch": 0.19372880347474267,
+      "grad_norm": 0.42514275612922164,
+      "learning_rate": 9.88393007841752e-06,
+      "loss": 0.3673,
+      "step": 814
+    },
+    {
+      "epoch": 0.19396679954780746,
+      "grad_norm": 0.4524089013354289,
+      "learning_rate": 9.88351679138968e-06,
+      "loss": 0.4277,
+      "step": 815
+    },
+    {
+      "epoch": 0.19420479562087226,
+      "grad_norm": 0.4380590877999043,
+      "learning_rate": 9.883102778550434e-06,
+      "loss": 0.3627,
+      "step": 816
+    },
+    {
+      "epoch": 0.19444279169393705,
+      "grad_norm": 0.4165421579055438,
+      "learning_rate": 9.882688039961312e-06,
+      "loss": 0.345,
+      "step": 817
+    },
+    {
+      "epoch": 0.19468078776700184,
+      "grad_norm": 0.43180936146351057,
+      "learning_rate": 9.882272575683956e-06,
+      "loss": 0.4111,
+      "step": 818
+    },
+    {
+      "epoch": 0.19491878384006664,
+      "grad_norm": 0.41648821424267424,
+      "learning_rate": 9.881856385780115e-06,
+      "loss": 0.4578,
+      "step": 819
+    },
+    {
+      "epoch": 0.19515677991313143,
+      "grad_norm": 0.4246404967531619,
+      "learning_rate": 9.881439470311642e-06,
+      "loss": 0.3842,
+      "step": 820
+    },
+    {
+      "epoch": 0.19539477598619623,
+      "grad_norm": 0.4158895838872154,
+      "learning_rate": 9.881021829340502e-06,
+      "loss": 0.41,
+      "step": 821
+    },
+    {
+      "epoch": 0.19563277205926102,
+      "grad_norm": 0.4149994996559407,
+      "learning_rate": 9.880603462928769e-06,
+      "loss": 0.4231,
+      "step": 822
+    },
+    {
+      "epoch": 0.19587076813232582,
+      "grad_norm": 0.4089944684040092,
+      "learning_rate": 9.880184371138621e-06,
+      "loss": 0.3609,
+      "step": 823
+    },
+    {
+      "epoch": 0.1961087642053906,
+      "grad_norm": 0.44441855579372713,
+      "learning_rate": 9.879764554032345e-06,
+      "loss": 0.3333,
+      "step": 824
+    },
+    {
+      "epoch": 0.1963467602784554,
+      "grad_norm": 0.4230323557425423,
+      "learning_rate": 9.879344011672337e-06,
+      "loss": 0.4026,
+      "step": 825
+    },
+    {
+      "epoch": 0.1965847563515202,
+      "grad_norm": 0.42605983334806596,
+      "learning_rate": 9.8789227441211e-06,
+      "loss": 0.4439,
+      "step": 826
+    },
+    {
+      "epoch": 0.196822752424585,
+      "grad_norm": 0.44788672346466435,
+      "learning_rate": 9.878500751441244e-06,
+      "loss": 0.3576,
+      "step": 827
+    },
+    {
+      "epoch": 0.1970607484976498,
+      "grad_norm": 0.42604961830620175,
+      "learning_rate": 9.878078033695488e-06,
+      "loss": 0.3765,
+      "step": 828
+    },
+    {
+      "epoch": 0.19729874457071458,
+      "grad_norm": 0.4348482551029076,
+      "learning_rate": 9.877654590946659e-06,
+      "loss": 0.4327,
+      "step": 829
+    },
+    {
+      "epoch": 0.19753674064377938,
+      "grad_norm": 0.41378312933719785,
+      "learning_rate": 9.87723042325769e-06,
+      "loss": 0.4134,
+      "step": 830
+    },
+    {
+      "epoch": 0.19777473671684417,
+      "grad_norm": 0.473934565450725,
+      "learning_rate": 9.876805530691622e-06,
+      "loss": 0.3382,
+      "step": 831
+    },
+    {
+      "epoch": 0.19801273278990897,
+      "grad_norm": 0.45241327431142664,
+      "learning_rate": 9.876379913311607e-06,
+      "loss": 0.3837,
+      "step": 832
+    },
+    {
+      "epoch": 0.19825072886297376,
+      "grad_norm": 0.4046931569037168,
+      "learning_rate": 9.875953571180901e-06,
+      "loss": 0.4148,
+      "step": 833
+    },
+    {
+      "epoch": 0.19848872493603856,
+      "grad_norm": 0.4515187001854004,
+      "learning_rate": 9.875526504362868e-06,
+      "loss": 0.3804,
+      "step": 834
+    },
+    {
+      "epoch": 0.19872672100910335,
+      "grad_norm": 0.4533213069296073,
+      "learning_rate": 9.875098712920983e-06,
+      "loss": 0.3442,
+      "step": 835
+    },
+    {
+      "epoch": 0.19896471708216815,
+      "grad_norm": 0.4523183437793231,
+      "learning_rate": 9.874670196918824e-06,
+      "loss": 0.4091,
+      "step": 836
+    },
+    {
+      "epoch": 0.19920271315523294,
+      "grad_norm": 0.4119738644745412,
+      "learning_rate": 9.874240956420082e-06,
+      "loss": 0.4687,
+      "step": 837
+    },
+    {
+      "epoch": 0.19944070922829774,
+      "grad_norm": 0.4699475156689954,
+      "learning_rate": 9.87381099148855e-06,
+      "loss": 0.3553,
+      "step": 838
+    },
+    {
+      "epoch": 0.19967870530136253,
+      "grad_norm": 0.41221446850629756,
+      "learning_rate": 9.873380302188133e-06,
+      "loss": 0.3856,
+      "step": 839
+    },
+    {
+      "epoch": 0.19991670137442732,
+      "grad_norm": 0.3967941653688478,
+      "learning_rate": 9.87294888858284e-06,
+      "loss": 0.4608,
+      "step": 840
+    },
+    {
+      "epoch": 0.20015469744749212,
+      "grad_norm": 0.4446657496487925,
+      "learning_rate": 9.872516750736793e-06,
+      "loss": 0.4013,
+      "step": 841
+    },
+    {
+      "epoch": 0.2003926935205569,
+      "grad_norm": 0.4896648889382981,
+      "learning_rate": 9.872083888714217e-06,
+      "loss": 0.356,
+      "step": 842
+    },
+    {
+      "epoch": 0.2006306895936217,
+      "grad_norm": 0.4087173120218746,
+      "learning_rate": 9.871650302579443e-06,
+      "loss": 0.3968,
+      "step": 843
+    },
+    {
+      "epoch": 0.2008686856666865,
+      "grad_norm": 0.4824564537672377,
+      "learning_rate": 9.871215992396917e-06,
+      "loss": 0.4077,
+      "step": 844
+    },
+    {
+      "epoch": 0.2011066817397513,
+      "grad_norm": 0.44137153458158945,
+      "learning_rate": 9.870780958231186e-06,
+      "loss": 0.3466,
+      "step": 845
+    },
+    {
+      "epoch": 0.2013446778128161,
+      "grad_norm": 0.5156882053665172,
+      "learning_rate": 9.870345200146907e-06,
+      "loss": 0.375,
+      "step": 846
+    },
+    {
+      "epoch": 0.2015826738858809,
+      "grad_norm": 0.3651462313643267,
+      "learning_rate": 9.869908718208845e-06,
+      "loss": 0.4293,
+      "step": 847
+    },
+    {
+      "epoch": 0.20182066995894568,
+      "grad_norm": 0.4248226402638282,
+      "learning_rate": 9.869471512481872e-06,
+      "loss": 0.4093,
+      "step": 848
+    },
+    {
+      "epoch": 0.20205866603201048,
+      "grad_norm": 0.4715061071976561,
+      "learning_rate": 9.869033583030967e-06,
+      "loss": 0.3187,
+      "step": 849
+    },
+    {
+      "epoch": 0.20229666210507527,
+      "grad_norm": 0.4360335200519038,
+      "learning_rate": 9.868594929921217e-06,
+      "loss": 0.3947,
+      "step": 850
+    },
+    {
+      "epoch": 0.20253465817814006,
+      "grad_norm": 0.4128931217238059,
+      "learning_rate": 9.86815555321782e-06,
+      "loss": 0.4361,
+      "step": 851
+    },
+    {
+      "epoch": 0.20277265425120486,
+      "grad_norm": 0.46321878255592136,
+      "learning_rate": 9.867715452986073e-06,
+      "loss": 0.3231,
+      "step": 852
+    },
+    {
+      "epoch": 0.20301065032426965,
+      "grad_norm": 0.4537539113370709,
+      "learning_rate": 9.867274629291387e-06,
+      "loss": 0.3619,
+      "step": 853
+    },
+    {
+      "epoch": 0.20324864639733445,
+      "grad_norm": 0.38249330366239725,
+      "learning_rate": 9.866833082199283e-06,
+      "loss": 0.3993,
+      "step": 854
+    },
+    {
+      "epoch": 0.20348664247039924,
+      "grad_norm": 0.43771271278034846,
+      "learning_rate": 9.866390811775382e-06,
+      "loss": 0.3913,
+      "step": 855
+    },
+    {
+      "epoch": 0.20372463854346404,
+      "grad_norm": 0.46356127634456645,
+      "learning_rate": 9.865947818085417e-06,
+      "loss": 0.3215,
+      "step": 856
+    },
+    {
+      "epoch": 0.20396263461652883,
+      "grad_norm": 0.495299460550108,
+      "learning_rate": 9.86550410119523e-06,
+      "loss": 0.3485,
+      "step": 857
+    },
+    {
+      "epoch": 0.20420063068959363,
+      "grad_norm": 0.4143122108558567,
+      "learning_rate": 9.865059661170767e-06,
+      "loss": 0.4334,
+      "step": 858
+    },
+    {
+      "epoch": 0.20443862676265842,
+      "grad_norm": 0.4325780620093191,
+      "learning_rate": 9.86461449807808e-06,
+      "loss": 0.379,
+      "step": 859
+    },
+    {
+      "epoch": 0.20467662283572322,
+      "grad_norm": 0.5084472841414205,
+      "learning_rate": 9.864168611983336e-06,
+      "loss": 0.3371,
+      "step": 860
+    },
+    {
+      "epoch": 0.204914618908788,
+      "grad_norm": 0.4451974603394198,
+      "learning_rate": 9.863722002952803e-06,
+      "loss": 0.3918,
+      "step": 861
+    },
+    {
+      "epoch": 0.2051526149818528,
+      "grad_norm": 0.4293240665529044,
+      "learning_rate": 9.863274671052857e-06,
+      "loss": 0.4285,
+      "step": 862
+    },
+    {
+      "epoch": 0.2053906110549176,
+      "grad_norm": 0.42020107235283133,
+      "learning_rate": 9.862826616349981e-06,
+      "loss": 0.3356,
+      "step": 863
+    },
+    {
+      "epoch": 0.2056286071279824,
+      "grad_norm": 0.45803176046948774,
+      "learning_rate": 9.862377838910771e-06,
+      "loss": 0.4254,
+      "step": 864
+    },
+    {
+      "epoch": 0.2058666032010472,
+      "grad_norm": 0.3886605229465983,
+      "learning_rate": 9.861928338801926e-06,
+      "loss": 0.4262,
+      "step": 865
+    },
+    {
+      "epoch": 0.20610459927411198,
+      "grad_norm": 0.43472442169007486,
+      "learning_rate": 9.86147811609025e-06,
+      "loss": 0.3713,
+      "step": 866
+    },
+    {
+      "epoch": 0.20634259534717678,
+      "grad_norm": 0.4414211701916826,
+      "learning_rate": 9.861027170842659e-06,
+      "loss": 0.3361,
+      "step": 867
+    },
+    {
+      "epoch": 0.20658059142024157,
+      "grad_norm": 0.408979611759581,
+      "learning_rate": 9.860575503126175e-06,
+      "loss": 0.4036,
+      "step": 868
+    },
+    {
+      "epoch": 0.20681858749330637,
+      "grad_norm": 0.446536065672617,
+      "learning_rate": 9.860123113007928e-06,
+      "loss": 0.4488,
+      "step": 869
+    },
+    {
+      "epoch": 0.20705658356637116,
+      "grad_norm": 0.436694557262441,
+      "learning_rate": 9.85967000055515e-06,
+      "loss": 0.3634,
+      "step": 870
+    },
+    {
+      "epoch": 0.20729457963943596,
+      "grad_norm": 0.4541956461930287,
+      "learning_rate": 9.859216165835188e-06,
+      "loss": 0.3835,
+      "step": 871
+    },
+    {
+      "epoch": 0.20753257571250075,
+      "grad_norm": 0.4213848003370073,
+      "learning_rate": 9.858761608915492e-06,
+      "loss": 0.4334,
+      "step": 872
+    },
+    {
+      "epoch": 0.20777057178556554,
+      "grad_norm": 0.3930202275134771,
+      "learning_rate": 9.858306329863623e-06,
+      "loss": 0.3955,
+      "step": 873
+    },
+    {
+      "epoch": 0.20800856785863034,
+      "grad_norm": 0.427845820752473,
+      "learning_rate": 9.857850328747243e-06,
+      "loss": 0.3356,
+      "step": 874
+    },
+    {
+      "epoch": 0.20824656393169513,
+      "grad_norm": 0.4116624892541286,
+      "learning_rate": 9.857393605634126e-06,
+      "loss": 0.3699,
+      "step": 875
+    },
+    {
+      "epoch": 0.20848456000475993,
+      "grad_norm": 0.3935251767233762,
+      "learning_rate": 9.856936160592155e-06,
+      "loss": 0.4307,
+      "step": 876
+    },
+    {
+      "epoch": 0.20872255607782472,
+      "grad_norm": 0.42170977712231184,
+      "learning_rate": 9.856477993689316e-06,
+      "loss": 0.3545,
+      "step": 877
+    },
+    {
+      "epoch": 0.20896055215088952,
+      "grad_norm": 0.41048028309068124,
+      "learning_rate": 9.856019104993702e-06,
+      "loss": 0.3381,
+      "step": 878
+    },
+    {
+      "epoch": 0.2091985482239543,
+      "grad_norm": 0.40569183356117694,
+      "learning_rate": 9.855559494573517e-06,
+      "loss": 0.4362,
+      "step": 879
+    },
+    {
+      "epoch": 0.2094365442970191,
+      "grad_norm": 0.42666793677125886,
+      "learning_rate": 9.855099162497071e-06,
+      "loss": 0.3783,
+      "step": 880
+    },
+    {
+      "epoch": 0.2096745403700839,
+      "grad_norm": 0.43438959276226846,
+      "learning_rate": 9.854638108832781e-06,
+      "loss": 0.3769,
+      "step": 881
+    },
+    {
+      "epoch": 0.2099125364431487,
+      "grad_norm": 0.44294773883559535,
+      "learning_rate": 9.854176333649169e-06,
+      "loss": 0.401,
+      "step": 882
+    },
+    {
+      "epoch": 0.2101505325162135,
+      "grad_norm": 0.40033962525880623,
+      "learning_rate": 9.853713837014867e-06,
+      "loss": 0.4307,
+      "step": 883
+    },
+    {
+      "epoch": 0.21038852858927828,
+      "grad_norm": 0.42544362225337135,
+      "learning_rate": 9.853250618998612e-06,
+      "loss": 0.3605,
+      "step": 884
+    },
+    {
+      "epoch": 0.21062652466234308,
+      "grad_norm": 0.43069551338393347,
+      "learning_rate": 9.852786679669256e-06,
+      "loss": 0.3538,
+      "step": 885
+    },
+    {
+      "epoch": 0.21086452073540787,
+      "grad_norm": 0.4061340048491632,
+      "learning_rate": 9.852322019095744e-06,
+      "loss": 0.3855,
+      "step": 886
+    },
+    {
+      "epoch": 0.21110251680847267,
+      "grad_norm": 0.4200438920047967,
+      "learning_rate": 9.85185663734714e-06,
+      "loss": 0.4657,
+      "step": 887
+    },
+    {
+      "epoch": 0.21134051288153746,
+      "grad_norm": 0.40315771492500546,
+      "learning_rate": 9.85139053449261e-06,
+      "loss": 0.323,
+      "step": 888
+    },
+    {
+      "epoch": 0.21157850895460226,
+      "grad_norm": 0.451869373976093,
+      "learning_rate": 9.85092371060143e-06,
+      "loss": 0.4123,
+      "step": 889
+    },
+    {
+      "epoch": 0.21181650502766705,
+      "grad_norm": 0.41009915304865363,
+      "learning_rate": 9.85045616574298e-06,
+      "loss": 0.4342,
+      "step": 890
+    },
+    {
+      "epoch": 0.21205450110073185,
+      "grad_norm": 0.41018214294921573,
+      "learning_rate": 9.84998789998675e-06,
+      "loss": 0.4131,
+      "step": 891
+    },
+    {
+      "epoch": 0.21229249717379664,
+      "grad_norm": 0.41395012445520085,
+      "learning_rate": 9.849518913402334e-06,
+      "loss": 0.3472,
+      "step": 892
+    },
+    {
+      "epoch": 0.21253049324686143,
+      "grad_norm": 0.41478482807464195,
+      "learning_rate": 9.849049206059435e-06,
+      "loss": 0.4359,
+      "step": 893
+    },
+    {
+      "epoch": 0.21276848931992623,
+      "grad_norm": 0.3969815086554544,
+      "learning_rate": 9.848578778027867e-06,
+      "loss": 0.4184,
+      "step": 894
+    },
+    {
+      "epoch": 0.21300648539299102,
+      "grad_norm": 0.4297335515729988,
+      "learning_rate": 9.848107629377544e-06,
+      "loss": 0.3719,
+      "step": 895
+    },
+    {
+      "epoch": 0.21324448146605582,
+      "grad_norm": 0.3855698648491326,
+      "learning_rate": 9.84763576017849e-06,
+      "loss": 0.3658,
+      "step": 896
+    },
+    {
+      "epoch": 0.2134824775391206,
+      "grad_norm": 0.40093853952594977,
+      "learning_rate": 9.847163170500837e-06,
+      "loss": 0.4394,
+      "step": 897
+    },
+    {
+      "epoch": 0.2137204736121854,
+      "grad_norm": 0.4485624156969946,
+      "learning_rate": 9.846689860414824e-06,
+      "loss": 0.3776,
+      "step": 898
+    },
+    {
+      "epoch": 0.2139584696852502,
+      "grad_norm": 0.3794469849541886,
+      "learning_rate": 9.846215829990797e-06,
+      "loss": 0.3115,
+      "step": 899
+    },
+    {
+      "epoch": 0.214196465758315,
+      "grad_norm": 0.3972841317444796,
+      "learning_rate": 9.84574107929921e-06,
+      "loss": 0.3894,
+      "step": 900
+    },
+    {
+      "epoch": 0.2144344618313798,
+      "grad_norm": 0.38348004951058207,
+      "learning_rate": 9.845265608410616e-06,
+      "loss": 0.4287,
+      "step": 901
+    },
+    {
+      "epoch": 0.21467245790444459,
+      "grad_norm": 0.41303667553057977,
+      "learning_rate": 9.84478941739569e-06,
+      "loss": 0.3239,
+      "step": 902
+    },
+    {
+      "epoch": 0.21491045397750938,
+      "grad_norm": 0.40879152143826697,
+      "learning_rate": 9.844312506325202e-06,
+      "loss": 0.3797,
+      "step": 903
+    },
+    {
+      "epoch": 0.21514845005057417,
+      "grad_norm": 0.4604571778006916,
+      "learning_rate": 9.843834875270032e-06,
+      "loss": 0.4199,
+      "step": 904
+    },
+    {
+      "epoch": 0.21538644612363897,
+      "grad_norm": 0.4152085389985221,
+      "learning_rate": 9.84335652430117e-06,
+      "loss": 0.3847,
+      "step": 905
+    },
+    {
+      "epoch": 0.21562444219670376,
+      "grad_norm": 0.4149816757480418,
+      "learning_rate": 9.842877453489708e-06,
+      "loss": 0.3592,
+      "step": 906
+    },
+    {
+      "epoch": 0.21586243826976856,
+      "grad_norm": 0.4299620220648408,
+      "learning_rate": 9.84239766290685e-06,
+      "loss": 0.4043,
+      "step": 907
+    },
+    {
+      "epoch": 0.21610043434283335,
+      "grad_norm": 0.41035423107133623,
+      "learning_rate": 9.841917152623905e-06,
+      "loss": 0.4585,
+      "step": 908
+    },
+    {
+      "epoch": 0.21633843041589815,
+      "grad_norm": 0.44913348339639636,
+      "learning_rate": 9.841435922712288e-06,
+      "loss": 0.3911,
+      "step": 909
+    },
+    {
+      "epoch": 0.21657642648896294,
+      "grad_norm": 0.405292221298209,
+      "learning_rate": 9.84095397324352e-06,
+      "loss": 0.3582,
+      "step": 910
+    },
+    {
+      "epoch": 0.21681442256202774,
+      "grad_norm": 0.38536854563574374,
+      "learning_rate": 9.840471304289233e-06,
+      "loss": 0.4007,
+      "step": 911
+    },
+    {
+      "epoch": 0.21705241863509253,
+      "grad_norm": 0.47808379504573534,
+      "learning_rate": 9.839987915921163e-06,
+      "loss": 0.4611,
+      "step": 912
+    },
+    {
+      "epoch": 0.21729041470815733,
+      "grad_norm": 0.3934343577154057,
+      "learning_rate": 9.839503808211153e-06,
+      "loss": 0.3176,
+      "step": 913
+    },
+    {
+      "epoch": 0.21752841078122212,
+      "grad_norm": 0.4253557198903114,
+      "learning_rate": 9.839018981231151e-06,
+      "loss": 0.3818,
+      "step": 914
+    },
+    {
+      "epoch": 0.21776640685428691,
+      "grad_norm": 0.4060612395908892,
+      "learning_rate": 9.838533435053221e-06,
+      "loss": 0.4327,
+      "step": 915
+    },
+    {
+      "epoch": 0.2180044029273517,
+      "grad_norm": 0.4044627954830839,
+      "learning_rate": 9.83804716974952e-06,
+      "loss": 0.3697,
+      "step": 916
+    },
+    {
+      "epoch": 0.2182423990004165,
+      "grad_norm": 0.3898450504134681,
+      "learning_rate": 9.837560185392325e-06,
+      "loss": 0.3444,
+      "step": 917
+    },
+    {
+      "epoch": 0.2184803950734813,
+      "grad_norm": 0.41317608963767616,
+      "learning_rate": 9.837072482054009e-06,
+      "loss": 0.4076,
+      "step": 918
+    },
+    {
+      "epoch": 0.2187183911465461,
+      "grad_norm": 0.4138083448258549,
+      "learning_rate": 9.83658405980706e-06,
+      "loss": 0.4545,
+      "step": 919
+    },
+    {
+      "epoch": 0.2189563872196109,
+      "grad_norm": 0.41247702913642026,
+      "learning_rate": 9.836094918724067e-06,
+      "loss": 0.3613,
+      "step": 920
+    },
+    {
+      "epoch": 0.21919438329267568,
+      "grad_norm": 0.4121156299354065,
+      "learning_rate": 9.83560505887773e-06,
+      "loss": 0.3765,
+      "step": 921
+    },
+    {
+      "epoch": 0.21943237936574048,
+      "grad_norm": 0.43268431314703193,
+      "learning_rate": 9.835114480340855e-06,
+      "loss": 0.4259,
+      "step": 922
+    },
+    {
+      "epoch": 0.21967037543880527,
+      "grad_norm": 0.38798899747199356,
+      "learning_rate": 9.834623183186352e-06,
+      "loss": 0.3945,
+      "step": 923
+    },
+    {
+      "epoch": 0.21990837151187007,
+      "grad_norm": 0.40463819440452653,
+      "learning_rate": 9.834131167487241e-06,
+      "loss": 0.3419,
+      "step": 924
+    },
+    {
+      "epoch": 0.22014636758493486,
+      "grad_norm": 0.4158947107356108,
+      "learning_rate": 9.833638433316647e-06,
+      "loss": 0.3918,
+      "step": 925
+    },
+    {
+      "epoch": 0.22038436365799965,
+      "grad_norm": 0.3951265208744111,
+      "learning_rate": 9.833144980747806e-06,
+      "loss": 0.4467,
+      "step": 926
+    },
+    {
+      "epoch": 0.22062235973106445,
+      "grad_norm": 0.3866457296285503,
+      "learning_rate": 9.832650809854054e-06,
+      "loss": 0.3626,
+      "step": 927
+    },
+    {
+      "epoch": 0.22086035580412924,
+      "grad_norm": 0.4148351326369386,
+      "learning_rate": 9.832155920708838e-06,
+      "loss": 0.3362,
+      "step": 928
+    },
+    {
+      "epoch": 0.22109835187719404,
+      "grad_norm": 0.3976030627374797,
+      "learning_rate": 9.831660313385709e-06,
+      "loss": 0.4314,
+      "step": 929
+    },
+    {
+      "epoch": 0.22133634795025883,
+      "grad_norm": 0.4157333689112259,
+      "learning_rate": 9.831163987958329e-06,
+      "loss": 0.4044,
+      "step": 930
+    },
+    {
+      "epoch": 0.22157434402332363,
+      "grad_norm": 0.4408772739009037,
+      "learning_rate": 9.830666944500462e-06,
+      "loss": 0.3282,
+      "step": 931
+    },
+    {
+      "epoch": 0.22181234009638842,
+      "grad_norm": 0.41164918567935477,
+      "learning_rate": 9.830169183085983e-06,
+      "loss": 0.381,
+      "step": 932
+    },
+    {
+      "epoch": 0.22205033616945322,
+      "grad_norm": 0.3644162693747717,
+      "learning_rate": 9.829670703788873e-06,
+      "loss": 0.4151,
+      "step": 933
+    },
+    {
+      "epoch": 0.222288332242518,
+      "grad_norm": 0.4207226846601998,
+      "learning_rate": 9.829171506683211e-06,
+      "loss": 0.3884,
+      "step": 934
+    },
+    {
+      "epoch": 0.2225263283155828,
+      "grad_norm": 0.45066402524896876,
+      "learning_rate": 9.828671591843198e-06,
+      "loss": 0.3369,
+      "step": 935
+    },
+    {
+      "epoch": 0.2227643243886476,
+      "grad_norm": 0.4168689897275966,
+      "learning_rate": 9.828170959343131e-06,
+      "loss": 0.3938,
+      "step": 936
+    },
+    {
+      "epoch": 0.2230023204617124,
+      "grad_norm": 0.44074947992378183,
+      "learning_rate": 9.827669609257417e-06,
+      "loss": 0.433,
+      "step": 937
+    },
+    {
+      "epoch": 0.2232403165347772,
+      "grad_norm": 0.47103576178023904,
+      "learning_rate": 9.827167541660568e-06,
+      "loss": 0.3565,
+      "step": 938
+    },
+    {
+      "epoch": 0.22347831260784198,
+      "grad_norm": 0.4532327477806086,
+      "learning_rate": 9.826664756627202e-06,
+      "loss": 0.3961,
+      "step": 939
+    },
+    {
+      "epoch": 0.22371630868090678,
+      "grad_norm": 0.39014864862033083,
+      "learning_rate": 9.826161254232048e-06,
+      "loss": 0.4257,
+      "step": 940
+    },
+    {
+      "epoch": 0.22395430475397157,
+      "grad_norm": 0.40042087107333996,
+      "learning_rate": 9.825657034549939e-06,
+      "loss": 0.3866,
+      "step": 941
+    },
+    {
+      "epoch": 0.22419230082703637,
+      "grad_norm": 0.44006439544231385,
+      "learning_rate": 9.825152097655813e-06,
+      "loss": 0.3589,
+      "step": 942
+    },
+    {
+      "epoch": 0.22443029690010116,
+      "grad_norm": 0.3999068012035896,
+      "learning_rate": 9.824646443624717e-06,
+      "loss": 0.3891,
+      "step": 943
+    },
+    {
+      "epoch": 0.22466829297316596,
+      "grad_norm": 0.4539949228586678,
+      "learning_rate": 9.824140072531805e-06,
+      "loss": 0.4532,
+      "step": 944
+    },
+    {
+      "epoch": 0.22490628904623075,
+      "grad_norm": 0.3981879315816523,
+      "learning_rate": 9.823632984452331e-06,
+      "loss": 0.3465,
+      "step": 945
+    },
+    {
+      "epoch": 0.22514428511929555,
+      "grad_norm": 0.6439992315492964,
+      "learning_rate": 9.823125179461668e-06,
+      "loss": 0.376,
+      "step": 946
+    },
+    {
+      "epoch": 0.22538228119236034,
+      "grad_norm": 0.404887917419206,
+      "learning_rate": 9.822616657635284e-06,
+      "loss": 0.4472,
+      "step": 947
+    },
+    {
+      "epoch": 0.22562027726542513,
+      "grad_norm": 0.37708651982188657,
+      "learning_rate": 9.822107419048759e-06,
+      "loss": 0.4146,
+      "step": 948
+    },
+    {
+      "epoch": 0.2258582733384899,
+      "grad_norm": 0.44302674607158987,
+      "learning_rate": 9.821597463777779e-06,
+      "loss": 0.321,
+      "step": 949
+    },
+    {
+      "epoch": 0.2260962694115547,
+      "grad_norm": 0.412240348327169,
+      "learning_rate": 9.821086791898133e-06,
+      "loss": 0.3731,
+      "step": 950
+    },
+    {
+      "epoch": 0.2263342654846195,
+      "grad_norm": 0.40579745891438523,
+      "learning_rate": 9.820575403485724e-06,
+      "loss": 0.4699,
+      "step": 951
+    },
+    {
+      "epoch": 0.22657226155768428,
+      "grad_norm": 0.415668474457593,
+      "learning_rate": 9.820063298616553e-06,
+      "loss": 0.3718,
+      "step": 952
+    },
+    {
+      "epoch": 0.22681025763074908,
+      "grad_norm": 0.416764959736151,
+      "learning_rate": 9.819550477366735e-06,
+      "loss": 0.3393,
+      "step": 953
+    },
+    {
+      "epoch": 0.22704825370381387,
+      "grad_norm": 0.44453596559445757,
+      "learning_rate": 9.819036939812485e-06,
+      "loss": 0.4483,
+      "step": 954
+    },
+    {
+      "epoch": 0.22728624977687867,
+      "grad_norm": 0.3709861408947193,
+      "learning_rate": 9.818522686030127e-06,
+      "loss": 0.4047,
+      "step": 955
+    },
+    {
+      "epoch": 0.22752424584994346,
+      "grad_norm": 0.40210692784551644,
+      "learning_rate": 9.818007716096096e-06,
+      "loss": 0.3185,
+      "step": 956
+    },
+    {
+      "epoch": 0.22776224192300826,
+      "grad_norm": 0.3862126896605156,
+      "learning_rate": 9.817492030086926e-06,
+      "loss": 0.3905,
+      "step": 957
+    },
+    {
+      "epoch": 0.22800023799607305,
+      "grad_norm": 0.43152498874354184,
+      "learning_rate": 9.816975628079261e-06,
+      "loss": 0.4296,
+      "step": 958
+    },
+    {
+      "epoch": 0.22823823406913785,
+      "grad_norm": 0.38525294582007225,
+      "learning_rate": 9.816458510149852e-06,
+      "loss": 0.3463,
+      "step": 959
+    },
+    {
+      "epoch": 0.22847623014220264,
+      "grad_norm": 0.47012279193131934,
+      "learning_rate": 9.815940676375554e-06,
+      "loss": 0.3733,
+      "step": 960
+    },
+    {
+      "epoch": 0.22871422621526744,
+      "grad_norm": 0.4002528394439282,
+      "learning_rate": 9.815422126833332e-06,
+      "loss": 0.4243,
+      "step": 961
+    },
+    {
+      "epoch": 0.22895222228833223,
+      "grad_norm": 0.4248826037103133,
+      "learning_rate": 9.814902861600252e-06,
+      "loss": 0.4592,
+      "step": 962
+    },
+    {
+      "epoch": 0.22919021836139702,
+      "grad_norm": 0.4022803320480314,
+      "learning_rate": 9.814382880753493e-06,
+      "loss": 0.361,
+      "step": 963
+    },
+    {
+      "epoch": 0.22942821443446182,
+      "grad_norm": 0.424538658023901,
+      "learning_rate": 9.813862184370338e-06,
+      "loss": 0.3863,
+      "step": 964
+    },
+    {
+      "epoch": 0.2296662105075266,
+      "grad_norm": 0.44976984034315987,
+      "learning_rate": 9.81334077252817e-06,
+      "loss": 0.4503,
+      "step": 965
+    },
+    {
+      "epoch": 0.2299042065805914,
+      "grad_norm": 0.406304483755596,
+      "learning_rate": 9.812818645304488e-06,
+      "loss": 0.3875,
+      "step": 966
+    },
+    {
+      "epoch": 0.2301422026536562,
+      "grad_norm": 0.48831090846163266,
+      "learning_rate": 9.812295802776893e-06,
+      "loss": 0.3351,
+      "step": 967
+    },
+    {
+      "epoch": 0.230380198726721,
+      "grad_norm": 0.4010981600344825,
+      "learning_rate": 9.81177224502309e-06,
+      "loss": 0.4074,
+      "step": 968
+    },
+    {
+      "epoch": 0.2306181947997858,
+      "grad_norm": 0.39821279677246363,
+      "learning_rate": 9.811247972120895e-06,
+      "loss": 0.4441,
+      "step": 969
+    },
+    {
+      "epoch": 0.2308561908728506,
+      "grad_norm": 0.40402056505396167,
+      "learning_rate": 9.810722984148224e-06,
+      "loss": 0.3292,
+      "step": 970
+    },
+    {
+      "epoch": 0.23109418694591538,
+      "grad_norm": 0.4031749039115374,
+      "learning_rate": 9.810197281183109e-06,
+      "loss": 0.3967,
+      "step": 971
+    },
+    {
+      "epoch": 0.23133218301898018,
+      "grad_norm": 0.4254043618777442,
+      "learning_rate": 9.809670863303678e-06,
+      "loss": 0.4345,
+      "step": 972
+    },
+    {
+      "epoch": 0.23157017909204497,
+      "grad_norm": 0.41821250677571714,
+      "learning_rate": 9.809143730588172e-06,
+      "loss": 0.3698,
+      "step": 973
+    },
+    {
+      "epoch": 0.23180817516510976,
+      "grad_norm": 0.4186518496968692,
+      "learning_rate": 9.808615883114935e-06,
+      "loss": 0.3435,
+      "step": 974
+    },
+    {
+      "epoch": 0.23204617123817456,
+      "grad_norm": 0.447054433602328,
+      "learning_rate": 9.808087320962418e-06,
+      "loss": 0.3868,
+      "step": 975
+    },
+    {
+      "epoch": 0.23228416731123935,
+      "grad_norm": 0.40528303959506257,
+      "learning_rate": 9.807558044209178e-06,
+      "loss": 0.4353,
+      "step": 976
+    },
+    {
+      "epoch": 0.23252216338430415,
+      "grad_norm": 0.4201076442750277,
+      "learning_rate": 9.80702805293388e-06,
+      "loss": 0.3498,
+      "step": 977
+    },
+    {
+      "epoch": 0.23276015945736894,
+      "grad_norm": 0.43371214248349416,
+      "learning_rate": 9.806497347215294e-06,
+      "loss": 0.3148,
+      "step": 978
+    },
+    {
+      "epoch": 0.23299815553043374,
+      "grad_norm": 0.41147178302674675,
+      "learning_rate": 9.805965927132294e-06,
+      "loss": 0.4017,
+      "step": 979
+    },
+    {
+      "epoch": 0.23323615160349853,
+      "grad_norm": 0.4447084847180371,
+      "learning_rate": 9.805433792763866e-06,
+      "loss": 0.4182,
+      "step": 980
+    },
+    {
+      "epoch": 0.23347414767656333,
+      "grad_norm": 0.41126728345921754,
+      "learning_rate": 9.804900944189093e-06,
+      "loss": 0.3648,
+      "step": 981
+    },
+    {
+      "epoch": 0.23371214374962812,
+      "grad_norm": 0.41273761768511535,
+      "learning_rate": 9.804367381487172e-06,
+      "loss": 0.385,
+      "step": 982
+    },
+    {
+      "epoch": 0.23395013982269292,
+      "grad_norm": 0.39008444226884603,
+      "learning_rate": 9.803833104737406e-06,
+      "loss": 0.4549,
+      "step": 983
+    },
+    {
+      "epoch": 0.2341881358957577,
+      "grad_norm": 0.40144010395161167,
+      "learning_rate": 9.803298114019198e-06,
+      "loss": 0.371,
+      "step": 984
+    },
+    {
+      "epoch": 0.2344261319688225,
+      "grad_norm": 0.39554229899897064,
+      "learning_rate": 9.802762409412062e-06,
+      "loss": 0.3357,
+      "step": 985
+    },
+    {
+      "epoch": 0.2346641280418873,
+      "grad_norm": 0.7571829198075979,
+      "learning_rate": 9.802225990995618e-06,
+      "loss": 0.3972,
+      "step": 986
+    },
+    {
+      "epoch": 0.2349021241149521,
+      "grad_norm": 0.4175979906090598,
+      "learning_rate": 9.801688858849589e-06,
+      "loss": 0.4162,
+      "step": 987
+    },
+    {
+      "epoch": 0.2351401201880169,
+      "grad_norm": 0.39744051393120755,
+      "learning_rate": 9.80115101305381e-06,
+      "loss": 0.3587,
+      "step": 988
+    },
+    {
+      "epoch": 0.23537811626108168,
+      "grad_norm": 0.3999755158754268,
+      "learning_rate": 9.800612453688214e-06,
+      "loss": 0.3679,
+      "step": 989
+    },
+    {
+      "epoch": 0.23561611233414648,
+      "grad_norm": 0.41712508601670123,
+      "learning_rate": 9.800073180832848e-06,
+      "loss": 0.4463,
+      "step": 990
+    },
+    {
+      "epoch": 0.23585410840721127,
+      "grad_norm": 0.4401207780378245,
+      "learning_rate": 9.799533194567856e-06,
+      "loss": 0.3739,
+      "step": 991
+    },
+    {
+      "epoch": 0.23609210448027607,
+      "grad_norm": 0.47720852526507335,
+      "learning_rate": 9.7989924949735e-06,
+      "loss": 0.3634,
+      "step": 992
+    },
+    {
+      "epoch": 0.23633010055334086,
+      "grad_norm": 0.41441117101780556,
+      "learning_rate": 9.798451082130136e-06,
+      "loss": 0.4021,
+      "step": 993
+    },
+    {
+      "epoch": 0.23656809662640566,
+      "grad_norm": 0.4079927127031793,
+      "learning_rate": 9.797908956118233e-06,
+      "loss": 0.4398,
+      "step": 994
+    },
+    {
+      "epoch": 0.23680609269947045,
+      "grad_norm": 0.4514084993876973,
+      "learning_rate": 9.797366117018365e-06,
+      "loss": 0.353,
+      "step": 995
+    },
+    {
+      "epoch": 0.23704408877253524,
+      "grad_norm": 0.41156120093358467,
+      "learning_rate": 9.79682256491121e-06,
+      "loss": 0.3714,
+      "step": 996
+    },
+    {
+      "epoch": 0.23728208484560004,
+      "grad_norm": 0.39736420037308545,
+      "learning_rate": 9.796278299877556e-06,
+      "loss": 0.4682,
+      "step": 997
+    },
+    {
+      "epoch": 0.23752008091866483,
+      "grad_norm": 0.419750377621859,
+      "learning_rate": 9.795733321998291e-06,
+      "loss": 0.3953,
+      "step": 998
+    },
+    {
+      "epoch": 0.23775807699172963,
+      "grad_norm": 0.4523615063724095,
+      "learning_rate": 9.795187631354415e-06,
+      "loss": 0.3176,
+      "step": 999
+    },
+    {
+      "epoch": 0.23799607306479442,
+      "grad_norm": 0.43280687276990515,
+      "learning_rate": 9.794641228027029e-06,
+      "loss": 0.3767,
+      "step": 1000
+    },
+    {
+      "epoch": 0.23823406913785922,
+      "grad_norm": 0.41701831566090125,
+      "learning_rate": 9.794094112097342e-06,
+      "loss": 0.4651,
+      "step": 1001
+    },
+    {
+      "epoch": 0.238472065210924,
+      "grad_norm": 0.4670265558164397,
+      "learning_rate": 9.793546283646671e-06,
+      "loss": 0.3804,
+      "step": 1002
+    },
+    {
+      "epoch": 0.2387100612839888,
+      "grad_norm": 0.4073253123192443,
+      "learning_rate": 9.792997742756433e-06,
+      "loss": 0.3447,
+      "step": 1003
+    },
+    {
+      "epoch": 0.2389480573570536,
+      "grad_norm": 0.41912347417284207,
+      "learning_rate": 9.792448489508161e-06,
+      "loss": 0.4335,
+      "step": 1004
+    },
+    {
+      "epoch": 0.2391860534301184,
+      "grad_norm": 0.4601404438983729,
+      "learning_rate": 9.791898523983483e-06,
+      "loss": 0.3913,
+      "step": 1005
+    },
+    {
+      "epoch": 0.2394240495031832,
+      "grad_norm": 0.40669062578592385,
+      "learning_rate": 9.791347846264137e-06,
+      "loss": 0.3174,
+      "step": 1006
+    },
+    {
+      "epoch": 0.23966204557624798,
+      "grad_norm": 0.40618171733993275,
+      "learning_rate": 9.790796456431971e-06,
+      "loss": 0.4043,
+      "step": 1007
+    },
+    {
+      "epoch": 0.23990004164931278,
+      "grad_norm": 0.41827989422393813,
+      "learning_rate": 9.79024435456893e-06,
+      "loss": 0.4413,
+      "step": 1008
+    },
+    {
+      "epoch": 0.24013803772237757,
+      "grad_norm": 0.431949416274619,
+      "learning_rate": 9.789691540757076e-06,
+      "loss": 0.3945,
+      "step": 1009
+    },
+    {
+      "epoch": 0.24037603379544237,
+      "grad_norm": 0.41413010096326724,
+      "learning_rate": 9.789138015078565e-06,
+      "loss": 0.3339,
+      "step": 1010
+    },
+    {
+      "epoch": 0.24061402986850716,
+      "grad_norm": 0.4140278014836118,
+      "learning_rate": 9.78858377761567e-06,
+      "loss": 0.4029,
+      "step": 1011
+    },
+    {
+      "epoch": 0.24085202594157196,
+      "grad_norm": 0.4285228881221896,
+      "learning_rate": 9.78802882845076e-06,
+      "loss": 0.4099,
+      "step": 1012
+    },
+    {
+      "epoch": 0.24109002201463675,
+      "grad_norm": 0.4491959646131581,
+      "learning_rate": 9.787473167666316e-06,
+      "loss": 0.3299,
+      "step": 1013
+    },
+    {
+      "epoch": 0.24132801808770155,
+      "grad_norm": 0.45078633145305924,
+      "learning_rate": 9.786916795344925e-06,
+      "loss": 0.3761,
+      "step": 1014
+    },
+    {
+      "epoch": 0.24156601416076634,
+      "grad_norm": 0.40484830252152926,
+      "learning_rate": 9.786359711569273e-06,
+      "loss": 0.4174,
+      "step": 1015
+    },
+    {
+      "epoch": 0.24180401023383113,
+      "grad_norm": 0.4129343063024655,
+      "learning_rate": 9.785801916422162e-06,
+      "loss": 0.3641,
+      "step": 1016
+    },
+    {
+      "epoch": 0.24204200630689593,
+      "grad_norm": 0.3896049676049958,
+      "learning_rate": 9.78524340998649e-06,
+      "loss": 0.3469,
+      "step": 1017
+    },
+    {
+      "epoch": 0.24228000237996072,
+      "grad_norm": 0.397046837899395,
+      "learning_rate": 9.784684192345264e-06,
+      "loss": 0.3873,
+      "step": 1018
+    },
+    {
+      "epoch": 0.24251799845302552,
+      "grad_norm": 0.46413478846406936,
+      "learning_rate": 9.7841242635816e-06,
+      "loss": 0.4772,
+      "step": 1019
+    },
+    {
+      "epoch": 0.2427559945260903,
+      "grad_norm": 0.4277807175436547,
+      "learning_rate": 9.78356362377872e-06,
+      "loss": 0.3502,
+      "step": 1020
+    },
+    {
+      "epoch": 0.2429939905991551,
+      "grad_norm": 0.40246848535234875,
+      "learning_rate": 9.783002273019942e-06,
+      "loss": 0.3317,
+      "step": 1021
+    },
+    {
+      "epoch": 0.2432319866722199,
+      "grad_norm": 0.40816497903302856,
+      "learning_rate": 9.782440211388703e-06,
+      "loss": 0.4474,
+      "step": 1022
+    },
+    {
+      "epoch": 0.2434699827452847,
+      "grad_norm": 0.43384610531907175,
+      "learning_rate": 9.781877438968536e-06,
+      "loss": 0.3799,
+      "step": 1023
+    },
+    {
+      "epoch": 0.2437079788183495,
+      "grad_norm": 0.425087792025929,
+      "learning_rate": 9.781313955843084e-06,
+      "loss": 0.3573,
+      "step": 1024
+    },
+    {
+      "epoch": 0.24394597489141429,
+      "grad_norm": 0.4026711207325404,
+      "learning_rate": 9.780749762096093e-06,
+      "loss": 0.3878,
+      "step": 1025
+    },
+    {
+      "epoch": 0.24418397096447908,
+      "grad_norm": 0.4098774528539441,
+      "learning_rate": 9.780184857811419e-06,
+      "loss": 0.4786,
+      "step": 1026
+    },
+    {
+      "epoch": 0.24442196703754387,
+      "grad_norm": 0.4396169899920244,
+      "learning_rate": 9.779619243073017e-06,
+      "loss": 0.3444,
+      "step": 1027
+    },
+    {
+      "epoch": 0.24465996311060867,
+      "grad_norm": 0.4136758278535869,
+      "learning_rate": 9.779052917964955e-06,
+      "loss": 0.3948,
+      "step": 1028
+    },
+    {
+      "epoch": 0.24489795918367346,
+      "grad_norm": 0.3739736169388403,
+      "learning_rate": 9.7784858825714e-06,
+      "loss": 0.3961,
+      "step": 1029
+    },
+    {
+      "epoch": 0.24513595525673826,
+      "grad_norm": 0.4184534974361922,
+      "learning_rate": 9.77791813697663e-06,
+      "loss": 0.4134,
+      "step": 1030
+    },
+    {
+      "epoch": 0.24537395132980305,
+      "grad_norm": 0.4728932911361564,
+      "learning_rate": 9.777349681265024e-06,
+      "loss": 0.3557,
+      "step": 1031
+    },
+    {
+      "epoch": 0.24561194740286785,
+      "grad_norm": 0.41399344692303164,
+      "learning_rate": 9.77678051552107e-06,
+      "loss": 0.3869,
+      "step": 1032
+    },
+    {
+      "epoch": 0.24584994347593264,
+      "grad_norm": 0.49034199619398566,
+      "learning_rate": 9.77621063982936e-06,
+      "loss": 0.4363,
+      "step": 1033
+    },
+    {
+      "epoch": 0.24608793954899744,
+      "grad_norm": 0.4165647555586178,
+      "learning_rate": 9.775640054274591e-06,
+      "loss": 0.3841,
+      "step": 1034
+    },
+    {
+      "epoch": 0.24632593562206223,
+      "grad_norm": 0.4360715981984538,
+      "learning_rate": 9.775068758941566e-06,
+      "loss": 0.3205,
+      "step": 1035
+    },
+    {
+      "epoch": 0.24656393169512703,
+      "grad_norm": 0.389169460851251,
+      "learning_rate": 9.774496753915193e-06,
+      "loss": 0.3785,
+      "step": 1036
+    },
+    {
+      "epoch": 0.24680192776819182,
+      "grad_norm": 0.39337607363580057,
+      "learning_rate": 9.773924039280488e-06,
+      "loss": 0.4302,
+      "step": 1037
+    },
+    {
+      "epoch": 0.24703992384125661,
+      "grad_norm": 0.39250700422998375,
+      "learning_rate": 9.77335061512257e-06,
+      "loss": 0.3382,
+      "step": 1038
+    },
+    {
+      "epoch": 0.2472779199143214,
+      "grad_norm": 0.43938604330305114,
+      "learning_rate": 9.772776481526662e-06,
+      "loss": 0.3888,
+      "step": 1039
+    },
+    {
+      "epoch": 0.2475159159873862,
+      "grad_norm": 0.39206815609324325,
+      "learning_rate": 9.772201638578099e-06,
+      "loss": 0.4238,
+      "step": 1040
+    },
+    {
+      "epoch": 0.247753912060451,
+      "grad_norm": 0.4053806301912741,
+      "learning_rate": 9.77162608636231e-06,
+      "loss": 0.3748,
+      "step": 1041
+    },
+    {
+      "epoch": 0.2479919081335158,
+      "grad_norm": 0.4413688344206532,
+      "learning_rate": 9.771049824964843e-06,
+      "loss": 0.3448,
+      "step": 1042
+    },
+    {
+      "epoch": 0.2482299042065806,
+      "grad_norm": 0.4272845904110967,
+      "learning_rate": 9.77047285447134e-06,
+      "loss": 0.4019,
+      "step": 1043
+    },
+    {
+      "epoch": 0.24846790027964538,
+      "grad_norm": 0.41367780286322975,
+      "learning_rate": 9.769895174967556e-06,
+      "loss": 0.4339,
+      "step": 1044
+    },
+    {
+      "epoch": 0.24870589635271018,
+      "grad_norm": 0.41454713221485506,
+      "learning_rate": 9.769316786539349e-06,
+      "loss": 0.368,
+      "step": 1045
+    },
+    {
+      "epoch": 0.24894389242577497,
+      "grad_norm": 0.40425323505335387,
+      "learning_rate": 9.768737689272678e-06,
+      "loss": 0.3679,
+      "step": 1046
+    },
+    {
+      "epoch": 0.24918188849883977,
+      "grad_norm": 0.38710204326010633,
+      "learning_rate": 9.768157883253616e-06,
+      "loss": 0.4258,
+      "step": 1047
+    },
+    {
+      "epoch": 0.24941988457190456,
+      "grad_norm": 0.40041946638737086,
+      "learning_rate": 9.76757736856833e-06,
+      "loss": 0.4006,
+      "step": 1048
+    },
+    {
+      "epoch": 0.24965788064496935,
+      "grad_norm": 0.36660667068666375,
+      "learning_rate": 9.766996145303107e-06,
+      "loss": 0.3208,
+      "step": 1049
+    },
+    {
+      "epoch": 0.24989587671803415,
+      "grad_norm": 0.40641295360789265,
+      "learning_rate": 9.766414213544325e-06,
+      "loss": 0.3991,
+      "step": 1050
+    },
+    {
+      "epoch": 0.25013387279109894,
+      "grad_norm": 0.3963317481756286,
+      "learning_rate": 9.765831573378474e-06,
+      "loss": 0.467,
+      "step": 1051
+    },
+    {
+      "epoch": 0.25037186886416374,
+      "grad_norm": 0.4102266739054517,
+      "learning_rate": 9.765248224892152e-06,
+      "loss": 0.3432,
+      "step": 1052
+    },
+    {
+      "epoch": 0.25060986493722853,
+      "grad_norm": 0.5225965842019219,
+      "learning_rate": 9.764664168172057e-06,
+      "loss": 0.3248,
+      "step": 1053
+    },
+    {
+      "epoch": 0.2508478610102933,
+      "grad_norm": 0.39422579626773063,
+      "learning_rate": 9.764079403304996e-06,
+      "loss": 0.441,
+      "step": 1054
+    },
+    {
+      "epoch": 0.2510858570833581,
+      "grad_norm": 0.43871315920548154,
+      "learning_rate": 9.763493930377877e-06,
+      "loss": 0.4206,
+      "step": 1055
+    },
+    {
+      "epoch": 0.2513238531564229,
+      "grad_norm": 0.39618156243497527,
+      "learning_rate": 9.762907749477717e-06,
+      "loss": 0.3543,
+      "step": 1056
+    },
+    {
+      "epoch": 0.2515618492294877,
+      "grad_norm": 0.40119944869118124,
+      "learning_rate": 9.762320860691636e-06,
+      "loss": 0.3664,
+      "step": 1057
+    },
+    {
+      "epoch": 0.2517998453025525,
+      "grad_norm": 0.4448657055439697,
+      "learning_rate": 9.76173326410686e-06,
+      "loss": 0.4404,
+      "step": 1058
+    },
+    {
+      "epoch": 0.2520378413756173,
+      "grad_norm": 0.4504360863126996,
+      "learning_rate": 9.761144959810723e-06,
+      "loss": 0.3729,
+      "step": 1059
+    },
+    {
+      "epoch": 0.2522758374486821,
+      "grad_norm": 0.3858257157958738,
+      "learning_rate": 9.760555947890659e-06,
+      "loss": 0.3264,
+      "step": 1060
+    },
+    {
+      "epoch": 0.2525138335217469,
+      "grad_norm": 0.42098269355466916,
+      "learning_rate": 9.759966228434212e-06,
+      "loss": 0.4308,
+      "step": 1061
+    },
+    {
+      "epoch": 0.2527518295948117,
+      "grad_norm": 0.4414996944483038,
+      "learning_rate": 9.759375801529026e-06,
+      "loss": 0.4122,
+      "step": 1062
+    },
+    {
+      "epoch": 0.2529898256678765,
+      "grad_norm": 0.4083823838111461,
+      "learning_rate": 9.758784667262856e-06,
+      "loss": 0.3455,
+      "step": 1063
+    },
+    {
+      "epoch": 0.2532278217409413,
+      "grad_norm": 0.39532890350795247,
+      "learning_rate": 9.758192825723556e-06,
+      "loss": 0.3742,
+      "step": 1064
+    },
+    {
+      "epoch": 0.25346581781400607,
+      "grad_norm": 0.7090467989608195,
+      "learning_rate": 9.757600276999092e-06,
+      "loss": 0.4249,
+      "step": 1065
+    },
+    {
+      "epoch": 0.25370381388707086,
+      "grad_norm": 0.40621648230909274,
+      "learning_rate": 9.757007021177529e-06,
+      "loss": 0.3793,
+      "step": 1066
+    },
+    {
+      "epoch": 0.25394180996013566,
+      "grad_norm": 0.4208314833352698,
+      "learning_rate": 9.756413058347039e-06,
+      "loss": 0.3424,
+      "step": 1067
+    },
+    {
+      "epoch": 0.25417980603320045,
+      "grad_norm": 0.377078139573875,
+      "learning_rate": 9.7558183885959e-06,
+      "loss": 0.3902,
+      "step": 1068
+    },
+    {
+      "epoch": 0.25441780210626525,
+      "grad_norm": 0.44783916724457257,
+      "learning_rate": 9.755223012012498e-06,
+      "loss": 0.4407,
+      "step": 1069
+    },
+    {
+      "epoch": 0.25465579817933004,
+      "grad_norm": 0.39045199889729937,
+      "learning_rate": 9.754626928685315e-06,
+      "loss": 0.3372,
+      "step": 1070
+    },
+    {
+      "epoch": 0.25489379425239483,
+      "grad_norm": 0.4319090741276457,
+      "learning_rate": 9.754030138702948e-06,
+      "loss": 0.3226,
+      "step": 1071
+    },
+    {
+      "epoch": 0.25513179032545963,
+      "grad_norm": 0.3789725941806326,
+      "learning_rate": 9.753432642154093e-06,
+      "loss": 0.4311,
+      "step": 1072
+    },
+    {
+      "epoch": 0.2553697863985244,
+      "grad_norm": 0.4119736475459839,
+      "learning_rate": 9.752834439127554e-06,
+      "loss": 0.4179,
+      "step": 1073
+    },
+    {
+      "epoch": 0.2556077824715892,
+      "grad_norm": 0.46500296910275435,
+      "learning_rate": 9.752235529712237e-06,
+      "loss": 0.3472,
+      "step": 1074
+    },
+    {
+      "epoch": 0.255845778544654,
+      "grad_norm": 0.410321050312457,
+      "learning_rate": 9.751635913997155e-06,
+      "loss": 0.3981,
+      "step": 1075
+    },
+    {
+      "epoch": 0.2560837746177188,
+      "grad_norm": 0.41821928052237395,
+      "learning_rate": 9.751035592071427e-06,
+      "loss": 0.4358,
+      "step": 1076
+    },
+    {
+      "epoch": 0.2563217706907836,
+      "grad_norm": 0.4250687984115567,
+      "learning_rate": 9.750434564024276e-06,
+      "loss": 0.3738,
+      "step": 1077
+    },
+    {
+      "epoch": 0.2565597667638484,
+      "grad_norm": 0.4407759908452419,
+      "learning_rate": 9.749832829945027e-06,
+      "loss": 0.3292,
+      "step": 1078
+    },
+    {
+      "epoch": 0.2567977628369132,
+      "grad_norm": 0.36605112835248904,
+      "learning_rate": 9.749230389923117e-06,
+      "loss": 0.4139,
+      "step": 1079
+    },
+    {
+      "epoch": 0.257035758909978,
+      "grad_norm": 0.46252554242389143,
+      "learning_rate": 9.748627244048077e-06,
+      "loss": 0.4072,
+      "step": 1080
+    },
+    {
+      "epoch": 0.2572737549830428,
+      "grad_norm": 0.4131656517840768,
+      "learning_rate": 9.748023392409556e-06,
+      "loss": 0.3444,
+      "step": 1081
+    },
+    {
+      "epoch": 0.2575117510561076,
+      "grad_norm": 0.4244606448210475,
+      "learning_rate": 9.747418835097298e-06,
+      "loss": 0.385,
+      "step": 1082
+    },
+    {
+      "epoch": 0.25774974712917237,
+      "grad_norm": 0.39038280156155775,
+      "learning_rate": 9.746813572201154e-06,
+      "loss": 0.442,
+      "step": 1083
+    },
+    {
+      "epoch": 0.25798774320223716,
+      "grad_norm": 0.4367941002506421,
+      "learning_rate": 9.746207603811085e-06,
+      "loss": 0.3848,
+      "step": 1084
+    },
+    {
+      "epoch": 0.25822573927530196,
+      "grad_norm": 0.43484071040684835,
+      "learning_rate": 9.74560093001715e-06,
+      "loss": 0.317,
+      "step": 1085
+    },
+    {
+      "epoch": 0.25846373534836675,
+      "grad_norm": 0.35367482678247547,
+      "learning_rate": 9.744993550909514e-06,
+      "loss": 0.3873,
+      "step": 1086
+    },
+    {
+      "epoch": 0.25870173142143155,
+      "grad_norm": 0.4941729589297459,
+      "learning_rate": 9.744385466578453e-06,
+      "loss": 0.4777,
+      "step": 1087
+    },
+    {
+      "epoch": 0.25893972749449634,
+      "grad_norm": 0.4272644514387727,
+      "learning_rate": 9.74377667711434e-06,
+      "loss": 0.3518,
+      "step": 1088
+    },
+    {
+      "epoch": 0.25917772356756114,
+      "grad_norm": 0.4169441718338264,
+      "learning_rate": 9.743167182607659e-06,
+      "loss": 0.3746,
+      "step": 1089
+    },
+    {
+      "epoch": 0.25941571964062593,
+      "grad_norm": 0.4436977383688021,
+      "learning_rate": 9.742556983148994e-06,
+      "loss": 0.4359,
+      "step": 1090
+    },
+    {
+      "epoch": 0.2596537157136907,
+      "grad_norm": 0.44907822644616785,
+      "learning_rate": 9.741946078829035e-06,
+      "loss": 0.3661,
+      "step": 1091
+    },
+    {
+      "epoch": 0.2598917117867555,
+      "grad_norm": 0.5085499247152285,
+      "learning_rate": 9.74133446973858e-06,
+      "loss": 0.3236,
+      "step": 1092
+    },
+    {
+      "epoch": 0.2601297078598203,
+      "grad_norm": 0.4095219628657235,
+      "learning_rate": 9.740722155968527e-06,
+      "loss": 0.3725,
+      "step": 1093
+    },
+    {
+      "epoch": 0.2603677039328851,
+      "grad_norm": 0.4492820629768859,
+      "learning_rate": 9.74010913760988e-06,
+      "loss": 0.4523,
+      "step": 1094
+    },
+    {
+      "epoch": 0.2606057000059499,
+      "grad_norm": 0.4632589956141342,
+      "learning_rate": 9.739495414753754e-06,
+      "loss": 0.3544,
+      "step": 1095
+    },
+    {
+      "epoch": 0.2608436960790147,
+      "grad_norm": 0.43509465444418566,
+      "learning_rate": 9.738880987491357e-06,
+      "loss": 0.3868,
+      "step": 1096
+    },
+    {
+      "epoch": 0.2610816921520795,
+      "grad_norm": 0.4155977515794823,
+      "learning_rate": 9.738265855914014e-06,
+      "loss": 0.413,
+      "step": 1097
+    },
+    {
+      "epoch": 0.2613196882251443,
+      "grad_norm": 0.4944891223357048,
+      "learning_rate": 9.737650020113143e-06,
+      "loss": 0.385,
+      "step": 1098
+    },
+    {
+      "epoch": 0.2615576842982091,
+      "grad_norm": 0.4408252014561249,
+      "learning_rate": 9.737033480180276e-06,
+      "loss": 0.3339,
+      "step": 1099
+    },
+    {
+      "epoch": 0.2617956803712739,
+      "grad_norm": 0.40705862767289636,
+      "learning_rate": 9.736416236207046e-06,
+      "loss": 0.3871,
+      "step": 1100
+    },
+    {
+      "epoch": 0.26203367644433867,
+      "grad_norm": 0.4087671198515162,
+      "learning_rate": 9.735798288285189e-06,
+      "loss": 0.4766,
+      "step": 1101
+    },
+    {
+      "epoch": 0.26227167251740346,
+      "grad_norm": 0.4129939242716965,
+      "learning_rate": 9.73517963650655e-06,
+      "loss": 0.3562,
+      "step": 1102
+    },
+    {
+      "epoch": 0.26250966859046826,
+      "grad_norm": 0.3870745733554611,
+      "learning_rate": 9.734560280963072e-06,
+      "loss": 0.3615,
+      "step": 1103
+    },
+    {
+      "epoch": 0.26274766466353305,
+      "grad_norm": 0.39750563414632717,
+      "learning_rate": 9.73394022174681e-06,
+      "loss": 0.4273,
+      "step": 1104
+    },
+    {
+      "epoch": 0.26298566073659785,
+      "grad_norm": 0.4523534366058027,
+      "learning_rate": 9.73331945894992e-06,
+      "loss": 0.3872,
+      "step": 1105
+    },
+    {
+      "epoch": 0.26322365680966264,
+      "grad_norm": 0.48990020782286,
+      "learning_rate": 9.73269799266466e-06,
+      "loss": 0.3577,
+      "step": 1106
+    },
+    {
+      "epoch": 0.26346165288272744,
+      "grad_norm": 0.43917343471689085,
+      "learning_rate": 9.732075822983398e-06,
+      "loss": 0.3667,
+      "step": 1107
+    },
+    {
+      "epoch": 0.26369964895579223,
+      "grad_norm": 0.4263835822823487,
+      "learning_rate": 9.731452949998603e-06,
+      "loss": 0.4421,
+      "step": 1108
+    },
+    {
+      "epoch": 0.263937645028857,
+      "grad_norm": 0.46278837021200253,
+      "learning_rate": 9.730829373802847e-06,
+      "loss": 0.3615,
+      "step": 1109
+    },
+    {
+      "epoch": 0.2641756411019218,
+      "grad_norm": 0.40000848194964195,
+      "learning_rate": 9.730205094488813e-06,
+      "loss": 0.3516,
+      "step": 1110
+    },
+    {
+      "epoch": 0.2644136371749866,
+      "grad_norm": 0.40932668897348323,
+      "learning_rate": 9.729580112149283e-06,
+      "loss": 0.3911,
+      "step": 1111
+    },
+    {
+      "epoch": 0.2646516332480514,
+      "grad_norm": 0.42423822678370915,
+      "learning_rate": 9.728954426877142e-06,
+      "loss": 0.3937,
+      "step": 1112
+    },
+    {
+      "epoch": 0.2648896293211162,
+      "grad_norm": 0.4536666880057892,
+      "learning_rate": 9.728328038765387e-06,
+      "loss": 0.3364,
+      "step": 1113
+    },
+    {
+      "epoch": 0.265127625394181,
+      "grad_norm": 0.4096764042742354,
+      "learning_rate": 9.72770094790711e-06,
+      "loss": 0.3995,
+      "step": 1114
+    },
+    {
+      "epoch": 0.2653656214672458,
+      "grad_norm": 0.41042568452857936,
+      "learning_rate": 9.727073154395516e-06,
+      "loss": 0.4235,
+      "step": 1115
+    },
+    {
+      "epoch": 0.2656036175403106,
+      "grad_norm": 0.4929851761073694,
+      "learning_rate": 9.726444658323908e-06,
+      "loss": 0.3269,
+      "step": 1116
+    },
+    {
+      "epoch": 0.2658416136133754,
+      "grad_norm": 0.44283586974793293,
+      "learning_rate": 9.725815459785696e-06,
+      "loss": 0.3496,
+      "step": 1117
+    },
+    {
+      "epoch": 0.2660796096864402,
+      "grad_norm": 0.4146273039324895,
+      "learning_rate": 9.725185558874399e-06,
+      "loss": 0.3658,
+      "step": 1118
+    },
+    {
+      "epoch": 0.26631760575950497,
+      "grad_norm": 0.4319382476751413,
+      "learning_rate": 9.72455495568363e-06,
+      "loss": 0.4564,
+      "step": 1119
+    },
+    {
+      "epoch": 0.26655560183256977,
+      "grad_norm": 0.48931654146756204,
+      "learning_rate": 9.723923650307116e-06,
+      "loss": 0.3553,
+      "step": 1120
+    },
+    {
+      "epoch": 0.26679359790563456,
+      "grad_norm": 0.4701876581191389,
+      "learning_rate": 9.723291642838682e-06,
+      "loss": 0.3524,
+      "step": 1121
+    },
+    {
+      "epoch": 0.26703159397869936,
+      "grad_norm": 0.39377148630113995,
+      "learning_rate": 9.722658933372262e-06,
+      "loss": 0.4599,
+      "step": 1122
+    },
+    {
+      "epoch": 0.26726959005176415,
+      "grad_norm": 0.4558315940897939,
+      "learning_rate": 9.722025522001892e-06,
+      "loss": 0.389,
+      "step": 1123
+    },
+    {
+      "epoch": 0.26750758612482894,
+      "grad_norm": 0.4215971407730172,
+      "learning_rate": 9.721391408821713e-06,
+      "loss": 0.3064,
+      "step": 1124
+    },
+    {
+      "epoch": 0.26774558219789374,
+      "grad_norm": 0.3911192317304776,
+      "learning_rate": 9.720756593925967e-06,
+      "loss": 0.424,
+      "step": 1125
+    },
+    {
+      "epoch": 0.26798357827095853,
+      "grad_norm": 0.3867608031443001,
+      "learning_rate": 9.720121077409006e-06,
+      "loss": 0.4329,
+      "step": 1126
+    },
+    {
+      "epoch": 0.26822157434402333,
+      "grad_norm": 0.4335469962868136,
+      "learning_rate": 9.719484859365283e-06,
+      "loss": 0.3384,
+      "step": 1127
+    },
+    {
+      "epoch": 0.2684595704170881,
+      "grad_norm": 0.41163503862716694,
+      "learning_rate": 9.718847939889354e-06,
+      "loss": 0.3526,
+      "step": 1128
+    },
+    {
+      "epoch": 0.2686975664901529,
+      "grad_norm": 0.3892507716482463,
+      "learning_rate": 9.718210319075883e-06,
+      "loss": 0.428,
+      "step": 1129
+    },
+    {
+      "epoch": 0.2689355625632177,
+      "grad_norm": 0.4116306866869224,
+      "learning_rate": 9.717571997019637e-06,
+      "loss": 0.3982,
+      "step": 1130
+    },
+    {
+      "epoch": 0.2691735586362825,
+      "grad_norm": 0.39033870830232814,
+      "learning_rate": 9.716932973815485e-06,
+      "loss": 0.3799,
+      "step": 1131
+    },
+    {
+      "epoch": 0.2694115547093473,
+      "grad_norm": 0.40225942909727946,
+      "learning_rate": 9.716293249558401e-06,
+      "loss": 0.3807,
+      "step": 1132
+    },
+    {
+      "epoch": 0.2696495507824121,
+      "grad_norm": 0.39112529026808723,
+      "learning_rate": 9.715652824343465e-06,
+      "loss": 0.4281,
+      "step": 1133
+    },
+    {
+      "epoch": 0.2698875468554769,
+      "grad_norm": 0.3851730979067281,
+      "learning_rate": 9.71501169826586e-06,
+      "loss": 0.3295,
+      "step": 1134
+    },
+    {
+      "epoch": 0.2701255429285417,
+      "grad_norm": 0.41373661540628287,
+      "learning_rate": 9.714369871420872e-06,
+      "loss": 0.3928,
+      "step": 1135
+    },
+    {
+      "epoch": 0.2703635390016065,
+      "grad_norm": 0.4461170158222069,
+      "learning_rate": 9.713727343903893e-06,
+      "loss": 0.4109,
+      "step": 1136
+    },
+    {
+      "epoch": 0.2706015350746713,
+      "grad_norm": 0.4212053908749594,
+      "learning_rate": 9.71308411581042e-06,
+      "loss": 0.4206,
+      "step": 1137
+    },
+    {
+      "epoch": 0.27083953114773607,
+      "grad_norm": 0.4563483334376799,
+      "learning_rate": 9.71244018723605e-06,
+      "loss": 0.3494,
+      "step": 1138
+    },
+    {
+      "epoch": 0.27107752722080086,
+      "grad_norm": 0.4063198246744073,
+      "learning_rate": 9.711795558276489e-06,
+      "loss": 0.3797,
+      "step": 1139
+    },
+    {
+      "epoch": 0.27131552329386566,
+      "grad_norm": 0.4766611835404524,
+      "learning_rate": 9.711150229027544e-06,
+      "loss": 0.4136,
+      "step": 1140
+    },
+    {
+      "epoch": 0.27155351936693045,
+      "grad_norm": 0.43835373298275027,
+      "learning_rate": 9.710504199585127e-06,
+      "loss": 0.3632,
+      "step": 1141
+    },
+    {
+      "epoch": 0.27179151543999525,
+      "grad_norm": 0.4164970849311708,
+      "learning_rate": 9.709857470045251e-06,
+      "loss": 0.329,
+      "step": 1142
+    },
+    {
+      "epoch": 0.27202951151306004,
+      "grad_norm": 0.40401054483859816,
+      "learning_rate": 9.709210040504042e-06,
+      "loss": 0.3835,
+      "step": 1143
+    },
+    {
+      "epoch": 0.27226750758612484,
+      "grad_norm": 0.39714074001954824,
+      "learning_rate": 9.708561911057719e-06,
+      "loss": 0.4795,
+      "step": 1144
+    },
+    {
+      "epoch": 0.27250550365918963,
+      "grad_norm": 0.4680110103103199,
+      "learning_rate": 9.707913081802613e-06,
+      "loss": 0.341,
+      "step": 1145
+    },
+    {
+      "epoch": 0.2727434997322544,
+      "grad_norm": 0.4089665368213675,
+      "learning_rate": 9.707263552835153e-06,
+      "loss": 0.3453,
+      "step": 1146
+    },
+    {
+      "epoch": 0.2729814958053192,
+      "grad_norm": 0.39910736552697396,
+      "learning_rate": 9.70661332425188e-06,
+      "loss": 0.3954,
+      "step": 1147
+    },
+    {
+      "epoch": 0.273219491878384,
+      "grad_norm": 0.4278429044590152,
+      "learning_rate": 9.705962396149428e-06,
+      "loss": 0.3727,
+      "step": 1148
+    },
+    {
+      "epoch": 0.2734574879514488,
+      "grad_norm": 0.4381769229333954,
+      "learning_rate": 9.705310768624545e-06,
+      "loss": 0.3003,
+      "step": 1149
+    },
+    {
+      "epoch": 0.2736954840245136,
+      "grad_norm": 0.3892819098989479,
+      "learning_rate": 9.704658441774078e-06,
+      "loss": 0.3448,
+      "step": 1150
+    },
+    {
+      "epoch": 0.2739334800975784,
+      "grad_norm": 0.45797697785561536,
+      "learning_rate": 9.704005415694979e-06,
+      "loss": 0.4175,
+      "step": 1151
+    },
+    {
+      "epoch": 0.2741714761706432,
+      "grad_norm": 0.4458801671185096,
+      "learning_rate": 9.703351690484305e-06,
+      "loss": 0.3442,
+      "step": 1152
+    },
+    {
+      "epoch": 0.274409472243708,
+      "grad_norm": 0.43193201208559157,
+      "learning_rate": 9.702697266239211e-06,
+      "loss": 0.3373,
+      "step": 1153
+    },
+    {
+      "epoch": 0.2746474683167728,
+      "grad_norm": 0.41679902408312985,
+      "learning_rate": 9.702042143056966e-06,
+      "loss": 0.3958,
+      "step": 1154
+    },
+    {
+      "epoch": 0.2748854643898376,
+      "grad_norm": 0.443366075981643,
+      "learning_rate": 9.701386321034937e-06,
+      "loss": 0.4387,
+      "step": 1155
+    },
+    {
+      "epoch": 0.27512346046290237,
+      "grad_norm": 0.4278061958287098,
+      "learning_rate": 9.700729800270592e-06,
+      "loss": 0.3615,
+      "step": 1156
+    },
+    {
+      "epoch": 0.27536145653596716,
+      "grad_norm": 0.4281403145666002,
+      "learning_rate": 9.700072580861511e-06,
+      "loss": 0.3827,
+      "step": 1157
+    },
+    {
+      "epoch": 0.27559945260903196,
+      "grad_norm": 0.4404470075524074,
+      "learning_rate": 9.699414662905368e-06,
+      "loss": 0.4316,
+      "step": 1158
+    },
+    {
+      "epoch": 0.27583744868209675,
+      "grad_norm": 0.44475673605890376,
+      "learning_rate": 9.698756046499948e-06,
+      "loss": 0.3357,
+      "step": 1159
+    },
+    {
+      "epoch": 0.27607544475516155,
+      "grad_norm": 0.43252690813011324,
+      "learning_rate": 9.698096731743139e-06,
+      "loss": 0.3298,
+      "step": 1160
+    },
+    {
+      "epoch": 0.27631344082822634,
+      "grad_norm": 0.40596838215176706,
+      "learning_rate": 9.69743671873293e-06,
+      "loss": 0.3876,
+      "step": 1161
+    },
+    {
+      "epoch": 0.27655143690129114,
+      "grad_norm": 0.38244031961085745,
+      "learning_rate": 9.696776007567414e-06,
+      "loss": 0.42,
+      "step": 1162
+    },
+    {
+      "epoch": 0.27678943297435593,
+      "grad_norm": 0.39348329580696617,
+      "learning_rate": 9.696114598344794e-06,
+      "loss": 0.3526,
+      "step": 1163
+    },
+    {
+      "epoch": 0.2770274290474207,
+      "grad_norm": 0.40702007665049295,
+      "learning_rate": 9.695452491163367e-06,
+      "loss": 0.3586,
+      "step": 1164
+    },
+    {
+      "epoch": 0.2772654251204855,
+      "grad_norm": 0.3806659463696473,
+      "learning_rate": 9.694789686121538e-06,
+      "loss": 0.4213,
+      "step": 1165
+    },
+    {
+      "epoch": 0.2775034211935503,
+      "grad_norm": 0.526039242215381,
+      "learning_rate": 9.69412618331782e-06,
+      "loss": 0.3702,
+      "step": 1166
+    },
+    {
+      "epoch": 0.2777414172666151,
+      "grad_norm": 0.4034865628814754,
+      "learning_rate": 9.693461982850824e-06,
+      "loss": 0.3582,
+      "step": 1167
+    },
+    {
+      "epoch": 0.2779794133396799,
+      "grad_norm": 0.36522029047320065,
+      "learning_rate": 9.692797084819265e-06,
+      "loss": 0.3887,
+      "step": 1168
+    },
+    {
+      "epoch": 0.2782174094127447,
+      "grad_norm": 0.40180604628740285,
+      "learning_rate": 9.692131489321968e-06,
+      "loss": 0.4349,
+      "step": 1169
+    },
+    {
+      "epoch": 0.2784554054858095,
+      "grad_norm": 0.4153489962964815,
+      "learning_rate": 9.691465196457852e-06,
+      "loss": 0.3318,
+      "step": 1170
+    },
+    {
+      "epoch": 0.2786934015588743,
+      "grad_norm": 0.3762423285238886,
+      "learning_rate": 9.690798206325947e-06,
+      "loss": 0.3601,
+      "step": 1171
+    },
+    {
+      "epoch": 0.2789313976319391,
+      "grad_norm": 0.41201247864367047,
+      "learning_rate": 9.690130519025382e-06,
+      "loss": 0.444,
+      "step": 1172
+    },
+    {
+      "epoch": 0.2791693937050039,
+      "grad_norm": 0.4139424779362683,
+      "learning_rate": 9.689462134655396e-06,
+      "loss": 0.4238,
+      "step": 1173
+    },
+    {
+      "epoch": 0.27940738977806867,
+      "grad_norm": 0.4213437892645641,
+      "learning_rate": 9.688793053315324e-06,
+      "loss": 0.3575,
+      "step": 1174
+    },
+    {
+      "epoch": 0.27964538585113347,
+      "grad_norm": 0.42068466169572954,
+      "learning_rate": 9.688123275104611e-06,
+      "loss": 0.3755,
+      "step": 1175
+    },
+    {
+      "epoch": 0.27988338192419826,
+      "grad_norm": 0.44836158304136864,
+      "learning_rate": 9.6874528001228e-06,
+      "loss": 0.4326,
+      "step": 1176
+    },
+    {
+      "epoch": 0.28012137799726305,
+      "grad_norm": 0.4033459951159842,
+      "learning_rate": 9.686781628469537e-06,
+      "loss": 0.366,
+      "step": 1177
+    },
+    {
+      "epoch": 0.28035937407032785,
+      "grad_norm": 0.427641362560405,
+      "learning_rate": 9.686109760244583e-06,
+      "loss": 0.3167,
+      "step": 1178
+    },
+    {
+      "epoch": 0.28059737014339264,
+      "grad_norm": 0.4312381613704766,
+      "learning_rate": 9.685437195547788e-06,
+      "loss": 0.4068,
+      "step": 1179
+    },
+    {
+      "epoch": 0.28083536621645744,
+      "grad_norm": 0.38271427698464,
+      "learning_rate": 9.684763934479116e-06,
+      "loss": 0.3542,
+      "step": 1180
+    },
+    {
+      "epoch": 0.28107336228952223,
+      "grad_norm": 0.40970093005177965,
+      "learning_rate": 9.684089977138625e-06,
+      "loss": 0.3241,
+      "step": 1181
+    },
+    {
+      "epoch": 0.281311358362587,
+      "grad_norm": 0.47135849095032084,
+      "learning_rate": 9.683415323626487e-06,
+      "loss": 0.3948,
+      "step": 1182
+    },
+    {
+      "epoch": 0.2815493544356518,
+      "grad_norm": 0.4150750744550006,
+      "learning_rate": 9.682739974042967e-06,
+      "loss": 0.4717,
+      "step": 1183
+    },
+    {
+      "epoch": 0.2817873505087166,
+      "grad_norm": 0.3988087657033761,
+      "learning_rate": 9.682063928488444e-06,
+      "loss": 0.3406,
+      "step": 1184
+    },
+    {
+      "epoch": 0.2820253465817814,
+      "grad_norm": 0.3730302298116086,
+      "learning_rate": 9.68138718706339e-06,
+      "loss": 0.3291,
+      "step": 1185
+    },
+    {
+      "epoch": 0.2822633426548462,
+      "grad_norm": 0.41362346247623033,
+      "learning_rate": 9.68070974986839e-06,
+      "loss": 0.4121,
+      "step": 1186
+    },
+    {
+      "epoch": 0.282501338727911,
+      "grad_norm": 0.4049749435958661,
+      "learning_rate": 9.680031617004127e-06,
+      "loss": 0.4326,
+      "step": 1187
+    },
+    {
+      "epoch": 0.2827393348009758,
+      "grad_norm": 0.43281535830805595,
+      "learning_rate": 9.679352788571385e-06,
+      "loss": 0.3082,
+      "step": 1188
+    },
+    {
+      "epoch": 0.2829773308740406,
+      "grad_norm": 0.38623462120774466,
+      "learning_rate": 9.678673264671057e-06,
+      "loss": 0.3458,
+      "step": 1189
+    },
+    {
+      "epoch": 0.2832153269471054,
+      "grad_norm": 0.40708256399448933,
+      "learning_rate": 9.677993045404138e-06,
+      "loss": 0.4467,
+      "step": 1190
+    },
+    {
+      "epoch": 0.2834533230201702,
+      "grad_norm": 0.3996160590648185,
+      "learning_rate": 9.677312130871724e-06,
+      "loss": 0.3843,
+      "step": 1191
+    },
+    {
+      "epoch": 0.283691319093235,
+      "grad_norm": 0.43448067080567276,
+      "learning_rate": 9.676630521175017e-06,
+      "loss": 0.3182,
+      "step": 1192
+    },
+    {
+      "epoch": 0.28392931516629977,
+      "grad_norm": 0.42327090704304354,
+      "learning_rate": 9.67594821641532e-06,
+      "loss": 0.3932,
+      "step": 1193
+    },
+    {
+      "epoch": 0.28416731123936456,
+      "grad_norm": 0.39790381975349265,
+      "learning_rate": 9.675265216694041e-06,
+      "loss": 0.4357,
+      "step": 1194
+    },
+    {
+      "epoch": 0.28440530731242936,
+      "grad_norm": 0.42504381205655906,
+      "learning_rate": 9.67458152211269e-06,
+      "loss": 0.3491,
+      "step": 1195
+    },
+    {
+      "epoch": 0.28464330338549415,
+      "grad_norm": 0.37577693430446824,
+      "learning_rate": 9.673897132772881e-06,
+      "loss": 0.3572,
+      "step": 1196
+    },
+    {
+      "epoch": 0.28488129945855895,
+      "grad_norm": 0.38908763308437694,
+      "learning_rate": 9.67321204877633e-06,
+      "loss": 0.4457,
+      "step": 1197
+    },
+    {
+      "epoch": 0.28511929553162374,
+      "grad_norm": 0.4153450452315847,
+      "learning_rate": 9.672526270224861e-06,
+      "loss": 0.4017,
+      "step": 1198
+    },
+    {
+      "epoch": 0.28535729160468853,
+      "grad_norm": 0.3652029600299754,
+      "learning_rate": 9.671839797220394e-06,
+      "loss": 0.312,
+      "step": 1199
+    },
+    {
+      "epoch": 0.28559528767775333,
+      "grad_norm": 0.3965275761201531,
+      "learning_rate": 9.67115262986496e-06,
+      "loss": 0.3594,
+      "step": 1200
+    },
+    {
+      "epoch": 0.2858332837508181,
+      "grad_norm": 0.3978539830135749,
+      "learning_rate": 9.670464768260684e-06,
+      "loss": 0.433,
+      "step": 1201
+    },
+    {
+      "epoch": 0.2860712798238829,
+      "grad_norm": 0.3911610545469143,
+      "learning_rate": 9.669776212509802e-06,
+      "loss": 0.3459,
+      "step": 1202
+    },
+    {
+      "epoch": 0.2863092758969477,
+      "grad_norm": 0.4563276723079206,
+      "learning_rate": 9.669086962714651e-06,
+      "loss": 0.3656,
+      "step": 1203
+    },
+    {
+      "epoch": 0.2865472719700125,
+      "grad_norm": 0.37157585507019325,
+      "learning_rate": 9.66839701897767e-06,
+      "loss": 0.4093,
+      "step": 1204
+    },
+    {
+      "epoch": 0.2867852680430773,
+      "grad_norm": 0.43109471368785796,
+      "learning_rate": 9.667706381401401e-06,
+      "loss": 0.4173,
+      "step": 1205
+    },
+    {
+      "epoch": 0.2870232641161421,
+      "grad_norm": 0.4106000046269324,
+      "learning_rate": 9.667015050088489e-06,
+      "loss": 0.3157,
+      "step": 1206
+    },
+    {
+      "epoch": 0.2872612601892069,
+      "grad_norm": 0.49437005628521224,
+      "learning_rate": 9.666323025141687e-06,
+      "loss": 0.4168,
+      "step": 1207
+    },
+    {
+      "epoch": 0.2874992562622717,
+      "grad_norm": 0.423786447227811,
+      "learning_rate": 9.66563030666384e-06,
+      "loss": 0.4489,
+      "step": 1208
+    },
+    {
+      "epoch": 0.2877372523353365,
+      "grad_norm": 0.41056267338791924,
+      "learning_rate": 9.66493689475791e-06,
+      "loss": 0.3675,
+      "step": 1209
+    },
+    {
+      "epoch": 0.2879752484084013,
+      "grad_norm": 0.38996061322766,
+      "learning_rate": 9.664242789526952e-06,
+      "loss": 0.3332,
+      "step": 1210
+    },
+    {
+      "epoch": 0.28821324448146607,
+      "grad_norm": 0.38150783205902883,
+      "learning_rate": 9.663547991074129e-06,
+      "loss": 0.3972,
+      "step": 1211
+    },
+    {
+      "epoch": 0.28845124055453086,
+      "grad_norm": 0.4577771253949555,
+      "learning_rate": 9.662852499502702e-06,
+      "loss": 0.4332,
+      "step": 1212
+    },
+    {
+      "epoch": 0.28868923662759566,
+      "grad_norm": 0.39041968301837926,
+      "learning_rate": 9.66215631491604e-06,
+      "loss": 0.3268,
+      "step": 1213
+    },
+    {
+      "epoch": 0.28892723270066045,
+      "grad_norm": 0.39588825022995383,
+      "learning_rate": 9.661459437417616e-06,
+      "loss": 0.3653,
+      "step": 1214
+    },
+    {
+      "epoch": 0.28916522877372525,
+      "grad_norm": 0.38238412045047987,
+      "learning_rate": 9.660761867110997e-06,
+      "loss": 0.4436,
+      "step": 1215
+    },
+    {
+      "epoch": 0.28940322484679004,
+      "grad_norm": 0.4388597314826315,
+      "learning_rate": 9.660063604099866e-06,
+      "loss": 0.3574,
+      "step": 1216
+    },
+    {
+      "epoch": 0.28964122091985484,
+      "grad_norm": 0.3979764946768326,
+      "learning_rate": 9.659364648487997e-06,
+      "loss": 0.3016,
+      "step": 1217
+    },
+    {
+      "epoch": 0.28987921699291963,
+      "grad_norm": 0.38473163200776395,
+      "learning_rate": 9.658665000379275e-06,
+      "loss": 0.4002,
+      "step": 1218
+    },
+    {
+      "epoch": 0.2901172130659844,
+      "grad_norm": 0.410009814193448,
+      "learning_rate": 9.657964659877683e-06,
+      "loss": 0.4234,
+      "step": 1219
+    },
+    {
+      "epoch": 0.2903552091390492,
+      "grad_norm": 0.44464576009309903,
+      "learning_rate": 9.657263627087312e-06,
+      "loss": 0.3546,
+      "step": 1220
+    },
+    {
+      "epoch": 0.290593205212114,
+      "grad_norm": 0.3867603419054925,
+      "learning_rate": 9.656561902112349e-06,
+      "loss": 0.3863,
+      "step": 1221
+    },
+    {
+      "epoch": 0.2908312012851788,
+      "grad_norm": 0.3979315475026651,
+      "learning_rate": 9.655859485057091e-06,
+      "loss": 0.4364,
+      "step": 1222
+    },
+    {
+      "epoch": 0.2910691973582436,
+      "grad_norm": 0.3855514558708465,
+      "learning_rate": 9.655156376025932e-06,
+      "loss": 0.3804,
+      "step": 1223
+    },
+    {
+      "epoch": 0.2913071934313084,
+      "grad_norm": 0.39833146644821477,
+      "learning_rate": 9.654452575123373e-06,
+      "loss": 0.3113,
+      "step": 1224
+    },
+    {
+      "epoch": 0.2915451895043732,
+      "grad_norm": 0.3873575017477939,
+      "learning_rate": 9.653748082454016e-06,
+      "loss": 0.4139,
+      "step": 1225
+    },
+    {
+      "epoch": 0.291783185577438,
+      "grad_norm": 0.4339801773731958,
+      "learning_rate": 9.653042898122565e-06,
+      "loss": 0.4278,
+      "step": 1226
+    },
+    {
+      "epoch": 0.2920211816505028,
+      "grad_norm": 0.4448702833601803,
+      "learning_rate": 9.652337022233829e-06,
+      "loss": 0.3272,
+      "step": 1227
+    },
+    {
+      "epoch": 0.2922591777235676,
+      "grad_norm": 0.42298609998301084,
+      "learning_rate": 9.651630454892718e-06,
+      "loss": 0.3342,
+      "step": 1228
+    },
+    {
+      "epoch": 0.29249717379663237,
+      "grad_norm": 0.386945622028957,
+      "learning_rate": 9.650923196204248e-06,
+      "loss": 0.4296,
+      "step": 1229
+    },
+    {
+      "epoch": 0.29273516986969716,
+      "grad_norm": 0.41831065108508075,
+      "learning_rate": 9.650215246273529e-06,
+      "loss": 0.4312,
+      "step": 1230
+    },
+    {
+      "epoch": 0.29297316594276196,
+      "grad_norm": 0.39310689224557216,
+      "learning_rate": 9.649506605205786e-06,
+      "loss": 0.3573,
+      "step": 1231
+    },
+    {
+      "epoch": 0.29321116201582675,
+      "grad_norm": 0.40652348372519026,
+      "learning_rate": 9.648797273106338e-06,
+      "loss": 0.3865,
+      "step": 1232
+    },
+    {
+      "epoch": 0.29344915808889155,
+      "grad_norm": 0.38816699963135914,
+      "learning_rate": 9.648087250080609e-06,
+      "loss": 0.4138,
+      "step": 1233
+    },
+    {
+      "epoch": 0.29368715416195634,
+      "grad_norm": 0.42050242112991676,
+      "learning_rate": 9.647376536234126e-06,
+      "loss": 0.3695,
+      "step": 1234
+    },
+    {
+      "epoch": 0.29392515023502114,
+      "grad_norm": 0.45027811269540524,
+      "learning_rate": 9.64666513167252e-06,
+      "loss": 0.3329,
+      "step": 1235
+    },
+    {
+      "epoch": 0.29416314630808593,
+      "grad_norm": 0.4324780377574711,
+      "learning_rate": 9.645953036501521e-06,
+      "loss": 0.412,
+      "step": 1236
+    },
+    {
+      "epoch": 0.2944011423811507,
+      "grad_norm": 0.4229530775773719,
+      "learning_rate": 9.645240250826969e-06,
+      "loss": 0.4417,
+      "step": 1237
+    },
+    {
+      "epoch": 0.2946391384542155,
+      "grad_norm": 0.5100450045808651,
+      "learning_rate": 9.644526774754794e-06,
+      "loss": 0.3331,
+      "step": 1238
+    },
+    {
+      "epoch": 0.2948771345272803,
+      "grad_norm": 0.39386712906627963,
+      "learning_rate": 9.643812608391042e-06,
+      "loss": 0.4048,
+      "step": 1239
+    },
+    {
+      "epoch": 0.2951151306003451,
+      "grad_norm": 0.4161776575953358,
+      "learning_rate": 9.643097751841854e-06,
+      "loss": 0.4666,
+      "step": 1240
+    },
+    {
+      "epoch": 0.2953531266734099,
+      "grad_norm": 0.40609935662404656,
+      "learning_rate": 9.642382205213476e-06,
+      "loss": 0.3557,
+      "step": 1241
+    },
+    {
+      "epoch": 0.2955911227464747,
+      "grad_norm": 0.40786129971537194,
+      "learning_rate": 9.641665968612254e-06,
+      "loss": 0.368,
+      "step": 1242
+    },
+    {
+      "epoch": 0.2958291188195395,
+      "grad_norm": 0.39892048689403614,
+      "learning_rate": 9.640949042144641e-06,
+      "loss": 0.374,
+      "step": 1243
+    },
+    {
+      "epoch": 0.2960671148926043,
+      "grad_norm": 0.42072884743876504,
+      "learning_rate": 9.640231425917186e-06,
+      "loss": 0.4409,
+      "step": 1244
+    },
+    {
+      "epoch": 0.2963051109656691,
+      "grad_norm": 0.3809774278312238,
+      "learning_rate": 9.63951312003655e-06,
+      "loss": 0.3276,
+      "step": 1245
+    },
+    {
+      "epoch": 0.2965431070387339,
+      "grad_norm": 0.4112816053420641,
+      "learning_rate": 9.638794124609487e-06,
+      "loss": 0.3342,
+      "step": 1246
+    },
+    {
+      "epoch": 0.29678110311179867,
+      "grad_norm": 0.41137858878488315,
+      "learning_rate": 9.63807443974286e-06,
+      "loss": 0.4281,
+      "step": 1247
+    },
+    {
+      "epoch": 0.29701909918486347,
+      "grad_norm": 0.4301212258272213,
+      "learning_rate": 9.63735406554363e-06,
+      "loss": 0.333,
+      "step": 1248
+    },
+    {
+      "epoch": 0.29725709525792826,
+      "grad_norm": 0.47855201036627526,
+      "learning_rate": 9.636633002118865e-06,
+      "loss": 0.3555,
+      "step": 1249
+    },
+    {
+      "epoch": 0.29749509133099306,
+      "grad_norm": 0.3926815914537034,
+      "learning_rate": 9.635911249575729e-06,
+      "loss": 0.3723,
+      "step": 1250
+    },
+    {
+      "epoch": 0.29773308740405785,
+      "grad_norm": 0.41169709632873663,
+      "learning_rate": 9.635188808021496e-06,
+      "loss": 0.443,
+      "step": 1251
+    },
+    {
+      "epoch": 0.29797108347712264,
+      "grad_norm": 0.41928978072848383,
+      "learning_rate": 9.634465677563537e-06,
+      "loss": 0.3731,
+      "step": 1252
+    },
+    {
+      "epoch": 0.29820907955018744,
+      "grad_norm": 0.4466280351365597,
+      "learning_rate": 9.633741858309325e-06,
+      "loss": 0.3359,
+      "step": 1253
+    },
+    {
+      "epoch": 0.29844707562325223,
+      "grad_norm": 0.5973113828554666,
+      "learning_rate": 9.633017350366441e-06,
+      "loss": 0.4,
+      "step": 1254
+    },
+    {
+      "epoch": 0.29868507169631703,
+      "grad_norm": 0.4181046582171295,
+      "learning_rate": 9.632292153842565e-06,
+      "loss": 0.4306,
+      "step": 1255
+    },
+    {
+      "epoch": 0.2989230677693818,
+      "grad_norm": 0.4310021297808129,
+      "learning_rate": 9.631566268845476e-06,
+      "loss": 0.3564,
+      "step": 1256
+    },
+    {
+      "epoch": 0.2991610638424466,
+      "grad_norm": 0.38843714102902427,
+      "learning_rate": 9.630839695483059e-06,
+      "loss": 0.3675,
+      "step": 1257
+    },
+    {
+      "epoch": 0.2993990599155114,
+      "grad_norm": 0.40697877470197674,
+      "learning_rate": 9.630112433863304e-06,
+      "loss": 0.4735,
+      "step": 1258
+    },
+    {
+      "epoch": 0.2996370559885762,
+      "grad_norm": 0.4383586613694374,
+      "learning_rate": 9.629384484094296e-06,
+      "loss": 0.3543,
+      "step": 1259
+    },
+    {
+      "epoch": 0.299875052061641,
+      "grad_norm": 0.44985065245385947,
+      "learning_rate": 9.628655846284228e-06,
+      "loss": 0.3336,
+      "step": 1260
+    },
+    {
+      "epoch": 0.3001130481347058,
+      "grad_norm": 0.394156383482113,
+      "learning_rate": 9.627926520541395e-06,
+      "loss": 0.3836,
+      "step": 1261
+    },
+    {
+      "epoch": 0.3003510442077706,
+      "grad_norm": 0.48348036503403014,
+      "learning_rate": 9.627196506974192e-06,
+      "loss": 0.4122,
+      "step": 1262
+    },
+    {
+      "epoch": 0.3005890402808354,
+      "grad_norm": 0.4046212158723506,
+      "learning_rate": 9.626465805691117e-06,
+      "loss": 0.3412,
+      "step": 1263
+    },
+    {
+      "epoch": 0.3008270363539002,
+      "grad_norm": 0.4637536508877546,
+      "learning_rate": 9.625734416800768e-06,
+      "loss": 0.3617,
+      "step": 1264
+    },
+    {
+      "epoch": 0.301065032426965,
+      "grad_norm": 0.3696735431596717,
+      "learning_rate": 9.625002340411851e-06,
+      "loss": 0.421,
+      "step": 1265
+    },
+    {
+      "epoch": 0.30130302850002977,
+      "grad_norm": 0.44039294060962975,
+      "learning_rate": 9.624269576633168e-06,
+      "loss": 0.342,
+      "step": 1266
+    },
+    {
+      "epoch": 0.30154102457309456,
+      "grad_norm": 0.4066558094391624,
+      "learning_rate": 9.623536125573628e-06,
+      "loss": 0.3259,
+      "step": 1267
+    },
+    {
+      "epoch": 0.30177902064615936,
+      "grad_norm": 0.4228494081970754,
+      "learning_rate": 9.622801987342239e-06,
+      "loss": 0.3966,
+      "step": 1268
+    },
+    {
+      "epoch": 0.30201701671922415,
+      "grad_norm": 0.3929044823680207,
+      "learning_rate": 9.622067162048111e-06,
+      "loss": 0.4214,
+      "step": 1269
+    },
+    {
+      "epoch": 0.30225501279228895,
+      "grad_norm": 0.4364618202087389,
+      "learning_rate": 9.62133164980046e-06,
+      "loss": 0.3539,
+      "step": 1270
+    },
+    {
+      "epoch": 0.30249300886535374,
+      "grad_norm": 0.3973240447603393,
+      "learning_rate": 9.620595450708598e-06,
+      "loss": 0.3228,
+      "step": 1271
+    },
+    {
+      "epoch": 0.30273100493841854,
+      "grad_norm": 0.423749076343449,
+      "learning_rate": 9.619858564881945e-06,
+      "loss": 0.4197,
+      "step": 1272
+    },
+    {
+      "epoch": 0.30296900101148333,
+      "grad_norm": 0.42928761287682166,
+      "learning_rate": 9.61912099243002e-06,
+      "loss": 0.3675,
+      "step": 1273
+    },
+    {
+      "epoch": 0.3032069970845481,
+      "grad_norm": 0.4262017901034838,
+      "learning_rate": 9.618382733462443e-06,
+      "loss": 0.3333,
+      "step": 1274
+    },
+    {
+      "epoch": 0.3034449931576129,
+      "grad_norm": 0.41851797958081777,
+      "learning_rate": 9.617643788088938e-06,
+      "loss": 0.4031,
+      "step": 1275
+    },
+    {
+      "epoch": 0.3036829892306777,
+      "grad_norm": 0.44040318016484536,
+      "learning_rate": 9.616904156419332e-06,
+      "loss": 0.4259,
+      "step": 1276
+    },
+    {
+      "epoch": 0.3039209853037425,
+      "grad_norm": 0.41176036786394593,
+      "learning_rate": 9.616163838563551e-06,
+      "loss": 0.323,
+      "step": 1277
+    },
+    {
+      "epoch": 0.3041589813768073,
+      "grad_norm": 0.39969851565063835,
+      "learning_rate": 9.615422834631627e-06,
+      "loss": 0.311,
+      "step": 1278
+    },
+    {
+      "epoch": 0.3043969774498721,
+      "grad_norm": 0.4226290204089705,
+      "learning_rate": 9.614681144733688e-06,
+      "loss": 0.4125,
+      "step": 1279
+    },
+    {
+      "epoch": 0.3046349735229369,
+      "grad_norm": 0.43676015567681714,
+      "learning_rate": 9.61393876897997e-06,
+      "loss": 0.4041,
+      "step": 1280
+    },
+    {
+      "epoch": 0.3048729695960017,
+      "grad_norm": 0.3863010410381727,
+      "learning_rate": 9.613195707480808e-06,
+      "loss": 0.3125,
+      "step": 1281
+    },
+    {
+      "epoch": 0.3051109656690665,
+      "grad_norm": 0.41088520843304066,
+      "learning_rate": 9.612451960346636e-06,
+      "loss": 0.3645,
+      "step": 1282
+    },
+    {
+      "epoch": 0.3053489617421313,
+      "grad_norm": 0.4163362500331567,
+      "learning_rate": 9.611707527688e-06,
+      "loss": 0.4114,
+      "step": 1283
+    },
+    {
+      "epoch": 0.30558695781519607,
+      "grad_norm": 0.4627242806057053,
+      "learning_rate": 9.610962409615534e-06,
+      "loss": 0.3343,
+      "step": 1284
+    },
+    {
+      "epoch": 0.30582495388826086,
+      "grad_norm": 0.4027539448469817,
+      "learning_rate": 9.610216606239987e-06,
+      "loss": 0.3279,
+      "step": 1285
+    },
+    {
+      "epoch": 0.30606294996132566,
+      "grad_norm": 0.4145181397037847,
+      "learning_rate": 9.609470117672199e-06,
+      "loss": 0.4067,
+      "step": 1286
+    },
+    {
+      "epoch": 0.30630094603439045,
+      "grad_norm": 0.41297813308066056,
+      "learning_rate": 9.608722944023119e-06,
+      "loss": 0.3966,
+      "step": 1287
+    },
+    {
+      "epoch": 0.30653894210745525,
+      "grad_norm": 0.410910470454594,
+      "learning_rate": 9.607975085403796e-06,
+      "loss": 0.3297,
+      "step": 1288
+    },
+    {
+      "epoch": 0.30677693818052004,
+      "grad_norm": 0.3854845755046187,
+      "learning_rate": 9.607226541925379e-06,
+      "loss": 0.3767,
+      "step": 1289
+    },
+    {
+      "epoch": 0.30701493425358484,
+      "grad_norm": 0.4443276950244607,
+      "learning_rate": 9.60647731369912e-06,
+      "loss": 0.4304,
+      "step": 1290
+    },
+    {
+      "epoch": 0.30725293032664963,
+      "grad_norm": 0.48085590127195105,
+      "learning_rate": 9.605727400836373e-06,
+      "loss": 0.3869,
+      "step": 1291
+    },
+    {
+      "epoch": 0.3074909263997144,
+      "grad_norm": 0.421202050029993,
+      "learning_rate": 9.604976803448596e-06,
+      "loss": 0.3321,
+      "step": 1292
+    },
+    {
+      "epoch": 0.3077289224727792,
+      "grad_norm": 0.4177390717152937,
+      "learning_rate": 9.604225521647343e-06,
+      "loss": 0.4087,
+      "step": 1293
+    },
+    {
+      "epoch": 0.307966918545844,
+      "grad_norm": 0.38933105632433157,
+      "learning_rate": 9.603473555544277e-06,
+      "loss": 0.4267,
+      "step": 1294
+    },
+    {
+      "epoch": 0.3082049146189088,
+      "grad_norm": 0.4249818846999146,
+      "learning_rate": 9.602720905251153e-06,
+      "loss": 0.3674,
+      "step": 1295
+    },
+    {
+      "epoch": 0.3084429106919736,
+      "grad_norm": 0.3917934403028404,
+      "learning_rate": 9.601967570879837e-06,
+      "loss": 0.3504,
+      "step": 1296
+    },
+    {
+      "epoch": 0.3086809067650384,
+      "grad_norm": 0.4115539403873786,
+      "learning_rate": 9.601213552542295e-06,
+      "loss": 0.4318,
+      "step": 1297
+    },
+    {
+      "epoch": 0.3089189028381032,
+      "grad_norm": 0.40804809857958424,
+      "learning_rate": 9.600458850350588e-06,
+      "loss": 0.3798,
+      "step": 1298
+    },
+    {
+      "epoch": 0.309156898911168,
+      "grad_norm": 0.4101236081509535,
+      "learning_rate": 9.599703464416888e-06,
+      "loss": 0.322,
+      "step": 1299
+    },
+    {
+      "epoch": 0.3093948949842328,
+      "grad_norm": 0.40151490793067635,
+      "learning_rate": 9.598947394853459e-06,
+      "loss": 0.3762,
+      "step": 1300
+    },
+    {
+      "epoch": 0.3096328910572976,
+      "grad_norm": 0.4266339246416687,
+      "learning_rate": 9.598190641772678e-06,
+      "loss": 0.4228,
+      "step": 1301
+    },
+    {
+      "epoch": 0.30987088713036237,
+      "grad_norm": 0.43973887455955873,
+      "learning_rate": 9.597433205287013e-06,
+      "loss": 0.3686,
+      "step": 1302
+    },
+    {
+      "epoch": 0.31010888320342717,
+      "grad_norm": 0.3914498413085817,
+      "learning_rate": 9.596675085509037e-06,
+      "loss": 0.3618,
+      "step": 1303
+    },
+    {
+      "epoch": 0.31034687927649196,
+      "grad_norm": 0.3984178847846324,
+      "learning_rate": 9.595916282551429e-06,
+      "loss": 0.4209,
+      "step": 1304
+    },
+    {
+      "epoch": 0.31058487534955675,
+      "grad_norm": 0.39984926069708504,
+      "learning_rate": 9.595156796526963e-06,
+      "loss": 0.4215,
+      "step": 1305
+    },
+    {
+      "epoch": 0.31082287142262155,
+      "grad_norm": 0.4070793173356604,
+      "learning_rate": 9.59439662754852e-06,
+      "loss": 0.3554,
+      "step": 1306
+    },
+    {
+      "epoch": 0.31106086749568634,
+      "grad_norm": 0.3874699043146027,
+      "learning_rate": 9.593635775729075e-06,
+      "loss": 0.3398,
+      "step": 1307
+    },
+    {
+      "epoch": 0.31129886356875114,
+      "grad_norm": 0.39786057856428775,
+      "learning_rate": 9.592874241181715e-06,
+      "loss": 0.4606,
+      "step": 1308
+    },
+    {
+      "epoch": 0.31153685964181593,
+      "grad_norm": 0.4280746953470422,
+      "learning_rate": 9.59211202401962e-06,
+      "loss": 0.3604,
+      "step": 1309
+    },
+    {
+      "epoch": 0.3117748557148807,
+      "grad_norm": 0.4021864469571394,
+      "learning_rate": 9.591349124356075e-06,
+      "loss": 0.3332,
+      "step": 1310
+    },
+    {
+      "epoch": 0.3120128517879455,
+      "grad_norm": 0.39554183184913133,
+      "learning_rate": 9.590585542304466e-06,
+      "loss": 0.3999,
+      "step": 1311
+    },
+    {
+      "epoch": 0.3122508478610103,
+      "grad_norm": 0.4424536440928366,
+      "learning_rate": 9.58982127797828e-06,
+      "loss": 0.4063,
+      "step": 1312
+    },
+    {
+      "epoch": 0.3124888439340751,
+      "grad_norm": 0.45440939559298216,
+      "learning_rate": 9.589056331491103e-06,
+      "loss": 0.3491,
+      "step": 1313
+    },
+    {
+      "epoch": 0.3127268400071399,
+      "grad_norm": 0.39330929295389805,
+      "learning_rate": 9.58829070295663e-06,
+      "loss": 0.3901,
+      "step": 1314
+    },
+    {
+      "epoch": 0.3129648360802047,
+      "grad_norm": 0.38833530931998134,
+      "learning_rate": 9.587524392488647e-06,
+      "loss": 0.4466,
+      "step": 1315
+    },
+    {
+      "epoch": 0.3132028321532695,
+      "grad_norm": 0.44664852368249736,
+      "learning_rate": 9.586757400201052e-06,
+      "loss": 0.3538,
+      "step": 1316
+    },
+    {
+      "epoch": 0.3134408282263343,
+      "grad_norm": 0.40740414428696925,
+      "learning_rate": 9.585989726207837e-06,
+      "loss": 0.3446,
+      "step": 1317
+    },
+    {
+      "epoch": 0.3136788242993991,
+      "grad_norm": 0.38148457875580716,
+      "learning_rate": 9.585221370623095e-06,
+      "loss": 0.3682,
+      "step": 1318
+    },
+    {
+      "epoch": 0.3139168203724639,
+      "grad_norm": 0.43478324832719745,
+      "learning_rate": 9.584452333561024e-06,
+      "loss": 0.424,
+      "step": 1319
+    },
+    {
+      "epoch": 0.3141548164455287,
+      "grad_norm": 0.4000748087214559,
+      "learning_rate": 9.583682615135923e-06,
+      "loss": 0.3392,
+      "step": 1320
+    },
+    {
+      "epoch": 0.31439281251859347,
+      "grad_norm": 0.3835145914512324,
+      "learning_rate": 9.58291221546219e-06,
+      "loss": 0.3482,
+      "step": 1321
+    },
+    {
+      "epoch": 0.31463080859165826,
+      "grad_norm": 0.45176179860066545,
+      "learning_rate": 9.582141134654327e-06,
+      "loss": 0.4232,
+      "step": 1322
+    },
+    {
+      "epoch": 0.31486880466472306,
+      "grad_norm": 0.4268907832728068,
+      "learning_rate": 9.581369372826933e-06,
+      "loss": 0.4082,
+      "step": 1323
+    },
+    {
+      "epoch": 0.31510680073778785,
+      "grad_norm": 0.41052383905112105,
+      "learning_rate": 9.580596930094716e-06,
+      "loss": 0.3438,
+      "step": 1324
+    },
+    {
+      "epoch": 0.31534479681085265,
+      "grad_norm": 0.37945978663945285,
+      "learning_rate": 9.579823806572474e-06,
+      "loss": 0.4025,
+      "step": 1325
+    },
+    {
+      "epoch": 0.31558279288391744,
+      "grad_norm": 0.46441757305264597,
+      "learning_rate": 9.579050002375115e-06,
+      "loss": 0.4172,
+      "step": 1326
+    },
+    {
+      "epoch": 0.31582078895698223,
+      "grad_norm": 0.4393276338549456,
+      "learning_rate": 9.578275517617646e-06,
+      "loss": 0.3395,
+      "step": 1327
+    },
+    {
+      "epoch": 0.31605878503004703,
+      "grad_norm": 0.41000616260473466,
+      "learning_rate": 9.577500352415174e-06,
+      "loss": 0.3254,
+      "step": 1328
+    },
+    {
+      "epoch": 0.3162967811031118,
+      "grad_norm": 0.420002315041422,
+      "learning_rate": 9.576724506882908e-06,
+      "loss": 0.4023,
+      "step": 1329
+    },
+    {
+      "epoch": 0.3165347771761766,
+      "grad_norm": 0.4236036611226363,
+      "learning_rate": 9.575947981136158e-06,
+      "loss": 0.367,
+      "step": 1330
+    },
+    {
+      "epoch": 0.3167727732492414,
+      "grad_norm": 0.4483840411549537,
+      "learning_rate": 9.575170775290333e-06,
+      "loss": 0.366,
+      "step": 1331
+    },
+    {
+      "epoch": 0.3170107693223062,
+      "grad_norm": 0.4313647137240431,
+      "learning_rate": 9.574392889460947e-06,
+      "loss": 0.3813,
+      "step": 1332
+    },
+    {
+      "epoch": 0.317248765395371,
+      "grad_norm": 0.38650979945066577,
+      "learning_rate": 9.573614323763613e-06,
+      "loss": 0.4281,
+      "step": 1333
+    },
+    {
+      "epoch": 0.3174867614684358,
+      "grad_norm": 0.41737500297562424,
+      "learning_rate": 9.572835078314044e-06,
+      "loss": 0.3513,
+      "step": 1334
+    },
+    {
+      "epoch": 0.3177247575415006,
+      "grad_norm": 0.4348449187581223,
+      "learning_rate": 9.572055153228056e-06,
+      "loss": 0.3564,
+      "step": 1335
+    },
+    {
+      "epoch": 0.3179627536145654,
+      "grad_norm": 0.39019537145273137,
+      "learning_rate": 9.571274548621566e-06,
+      "loss": 0.3876,
+      "step": 1336
+    },
+    {
+      "epoch": 0.3182007496876302,
+      "grad_norm": 0.4246028933708349,
+      "learning_rate": 9.570493264610589e-06,
+      "loss": 0.397,
+      "step": 1337
+    },
+    {
+      "epoch": 0.318438745760695,
+      "grad_norm": 0.39930253543441263,
+      "learning_rate": 9.569711301311247e-06,
+      "loss": 0.3201,
+      "step": 1338
+    },
+    {
+      "epoch": 0.31867674183375977,
+      "grad_norm": 0.45407334734341276,
+      "learning_rate": 9.568928658839754e-06,
+      "loss": 0.3434,
+      "step": 1339
+    },
+    {
+      "epoch": 0.31891473790682456,
+      "grad_norm": 0.39643025758667066,
+      "learning_rate": 9.568145337312432e-06,
+      "loss": 0.4361,
+      "step": 1340
+    },
+    {
+      "epoch": 0.31915273397988936,
+      "grad_norm": 0.421815069567987,
+      "learning_rate": 9.567361336845704e-06,
+      "loss": 0.3701,
+      "step": 1341
+    },
+    {
+      "epoch": 0.31939073005295415,
+      "grad_norm": 0.38833337925881534,
+      "learning_rate": 9.566576657556089e-06,
+      "loss": 0.3495,
+      "step": 1342
+    },
+    {
+      "epoch": 0.31962872612601895,
+      "grad_norm": 0.4180768293262518,
+      "learning_rate": 9.565791299560211e-06,
+      "loss": 0.397,
+      "step": 1343
+    },
+    {
+      "epoch": 0.31986672219908374,
+      "grad_norm": 0.4164700070432731,
+      "learning_rate": 9.565005262974795e-06,
+      "loss": 0.4271,
+      "step": 1344
+    },
+    {
+      "epoch": 0.32010471827214854,
+      "grad_norm": 0.4468347817517033,
+      "learning_rate": 9.564218547916664e-06,
+      "loss": 0.3495,
+      "step": 1345
+    },
+    {
+      "epoch": 0.32034271434521333,
+      "grad_norm": 0.40663558186072163,
+      "learning_rate": 9.563431154502742e-06,
+      "loss": 0.3238,
+      "step": 1346
+    },
+    {
+      "epoch": 0.3205807104182781,
+      "grad_norm": 0.3817636234384891,
+      "learning_rate": 9.562643082850058e-06,
+      "loss": 0.4335,
+      "step": 1347
+    },
+    {
+      "epoch": 0.3208187064913429,
+      "grad_norm": 0.4623172946195981,
+      "learning_rate": 9.561854333075737e-06,
+      "loss": 0.3936,
+      "step": 1348
+    },
+    {
+      "epoch": 0.3210567025644077,
+      "grad_norm": 0.41347372584642295,
+      "learning_rate": 9.561064905297007e-06,
+      "loss": 0.3192,
+      "step": 1349
+    },
+    {
+      "epoch": 0.3212946986374725,
+      "grad_norm": 0.45698235146730654,
+      "learning_rate": 9.560274799631196e-06,
+      "loss": 0.4274,
+      "step": 1350
+    },
+    {
+      "epoch": 0.3215326947105373,
+      "grad_norm": 0.4125490994148022,
+      "learning_rate": 9.559484016195734e-06,
+      "loss": 0.4421,
+      "step": 1351
+    },
+    {
+      "epoch": 0.3217706907836021,
+      "grad_norm": 0.3839103030987149,
+      "learning_rate": 9.558692555108153e-06,
+      "loss": 0.3393,
+      "step": 1352
+    },
+    {
+      "epoch": 0.3220086868566669,
+      "grad_norm": 0.4091260775678065,
+      "learning_rate": 9.557900416486082e-06,
+      "loss": 0.3491,
+      "step": 1353
+    },
+    {
+      "epoch": 0.3222466829297317,
+      "grad_norm": 0.3807018578165834,
+      "learning_rate": 9.55710760044725e-06,
+      "loss": 0.4185,
+      "step": 1354
+    },
+    {
+      "epoch": 0.3224846790027965,
+      "grad_norm": 0.40209441835271953,
+      "learning_rate": 9.556314107109492e-06,
+      "loss": 0.4234,
+      "step": 1355
+    },
+    {
+      "epoch": 0.3227226750758612,
+      "grad_norm": 0.40937014075451966,
+      "learning_rate": 9.555519936590739e-06,
+      "loss": 0.3316,
+      "step": 1356
+    },
+    {
+      "epoch": 0.322960671148926,
+      "grad_norm": 0.3884304266780269,
+      "learning_rate": 9.554725089009028e-06,
+      "loss": 0.4138,
+      "step": 1357
+    },
+    {
+      "epoch": 0.3231986672219908,
+      "grad_norm": 0.44485363699581787,
+      "learning_rate": 9.553929564482486e-06,
+      "loss": 0.4333,
+      "step": 1358
+    },
+    {
+      "epoch": 0.3234366632950556,
+      "grad_norm": 0.3994706589980082,
+      "learning_rate": 9.553133363129354e-06,
+      "loss": 0.3354,
+      "step": 1359
+    },
+    {
+      "epoch": 0.3236746593681204,
+      "grad_norm": 0.38575800184976405,
+      "learning_rate": 9.552336485067966e-06,
+      "loss": 0.3318,
+      "step": 1360
+    },
+    {
+      "epoch": 0.3239126554411852,
+      "grad_norm": 0.38907588895075834,
+      "learning_rate": 9.551538930416757e-06,
+      "loss": 0.3931,
+      "step": 1361
+    },
+    {
+      "epoch": 0.32415065151425,
+      "grad_norm": 0.4005989165863969,
+      "learning_rate": 9.550740699294263e-06,
+      "loss": 0.4475,
+      "step": 1362
+    },
+    {
+      "epoch": 0.3243886475873148,
+      "grad_norm": 0.3978604964741817,
+      "learning_rate": 9.54994179181912e-06,
+      "loss": 0.3326,
+      "step": 1363
+    },
+    {
+      "epoch": 0.3246266436603796,
+      "grad_norm": 0.41103874966576015,
+      "learning_rate": 9.549142208110069e-06,
+      "loss": 0.3415,
+      "step": 1364
+    },
+    {
+      "epoch": 0.32486463973344437,
+      "grad_norm": 0.407764483418257,
+      "learning_rate": 9.548341948285945e-06,
+      "loss": 0.4316,
+      "step": 1365
+    },
+    {
+      "epoch": 0.32510263580650917,
+      "grad_norm": 0.41035785125072904,
+      "learning_rate": 9.547541012465684e-06,
+      "loss": 0.3737,
+      "step": 1366
+    },
+    {
+      "epoch": 0.32534063187957396,
+      "grad_norm": 0.43499657258118146,
+      "learning_rate": 9.54673940076833e-06,
+      "loss": 0.3175,
+      "step": 1367
+    },
+    {
+      "epoch": 0.32557862795263875,
+      "grad_norm": 0.4004695765130625,
+      "learning_rate": 9.545937113313019e-06,
+      "loss": 0.3881,
+      "step": 1368
+    },
+    {
+      "epoch": 0.32581662402570355,
+      "grad_norm": 0.4020663184428851,
+      "learning_rate": 9.545134150218993e-06,
+      "loss": 0.4156,
+      "step": 1369
+    },
+    {
+      "epoch": 0.32605462009876834,
+      "grad_norm": 0.40642741251741493,
+      "learning_rate": 9.544330511605591e-06,
+      "loss": 0.3247,
+      "step": 1370
+    },
+    {
+      "epoch": 0.32629261617183314,
+      "grad_norm": 0.3641358027632288,
+      "learning_rate": 9.543526197592255e-06,
+      "loss": 0.345,
+      "step": 1371
+    },
+    {
+      "epoch": 0.32653061224489793,
+      "grad_norm": 0.3989061307234226,
+      "learning_rate": 9.542721208298522e-06,
+      "loss": 0.4211,
+      "step": 1372
+    },
+    {
+      "epoch": 0.3267686083179627,
+      "grad_norm": 0.38215478134894165,
+      "learning_rate": 9.541915543844036e-06,
+      "loss": 0.3929,
+      "step": 1373
+    },
+    {
+      "epoch": 0.3270066043910275,
+      "grad_norm": 0.39691981690953293,
+      "learning_rate": 9.54110920434854e-06,
+      "loss": 0.3495,
+      "step": 1374
+    },
+    {
+      "epoch": 0.3272446004640923,
+      "grad_norm": 0.38215194964724575,
+      "learning_rate": 9.540302189931874e-06,
+      "loss": 0.385,
+      "step": 1375
+    },
+    {
+      "epoch": 0.3274825965371571,
+      "grad_norm": 0.3606406479597386,
+      "learning_rate": 9.539494500713982e-06,
+      "loss": 0.4526,
+      "step": 1376
+    },
+    {
+      "epoch": 0.3277205926102219,
+      "grad_norm": 0.40105227105638763,
+      "learning_rate": 9.538686136814905e-06,
+      "loss": 0.3464,
+      "step": 1377
+    },
+    {
+      "epoch": 0.3279585886832867,
+      "grad_norm": 0.41070873761480936,
+      "learning_rate": 9.537877098354787e-06,
+      "loss": 0.3559,
+      "step": 1378
+    },
+    {
+      "epoch": 0.3281965847563515,
+      "grad_norm": 0.40931042414878005,
+      "learning_rate": 9.53706738545387e-06,
+      "loss": 0.3865,
+      "step": 1379
+    },
+    {
+      "epoch": 0.3284345808294163,
+      "grad_norm": 0.4254957758069491,
+      "learning_rate": 9.536256998232496e-06,
+      "loss": 0.3928,
+      "step": 1380
+    },
+    {
+      "epoch": 0.3286725769024811,
+      "grad_norm": 0.4198060992786074,
+      "learning_rate": 9.535445936811111e-06,
+      "loss": 0.3368,
+      "step": 1381
+    },
+    {
+      "epoch": 0.3289105729755459,
+      "grad_norm": 0.39019263851197356,
+      "learning_rate": 9.53463420131026e-06,
+      "loss": 0.3788,
+      "step": 1382
+    },
+    {
+      "epoch": 0.3291485690486107,
+      "grad_norm": 0.3881968657734908,
+      "learning_rate": 9.533821791850585e-06,
+      "loss": 0.4478,
+      "step": 1383
+    },
+    {
+      "epoch": 0.32938656512167547,
+      "grad_norm": 0.41104330210033657,
+      "learning_rate": 9.533008708552829e-06,
+      "loss": 0.3468,
+      "step": 1384
+    },
+    {
+      "epoch": 0.32962456119474026,
+      "grad_norm": 0.4668376033750086,
+      "learning_rate": 9.532194951537838e-06,
+      "loss": 0.339,
+      "step": 1385
+    },
+    {
+      "epoch": 0.32986255726780506,
+      "grad_norm": 0.41125619080230763,
+      "learning_rate": 9.531380520926559e-06,
+      "loss": 0.3966,
+      "step": 1386
+    },
+    {
+      "epoch": 0.33010055334086985,
+      "grad_norm": 0.4471468164842655,
+      "learning_rate": 9.53056541684003e-06,
+      "loss": 0.4227,
+      "step": 1387
+    },
+    {
+      "epoch": 0.33033854941393465,
+      "grad_norm": 0.3796087714093602,
+      "learning_rate": 9.5297496393994e-06,
+      "loss": 0.3417,
+      "step": 1388
+    },
+    {
+      "epoch": 0.33057654548699944,
+      "grad_norm": 0.4097059136074371,
+      "learning_rate": 9.528933188725913e-06,
+      "loss": 0.3784,
+      "step": 1389
+    },
+    {
+      "epoch": 0.33081454156006423,
+      "grad_norm": 0.41994702732943573,
+      "learning_rate": 9.528116064940915e-06,
+      "loss": 0.4296,
+      "step": 1390
+    },
+    {
+      "epoch": 0.33105253763312903,
+      "grad_norm": 0.43876123590734234,
+      "learning_rate": 9.52729826816585e-06,
+      "loss": 0.3459,
+      "step": 1391
+    },
+    {
+      "epoch": 0.3312905337061938,
+      "grad_norm": 0.44759563573787553,
+      "learning_rate": 9.526479798522261e-06,
+      "loss": 0.3338,
+      "step": 1392
+    },
+    {
+      "epoch": 0.3315285297792586,
+      "grad_norm": 0.39997856436860535,
+      "learning_rate": 9.525660656131794e-06,
+      "loss": 0.3782,
+      "step": 1393
+    },
+    {
+      "epoch": 0.3317665258523234,
+      "grad_norm": 0.4496886369666572,
+      "learning_rate": 9.524840841116194e-06,
+      "loss": 0.4685,
+      "step": 1394
+    },
+    {
+      "epoch": 0.3320045219253882,
+      "grad_norm": 0.4074594148653707,
+      "learning_rate": 9.524020353597306e-06,
+      "loss": 0.3464,
+      "step": 1395
+    },
+    {
+      "epoch": 0.332242517998453,
+      "grad_norm": 0.4240780629108365,
+      "learning_rate": 9.523199193697076e-06,
+      "loss": 0.3664,
+      "step": 1396
+    },
+    {
+      "epoch": 0.3324805140715178,
+      "grad_norm": 0.3724547733959635,
+      "learning_rate": 9.522377361537546e-06,
+      "loss": 0.4263,
+      "step": 1397
+    },
+    {
+      "epoch": 0.3327185101445826,
+      "grad_norm": 0.3848768663283912,
+      "learning_rate": 9.521554857240863e-06,
+      "loss": 0.3997,
+      "step": 1398
+    },
+    {
+      "epoch": 0.3329565062176474,
+      "grad_norm": 0.3747787306372805,
+      "learning_rate": 9.520731680929268e-06,
+      "loss": 0.3324,
+      "step": 1399
+    },
+    {
+      "epoch": 0.3331945022907122,
+      "grad_norm": 0.40168499777226635,
+      "learning_rate": 9.51990783272511e-06,
+      "loss": 0.3988,
+      "step": 1400
+    },
+    {
+      "epoch": 0.333432498363777,
+      "grad_norm": 0.37263893468315484,
+      "learning_rate": 9.519083312750829e-06,
+      "loss": 0.433,
+      "step": 1401
+    },
+    {
+      "epoch": 0.33367049443684177,
+      "grad_norm": 0.39450747544417814,
+      "learning_rate": 9.518258121128971e-06,
+      "loss": 0.3332,
+      "step": 1402
+    },
+    {
+      "epoch": 0.33390849050990656,
+      "grad_norm": 0.4159109527655681,
+      "learning_rate": 9.517432257982182e-06,
+      "loss": 0.3519,
+      "step": 1403
+    },
+    {
+      "epoch": 0.33414648658297136,
+      "grad_norm": 0.3758964436379061,
+      "learning_rate": 9.516605723433202e-06,
+      "loss": 0.404,
+      "step": 1404
+    },
+    {
+      "epoch": 0.33438448265603615,
+      "grad_norm": 0.4167452335375169,
+      "learning_rate": 9.515778517604876e-06,
+      "loss": 0.3771,
+      "step": 1405
+    },
+    {
+      "epoch": 0.33462247872910095,
+      "grad_norm": 0.4231708121211018,
+      "learning_rate": 9.514950640620148e-06,
+      "loss": 0.3153,
+      "step": 1406
+    },
+    {
+      "epoch": 0.33486047480216574,
+      "grad_norm": 0.4358774633836512,
+      "learning_rate": 9.514122092602061e-06,
+      "loss": 0.3527,
+      "step": 1407
+    },
+    {
+      "epoch": 0.33509847087523054,
+      "grad_norm": 0.3773178127174748,
+      "learning_rate": 9.513292873673757e-06,
+      "loss": 0.4515,
+      "step": 1408
+    },
+    {
+      "epoch": 0.33533646694829533,
+      "grad_norm": 0.39844835991429833,
+      "learning_rate": 9.512462983958476e-06,
+      "loss": 0.3841,
+      "step": 1409
+    },
+    {
+      "epoch": 0.3355744630213601,
+      "grad_norm": 0.41246380594274273,
+      "learning_rate": 9.511632423579564e-06,
+      "loss": 0.312,
+      "step": 1410
+    },
+    {
+      "epoch": 0.3358124590944249,
+      "grad_norm": 0.39771084162538206,
+      "learning_rate": 9.510801192660463e-06,
+      "loss": 0.4066,
+      "step": 1411
+    },
+    {
+      "epoch": 0.3360504551674897,
+      "grad_norm": 0.40735314602043377,
+      "learning_rate": 9.509969291324711e-06,
+      "loss": 0.4026,
+      "step": 1412
+    },
+    {
+      "epoch": 0.3362884512405545,
+      "grad_norm": 0.4306392515796982,
+      "learning_rate": 9.509136719695952e-06,
+      "loss": 0.3391,
+      "step": 1413
+    },
+    {
+      "epoch": 0.3365264473136193,
+      "grad_norm": 0.38787745703798865,
+      "learning_rate": 9.508303477897925e-06,
+      "loss": 0.3623,
+      "step": 1414
+    },
+    {
+      "epoch": 0.3367644433866841,
+      "grad_norm": 0.38221096427891993,
+      "learning_rate": 9.507469566054472e-06,
+      "loss": 0.4347,
+      "step": 1415
+    },
+    {
+      "epoch": 0.3370024394597489,
+      "grad_norm": 0.39794787975787377,
+      "learning_rate": 9.506634984289532e-06,
+      "loss": 0.3685,
+      "step": 1416
+    },
+    {
+      "epoch": 0.3372404355328137,
+      "grad_norm": 0.42203140787297544,
+      "learning_rate": 9.505799732727144e-06,
+      "loss": 0.3255,
+      "step": 1417
+    },
+    {
+      "epoch": 0.3374784316058785,
+      "grad_norm": 0.4130150729651241,
+      "learning_rate": 9.504963811491448e-06,
+      "loss": 0.3852,
+      "step": 1418
+    },
+    {
+      "epoch": 0.3377164276789433,
+      "grad_norm": 0.4035329355042889,
+      "learning_rate": 9.50412722070668e-06,
+      "loss": 0.4231,
+      "step": 1419
+    },
+    {
+      "epoch": 0.33795442375200807,
+      "grad_norm": 0.42264021258045936,
+      "learning_rate": 9.503289960497184e-06,
+      "loss": 0.3096,
+      "step": 1420
+    },
+    {
+      "epoch": 0.33819241982507287,
+      "grad_norm": 0.41033520433424203,
+      "learning_rate": 9.502452030987392e-06,
+      "loss": 0.3462,
+      "step": 1421
+    },
+    {
+      "epoch": 0.33843041589813766,
+      "grad_norm": 0.3734652547743581,
+      "learning_rate": 9.501613432301843e-06,
+      "loss": 0.4026,
+      "step": 1422
+    },
+    {
+      "epoch": 0.33866841197120245,
+      "grad_norm": 0.44488771827168927,
+      "learning_rate": 9.500774164565172e-06,
+      "loss": 0.403,
+      "step": 1423
+    },
+    {
+      "epoch": 0.33890640804426725,
+      "grad_norm": 0.41540692997233103,
+      "learning_rate": 9.499934227902119e-06,
+      "loss": 0.3038,
+      "step": 1424
+    },
+    {
+      "epoch": 0.33914440411733204,
+      "grad_norm": 0.3759433684152908,
+      "learning_rate": 9.499093622437516e-06,
+      "loss": 0.3459,
+      "step": 1425
+    },
+    {
+      "epoch": 0.33938240019039684,
+      "grad_norm": 0.3790112278132182,
+      "learning_rate": 9.498252348296298e-06,
+      "loss": 0.4454,
+      "step": 1426
+    },
+    {
+      "epoch": 0.33962039626346163,
+      "grad_norm": 0.4478868649304047,
+      "learning_rate": 9.4974104056035e-06,
+      "loss": 0.3263,
+      "step": 1427
+    },
+    {
+      "epoch": 0.3398583923365264,
+      "grad_norm": 0.3687692738652499,
+      "learning_rate": 9.496567794484254e-06,
+      "loss": 0.3447,
+      "step": 1428
+    },
+    {
+      "epoch": 0.3400963884095912,
+      "grad_norm": 0.3815394491514148,
+      "learning_rate": 9.495724515063795e-06,
+      "loss": 0.411,
+      "step": 1429
+    },
+    {
+      "epoch": 0.340334384482656,
+      "grad_norm": 0.41607171762494266,
+      "learning_rate": 9.494880567467454e-06,
+      "loss": 0.4161,
+      "step": 1430
+    },
+    {
+      "epoch": 0.3405723805557208,
+      "grad_norm": 0.39572328438968524,
+      "learning_rate": 9.494035951820662e-06,
+      "loss": 0.332,
+      "step": 1431
+    },
+    {
+      "epoch": 0.3408103766287856,
+      "grad_norm": 0.37481472394853776,
+      "learning_rate": 9.493190668248951e-06,
+      "loss": 0.3546,
+      "step": 1432
+    },
+    {
+      "epoch": 0.3410483727018504,
+      "grad_norm": 0.40603528903956776,
+      "learning_rate": 9.49234471687795e-06,
+      "loss": 0.4331,
+      "step": 1433
+    },
+    {
+      "epoch": 0.3412863687749152,
+      "grad_norm": 0.3562649954495181,
+      "learning_rate": 9.491498097833391e-06,
+      "loss": 0.3442,
+      "step": 1434
+    },
+    {
+      "epoch": 0.34152436484798,
+      "grad_norm": 0.4633907342034106,
+      "learning_rate": 9.490650811241098e-06,
+      "loss": 0.3301,
+      "step": 1435
+    },
+    {
+      "epoch": 0.3417623609210448,
+      "grad_norm": 0.4196029970855259,
+      "learning_rate": 9.489802857227001e-06,
+      "loss": 0.4032,
+      "step": 1436
+    },
+    {
+      "epoch": 0.3420003569941096,
+      "grad_norm": 0.4036864305636982,
+      "learning_rate": 9.488954235917129e-06,
+      "loss": 0.4371,
+      "step": 1437
+    },
+    {
+      "epoch": 0.3422383530671744,
+      "grad_norm": 0.4424631609113763,
+      "learning_rate": 9.488104947437606e-06,
+      "loss": 0.3401,
+      "step": 1438
+    },
+    {
+      "epoch": 0.34247634914023917,
+      "grad_norm": 0.3868950566216425,
+      "learning_rate": 9.487254991914655e-06,
+      "loss": 0.3546,
+      "step": 1439
+    },
+    {
+      "epoch": 0.34271434521330396,
+      "grad_norm": 0.3939081607602161,
+      "learning_rate": 9.486404369474605e-06,
+      "loss": 0.4171,
+      "step": 1440
+    },
+    {
+      "epoch": 0.34295234128636876,
+      "grad_norm": 0.42524765774393825,
+      "learning_rate": 9.485553080243877e-06,
+      "loss": 0.4112,
+      "step": 1441
+    },
+    {
+      "epoch": 0.34319033735943355,
+      "grad_norm": 0.40156763127279305,
+      "learning_rate": 9.484701124348994e-06,
+      "loss": 0.3195,
+      "step": 1442
+    },
+    {
+      "epoch": 0.34342833343249834,
+      "grad_norm": 0.3769325847215234,
+      "learning_rate": 9.483848501916578e-06,
+      "loss": 0.3893,
+      "step": 1443
+    },
+    {
+      "epoch": 0.34366632950556314,
+      "grad_norm": 0.38885175403043437,
+      "learning_rate": 9.482995213073349e-06,
+      "loss": 0.4137,
+      "step": 1444
+    },
+    {
+      "epoch": 0.34390432557862793,
+      "grad_norm": 0.41829331584548823,
+      "learning_rate": 9.482141257946128e-06,
+      "loss": 0.3256,
+      "step": 1445
+    },
+    {
+      "epoch": 0.34414232165169273,
+      "grad_norm": 0.39698859193541414,
+      "learning_rate": 9.481286636661832e-06,
+      "loss": 0.3461,
+      "step": 1446
+    },
+    {
+      "epoch": 0.3443803177247575,
+      "grad_norm": 0.40947635708434343,
+      "learning_rate": 9.480431349347482e-06,
+      "loss": 0.4253,
+      "step": 1447
+    },
+    {
+      "epoch": 0.3446183137978223,
+      "grad_norm": 0.4008658220372033,
+      "learning_rate": 9.479575396130192e-06,
+      "loss": 0.3822,
+      "step": 1448
+    },
+    {
+      "epoch": 0.3448563098708871,
+      "grad_norm": 0.38802614165640026,
+      "learning_rate": 9.478718777137178e-06,
+      "loss": 0.3252,
+      "step": 1449
+    },
+    {
+      "epoch": 0.3450943059439519,
+      "grad_norm": 0.40732799270557263,
+      "learning_rate": 9.477861492495757e-06,
+      "loss": 0.3823,
+      "step": 1450
+    },
+    {
+      "epoch": 0.3453323020170167,
+      "grad_norm": 0.41644066379485684,
+      "learning_rate": 9.47700354233334e-06,
+      "loss": 0.4282,
+      "step": 1451
+    },
+    {
+      "epoch": 0.3455702980900815,
+      "grad_norm": 0.4224474904646012,
+      "learning_rate": 9.476144926777441e-06,
+      "loss": 0.3412,
+      "step": 1452
+    },
+    {
+      "epoch": 0.3458082941631463,
+      "grad_norm": 0.3909474423211058,
+      "learning_rate": 9.475285645955672e-06,
+      "loss": 0.3453,
+      "step": 1453
+    },
+    {
+      "epoch": 0.3460462902362111,
+      "grad_norm": 0.3842803363987257,
+      "learning_rate": 9.474425699995741e-06,
+      "loss": 0.4122,
+      "step": 1454
+    },
+    {
+      "epoch": 0.3462842863092759,
+      "grad_norm": 0.3862238147346478,
+      "learning_rate": 9.473565089025463e-06,
+      "loss": 0.3803,
+      "step": 1455
+    },
+    {
+      "epoch": 0.3465222823823407,
+      "grad_norm": 0.39766464008162167,
+      "learning_rate": 9.472703813172739e-06,
+      "loss": 0.3174,
+      "step": 1456
+    },
+    {
+      "epoch": 0.34676027845540547,
+      "grad_norm": 0.37570231724988296,
+      "learning_rate": 9.47184187256558e-06,
+      "loss": 0.3791,
+      "step": 1457
+    },
+    {
+      "epoch": 0.34699827452847026,
+      "grad_norm": 0.4142156067032061,
+      "learning_rate": 9.47097926733209e-06,
+      "loss": 0.4024,
+      "step": 1458
+    },
+    {
+      "epoch": 0.34723627060153506,
+      "grad_norm": 0.398792512942407,
+      "learning_rate": 9.470115997600474e-06,
+      "loss": 0.3421,
+      "step": 1459
+    },
+    {
+      "epoch": 0.34747426667459985,
+      "grad_norm": 0.43012148252168575,
+      "learning_rate": 9.469252063499036e-06,
+      "loss": 0.3314,
+      "step": 1460
+    },
+    {
+      "epoch": 0.34771226274766465,
+      "grad_norm": 0.4927776656728705,
+      "learning_rate": 9.468387465156176e-06,
+      "loss": 0.4153,
+      "step": 1461
+    },
+    {
+      "epoch": 0.34795025882072944,
+      "grad_norm": 0.4096929455276192,
+      "learning_rate": 9.467522202700399e-06,
+      "loss": 0.4071,
+      "step": 1462
+    },
+    {
+      "epoch": 0.34818825489379424,
+      "grad_norm": 0.44407738784995165,
+      "learning_rate": 9.4666562762603e-06,
+      "loss": 0.3292,
+      "step": 1463
+    },
+    {
+      "epoch": 0.34842625096685903,
+      "grad_norm": 0.4261438734366603,
+      "learning_rate": 9.465789685964579e-06,
+      "loss": 0.3609,
+      "step": 1464
+    },
+    {
+      "epoch": 0.3486642470399238,
+      "grad_norm": 0.3842203365009242,
+      "learning_rate": 9.464922431942032e-06,
+      "loss": 0.4201,
+      "step": 1465
+    },
+    {
+      "epoch": 0.3489022431129886,
+      "grad_norm": 0.378083104719912,
+      "learning_rate": 9.464054514321554e-06,
+      "loss": 0.3694,
+      "step": 1466
+    },
+    {
+      "epoch": 0.3491402391860534,
+      "grad_norm": 0.4040049074787795,
+      "learning_rate": 9.46318593323214e-06,
+      "loss": 0.3564,
+      "step": 1467
+    },
+    {
+      "epoch": 0.3493782352591182,
+      "grad_norm": 0.425890416302795,
+      "learning_rate": 9.462316688802884e-06,
+      "loss": 0.389,
+      "step": 1468
+    },
+    {
+      "epoch": 0.349616231332183,
+      "grad_norm": 0.3803658953106735,
+      "learning_rate": 9.461446781162974e-06,
+      "loss": 0.4331,
+      "step": 1469
+    },
+    {
+      "epoch": 0.3498542274052478,
+      "grad_norm": 0.4060084138224938,
+      "learning_rate": 9.4605762104417e-06,
+      "loss": 0.3327,
+      "step": 1470
+    },
+    {
+      "epoch": 0.3500922234783126,
+      "grad_norm": 0.640815748060444,
+      "learning_rate": 9.459704976768455e-06,
+      "loss": 0.3434,
+      "step": 1471
+    },
+    {
+      "epoch": 0.3503302195513774,
+      "grad_norm": 0.42837101898379226,
+      "learning_rate": 9.458833080272723e-06,
+      "loss": 0.424,
+      "step": 1472
+    },
+    {
+      "epoch": 0.3505682156244422,
+      "grad_norm": 0.4030658639433663,
+      "learning_rate": 9.457960521084087e-06,
+      "loss": 0.3982,
+      "step": 1473
+    },
+    {
+      "epoch": 0.350806211697507,
+      "grad_norm": 0.3984195226956701,
+      "learning_rate": 9.457087299332232e-06,
+      "loss": 0.3275,
+      "step": 1474
+    },
+    {
+      "epoch": 0.35104420777057177,
+      "grad_norm": 0.39898439381782036,
+      "learning_rate": 9.456213415146943e-06,
+      "loss": 0.3891,
+      "step": 1475
+    },
+    {
+      "epoch": 0.35128220384363656,
+      "grad_norm": 0.39964981876808653,
+      "learning_rate": 9.4553388686581e-06,
+      "loss": 0.4239,
+      "step": 1476
+    },
+    {
+      "epoch": 0.35152019991670136,
+      "grad_norm": 0.41473687287806216,
+      "learning_rate": 9.454463659995678e-06,
+      "loss": 0.3502,
+      "step": 1477
+    },
+    {
+      "epoch": 0.35175819598976615,
+      "grad_norm": 0.40780054376337166,
+      "learning_rate": 9.453587789289762e-06,
+      "loss": 0.3565,
+      "step": 1478
+    },
+    {
+      "epoch": 0.35199619206283095,
+      "grad_norm": 0.4040364947021049,
+      "learning_rate": 9.452711256670521e-06,
+      "loss": 0.4009,
+      "step": 1479
+    },
+    {
+      "epoch": 0.35223418813589574,
+      "grad_norm": 0.43117573330093506,
+      "learning_rate": 9.451834062268234e-06,
+      "loss": 0.3894,
+      "step": 1480
+    },
+    {
+      "epoch": 0.35247218420896054,
+      "grad_norm": 0.4115001277743943,
+      "learning_rate": 9.450956206213272e-06,
+      "loss": 0.3176,
+      "step": 1481
+    },
+    {
+      "epoch": 0.35271018028202533,
+      "grad_norm": 0.3928594010225446,
+      "learning_rate": 9.450077688636107e-06,
+      "loss": 0.3696,
+      "step": 1482
+    },
+    {
+      "epoch": 0.3529481763550901,
+      "grad_norm": 0.3924725511243915,
+      "learning_rate": 9.449198509667307e-06,
+      "loss": 0.429,
+      "step": 1483
+    },
+    {
+      "epoch": 0.3531861724281549,
+      "grad_norm": 0.4026567257458122,
+      "learning_rate": 9.448318669437541e-06,
+      "loss": 0.3303,
+      "step": 1484
+    },
+    {
+      "epoch": 0.3534241685012197,
+      "grad_norm": 0.38358768538776655,
+      "learning_rate": 9.447438168077574e-06,
+      "loss": 0.3477,
+      "step": 1485
+    },
+    {
+      "epoch": 0.3536621645742845,
+      "grad_norm": 0.38097335550299927,
+      "learning_rate": 9.446557005718271e-06,
+      "loss": 0.3577,
+      "step": 1486
+    },
+    {
+      "epoch": 0.3539001606473493,
+      "grad_norm": 0.3625205031776034,
+      "learning_rate": 9.445675182490594e-06,
+      "loss": 0.4126,
+      "step": 1487
+    },
+    {
+      "epoch": 0.3541381567204141,
+      "grad_norm": 0.3598134954454926,
+      "learning_rate": 9.444792698525606e-06,
+      "loss": 0.3341,
+      "step": 1488
+    },
+    {
+      "epoch": 0.3543761527934789,
+      "grad_norm": 0.4479832164810966,
+      "learning_rate": 9.443909553954463e-06,
+      "loss": 0.3451,
+      "step": 1489
+    },
+    {
+      "epoch": 0.3546141488665437,
+      "grad_norm": 0.38803228377964066,
+      "learning_rate": 9.443025748908423e-06,
+      "loss": 0.4234,
+      "step": 1490
+    },
+    {
+      "epoch": 0.3548521449396085,
+      "grad_norm": 0.4066075957863286,
+      "learning_rate": 9.442141283518842e-06,
+      "loss": 0.3549,
+      "step": 1491
+    },
+    {
+      "epoch": 0.3550901410126733,
+      "grad_norm": 0.42403239997363124,
+      "learning_rate": 9.441256157917174e-06,
+      "loss": 0.3344,
+      "step": 1492
+    },
+    {
+      "epoch": 0.35532813708573807,
+      "grad_norm": 0.38290202011308166,
+      "learning_rate": 9.440370372234968e-06,
+      "loss": 0.3861,
+      "step": 1493
+    },
+    {
+      "epoch": 0.35556613315880287,
+      "grad_norm": 0.3890748893970797,
+      "learning_rate": 9.439483926603876e-06,
+      "loss": 0.4325,
+      "step": 1494
+    },
+    {
+      "epoch": 0.35580412923186766,
+      "grad_norm": 0.41264784304893215,
+      "learning_rate": 9.438596821155644e-06,
+      "loss": 0.3185,
+      "step": 1495
+    },
+    {
+      "epoch": 0.35604212530493246,
+      "grad_norm": 0.4150247829602727,
+      "learning_rate": 9.43770905602212e-06,
+      "loss": 0.3659,
+      "step": 1496
+    },
+    {
+      "epoch": 0.35628012137799725,
+      "grad_norm": 0.3925170437737852,
+      "learning_rate": 9.436820631335245e-06,
+      "loss": 0.4141,
+      "step": 1497
+    },
+    {
+      "epoch": 0.35651811745106204,
+      "grad_norm": 0.4286671248087841,
+      "learning_rate": 9.435931547227064e-06,
+      "loss": 0.381,
+      "step": 1498
+    },
+    {
+      "epoch": 0.35675611352412684,
+      "grad_norm": 0.4388357656753384,
+      "learning_rate": 9.435041803829716e-06,
+      "loss": 0.3278,
+      "step": 1499
+    },
+    {
+      "epoch": 0.35699410959719163,
+      "grad_norm": 0.418842687865229,
+      "learning_rate": 9.434151401275436e-06,
+      "loss": 0.3708,
+      "step": 1500
+    },
+    {
+      "epoch": 0.35723210567025643,
+      "grad_norm": 0.4811664724649453,
+      "learning_rate": 9.433260339696564e-06,
+      "loss": 0.4451,
+      "step": 1501
+    },
+    {
+      "epoch": 0.3574701017433212,
+      "grad_norm": 0.4847264758830864,
+      "learning_rate": 9.432368619225532e-06,
+      "loss": 0.3434,
+      "step": 1502
+    },
+    {
+      "epoch": 0.357708097816386,
+      "grad_norm": 0.4157855613581623,
+      "learning_rate": 9.43147623999487e-06,
+      "loss": 0.3456,
+      "step": 1503
+    },
+    {
+      "epoch": 0.3579460938894508,
+      "grad_norm": 0.39518229863812704,
+      "learning_rate": 9.43058320213721e-06,
+      "loss": 0.3947,
+      "step": 1504
+    },
+    {
+      "epoch": 0.3581840899625156,
+      "grad_norm": 0.4171833220322581,
+      "learning_rate": 9.42968950578528e-06,
+      "loss": 0.3934,
+      "step": 1505
+    },
+    {
+      "epoch": 0.3584220860355804,
+      "grad_norm": 0.46935586691478637,
+      "learning_rate": 9.428795151071904e-06,
+      "loss": 0.335,
+      "step": 1506
+    },
+    {
+      "epoch": 0.3586600821086452,
+      "grad_norm": 0.4076327299219336,
+      "learning_rate": 9.427900138130005e-06,
+      "loss": 0.3517,
+      "step": 1507
+    },
+    {
+      "epoch": 0.35889807818171,
+      "grad_norm": 0.39393198965417636,
+      "learning_rate": 9.427004467092604e-06,
+      "loss": 0.4382,
+      "step": 1508
+    },
+    {
+      "epoch": 0.3591360742547748,
+      "grad_norm": 0.4165484912582241,
+      "learning_rate": 9.42610813809282e-06,
+      "loss": 0.3512,
+      "step": 1509
+    },
+    {
+      "epoch": 0.3593740703278396,
+      "grad_norm": 0.45933344516028873,
+      "learning_rate": 9.425211151263871e-06,
+      "loss": 0.3291,
+      "step": 1510
+    },
+    {
+      "epoch": 0.3596120664009044,
+      "grad_norm": 0.3722603137882553,
+      "learning_rate": 9.42431350673907e-06,
+      "loss": 0.3777,
+      "step": 1511
+    },
+    {
+      "epoch": 0.35985006247396917,
+      "grad_norm": 0.4539895853130569,
+      "learning_rate": 9.42341520465183e-06,
+      "loss": 0.4187,
+      "step": 1512
+    },
+    {
+      "epoch": 0.36008805854703396,
+      "grad_norm": 0.4630126425642641,
+      "learning_rate": 9.42251624513566e-06,
+      "loss": 0.3322,
+      "step": 1513
+    },
+    {
+      "epoch": 0.36032605462009876,
+      "grad_norm": 0.4571830447922558,
+      "learning_rate": 9.421616628324168e-06,
+      "loss": 0.376,
+      "step": 1514
+    },
+    {
+      "epoch": 0.36056405069316355,
+      "grad_norm": 0.3805355362272922,
+      "learning_rate": 9.42071635435106e-06,
+      "loss": 0.4204,
+      "step": 1515
+    },
+    {
+      "epoch": 0.36080204676622835,
+      "grad_norm": 0.43191632003281544,
+      "learning_rate": 9.41981542335014e-06,
+      "loss": 0.3545,
+      "step": 1516
+    },
+    {
+      "epoch": 0.36104004283929314,
+      "grad_norm": 0.40851897479644855,
+      "learning_rate": 9.418913835455306e-06,
+      "loss": 0.3339,
+      "step": 1517
+    },
+    {
+      "epoch": 0.36127803891235793,
+      "grad_norm": 0.37903195693910496,
+      "learning_rate": 9.418011590800556e-06,
+      "loss": 0.3913,
+      "step": 1518
+    },
+    {
+      "epoch": 0.36151603498542273,
+      "grad_norm": 0.4520503442337851,
+      "learning_rate": 9.41710868951999e-06,
+      "loss": 0.4169,
+      "step": 1519
+    },
+    {
+      "epoch": 0.3617540310584875,
+      "grad_norm": 0.43239012543775074,
+      "learning_rate": 9.416205131747796e-06,
+      "loss": 0.3376,
+      "step": 1520
+    },
+    {
+      "epoch": 0.3619920271315523,
+      "grad_norm": 0.41380758515497446,
+      "learning_rate": 9.415300917618269e-06,
+      "loss": 0.3672,
+      "step": 1521
+    },
+    {
+      "epoch": 0.3622300232046171,
+      "grad_norm": 0.3843142207624058,
+      "learning_rate": 9.414396047265797e-06,
+      "loss": 0.4091,
+      "step": 1522
+    },
+    {
+      "epoch": 0.3624680192776819,
+      "grad_norm": 0.37767641451983713,
+      "learning_rate": 9.413490520824864e-06,
+      "loss": 0.3759,
+      "step": 1523
+    },
+    {
+      "epoch": 0.3627060153507467,
+      "grad_norm": 0.3684023556846619,
+      "learning_rate": 9.412584338430056e-06,
+      "loss": 0.3418,
+      "step": 1524
+    },
+    {
+      "epoch": 0.3629440114238115,
+      "grad_norm": 0.5087384806908768,
+      "learning_rate": 9.411677500216053e-06,
+      "loss": 0.3908,
+      "step": 1525
+    },
+    {
+      "epoch": 0.3631820074968763,
+      "grad_norm": 0.36310042219702077,
+      "learning_rate": 9.410770006317634e-06,
+      "loss": 0.4402,
+      "step": 1526
+    },
+    {
+      "epoch": 0.3634200035699411,
+      "grad_norm": 0.36444105379216124,
+      "learning_rate": 9.409861856869676e-06,
+      "loss": 0.3427,
+      "step": 1527
+    },
+    {
+      "epoch": 0.3636579996430059,
+      "grad_norm": 0.48371401020099114,
+      "learning_rate": 9.40895305200715e-06,
+      "loss": 0.3294,
+      "step": 1528
+    },
+    {
+      "epoch": 0.3638959957160707,
+      "grad_norm": 0.4123934901314048,
+      "learning_rate": 9.408043591865129e-06,
+      "loss": 0.3941,
+      "step": 1529
+    },
+    {
+      "epoch": 0.36413399178913547,
+      "grad_norm": 0.39143950957309054,
+      "learning_rate": 9.407133476578778e-06,
+      "loss": 0.3836,
+      "step": 1530
+    },
+    {
+      "epoch": 0.36437198786220026,
+      "grad_norm": 0.4513000975255898,
+      "learning_rate": 9.406222706283368e-06,
+      "loss": 0.3188,
+      "step": 1531
+    },
+    {
+      "epoch": 0.36460998393526506,
+      "grad_norm": 0.39125708494077566,
+      "learning_rate": 9.405311281114258e-06,
+      "loss": 0.3845,
+      "step": 1532
+    },
+    {
+      "epoch": 0.36484798000832985,
+      "grad_norm": 0.39160593726214166,
+      "learning_rate": 9.404399201206908e-06,
+      "loss": 0.4172,
+      "step": 1533
+    },
+    {
+      "epoch": 0.36508597608139465,
+      "grad_norm": 0.46690194148815245,
+      "learning_rate": 9.40348646669688e-06,
+      "loss": 0.3384,
+      "step": 1534
+    },
+    {
+      "epoch": 0.36532397215445944,
+      "grad_norm": 0.43376276227001326,
+      "learning_rate": 9.402573077719825e-06,
+      "loss": 0.3107,
+      "step": 1535
+    },
+    {
+      "epoch": 0.36556196822752424,
+      "grad_norm": 0.3748979225551678,
+      "learning_rate": 9.401659034411496e-06,
+      "loss": 0.3881,
+      "step": 1536
+    },
+    {
+      "epoch": 0.36579996430058903,
+      "grad_norm": 0.41357476791350894,
+      "learning_rate": 9.400744336907743e-06,
+      "loss": 0.4555,
+      "step": 1537
+    },
+    {
+      "epoch": 0.3660379603736538,
+      "grad_norm": 0.40289808520289133,
+      "learning_rate": 9.399828985344513e-06,
+      "loss": 0.3289,
+      "step": 1538
+    },
+    {
+      "epoch": 0.3662759564467186,
+      "grad_norm": 0.37195330338030785,
+      "learning_rate": 9.398912979857848e-06,
+      "loss": 0.3662,
+      "step": 1539
+    },
+    {
+      "epoch": 0.3665139525197834,
+      "grad_norm": 0.379044773247951,
+      "learning_rate": 9.39799632058389e-06,
+      "loss": 0.4099,
+      "step": 1540
+    },
+    {
+      "epoch": 0.3667519485928482,
+      "grad_norm": 0.40409817593250036,
+      "learning_rate": 9.397079007658878e-06,
+      "loss": 0.3912,
+      "step": 1541
+    },
+    {
+      "epoch": 0.366989944665913,
+      "grad_norm": 0.38093441918353926,
+      "learning_rate": 9.396161041219147e-06,
+      "loss": 0.3345,
+      "step": 1542
+    },
+    {
+      "epoch": 0.3672279407389778,
+      "grad_norm": 0.4006873027568648,
+      "learning_rate": 9.39524242140113e-06,
+      "loss": 0.386,
+      "step": 1543
+    },
+    {
+      "epoch": 0.3674659368120426,
+      "grad_norm": 0.407045960911348,
+      "learning_rate": 9.394323148341355e-06,
+      "loss": 0.4196,
+      "step": 1544
+    },
+    {
+      "epoch": 0.3677039328851074,
+      "grad_norm": 0.45641130604553176,
+      "learning_rate": 9.393403222176451e-06,
+      "loss": 0.3589,
+      "step": 1545
+    },
+    {
+      "epoch": 0.3679419289581722,
+      "grad_norm": 0.44633345529058044,
+      "learning_rate": 9.392482643043142e-06,
+      "loss": 0.3553,
+      "step": 1546
+    },
+    {
+      "epoch": 0.368179925031237,
+      "grad_norm": 0.3856503746950216,
+      "learning_rate": 9.391561411078245e-06,
+      "loss": 0.414,
+      "step": 1547
+    },
+    {
+      "epoch": 0.36841792110430177,
+      "grad_norm": 0.43687712395015343,
+      "learning_rate": 9.39063952641868e-06,
+      "loss": 0.3678,
+      "step": 1548
+    },
+    {
+      "epoch": 0.36865591717736657,
+      "grad_norm": 0.43545658546660954,
+      "learning_rate": 9.389716989201464e-06,
+      "loss": 0.3333,
+      "step": 1549
+    },
+    {
+      "epoch": 0.36889391325043136,
+      "grad_norm": 0.41050982682577164,
+      "learning_rate": 9.388793799563706e-06,
+      "loss": 0.392,
+      "step": 1550
+    },
+    {
+      "epoch": 0.36913190932349615,
+      "grad_norm": 0.37175911777408144,
+      "learning_rate": 9.387869957642616e-06,
+      "loss": 0.4177,
+      "step": 1551
+    },
+    {
+      "epoch": 0.36936990539656095,
+      "grad_norm": 0.4825250816525954,
+      "learning_rate": 9.3869454635755e-06,
+      "loss": 0.3373,
+      "step": 1552
+    },
+    {
+      "epoch": 0.36960790146962574,
+      "grad_norm": 0.46259012792132853,
+      "learning_rate": 9.38602031749976e-06,
+      "loss": 0.3139,
+      "step": 1553
+    },
+    {
+      "epoch": 0.36984589754269054,
+      "grad_norm": 0.43835976647878444,
+      "learning_rate": 9.385094519552896e-06,
+      "loss": 0.368,
+      "step": 1554
+    },
+    {
+      "epoch": 0.37008389361575533,
+      "grad_norm": 0.4476474289989832,
+      "learning_rate": 9.384168069872505e-06,
+      "loss": 0.4029,
+      "step": 1555
+    },
+    {
+      "epoch": 0.3703218896888201,
+      "grad_norm": 0.4670049546010331,
+      "learning_rate": 9.38324096859628e-06,
+      "loss": 0.3323,
+      "step": 1556
+    },
+    {
+      "epoch": 0.3705598857618849,
+      "grad_norm": 0.379594355730863,
+      "learning_rate": 9.382313215862009e-06,
+      "loss": 0.376,
+      "step": 1557
+    },
+    {
+      "epoch": 0.3707978818349497,
+      "grad_norm": 0.44935069232834013,
+      "learning_rate": 9.38138481180758e-06,
+      "loss": 0.4237,
+      "step": 1558
+    },
+    {
+      "epoch": 0.3710358779080145,
+      "grad_norm": 0.5090826653616003,
+      "learning_rate": 9.38045575657098e-06,
+      "loss": 0.3653,
+      "step": 1559
+    },
+    {
+      "epoch": 0.3712738739810793,
+      "grad_norm": 0.47245779020299394,
+      "learning_rate": 9.37952605029029e-06,
+      "loss": 0.3275,
+      "step": 1560
+    },
+    {
+      "epoch": 0.3715118700541441,
+      "grad_norm": 0.40765167623096316,
+      "learning_rate": 9.378595693103681e-06,
+      "loss": 0.4096,
+      "step": 1561
+    },
+    {
+      "epoch": 0.3717498661272089,
+      "grad_norm": 0.4459617175575773,
+      "learning_rate": 9.37766468514943e-06,
+      "loss": 0.4226,
+      "step": 1562
+    },
+    {
+      "epoch": 0.3719878622002737,
+      "grad_norm": 0.43777206352478504,
+      "learning_rate": 9.376733026565911e-06,
+      "loss": 0.3128,
+      "step": 1563
+    },
+    {
+      "epoch": 0.3722258582733385,
+      "grad_norm": 0.43171693526067784,
+      "learning_rate": 9.375800717491588e-06,
+      "loss": 0.3511,
+      "step": 1564
+    },
+    {
+      "epoch": 0.3724638543464033,
+      "grad_norm": 0.37666167095835446,
+      "learning_rate": 9.374867758065027e-06,
+      "loss": 0.4001,
+      "step": 1565
+    },
+    {
+      "epoch": 0.3727018504194681,
+      "grad_norm": 0.45130931365820925,
+      "learning_rate": 9.373934148424887e-06,
+      "loss": 0.371,
+      "step": 1566
+    },
+    {
+      "epoch": 0.37293984649253287,
+      "grad_norm": 0.4067510576017569,
+      "learning_rate": 9.372999888709927e-06,
+      "loss": 0.3598,
+      "step": 1567
+    },
+    {
+      "epoch": 0.37317784256559766,
+      "grad_norm": 0.3956103473115734,
+      "learning_rate": 9.372064979059001e-06,
+      "loss": 0.3735,
+      "step": 1568
+    },
+    {
+      "epoch": 0.37341583863866246,
+      "grad_norm": 0.38688699762267664,
+      "learning_rate": 9.371129419611059e-06,
+      "loss": 0.4281,
+      "step": 1569
+    },
+    {
+      "epoch": 0.37365383471172725,
+      "grad_norm": 0.42553445781670524,
+      "learning_rate": 9.37019321050515e-06,
+      "loss": 0.3071,
+      "step": 1570
+    },
+    {
+      "epoch": 0.37389183078479205,
+      "grad_norm": 0.43085996359356,
+      "learning_rate": 9.369256351880415e-06,
+      "loss": 0.3452,
+      "step": 1571
+    },
+    {
+      "epoch": 0.37412982685785684,
+      "grad_norm": 0.40686052760995856,
+      "learning_rate": 9.368318843876097e-06,
+      "loss": 0.4233,
+      "step": 1572
+    },
+    {
+      "epoch": 0.37436782293092163,
+      "grad_norm": 0.359734329282973,
+      "learning_rate": 9.36738068663153e-06,
+      "loss": 0.3651,
+      "step": 1573
+    },
+    {
+      "epoch": 0.37460581900398643,
+      "grad_norm": 0.3856761545147769,
+      "learning_rate": 9.36644188028615e-06,
+      "loss": 0.3182,
+      "step": 1574
+    },
+    {
+      "epoch": 0.3748438150770512,
+      "grad_norm": 0.3957759378316436,
+      "learning_rate": 9.365502424979488e-06,
+      "loss": 0.4013,
+      "step": 1575
+    },
+    {
+      "epoch": 0.375081811150116,
+      "grad_norm": 0.40850014589920525,
+      "learning_rate": 9.364562320851167e-06,
+      "loss": 0.4193,
+      "step": 1576
+    },
+    {
+      "epoch": 0.3753198072231808,
+      "grad_norm": 0.38661039932841545,
+      "learning_rate": 9.36362156804091e-06,
+      "loss": 0.3553,
+      "step": 1577
+    },
+    {
+      "epoch": 0.3755578032962456,
+      "grad_norm": 0.4003228203902698,
+      "learning_rate": 9.362680166688538e-06,
+      "loss": 0.3396,
+      "step": 1578
+    },
+    {
+      "epoch": 0.3757957993693104,
+      "grad_norm": 0.3831137546008049,
+      "learning_rate": 9.361738116933967e-06,
+      "loss": 0.3872,
+      "step": 1579
+    },
+    {
+      "epoch": 0.3760337954423752,
+      "grad_norm": 0.44314506167594236,
+      "learning_rate": 9.360795418917205e-06,
+      "loss": 0.3819,
+      "step": 1580
+    },
+    {
+      "epoch": 0.37627179151544,
+      "grad_norm": 0.46088081221952837,
+      "learning_rate": 9.359852072778365e-06,
+      "loss": 0.3196,
+      "step": 1581
+    },
+    {
+      "epoch": 0.3765097875885048,
+      "grad_norm": 0.481649845998446,
+      "learning_rate": 9.35890807865765e-06,
+      "loss": 0.3606,
+      "step": 1582
+    },
+    {
+      "epoch": 0.3767477836615696,
+      "grad_norm": 0.4324693237486104,
+      "learning_rate": 9.357963436695357e-06,
+      "loss": 0.4206,
+      "step": 1583
+    },
+    {
+      "epoch": 0.3769857797346344,
+      "grad_norm": 0.47833760040095646,
+      "learning_rate": 9.357018147031888e-06,
+      "loss": 0.3558,
+      "step": 1584
+    },
+    {
+      "epoch": 0.37722377580769917,
+      "grad_norm": 0.3956699015096483,
+      "learning_rate": 9.356072209807737e-06,
+      "loss": 0.3198,
+      "step": 1585
+    },
+    {
+      "epoch": 0.37746177188076396,
+      "grad_norm": 0.3864788423378403,
+      "learning_rate": 9.35512562516349e-06,
+      "loss": 0.4045,
+      "step": 1586
+    },
+    {
+      "epoch": 0.37769976795382876,
+      "grad_norm": 0.4160431742514611,
+      "learning_rate": 9.354178393239834e-06,
+      "loss": 0.4082,
+      "step": 1587
+    },
+    {
+      "epoch": 0.37793776402689355,
+      "grad_norm": 0.42646842694183207,
+      "learning_rate": 9.353230514177553e-06,
+      "loss": 0.3299,
+      "step": 1588
+    },
+    {
+      "epoch": 0.37817576009995835,
+      "grad_norm": 0.4132760310497568,
+      "learning_rate": 9.352281988117521e-06,
+      "loss": 0.4134,
+      "step": 1589
+    },
+    {
+      "epoch": 0.37841375617302314,
+      "grad_norm": 0.39054592818013045,
+      "learning_rate": 9.35133281520072e-06,
+      "loss": 0.4301,
+      "step": 1590
+    },
+    {
+      "epoch": 0.37865175224608794,
+      "grad_norm": 0.4404186715801395,
+      "learning_rate": 9.350382995568213e-06,
+      "loss": 0.3641,
+      "step": 1591
+    },
+    {
+      "epoch": 0.37888974831915273,
+      "grad_norm": 0.38490449743758554,
+      "learning_rate": 9.349432529361168e-06,
+      "loss": 0.3679,
+      "step": 1592
+    },
+    {
+      "epoch": 0.3791277443922175,
+      "grad_norm": 0.38256739427952213,
+      "learning_rate": 9.348481416720852e-06,
+      "loss": 0.4013,
+      "step": 1593
+    },
+    {
+      "epoch": 0.3793657404652823,
+      "grad_norm": 0.392281809140338,
+      "learning_rate": 9.34752965778862e-06,
+      "loss": 0.4262,
+      "step": 1594
+    },
+    {
+      "epoch": 0.3796037365383471,
+      "grad_norm": 0.3603426933987028,
+      "learning_rate": 9.346577252705929e-06,
+      "loss": 0.3434,
+      "step": 1595
+    },
+    {
+      "epoch": 0.3798417326114119,
+      "grad_norm": 0.3862895458269031,
+      "learning_rate": 9.345624201614328e-06,
+      "loss": 0.3273,
+      "step": 1596
+    },
+    {
+      "epoch": 0.3800797286844767,
+      "grad_norm": 0.4424213661034337,
+      "learning_rate": 9.344670504655466e-06,
+      "loss": 0.4163,
+      "step": 1597
+    },
+    {
+      "epoch": 0.3803177247575415,
+      "grad_norm": 0.3946042222856275,
+      "learning_rate": 9.343716161971084e-06,
+      "loss": 0.3853,
+      "step": 1598
+    },
+    {
+      "epoch": 0.3805557208306063,
+      "grad_norm": 0.48228594107422296,
+      "learning_rate": 9.342761173703023e-06,
+      "loss": 0.3121,
+      "step": 1599
+    },
+    {
+      "epoch": 0.3807937169036711,
+      "grad_norm": 0.4055270748007401,
+      "learning_rate": 9.341805539993216e-06,
+      "loss": 0.3568,
+      "step": 1600
+    },
+    {
+      "epoch": 0.3810317129767359,
+      "grad_norm": 0.37266993056107267,
+      "learning_rate": 9.340849260983695e-06,
+      "loss": 0.4455,
+      "step": 1601
+    },
+    {
+      "epoch": 0.3812697090498007,
+      "grad_norm": 0.4040859091869553,
+      "learning_rate": 9.339892336816587e-06,
+      "loss": 0.316,
+      "step": 1602
+    },
+    {
+      "epoch": 0.38150770512286547,
+      "grad_norm": 0.3984553047662032,
+      "learning_rate": 9.338934767634114e-06,
+      "loss": 0.3299,
+      "step": 1603
+    },
+    {
+      "epoch": 0.38174570119593026,
+      "grad_norm": 0.3937006805260035,
+      "learning_rate": 9.337976553578593e-06,
+      "loss": 0.4225,
+      "step": 1604
+    },
+    {
+      "epoch": 0.38198369726899506,
+      "grad_norm": 0.37712895414520553,
+      "learning_rate": 9.337017694792441e-06,
+      "loss": 0.3939,
+      "step": 1605
+    },
+    {
+      "epoch": 0.38222169334205985,
+      "grad_norm": 0.4069249598197311,
+      "learning_rate": 9.336058191418167e-06,
+      "loss": 0.3414,
+      "step": 1606
+    },
+    {
+      "epoch": 0.38245968941512465,
+      "grad_norm": 0.37857978246964663,
+      "learning_rate": 9.335098043598376e-06,
+      "loss": 0.3751,
+      "step": 1607
+    },
+    {
+      "epoch": 0.38269768548818944,
+      "grad_norm": 0.37160587062598166,
+      "learning_rate": 9.334137251475771e-06,
+      "loss": 0.4226,
+      "step": 1608
+    },
+    {
+      "epoch": 0.38293568156125424,
+      "grad_norm": 0.4083708370687077,
+      "learning_rate": 9.333175815193149e-06,
+      "loss": 0.3667,
+      "step": 1609
+    },
+    {
+      "epoch": 0.38317367763431903,
+      "grad_norm": 0.39386664440771063,
+      "learning_rate": 9.332213734893406e-06,
+      "loss": 0.3182,
+      "step": 1610
+    },
+    {
+      "epoch": 0.3834116737073838,
+      "grad_norm": 0.4343970310481392,
+      "learning_rate": 9.331251010719525e-06,
+      "loss": 0.4,
+      "step": 1611
+    },
+    {
+      "epoch": 0.3836496697804486,
+      "grad_norm": 0.43461958864883965,
+      "learning_rate": 9.330287642814593e-06,
+      "loss": 0.4267,
+      "step": 1612
+    },
+    {
+      "epoch": 0.3838876658535134,
+      "grad_norm": 0.41875602341308255,
+      "learning_rate": 9.329323631321793e-06,
+      "loss": 0.3314,
+      "step": 1613
+    },
+    {
+      "epoch": 0.3841256619265782,
+      "grad_norm": 0.3891910114438827,
+      "learning_rate": 9.328358976384398e-06,
+      "loss": 0.3693,
+      "step": 1614
+    },
+    {
+      "epoch": 0.384363657999643,
+      "grad_norm": 0.3988395541503049,
+      "learning_rate": 9.327393678145781e-06,
+      "loss": 0.4159,
+      "step": 1615
+    },
+    {
+      "epoch": 0.3846016540727078,
+      "grad_norm": 0.4273306553986297,
+      "learning_rate": 9.32642773674941e-06,
+      "loss": 0.3702,
+      "step": 1616
+    },
+    {
+      "epoch": 0.3848396501457726,
+      "grad_norm": 0.4297415267388038,
+      "learning_rate": 9.325461152338846e-06,
+      "loss": 0.3125,
+      "step": 1617
+    },
+    {
+      "epoch": 0.3850776462188374,
+      "grad_norm": 0.3713890087288812,
+      "learning_rate": 9.324493925057747e-06,
+      "loss": 0.3745,
+      "step": 1618
+    },
+    {
+      "epoch": 0.3853156422919022,
+      "grad_norm": 0.397582778645666,
+      "learning_rate": 9.32352605504987e-06,
+      "loss": 0.4028,
+      "step": 1619
+    },
+    {
+      "epoch": 0.385553638364967,
+      "grad_norm": 0.3879589456647183,
+      "learning_rate": 9.322557542459061e-06,
+      "loss": 0.3442,
+      "step": 1620
+    },
+    {
+      "epoch": 0.38579163443803177,
+      "grad_norm": 0.3734347689407556,
+      "learning_rate": 9.321588387429266e-06,
+      "loss": 0.3492,
+      "step": 1621
+    },
+    {
+      "epoch": 0.38602963051109657,
+      "grad_norm": 0.38372821085957537,
+      "learning_rate": 9.320618590104525e-06,
+      "loss": 0.4057,
+      "step": 1622
+    },
+    {
+      "epoch": 0.38626762658416136,
+      "grad_norm": 0.39004512408218667,
+      "learning_rate": 9.319648150628978e-06,
+      "loss": 0.3837,
+      "step": 1623
+    },
+    {
+      "epoch": 0.38650562265722616,
+      "grad_norm": 0.43356887512118736,
+      "learning_rate": 9.318677069146848e-06,
+      "loss": 0.3141,
+      "step": 1624
+    },
+    {
+      "epoch": 0.38674361873029095,
+      "grad_norm": 0.4399012491269116,
+      "learning_rate": 9.31770534580247e-06,
+      "loss": 0.4022,
+      "step": 1625
+    },
+    {
+      "epoch": 0.38698161480335574,
+      "grad_norm": 0.40492521303478335,
+      "learning_rate": 9.316732980740262e-06,
+      "loss": 0.419,
+      "step": 1626
+    },
+    {
+      "epoch": 0.38721961087642054,
+      "grad_norm": 0.4061119104072092,
+      "learning_rate": 9.315759974104741e-06,
+      "loss": 0.3166,
+      "step": 1627
+    },
+    {
+      "epoch": 0.38745760694948533,
+      "grad_norm": 0.3909427341832737,
+      "learning_rate": 9.314786326040523e-06,
+      "loss": 0.3332,
+      "step": 1628
+    },
+    {
+      "epoch": 0.38769560302255013,
+      "grad_norm": 0.42034544526119255,
+      "learning_rate": 9.313812036692314e-06,
+      "loss": 0.3961,
+      "step": 1629
+    },
+    {
+      "epoch": 0.3879335990956149,
+      "grad_norm": 0.43657367145331466,
+      "learning_rate": 9.312837106204916e-06,
+      "loss": 0.373,
+      "step": 1630
+    },
+    {
+      "epoch": 0.3881715951686797,
+      "grad_norm": 0.37018141911486274,
+      "learning_rate": 9.31186153472323e-06,
+      "loss": 0.3114,
+      "step": 1631
+    },
+    {
+      "epoch": 0.3884095912417445,
+      "grad_norm": 0.38503056353803755,
+      "learning_rate": 9.31088532239225e-06,
+      "loss": 0.386,
+      "step": 1632
+    },
+    {
+      "epoch": 0.3886475873148093,
+      "grad_norm": 0.39233949025594983,
+      "learning_rate": 9.309908469357067e-06,
+      "loss": 0.416,
+      "step": 1633
+    },
+    {
+      "epoch": 0.3888855833878741,
+      "grad_norm": 0.4433267496831673,
+      "learning_rate": 9.308930975762862e-06,
+      "loss": 0.3621,
+      "step": 1634
+    },
+    {
+      "epoch": 0.3891235794609389,
+      "grad_norm": 0.3952777289255298,
+      "learning_rate": 9.307952841754916e-06,
+      "loss": 0.3255,
+      "step": 1635
+    },
+    {
+      "epoch": 0.3893615755340037,
+      "grad_norm": 0.4022306208679625,
+      "learning_rate": 9.306974067478602e-06,
+      "loss": 0.3977,
+      "step": 1636
+    },
+    {
+      "epoch": 0.3895995716070685,
+      "grad_norm": 0.41753431253514406,
+      "learning_rate": 9.305994653079396e-06,
+      "loss": 0.4056,
+      "step": 1637
+    },
+    {
+      "epoch": 0.3898375676801333,
+      "grad_norm": 0.40246752404914526,
+      "learning_rate": 9.305014598702857e-06,
+      "loss": 0.3407,
+      "step": 1638
+    },
+    {
+      "epoch": 0.3900755637531981,
+      "grad_norm": 0.400365444870483,
+      "learning_rate": 9.304033904494649e-06,
+      "loss": 0.3742,
+      "step": 1639
+    },
+    {
+      "epoch": 0.39031355982626287,
+      "grad_norm": 0.4157295616816406,
+      "learning_rate": 9.303052570600524e-06,
+      "loss": 0.4111,
+      "step": 1640
+    },
+    {
+      "epoch": 0.39055155589932766,
+      "grad_norm": 0.4525080007717071,
+      "learning_rate": 9.302070597166337e-06,
+      "loss": 0.3378,
+      "step": 1641
+    },
+    {
+      "epoch": 0.39078955197239246,
+      "grad_norm": 0.3854211007538121,
+      "learning_rate": 9.301087984338029e-06,
+      "loss": 0.3288,
+      "step": 1642
+    },
+    {
+      "epoch": 0.39102754804545725,
+      "grad_norm": 0.4832368919472511,
+      "learning_rate": 9.300104732261645e-06,
+      "loss": 0.3952,
+      "step": 1643
+    },
+    {
+      "epoch": 0.39126554411852205,
+      "grad_norm": 0.42640640785474865,
+      "learning_rate": 9.299120841083317e-06,
+      "loss": 0.4237,
+      "step": 1644
+    },
+    {
+      "epoch": 0.39150354019158684,
+      "grad_norm": 0.41078488904477306,
+      "learning_rate": 9.298136310949278e-06,
+      "loss": 0.3299,
+      "step": 1645
+    },
+    {
+      "epoch": 0.39174153626465164,
+      "grad_norm": 0.3496318552591008,
+      "learning_rate": 9.297151142005852e-06,
+      "loss": 0.3274,
+      "step": 1646
+    },
+    {
+      "epoch": 0.39197953233771643,
+      "grad_norm": 0.45341271456511995,
+      "learning_rate": 9.296165334399458e-06,
+      "loss": 0.4078,
+      "step": 1647
+    },
+    {
+      "epoch": 0.3922175284107812,
+      "grad_norm": 0.4210029643414087,
+      "learning_rate": 9.295178888276615e-06,
+      "loss": 0.3651,
+      "step": 1648
+    },
+    {
+      "epoch": 0.392455524483846,
+      "grad_norm": 0.38599905665403594,
+      "learning_rate": 9.294191803783931e-06,
+      "loss": 0.3477,
+      "step": 1649
+    },
+    {
+      "epoch": 0.3926935205569108,
+      "grad_norm": 0.4189352579336025,
+      "learning_rate": 9.293204081068113e-06,
+      "loss": 0.3932,
+      "step": 1650
+    },
+    {
+      "epoch": 0.3929315166299756,
+      "grad_norm": 0.43257183707928304,
+      "learning_rate": 9.292215720275959e-06,
+      "loss": 0.422,
+      "step": 1651
+    },
+    {
+      "epoch": 0.3931695127030404,
+      "grad_norm": 0.4838474825444332,
+      "learning_rate": 9.291226721554364e-06,
+      "loss": 0.3335,
+      "step": 1652
+    },
+    {
+      "epoch": 0.3934075087761052,
+      "grad_norm": 0.36325017737845905,
+      "learning_rate": 9.290237085050318e-06,
+      "loss": 0.3438,
+      "step": 1653
+    },
+    {
+      "epoch": 0.39364550484917,
+      "grad_norm": 0.412707690424985,
+      "learning_rate": 9.289246810910909e-06,
+      "loss": 0.4129,
+      "step": 1654
+    },
+    {
+      "epoch": 0.3938835009222348,
+      "grad_norm": 0.41393730782449073,
+      "learning_rate": 9.288255899283309e-06,
+      "loss": 0.4021,
+      "step": 1655
+    },
+    {
+      "epoch": 0.3941214969952996,
+      "grad_norm": 0.4386727788302991,
+      "learning_rate": 9.287264350314797e-06,
+      "loss": 0.3621,
+      "step": 1656
+    },
+    {
+      "epoch": 0.3943594930683644,
+      "grad_norm": 0.3681925197909611,
+      "learning_rate": 9.286272164152744e-06,
+      "loss": 0.3769,
+      "step": 1657
+    },
+    {
+      "epoch": 0.39459748914142917,
+      "grad_norm": 0.4635910656535038,
+      "learning_rate": 9.285279340944607e-06,
+      "loss": 0.4101,
+      "step": 1658
+    },
+    {
+      "epoch": 0.39483548521449396,
+      "grad_norm": 0.4598290354163402,
+      "learning_rate": 9.284285880837947e-06,
+      "loss": 0.3532,
+      "step": 1659
+    },
+    {
+      "epoch": 0.39507348128755876,
+      "grad_norm": 0.3933380949916356,
+      "learning_rate": 9.283291783980418e-06,
+      "loss": 0.3303,
+      "step": 1660
+    },
+    {
+      "epoch": 0.39531147736062355,
+      "grad_norm": 0.39884801926811836,
+      "learning_rate": 9.282297050519767e-06,
+      "loss": 0.3897,
+      "step": 1661
+    },
+    {
+      "epoch": 0.39554947343368835,
+      "grad_norm": 0.5083918488666362,
+      "learning_rate": 9.281301680603834e-06,
+      "loss": 0.4365,
+      "step": 1662
+    },
+    {
+      "epoch": 0.39578746950675314,
+      "grad_norm": 0.5222377151000596,
+      "learning_rate": 9.280305674380558e-06,
+      "loss": 0.3347,
+      "step": 1663
+    },
+    {
+      "epoch": 0.39602546557981794,
+      "grad_norm": 0.40550870332567346,
+      "learning_rate": 9.279309031997968e-06,
+      "loss": 0.3609,
+      "step": 1664
+    },
+    {
+      "epoch": 0.39626346165288273,
+      "grad_norm": 0.4047710761598781,
+      "learning_rate": 9.278311753604192e-06,
+      "loss": 0.437,
+      "step": 1665
+    },
+    {
+      "epoch": 0.3965014577259475,
+      "grad_norm": 0.5151918734857975,
+      "learning_rate": 9.277313839347449e-06,
+      "loss": 0.3752,
+      "step": 1666
+    },
+    {
+      "epoch": 0.3967394537990123,
+      "grad_norm": 0.4697431600744339,
+      "learning_rate": 9.276315289376052e-06,
+      "loss": 0.3242,
+      "step": 1667
+    },
+    {
+      "epoch": 0.3969774498720771,
+      "grad_norm": 0.43709669423640757,
+      "learning_rate": 9.275316103838414e-06,
+      "loss": 0.3797,
+      "step": 1668
+    },
+    {
+      "epoch": 0.3972154459451419,
+      "grad_norm": 0.4192015191302519,
+      "learning_rate": 9.274316282883037e-06,
+      "loss": 0.4335,
+      "step": 1669
+    },
+    {
+      "epoch": 0.3974534420182067,
+      "grad_norm": 0.4117716004967471,
+      "learning_rate": 9.273315826658518e-06,
+      "loss": 0.3522,
+      "step": 1670
+    },
+    {
+      "epoch": 0.3976914380912715,
+      "grad_norm": 0.42633382432239425,
+      "learning_rate": 9.27231473531355e-06,
+      "loss": 0.3689,
+      "step": 1671
+    },
+    {
+      "epoch": 0.3979294341643363,
+      "grad_norm": 0.38633621983522715,
+      "learning_rate": 9.271313008996922e-06,
+      "loss": 0.4201,
+      "step": 1672
+    },
+    {
+      "epoch": 0.3981674302374011,
+      "grad_norm": 0.42827859761102866,
+      "learning_rate": 9.270310647857513e-06,
+      "loss": 0.3973,
+      "step": 1673
+    },
+    {
+      "epoch": 0.3984054263104659,
+      "grad_norm": 0.5003093170938062,
+      "learning_rate": 9.269307652044298e-06,
+      "loss": 0.326,
+      "step": 1674
+    },
+    {
+      "epoch": 0.3986434223835307,
+      "grad_norm": 0.43808182708918314,
+      "learning_rate": 9.26830402170635e-06,
+      "loss": 0.3888,
+      "step": 1675
+    },
+    {
+      "epoch": 0.39888141845659547,
+      "grad_norm": 0.41053386389897284,
+      "learning_rate": 9.267299756992829e-06,
+      "loss": 0.4315,
+      "step": 1676
+    },
+    {
+      "epoch": 0.39911941452966027,
+      "grad_norm": 0.4230397692191716,
+      "learning_rate": 9.266294858052998e-06,
+      "loss": 0.3346,
+      "step": 1677
+    },
+    {
+      "epoch": 0.39935741060272506,
+      "grad_norm": 0.431898235869182,
+      "learning_rate": 9.265289325036209e-06,
+      "loss": 0.3241,
+      "step": 1678
+    },
+    {
+      "epoch": 0.39959540667578985,
+      "grad_norm": 0.39729706567685696,
+      "learning_rate": 9.264283158091909e-06,
+      "loss": 0.4009,
+      "step": 1679
+    },
+    {
+      "epoch": 0.39983340274885465,
+      "grad_norm": 0.4004504040408847,
+      "learning_rate": 9.263276357369635e-06,
+      "loss": 0.3942,
+      "step": 1680
+    },
+    {
+      "epoch": 0.40007139882191944,
+      "grad_norm": 0.4300597724152983,
+      "learning_rate": 9.262268923019028e-06,
+      "loss": 0.3161,
+      "step": 1681
+    },
+    {
+      "epoch": 0.40030939489498424,
+      "grad_norm": 0.5008687287891455,
+      "learning_rate": 9.261260855189815e-06,
+      "loss": 0.3539,
+      "step": 1682
+    },
+    {
+      "epoch": 0.40054739096804903,
+      "grad_norm": 0.36119133039165174,
+      "learning_rate": 9.26025215403182e-06,
+      "loss": 0.4348,
+      "step": 1683
+    },
+    {
+      "epoch": 0.4007853870411138,
+      "grad_norm": 0.42846844104954446,
+      "learning_rate": 9.259242819694963e-06,
+      "loss": 0.3417,
+      "step": 1684
+    },
+    {
+      "epoch": 0.4010233831141786,
+      "grad_norm": 0.4165944881715832,
+      "learning_rate": 9.258232852329253e-06,
+      "loss": 0.317,
+      "step": 1685
+    },
+    {
+      "epoch": 0.4012613791872434,
+      "grad_norm": 0.3890173772649005,
+      "learning_rate": 9.257222252084798e-06,
+      "loss": 0.3775,
+      "step": 1686
+    },
+    {
+      "epoch": 0.4014993752603082,
+      "grad_norm": 0.38942060350209845,
+      "learning_rate": 9.256211019111799e-06,
+      "loss": 0.4156,
+      "step": 1687
+    },
+    {
+      "epoch": 0.401737371333373,
+      "grad_norm": 0.4134669751208099,
+      "learning_rate": 9.255199153560546e-06,
+      "loss": 0.3203,
+      "step": 1688
+    },
+    {
+      "epoch": 0.4019753674064378,
+      "grad_norm": 0.40855285972159666,
+      "learning_rate": 9.254186655581431e-06,
+      "loss": 0.3485,
+      "step": 1689
+    },
+    {
+      "epoch": 0.4022133634795026,
+      "grad_norm": 0.4066844623707453,
+      "learning_rate": 9.253173525324937e-06,
+      "loss": 0.448,
+      "step": 1690
+    },
+    {
+      "epoch": 0.4024513595525674,
+      "grad_norm": 0.40579610697102486,
+      "learning_rate": 9.252159762941638e-06,
+      "loss": 0.3493,
+      "step": 1691
+    },
+    {
+      "epoch": 0.4026893556256322,
+      "grad_norm": 0.4612752035521552,
+      "learning_rate": 9.251145368582204e-06,
+      "loss": 0.314,
+      "step": 1692
+    },
+    {
+      "epoch": 0.402927351698697,
+      "grad_norm": 0.38917703961071853,
+      "learning_rate": 9.2501303423974e-06,
+      "loss": 0.3809,
+      "step": 1693
+    },
+    {
+      "epoch": 0.4031653477717618,
+      "grad_norm": 0.42688170513775625,
+      "learning_rate": 9.249114684538087e-06,
+      "loss": 0.439,
+      "step": 1694
+    },
+    {
+      "epoch": 0.40340334384482657,
+      "grad_norm": 0.4270682354405054,
+      "learning_rate": 9.248098395155212e-06,
+      "loss": 0.3197,
+      "step": 1695
+    },
+    {
+      "epoch": 0.40364133991789136,
+      "grad_norm": 0.38427191838232466,
+      "learning_rate": 9.247081474399821e-06,
+      "loss": 0.3653,
+      "step": 1696
+    },
+    {
+      "epoch": 0.40387933599095616,
+      "grad_norm": 0.3635373879235468,
+      "learning_rate": 9.246063922423057e-06,
+      "loss": 0.3965,
+      "step": 1697
+    },
+    {
+      "epoch": 0.40411733206402095,
+      "grad_norm": 0.36800213725949676,
+      "learning_rate": 9.24504573937615e-06,
+      "loss": 0.3629,
+      "step": 1698
+    },
+    {
+      "epoch": 0.40435532813708575,
+      "grad_norm": 0.46951813533096415,
+      "learning_rate": 9.24402692541043e-06,
+      "loss": 0.3301,
+      "step": 1699
+    },
+    {
+      "epoch": 0.40459332421015054,
+      "grad_norm": 0.40731159678572715,
+      "learning_rate": 9.243007480677317e-06,
+      "loss": 0.3892,
+      "step": 1700
+    },
+    {
+      "epoch": 0.40483132028321533,
+      "grad_norm": 0.39470797179107836,
+      "learning_rate": 9.241987405328325e-06,
+      "loss": 0.4199,
+      "step": 1701
+    },
+    {
+      "epoch": 0.40506931635628013,
+      "grad_norm": 0.379892047471591,
+      "learning_rate": 9.240966699515062e-06,
+      "loss": 0.339,
+      "step": 1702
+    },
+    {
+      "epoch": 0.4053073124293449,
+      "grad_norm": 0.4299870200864259,
+      "learning_rate": 9.239945363389233e-06,
+      "loss": 0.3226,
+      "step": 1703
+    },
+    {
+      "epoch": 0.4055453085024097,
+      "grad_norm": 0.38016909458014425,
+      "learning_rate": 9.238923397102629e-06,
+      "loss": 0.4155,
+      "step": 1704
+    },
+    {
+      "epoch": 0.4057833045754745,
+      "grad_norm": 0.41162849886265007,
+      "learning_rate": 9.237900800807144e-06,
+      "loss": 0.3961,
+      "step": 1705
+    },
+    {
+      "epoch": 0.4060213006485393,
+      "grad_norm": 0.41727989058008336,
+      "learning_rate": 9.23687757465476e-06,
+      "loss": 0.3234,
+      "step": 1706
+    },
+    {
+      "epoch": 0.4062592967216041,
+      "grad_norm": 0.4257234199843476,
+      "learning_rate": 9.235853718797552e-06,
+      "loss": 0.3652,
+      "step": 1707
+    },
+    {
+      "epoch": 0.4064972927946689,
+      "grad_norm": 0.38093238010454294,
+      "learning_rate": 9.234829233387692e-06,
+      "loss": 0.4513,
+      "step": 1708
+    },
+    {
+      "epoch": 0.4067352888677337,
+      "grad_norm": 0.42134937392720456,
+      "learning_rate": 9.233804118577442e-06,
+      "loss": 0.3595,
+      "step": 1709
+    },
+    {
+      "epoch": 0.4069732849407985,
+      "grad_norm": 0.36011012728718184,
+      "learning_rate": 9.232778374519162e-06,
+      "loss": 0.3139,
+      "step": 1710
+    },
+    {
+      "epoch": 0.4072112810138633,
+      "grad_norm": 0.3848429350956219,
+      "learning_rate": 9.231752001365301e-06,
+      "loss": 0.3874,
+      "step": 1711
+    },
+    {
+      "epoch": 0.4074492770869281,
+      "grad_norm": 0.42989698919103714,
+      "learning_rate": 9.230724999268405e-06,
+      "loss": 0.4137,
+      "step": 1712
+    },
+    {
+      "epoch": 0.40768727315999287,
+      "grad_norm": 0.40356043597179625,
+      "learning_rate": 9.22969736838111e-06,
+      "loss": 0.3497,
+      "step": 1713
+    },
+    {
+      "epoch": 0.40792526923305766,
+      "grad_norm": 0.38088846927432707,
+      "learning_rate": 9.22866910885615e-06,
+      "loss": 0.3511,
+      "step": 1714
+    },
+    {
+      "epoch": 0.40816326530612246,
+      "grad_norm": 0.36991122947115884,
+      "learning_rate": 9.227640220846345e-06,
+      "loss": 0.4034,
+      "step": 1715
+    },
+    {
+      "epoch": 0.40840126137918725,
+      "grad_norm": 0.4121306552946537,
+      "learning_rate": 9.226610704504619e-06,
+      "loss": 0.3843,
+      "step": 1716
+    },
+    {
+      "epoch": 0.40863925745225205,
+      "grad_norm": 0.39375147468162974,
+      "learning_rate": 9.225580559983982e-06,
+      "loss": 0.3329,
+      "step": 1717
+    },
+    {
+      "epoch": 0.40887725352531684,
+      "grad_norm": 0.39350805991360777,
+      "learning_rate": 9.224549787437536e-06,
+      "loss": 0.398,
+      "step": 1718
+    },
+    {
+      "epoch": 0.40911524959838164,
+      "grad_norm": 0.40232745688418725,
+      "learning_rate": 9.223518387018481e-06,
+      "loss": 0.4178,
+      "step": 1719
+    },
+    {
+      "epoch": 0.40935324567144643,
+      "grad_norm": 0.39632952276294475,
+      "learning_rate": 9.22248635888011e-06,
+      "loss": 0.3531,
+      "step": 1720
+    },
+    {
+      "epoch": 0.4095912417445112,
+      "grad_norm": 0.38179777362704265,
+      "learning_rate": 9.221453703175805e-06,
+      "loss": 0.3252,
+      "step": 1721
+    },
+    {
+      "epoch": 0.409829237817576,
+      "grad_norm": 0.3717944610724434,
+      "learning_rate": 9.220420420059048e-06,
+      "loss": 0.4133,
+      "step": 1722
+    },
+    {
+      "epoch": 0.4100672338906408,
+      "grad_norm": 0.3996883821281656,
+      "learning_rate": 9.219386509683408e-06,
+      "loss": 0.38,
+      "step": 1723
+    },
+    {
+      "epoch": 0.4103052299637056,
+      "grad_norm": 0.47689973154409665,
+      "learning_rate": 9.21835197220255e-06,
+      "loss": 0.3705,
+      "step": 1724
+    },
+    {
+      "epoch": 0.4105432260367704,
+      "grad_norm": 0.40061030822431853,
+      "learning_rate": 9.217316807770232e-06,
+      "loss": 0.3798,
+      "step": 1725
+    },
+    {
+      "epoch": 0.4107812221098352,
+      "grad_norm": 0.3891661943835348,
+      "learning_rate": 9.216281016540305e-06,
+      "loss": 0.413,
+      "step": 1726
+    },
+    {
+      "epoch": 0.4110192181829,
+      "grad_norm": 0.40874575169057586,
+      "learning_rate": 9.215244598666712e-06,
+      "loss": 0.3299,
+      "step": 1727
+    },
+    {
+      "epoch": 0.4112572142559648,
+      "grad_norm": 0.4192275317369746,
+      "learning_rate": 9.214207554303492e-06,
+      "loss": 0.3232,
+      "step": 1728
+    },
+    {
+      "epoch": 0.4114952103290296,
+      "grad_norm": 0.39895282478246985,
+      "learning_rate": 9.213169883604776e-06,
+      "loss": 0.4003,
+      "step": 1729
+    },
+    {
+      "epoch": 0.4117332064020944,
+      "grad_norm": 0.4354043645900782,
+      "learning_rate": 9.212131586724787e-06,
+      "loss": 0.3975,
+      "step": 1730
+    },
+    {
+      "epoch": 0.41197120247515917,
+      "grad_norm": 0.4120078987709819,
+      "learning_rate": 9.211092663817839e-06,
+      "loss": 0.3097,
+      "step": 1731
+    },
+    {
+      "epoch": 0.41220919854822397,
+      "grad_norm": 0.389265114227728,
+      "learning_rate": 9.210053115038345e-06,
+      "loss": 0.3601,
+      "step": 1732
+    },
+    {
+      "epoch": 0.41244719462128876,
+      "grad_norm": 0.5583703836161701,
+      "learning_rate": 9.209012940540806e-06,
+      "loss": 0.4214,
+      "step": 1733
+    },
+    {
+      "epoch": 0.41268519069435355,
+      "grad_norm": 0.3795249796311343,
+      "learning_rate": 9.207972140479817e-06,
+      "loss": 0.3467,
+      "step": 1734
+    },
+    {
+      "epoch": 0.41292318676741835,
+      "grad_norm": 0.37887544114873867,
+      "learning_rate": 9.206930715010069e-06,
+      "loss": 0.3176,
+      "step": 1735
+    },
+    {
+      "epoch": 0.41316118284048314,
+      "grad_norm": 0.3908429675593901,
+      "learning_rate": 9.205888664286343e-06,
+      "loss": 0.3866,
+      "step": 1736
+    },
+    {
+      "epoch": 0.41339917891354794,
+      "grad_norm": 0.4053593309349944,
+      "learning_rate": 9.20484598846351e-06,
+      "loss": 0.4207,
+      "step": 1737
+    },
+    {
+      "epoch": 0.41363717498661273,
+      "grad_norm": 0.40431227924333274,
+      "learning_rate": 9.203802687696543e-06,
+      "loss": 0.2908,
+      "step": 1738
+    },
+    {
+      "epoch": 0.4138751710596775,
+      "grad_norm": 0.41350971082279164,
+      "learning_rate": 9.2027587621405e-06,
+      "loss": 0.3968,
+      "step": 1739
+    },
+    {
+      "epoch": 0.4141131671327423,
+      "grad_norm": 0.37481649448861415,
+      "learning_rate": 9.201714211950532e-06,
+      "loss": 0.4003,
+      "step": 1740
+    },
+    {
+      "epoch": 0.4143511632058071,
+      "grad_norm": 0.3800874685713264,
+      "learning_rate": 9.200669037281888e-06,
+      "loss": 0.3346,
+      "step": 1741
+    },
+    {
+      "epoch": 0.4145891592788719,
+      "grad_norm": 0.4080865424809329,
+      "learning_rate": 9.199623238289903e-06,
+      "loss": 0.3263,
+      "step": 1742
+    },
+    {
+      "epoch": 0.4148271553519367,
+      "grad_norm": 0.38620720186828605,
+      "learning_rate": 9.198576815130013e-06,
+      "loss": 0.3869,
+      "step": 1743
+    },
+    {
+      "epoch": 0.4150651514250015,
+      "grad_norm": 0.42012676664637677,
+      "learning_rate": 9.197529767957742e-06,
+      "loss": 0.4334,
+      "step": 1744
+    },
+    {
+      "epoch": 0.4153031474980663,
+      "grad_norm": 0.44141968095141493,
+      "learning_rate": 9.196482096928702e-06,
+      "loss": 0.3451,
+      "step": 1745
+    },
+    {
+      "epoch": 0.4155411435711311,
+      "grad_norm": 0.3868284133888196,
+      "learning_rate": 9.19543380219861e-06,
+      "loss": 0.3473,
+      "step": 1746
+    },
+    {
+      "epoch": 0.4157791396441959,
+      "grad_norm": 0.3881735751224557,
+      "learning_rate": 9.194384883923262e-06,
+      "loss": 0.4139,
+      "step": 1747
+    },
+    {
+      "epoch": 0.4160171357172607,
+      "grad_norm": 0.4022882687609258,
+      "learning_rate": 9.193335342258558e-06,
+      "loss": 0.3668,
+      "step": 1748
+    },
+    {
+      "epoch": 0.41625513179032547,
+      "grad_norm": 0.36310020776516483,
+      "learning_rate": 9.192285177360482e-06,
+      "loss": 0.3155,
+      "step": 1749
+    },
+    {
+      "epoch": 0.41649312786339027,
+      "grad_norm": 0.3999886426578984,
+      "learning_rate": 9.191234389385119e-06,
+      "loss": 0.3557,
+      "step": 1750
+    },
+    {
+      "epoch": 0.41673112393645506,
+      "grad_norm": 0.3616937980400372,
+      "learning_rate": 9.19018297848864e-06,
+      "loss": 0.4089,
+      "step": 1751
+    },
+    {
+      "epoch": 0.41696912000951986,
+      "grad_norm": 0.3834384131011674,
+      "learning_rate": 9.189130944827308e-06,
+      "loss": 0.3247,
+      "step": 1752
+    },
+    {
+      "epoch": 0.41720711608258465,
+      "grad_norm": 0.38320709923324137,
+      "learning_rate": 9.188078288557485e-06,
+      "loss": 0.3301,
+      "step": 1753
+    },
+    {
+      "epoch": 0.41744511215564944,
+      "grad_norm": 0.38222131872575443,
+      "learning_rate": 9.18702500983562e-06,
+      "loss": 0.4085,
+      "step": 1754
+    },
+    {
+      "epoch": 0.41768310822871424,
+      "grad_norm": 0.3824160404741556,
+      "learning_rate": 9.185971108818254e-06,
+      "loss": 0.4017,
+      "step": 1755
+    },
+    {
+      "epoch": 0.41792110430177903,
+      "grad_norm": 0.36169585842835983,
+      "learning_rate": 9.184916585662029e-06,
+      "loss": 0.3161,
+      "step": 1756
+    },
+    {
+      "epoch": 0.41815910037484383,
+      "grad_norm": 0.372151927298045,
+      "learning_rate": 9.183861440523667e-06,
+      "loss": 0.386,
+      "step": 1757
+    },
+    {
+      "epoch": 0.4183970964479086,
+      "grad_norm": 0.39912606436044784,
+      "learning_rate": 9.182805673559993e-06,
+      "loss": 0.4162,
+      "step": 1758
+    },
+    {
+      "epoch": 0.4186350925209734,
+      "grad_norm": 0.4088578633797585,
+      "learning_rate": 9.181749284927917e-06,
+      "loss": 0.3041,
+      "step": 1759
+    },
+    {
+      "epoch": 0.4188730885940382,
+      "grad_norm": 0.3872367130929281,
+      "learning_rate": 9.180692274784445e-06,
+      "loss": 0.3263,
+      "step": 1760
+    },
+    {
+      "epoch": 0.419111084667103,
+      "grad_norm": 0.3909524702284084,
+      "learning_rate": 9.179634643286677e-06,
+      "loss": 0.4072,
+      "step": 1761
+    },
+    {
+      "epoch": 0.4193490807401678,
+      "grad_norm": 0.39296104928130404,
+      "learning_rate": 9.178576390591803e-06,
+      "loss": 0.4156,
+      "step": 1762
+    },
+    {
+      "epoch": 0.4195870768132326,
+      "grad_norm": 0.39592073252457005,
+      "learning_rate": 9.177517516857102e-06,
+      "loss": 0.3354,
+      "step": 1763
+    },
+    {
+      "epoch": 0.4198250728862974,
+      "grad_norm": 0.4058493679855248,
+      "learning_rate": 9.176458022239954e-06,
+      "loss": 0.3626,
+      "step": 1764
+    },
+    {
+      "epoch": 0.4200630689593622,
+      "grad_norm": 0.3971663194653234,
+      "learning_rate": 9.175397906897821e-06,
+      "loss": 0.4147,
+      "step": 1765
+    },
+    {
+      "epoch": 0.420301065032427,
+      "grad_norm": 0.39799197552647586,
+      "learning_rate": 9.174337170988265e-06,
+      "loss": 0.3773,
+      "step": 1766
+    },
+    {
+      "epoch": 0.4205390611054918,
+      "grad_norm": 0.3674682210725563,
+      "learning_rate": 9.173275814668937e-06,
+      "loss": 0.3324,
+      "step": 1767
+    },
+    {
+      "epoch": 0.42077705717855657,
+      "grad_norm": 0.4437292514829072,
+      "learning_rate": 9.17221383809758e-06,
+      "loss": 0.3865,
+      "step": 1768
+    },
+    {
+      "epoch": 0.42101505325162136,
+      "grad_norm": 0.3802556994977754,
+      "learning_rate": 9.171151241432034e-06,
+      "loss": 0.4228,
+      "step": 1769
+    },
+    {
+      "epoch": 0.42125304932468616,
+      "grad_norm": 0.4701590991382848,
+      "learning_rate": 9.170088024830223e-06,
+      "loss": 0.3363,
+      "step": 1770
+    },
+    {
+      "epoch": 0.42149104539775095,
+      "grad_norm": 0.40536949953318796,
+      "learning_rate": 9.169024188450169e-06,
+      "loss": 0.3709,
+      "step": 1771
+    },
+    {
+      "epoch": 0.42172904147081575,
+      "grad_norm": 0.36959719177247113,
+      "learning_rate": 9.167959732449983e-06,
+      "loss": 0.407,
+      "step": 1772
+    },
+    {
+      "epoch": 0.42196703754388054,
+      "grad_norm": 0.4019427295156592,
+      "learning_rate": 9.16689465698787e-06,
+      "loss": 0.387,
+      "step": 1773
+    },
+    {
+      "epoch": 0.42220503361694534,
+      "grad_norm": 0.3795424586658338,
+      "learning_rate": 9.165828962222128e-06,
+      "loss": 0.3037,
+      "step": 1774
+    },
+    {
+      "epoch": 0.42244302969001013,
+      "grad_norm": 0.41841783370483765,
+      "learning_rate": 9.164762648311142e-06,
+      "loss": 0.3775,
+      "step": 1775
+    },
+    {
+      "epoch": 0.4226810257630749,
+      "grad_norm": 0.3622382328620447,
+      "learning_rate": 9.163695715413399e-06,
+      "loss": 0.3942,
+      "step": 1776
+    },
+    {
+      "epoch": 0.4229190218361397,
+      "grad_norm": 0.36992410907411066,
+      "learning_rate": 9.162628163687466e-06,
+      "loss": 0.3483,
+      "step": 1777
+    },
+    {
+      "epoch": 0.4231570179092045,
+      "grad_norm": 0.3847751360001051,
+      "learning_rate": 9.16155999329201e-06,
+      "loss": 0.3104,
+      "step": 1778
+    },
+    {
+      "epoch": 0.4233950139822693,
+      "grad_norm": 0.395246830815519,
+      "learning_rate": 9.160491204385786e-06,
+      "loss": 0.3947,
+      "step": 1779
+    },
+    {
+      "epoch": 0.4236330100553341,
+      "grad_norm": 0.38146783859148997,
+      "learning_rate": 9.159421797127643e-06,
+      "loss": 0.4296,
+      "step": 1780
+    },
+    {
+      "epoch": 0.4238710061283989,
+      "grad_norm": 0.4112992635868304,
+      "learning_rate": 9.158351771676523e-06,
+      "loss": 0.3382,
+      "step": 1781
+    },
+    {
+      "epoch": 0.4241090022014637,
+      "grad_norm": 0.36930776652809816,
+      "learning_rate": 9.157281128191458e-06,
+      "loss": 0.3593,
+      "step": 1782
+    },
+    {
+      "epoch": 0.4243469982745285,
+      "grad_norm": 0.3908100344030319,
+      "learning_rate": 9.156209866831568e-06,
+      "loss": 0.3928,
+      "step": 1783
+    },
+    {
+      "epoch": 0.4245849943475933,
+      "grad_norm": 0.36549619200757544,
+      "learning_rate": 9.155137987756075e-06,
+      "loss": 0.3465,
+      "step": 1784
+    },
+    {
+      "epoch": 0.4248229904206581,
+      "grad_norm": 0.3849426680408924,
+      "learning_rate": 9.154065491124284e-06,
+      "loss": 0.2999,
+      "step": 1785
+    },
+    {
+      "epoch": 0.42506098649372287,
+      "grad_norm": 0.3782937649081525,
+      "learning_rate": 9.152992377095594e-06,
+      "loss": 0.3892,
+      "step": 1786
+    },
+    {
+      "epoch": 0.42529898256678766,
+      "grad_norm": 0.4181419981500838,
+      "learning_rate": 9.151918645829495e-06,
+      "loss": 0.4155,
+      "step": 1787
+    },
+    {
+      "epoch": 0.42553697863985246,
+      "grad_norm": 0.40443588328082936,
+      "learning_rate": 9.150844297485573e-06,
+      "loss": 0.3487,
+      "step": 1788
+    },
+    {
+      "epoch": 0.42577497471291725,
+      "grad_norm": 0.4037548245270624,
+      "learning_rate": 9.149769332223502e-06,
+      "loss": 0.3502,
+      "step": 1789
+    },
+    {
+      "epoch": 0.42601297078598205,
+      "grad_norm": 0.37518033203980755,
+      "learning_rate": 9.148693750203046e-06,
+      "loss": 0.4035,
+      "step": 1790
+    },
+    {
+      "epoch": 0.42625096685904684,
+      "grad_norm": 0.4112116522406348,
+      "learning_rate": 9.147617551584066e-06,
+      "loss": 0.3702,
+      "step": 1791
+    },
+    {
+      "epoch": 0.42648896293211164,
+      "grad_norm": 0.403896574510639,
+      "learning_rate": 9.14654073652651e-06,
+      "loss": 0.3408,
+      "step": 1792
+    },
+    {
+      "epoch": 0.42672695900517643,
+      "grad_norm": 0.3757284298168529,
+      "learning_rate": 9.145463305190422e-06,
+      "loss": 0.3561,
+      "step": 1793
+    },
+    {
+      "epoch": 0.4269649550782412,
+      "grad_norm": 0.40025611305787007,
+      "learning_rate": 9.144385257735931e-06,
+      "loss": 0.407,
+      "step": 1794
+    },
+    {
+      "epoch": 0.427202951151306,
+      "grad_norm": 0.3825822712921448,
+      "learning_rate": 9.143306594323265e-06,
+      "loss": 0.3361,
+      "step": 1795
+    },
+    {
+      "epoch": 0.4274409472243708,
+      "grad_norm": 0.3979473641423701,
+      "learning_rate": 9.142227315112739e-06,
+      "loss": 0.3523,
+      "step": 1796
+    },
+    {
+      "epoch": 0.4276789432974356,
+      "grad_norm": 0.37047270991043285,
+      "learning_rate": 9.141147420264758e-06,
+      "loss": 0.3966,
+      "step": 1797
+    },
+    {
+      "epoch": 0.4279169393705004,
+      "grad_norm": 0.45550159986727345,
+      "learning_rate": 9.140066909939824e-06,
+      "loss": 0.3841,
+      "step": 1798
+    },
+    {
+      "epoch": 0.4281549354435652,
+      "grad_norm": 0.4044470476748523,
+      "learning_rate": 9.138985784298528e-06,
+      "loss": 0.3111,
+      "step": 1799
+    },
+    {
+      "epoch": 0.42839293151663,
+      "grad_norm": 0.41668001069625243,
+      "learning_rate": 9.13790404350155e-06,
+      "loss": 0.3681,
+      "step": 1800
+    },
+    {
+      "epoch": 0.4286309275896948,
+      "grad_norm": 0.36553533604315613,
+      "learning_rate": 9.136821687709664e-06,
+      "loss": 0.4334,
+      "step": 1801
+    },
+    {
+      "epoch": 0.4288689236627596,
+      "grad_norm": 0.37044586892206033,
+      "learning_rate": 9.135738717083738e-06,
+      "loss": 0.3309,
+      "step": 1802
+    },
+    {
+      "epoch": 0.4291069197358244,
+      "grad_norm": 0.40087470340481024,
+      "learning_rate": 9.134655131784723e-06,
+      "loss": 0.3188,
+      "step": 1803
+    },
+    {
+      "epoch": 0.42934491580888917,
+      "grad_norm": 0.35972133195166445,
+      "learning_rate": 9.133570931973668e-06,
+      "loss": 0.4269,
+      "step": 1804
+    },
+    {
+      "epoch": 0.42958291188195397,
+      "grad_norm": 0.4047830499287121,
+      "learning_rate": 9.132486117811715e-06,
+      "loss": 0.3887,
+      "step": 1805
+    },
+    {
+      "epoch": 0.42982090795501876,
+      "grad_norm": 0.372292118382629,
+      "learning_rate": 9.131400689460091e-06,
+      "loss": 0.3166,
+      "step": 1806
+    },
+    {
+      "epoch": 0.43005890402808356,
+      "grad_norm": 0.3801968581448093,
+      "learning_rate": 9.130314647080118e-06,
+      "loss": 0.3544,
+      "step": 1807
+    },
+    {
+      "epoch": 0.43029690010114835,
+      "grad_norm": 0.41390771036953183,
+      "learning_rate": 9.129227990833212e-06,
+      "loss": 0.4201,
+      "step": 1808
+    },
+    {
+      "epoch": 0.43053489617421314,
+      "grad_norm": 0.37114026857947074,
+      "learning_rate": 9.128140720880872e-06,
+      "loss": 0.3168,
+      "step": 1809
+    },
+    {
+      "epoch": 0.43077289224727794,
+      "grad_norm": 0.37943566827382635,
+      "learning_rate": 9.127052837384696e-06,
+      "loss": 0.3383,
+      "step": 1810
+    },
+    {
+      "epoch": 0.43101088832034273,
+      "grad_norm": 0.37039949864005134,
+      "learning_rate": 9.125964340506371e-06,
+      "loss": 0.3921,
+      "step": 1811
+    },
+    {
+      "epoch": 0.43124888439340753,
+      "grad_norm": 0.4009046077092876,
+      "learning_rate": 9.12487523040767e-06,
+      "loss": 0.4055,
+      "step": 1812
+    },
+    {
+      "epoch": 0.4314868804664723,
+      "grad_norm": 0.415891186146942,
+      "learning_rate": 9.12378550725047e-06,
+      "loss": 0.3205,
+      "step": 1813
+    },
+    {
+      "epoch": 0.4317248765395371,
+      "grad_norm": 0.39458867165479744,
+      "learning_rate": 9.122695171196724e-06,
+      "loss": 0.3616,
+      "step": 1814
+    },
+    {
+      "epoch": 0.4319628726126019,
+      "grad_norm": 0.36931772090998816,
+      "learning_rate": 9.121604222408484e-06,
+      "loss": 0.4181,
+      "step": 1815
+    },
+    {
+      "epoch": 0.4322008686856667,
+      "grad_norm": 0.4095231673961699,
+      "learning_rate": 9.120512661047895e-06,
+      "loss": 0.3514,
+      "step": 1816
+    },
+    {
+      "epoch": 0.4324388647587315,
+      "grad_norm": 0.4938169837089172,
+      "learning_rate": 9.119420487277186e-06,
+      "loss": 0.3169,
+      "step": 1817
+    },
+    {
+      "epoch": 0.4326768608317963,
+      "grad_norm": 0.41916255433374716,
+      "learning_rate": 9.118327701258685e-06,
+      "loss": 0.3944,
+      "step": 1818
+    },
+    {
+      "epoch": 0.4329148569048611,
+      "grad_norm": 0.40908642674760587,
+      "learning_rate": 9.117234303154806e-06,
+      "loss": 0.4355,
+      "step": 1819
+    },
+    {
+      "epoch": 0.4331528529779259,
+      "grad_norm": 0.3857608329887888,
+      "learning_rate": 9.11614029312805e-06,
+      "loss": 0.3259,
+      "step": 1820
+    },
+    {
+      "epoch": 0.4333908490509907,
+      "grad_norm": 0.4067680142939106,
+      "learning_rate": 9.11504567134102e-06,
+      "loss": 0.3557,
+      "step": 1821
+    },
+    {
+      "epoch": 0.4336288451240555,
+      "grad_norm": 0.37531284613422866,
+      "learning_rate": 9.113950437956403e-06,
+      "loss": 0.4032,
+      "step": 1822
+    },
+    {
+      "epoch": 0.43386684119712027,
+      "grad_norm": 0.39254495182058463,
+      "learning_rate": 9.112854593136976e-06,
+      "loss": 0.3923,
+      "step": 1823
+    },
+    {
+      "epoch": 0.43410483727018506,
+      "grad_norm": 0.4058157753281052,
+      "learning_rate": 9.111758137045609e-06,
+      "loss": 0.3266,
+      "step": 1824
+    },
+    {
+      "epoch": 0.43434283334324986,
+      "grad_norm": 0.36694701438635535,
+      "learning_rate": 9.110661069845263e-06,
+      "loss": 0.3684,
+      "step": 1825
+    },
+    {
+      "epoch": 0.43458082941631465,
+      "grad_norm": 0.36461266098587286,
+      "learning_rate": 9.10956339169899e-06,
+      "loss": 0.4108,
+      "step": 1826
+    },
+    {
+      "epoch": 0.43481882548937945,
+      "grad_norm": 0.39631841763000103,
+      "learning_rate": 9.10846510276993e-06,
+      "loss": 0.337,
+      "step": 1827
+    },
+    {
+      "epoch": 0.43505682156244424,
+      "grad_norm": 0.40343286423290825,
+      "learning_rate": 9.107366203221318e-06,
+      "loss": 0.344,
+      "step": 1828
+    },
+    {
+      "epoch": 0.43529481763550903,
+      "grad_norm": 0.36284461096995285,
+      "learning_rate": 9.106266693216477e-06,
+      "loss": 0.3818,
+      "step": 1829
+    },
+    {
+      "epoch": 0.43553281370857383,
+      "grad_norm": 0.42037283100640777,
+      "learning_rate": 9.10516657291882e-06,
+      "loss": 0.4134,
+      "step": 1830
+    },
+    {
+      "epoch": 0.4357708097816386,
+      "grad_norm": 0.4135556119364222,
+      "learning_rate": 9.104065842491854e-06,
+      "loss": 0.3355,
+      "step": 1831
+    },
+    {
+      "epoch": 0.4360088058547034,
+      "grad_norm": 0.5512601557394179,
+      "learning_rate": 9.102964502099175e-06,
+      "loss": 0.3644,
+      "step": 1832
+    },
+    {
+      "epoch": 0.4362468019277682,
+      "grad_norm": 0.3726460220199828,
+      "learning_rate": 9.101862551904467e-06,
+      "loss": 0.4127,
+      "step": 1833
+    },
+    {
+      "epoch": 0.436484798000833,
+      "grad_norm": 0.3889721147844331,
+      "learning_rate": 9.100759992071509e-06,
+      "loss": 0.3612,
+      "step": 1834
+    },
+    {
+      "epoch": 0.4367227940738978,
+      "grad_norm": 0.3827603318719862,
+      "learning_rate": 9.099656822764169e-06,
+      "loss": 0.2957,
+      "step": 1835
+    },
+    {
+      "epoch": 0.4369607901469626,
+      "grad_norm": 0.3637594579871201,
+      "learning_rate": 9.098553044146404e-06,
+      "loss": 0.3518,
+      "step": 1836
+    },
+    {
+      "epoch": 0.4371987862200274,
+      "grad_norm": 0.41257520091473093,
+      "learning_rate": 9.097448656382263e-06,
+      "loss": 0.4387,
+      "step": 1837
+    },
+    {
+      "epoch": 0.4374367822930922,
+      "grad_norm": 0.43864336163698847,
+      "learning_rate": 9.096343659635887e-06,
+      "loss": 0.3276,
+      "step": 1838
+    },
+    {
+      "epoch": 0.437674778366157,
+      "grad_norm": 0.3841715752329595,
+      "learning_rate": 9.095238054071505e-06,
+      "loss": 0.3416,
+      "step": 1839
+    },
+    {
+      "epoch": 0.4379127744392218,
+      "grad_norm": 0.370655417628842,
+      "learning_rate": 9.094131839853435e-06,
+      "loss": 0.3964,
+      "step": 1840
+    },
+    {
+      "epoch": 0.43815077051228657,
+      "grad_norm": 0.3513470462073168,
+      "learning_rate": 9.093025017146089e-06,
+      "loss": 0.3362,
+      "step": 1841
+    },
+    {
+      "epoch": 0.43838876658535136,
+      "grad_norm": 0.3802770131312354,
+      "learning_rate": 9.09191758611397e-06,
+      "loss": 0.3301,
+      "step": 1842
+    },
+    {
+      "epoch": 0.43862676265841616,
+      "grad_norm": 0.4226138218606067,
+      "learning_rate": 9.09080954692167e-06,
+      "loss": 0.3848,
+      "step": 1843
+    },
+    {
+      "epoch": 0.43886475873148095,
+      "grad_norm": 0.4535187719262108,
+      "learning_rate": 9.089700899733867e-06,
+      "loss": 0.4167,
+      "step": 1844
+    },
+    {
+      "epoch": 0.43910275480454575,
+      "grad_norm": 0.3702844760500393,
+      "learning_rate": 9.088591644715338e-06,
+      "loss": 0.3273,
+      "step": 1845
+    },
+    {
+      "epoch": 0.43934075087761054,
+      "grad_norm": 0.4038109651897949,
+      "learning_rate": 9.087481782030943e-06,
+      "loss": 0.3614,
+      "step": 1846
+    },
+    {
+      "epoch": 0.43957874695067534,
+      "grad_norm": 0.40604586580793806,
+      "learning_rate": 9.086371311845636e-06,
+      "loss": 0.4053,
+      "step": 1847
+    },
+    {
+      "epoch": 0.43981674302374013,
+      "grad_norm": 0.421721246391553,
+      "learning_rate": 9.08526023432446e-06,
+      "loss": 0.3641,
+      "step": 1848
+    },
+    {
+      "epoch": 0.4400547390968049,
+      "grad_norm": 0.38810442228840714,
+      "learning_rate": 9.084148549632547e-06,
+      "loss": 0.3291,
+      "step": 1849
+    },
+    {
+      "epoch": 0.4402927351698697,
+      "grad_norm": 0.4074211523419894,
+      "learning_rate": 9.083036257935125e-06,
+      "loss": 0.3818,
+      "step": 1850
+    },
+    {
+      "epoch": 0.4405307312429345,
+      "grad_norm": 0.3979947415081816,
+      "learning_rate": 9.081923359397504e-06,
+      "loss": 0.4428,
+      "step": 1851
+    },
+    {
+      "epoch": 0.4407687273159993,
+      "grad_norm": 0.4180778911918767,
+      "learning_rate": 9.080809854185091e-06,
+      "loss": 0.3094,
+      "step": 1852
+    },
+    {
+      "epoch": 0.4410067233890641,
+      "grad_norm": 0.44372997553941645,
+      "learning_rate": 9.07969574246338e-06,
+      "loss": 0.3535,
+      "step": 1853
+    },
+    {
+      "epoch": 0.4412447194621289,
+      "grad_norm": 0.41225588433239624,
+      "learning_rate": 9.078581024397952e-06,
+      "loss": 0.4067,
+      "step": 1854
+    },
+    {
+      "epoch": 0.4414827155351937,
+      "grad_norm": 0.4001859550814012,
+      "learning_rate": 9.077465700154487e-06,
+      "loss": 0.433,
+      "step": 1855
+    },
+    {
+      "epoch": 0.4417207116082585,
+      "grad_norm": 0.4167341678602161,
+      "learning_rate": 9.076349769898746e-06,
+      "loss": 0.3567,
+      "step": 1856
+    },
+    {
+      "epoch": 0.4419587076813233,
+      "grad_norm": 0.40392402299201724,
+      "learning_rate": 9.075233233796585e-06,
+      "loss": 0.3763,
+      "step": 1857
+    },
+    {
+      "epoch": 0.4421967037543881,
+      "grad_norm": 0.362779786516785,
+      "learning_rate": 9.074116092013952e-06,
+      "loss": 0.3903,
+      "step": 1858
+    },
+    {
+      "epoch": 0.44243469982745287,
+      "grad_norm": 0.3781623110186848,
+      "learning_rate": 9.072998344716875e-06,
+      "loss": 0.3343,
+      "step": 1859
+    },
+    {
+      "epoch": 0.44267269590051767,
+      "grad_norm": 0.38836103267593736,
+      "learning_rate": 9.071879992071484e-06,
+      "loss": 0.3055,
+      "step": 1860
+    },
+    {
+      "epoch": 0.44291069197358246,
+      "grad_norm": 0.37564508597971813,
+      "learning_rate": 9.070761034243995e-06,
+      "loss": 0.4162,
+      "step": 1861
+    },
+    {
+      "epoch": 0.44314868804664725,
+      "grad_norm": 0.381826072613129,
+      "learning_rate": 9.069641471400707e-06,
+      "loss": 0.3939,
+      "step": 1862
+    },
+    {
+      "epoch": 0.44338668411971205,
+      "grad_norm": 0.41067999322559867,
+      "learning_rate": 9.06852130370802e-06,
+      "loss": 0.314,
+      "step": 1863
+    },
+    {
+      "epoch": 0.44362468019277684,
+      "grad_norm": 0.4382314242050576,
+      "learning_rate": 9.067400531332418e-06,
+      "loss": 0.3397,
+      "step": 1864
+    },
+    {
+      "epoch": 0.44386267626584164,
+      "grad_norm": 0.3671784649407471,
+      "learning_rate": 9.066279154440474e-06,
+      "loss": 0.4134,
+      "step": 1865
+    },
+    {
+      "epoch": 0.44410067233890643,
+      "grad_norm": 0.38775612781990293,
+      "learning_rate": 9.065157173198852e-06,
+      "loss": 0.3365,
+      "step": 1866
+    },
+    {
+      "epoch": 0.4443386684119712,
+      "grad_norm": 0.399921745750612,
+      "learning_rate": 9.064034587774307e-06,
+      "loss": 0.3262,
+      "step": 1867
+    },
+    {
+      "epoch": 0.444576664485036,
+      "grad_norm": 0.3883120453391914,
+      "learning_rate": 9.062911398333682e-06,
+      "loss": 0.3783,
+      "step": 1868
+    },
+    {
+      "epoch": 0.4448146605581008,
+      "grad_norm": 0.42591617134823834,
+      "learning_rate": 9.061787605043913e-06,
+      "loss": 0.4329,
+      "step": 1869
+    },
+    {
+      "epoch": 0.4450526566311656,
+      "grad_norm": 0.4145202625576308,
+      "learning_rate": 9.060663208072022e-06,
+      "loss": 0.319,
+      "step": 1870
+    },
+    {
+      "epoch": 0.4452906527042304,
+      "grad_norm": 0.4256230711123469,
+      "learning_rate": 9.059538207585123e-06,
+      "loss": 0.3498,
+      "step": 1871
+    },
+    {
+      "epoch": 0.4455286487772952,
+      "grad_norm": 0.36200201893045636,
+      "learning_rate": 9.058412603750417e-06,
+      "loss": 0.4051,
+      "step": 1872
+    },
+    {
+      "epoch": 0.44576664485036,
+      "grad_norm": 0.4088077633298198,
+      "learning_rate": 9.057286396735198e-06,
+      "loss": 0.352,
+      "step": 1873
+    },
+    {
+      "epoch": 0.4460046409234248,
+      "grad_norm": 0.411767781342609,
+      "learning_rate": 9.056159586706847e-06,
+      "loss": 0.3159,
+      "step": 1874
+    },
+    {
+      "epoch": 0.4462426369964896,
+      "grad_norm": 0.38770515767105623,
+      "learning_rate": 9.055032173832838e-06,
+      "loss": 0.3435,
+      "step": 1875
+    },
+    {
+      "epoch": 0.4464806330695544,
+      "grad_norm": 0.4063020777640675,
+      "learning_rate": 9.053904158280731e-06,
+      "loss": 0.448,
+      "step": 1876
+    },
+    {
+      "epoch": 0.4467186291426192,
+      "grad_norm": 0.4108973315393291,
+      "learning_rate": 9.052775540218178e-06,
+      "loss": 0.3257,
+      "step": 1877
+    },
+    {
+      "epoch": 0.44695662521568397,
+      "grad_norm": 0.3962899590466375,
+      "learning_rate": 9.05164631981292e-06,
+      "loss": 0.3428,
+      "step": 1878
+    },
+    {
+      "epoch": 0.44719462128874876,
+      "grad_norm": 0.39154457859950453,
+      "learning_rate": 9.050516497232783e-06,
+      "loss": 0.4113,
+      "step": 1879
+    },
+    {
+      "epoch": 0.44743261736181356,
+      "grad_norm": 0.40521710300411384,
+      "learning_rate": 9.049386072645691e-06,
+      "loss": 0.3697,
+      "step": 1880
+    },
+    {
+      "epoch": 0.44767061343487835,
+      "grad_norm": 0.41448812938566526,
+      "learning_rate": 9.048255046219652e-06,
+      "loss": 0.3128,
+      "step": 1881
+    },
+    {
+      "epoch": 0.44790860950794315,
+      "grad_norm": 0.3732301223230006,
+      "learning_rate": 9.047123418122762e-06,
+      "loss": 0.3486,
+      "step": 1882
+    },
+    {
+      "epoch": 0.44814660558100794,
+      "grad_norm": 0.37645904480775877,
+      "learning_rate": 9.045991188523213e-06,
+      "loss": 0.4007,
+      "step": 1883
+    },
+    {
+      "epoch": 0.44838460165407273,
+      "grad_norm": 0.5589491366235092,
+      "learning_rate": 9.044858357589281e-06,
+      "loss": 0.3507,
+      "step": 1884
+    },
+    {
+      "epoch": 0.44862259772713753,
+      "grad_norm": 0.41677787585285414,
+      "learning_rate": 9.043724925489332e-06,
+      "loss": 0.3181,
+      "step": 1885
+    },
+    {
+      "epoch": 0.4488605938002023,
+      "grad_norm": 0.41414556643068295,
+      "learning_rate": 9.04259089239182e-06,
+      "loss": 0.4107,
+      "step": 1886
+    },
+    {
+      "epoch": 0.4490985898732671,
+      "grad_norm": 0.43327968208641654,
+      "learning_rate": 9.041456258465295e-06,
+      "loss": 0.4257,
+      "step": 1887
+    },
+    {
+      "epoch": 0.4493365859463319,
+      "grad_norm": 0.4245883014008134,
+      "learning_rate": 9.040321023878387e-06,
+      "loss": 0.3388,
+      "step": 1888
+    },
+    {
+      "epoch": 0.4495745820193967,
+      "grad_norm": 0.4421642723801354,
+      "learning_rate": 9.039185188799824e-06,
+      "loss": 0.3743,
+      "step": 1889
+    },
+    {
+      "epoch": 0.4498125780924615,
+      "grad_norm": 0.41567634181571017,
+      "learning_rate": 9.038048753398417e-06,
+      "loss": 0.4091,
+      "step": 1890
+    },
+    {
+      "epoch": 0.4500505741655263,
+      "grad_norm": 0.4133338214658035,
+      "learning_rate": 9.036911717843067e-06,
+      "loss": 0.3367,
+      "step": 1891
+    },
+    {
+      "epoch": 0.4502885702385911,
+      "grad_norm": 0.41679245676442966,
+      "learning_rate": 9.035774082302769e-06,
+      "loss": 0.3222,
+      "step": 1892
+    },
+    {
+      "epoch": 0.4505265663116559,
+      "grad_norm": 0.38537142529458723,
+      "learning_rate": 9.034635846946603e-06,
+      "loss": 0.3554,
+      "step": 1893
+    },
+    {
+      "epoch": 0.4507645623847207,
+      "grad_norm": 0.4135663142912014,
+      "learning_rate": 9.033497011943735e-06,
+      "loss": 0.4212,
+      "step": 1894
+    },
+    {
+      "epoch": 0.4510025584577855,
+      "grad_norm": 0.39185654824504884,
+      "learning_rate": 9.032357577463429e-06,
+      "loss": 0.3207,
+      "step": 1895
+    },
+    {
+      "epoch": 0.45124055453085027,
+      "grad_norm": 0.39385062485453165,
+      "learning_rate": 9.031217543675032e-06,
+      "loss": 0.33,
+      "step": 1896
+    },
+    {
+      "epoch": 0.45147855060391506,
+      "grad_norm": 0.41350024044695805,
+      "learning_rate": 9.03007691074798e-06,
+      "loss": 0.395,
+      "step": 1897
+    },
+    {
+      "epoch": 0.4517165466769798,
+      "grad_norm": 0.4048615581588889,
+      "learning_rate": 9.028935678851798e-06,
+      "loss": 0.3942,
+      "step": 1898
+    },
+    {
+      "epoch": 0.4519545427500446,
+      "grad_norm": 0.43192810625467914,
+      "learning_rate": 9.027793848156106e-06,
+      "loss": 0.2972,
+      "step": 1899
+    },
+    {
+      "epoch": 0.4521925388231094,
+      "grad_norm": 0.3638638789244606,
+      "learning_rate": 9.026651418830603e-06,
+      "loss": 0.3737,
+      "step": 1900
+    },
+    {
+      "epoch": 0.4524305348961742,
+      "grad_norm": 0.4178434063849281,
+      "learning_rate": 9.025508391045087e-06,
+      "loss": 0.4463,
+      "step": 1901
+    },
+    {
+      "epoch": 0.452668530969239,
+      "grad_norm": 0.43266874725956067,
+      "learning_rate": 9.024364764969435e-06,
+      "loss": 0.354,
+      "step": 1902
+    },
+    {
+      "epoch": 0.4529065270423038,
+      "grad_norm": 0.4386170116690222,
+      "learning_rate": 9.023220540773621e-06,
+      "loss": 0.3436,
+      "step": 1903
+    },
+    {
+      "epoch": 0.45314452311536857,
+      "grad_norm": 0.4154738684824685,
+      "learning_rate": 9.022075718627707e-06,
+      "loss": 0.3934,
+      "step": 1904
+    },
+    {
+      "epoch": 0.45338251918843336,
+      "grad_norm": 0.4336014698946849,
+      "learning_rate": 9.02093029870184e-06,
+      "loss": 0.376,
+      "step": 1905
+    },
+    {
+      "epoch": 0.45362051526149816,
+      "grad_norm": 0.4496211440899709,
+      "learning_rate": 9.019784281166255e-06,
+      "loss": 0.3017,
+      "step": 1906
+    },
+    {
+      "epoch": 0.45385851133456295,
+      "grad_norm": 0.438757831396024,
+      "learning_rate": 9.018637666191284e-06,
+      "loss": 0.3907,
+      "step": 1907
+    },
+    {
+      "epoch": 0.45409650740762775,
+      "grad_norm": 0.38217794195897165,
+      "learning_rate": 9.017490453947337e-06,
+      "loss": 0.4368,
+      "step": 1908
+    },
+    {
+      "epoch": 0.45433450348069254,
+      "grad_norm": 0.4045541745421685,
+      "learning_rate": 9.016342644604923e-06,
+      "loss": 0.3241,
+      "step": 1909
+    },
+    {
+      "epoch": 0.45457249955375734,
+      "grad_norm": 0.4749105178045131,
+      "learning_rate": 9.01519423833463e-06,
+      "loss": 0.3236,
+      "step": 1910
+    },
+    {
+      "epoch": 0.45481049562682213,
+      "grad_norm": 0.39441695061972637,
+      "learning_rate": 9.014045235307144e-06,
+      "loss": 0.3832,
+      "step": 1911
+    },
+    {
+      "epoch": 0.4550484916998869,
+      "grad_norm": 0.4006478226656663,
+      "learning_rate": 9.012895635693232e-06,
+      "loss": 0.3885,
+      "step": 1912
+    },
+    {
+      "epoch": 0.4552864877729517,
+      "grad_norm": 0.42235636070301646,
+      "learning_rate": 9.011745439663756e-06,
+      "loss": 0.3053,
+      "step": 1913
+    },
+    {
+      "epoch": 0.4555244838460165,
+      "grad_norm": 0.4330833366027642,
+      "learning_rate": 9.010594647389662e-06,
+      "loss": 0.3426,
+      "step": 1914
+    },
+    {
+      "epoch": 0.4557624799190813,
+      "grad_norm": 0.3935945827408916,
+      "learning_rate": 9.009443259041984e-06,
+      "loss": 0.438,
+      "step": 1915
+    },
+    {
+      "epoch": 0.4560004759921461,
+      "grad_norm": 0.4573862811142753,
+      "learning_rate": 9.008291274791849e-06,
+      "loss": 0.3511,
+      "step": 1916
+    },
+    {
+      "epoch": 0.4562384720652109,
+      "grad_norm": 0.4191822242792251,
+      "learning_rate": 9.00713869481047e-06,
+      "loss": 0.3424,
+      "step": 1917
+    },
+    {
+      "epoch": 0.4564764681382757,
+      "grad_norm": 0.3714154576618009,
+      "learning_rate": 9.005985519269151e-06,
+      "loss": 0.3758,
+      "step": 1918
+    },
+    {
+      "epoch": 0.4567144642113405,
+      "grad_norm": 0.3968866803688529,
+      "learning_rate": 9.00483174833928e-06,
+      "loss": 0.4399,
+      "step": 1919
+    },
+    {
+      "epoch": 0.4569524602844053,
+      "grad_norm": 0.4381863859987972,
+      "learning_rate": 9.003677382192337e-06,
+      "loss": 0.3317,
+      "step": 1920
+    },
+    {
+      "epoch": 0.4571904563574701,
+      "grad_norm": 0.3668542092189725,
+      "learning_rate": 9.002522420999887e-06,
+      "loss": 0.3302,
+      "step": 1921
+    },
+    {
+      "epoch": 0.45742845243053487,
+      "grad_norm": 0.40073715252923153,
+      "learning_rate": 9.00136686493359e-06,
+      "loss": 0.3946,
+      "step": 1922
+    },
+    {
+      "epoch": 0.45766644850359967,
+      "grad_norm": 0.422377185704182,
+      "learning_rate": 9.000210714165185e-06,
+      "loss": 0.3721,
+      "step": 1923
+    },
+    {
+      "epoch": 0.45790444457666446,
+      "grad_norm": 0.39892724787175154,
+      "learning_rate": 8.999053968866509e-06,
+      "loss": 0.3285,
+      "step": 1924
+    },
+    {
+      "epoch": 0.45814244064972925,
+      "grad_norm": 0.39420537454241245,
+      "learning_rate": 8.997896629209482e-06,
+      "loss": 0.3954,
+      "step": 1925
+    },
+    {
+      "epoch": 0.45838043672279405,
+      "grad_norm": 0.3875725349434963,
+      "learning_rate": 8.996738695366111e-06,
+      "loss": 0.4209,
+      "step": 1926
+    },
+    {
+      "epoch": 0.45861843279585884,
+      "grad_norm": 0.4320566083850361,
+      "learning_rate": 8.995580167508495e-06,
+      "loss": 0.3111,
+      "step": 1927
+    },
+    {
+      "epoch": 0.45885642886892364,
+      "grad_norm": 0.39786821137659234,
+      "learning_rate": 8.994421045808821e-06,
+      "loss": 0.3344,
+      "step": 1928
+    },
+    {
+      "epoch": 0.45909442494198843,
+      "grad_norm": 0.3864397258868153,
+      "learning_rate": 8.993261330439365e-06,
+      "loss": 0.4114,
+      "step": 1929
+    },
+    {
+      "epoch": 0.4593324210150532,
+      "grad_norm": 0.44366727708416975,
+      "learning_rate": 8.992101021572483e-06,
+      "loss": 0.3847,
+      "step": 1930
+    },
+    {
+      "epoch": 0.459570417088118,
+      "grad_norm": 0.42879150308085556,
+      "learning_rate": 8.990940119380632e-06,
+      "loss": 0.3251,
+      "step": 1931
+    },
+    {
+      "epoch": 0.4598084131611828,
+      "grad_norm": 0.4013669319373018,
+      "learning_rate": 8.989778624036346e-06,
+      "loss": 0.3751,
+      "step": 1932
+    },
+    {
+      "epoch": 0.4600464092342476,
+      "grad_norm": 0.38329016484024186,
+      "learning_rate": 8.988616535712255e-06,
+      "loss": 0.4322,
+      "step": 1933
+    },
+    {
+      "epoch": 0.4602844053073124,
+      "grad_norm": 0.4593810052561611,
+      "learning_rate": 8.987453854581074e-06,
+      "loss": 0.351,
+      "step": 1934
+    },
+    {
+      "epoch": 0.4605224013803772,
+      "grad_norm": 0.4290257106099543,
+      "learning_rate": 8.986290580815605e-06,
+      "loss": 0.2975,
+      "step": 1935
+    },
+    {
+      "epoch": 0.460760397453442,
+      "grad_norm": 0.38235899911292076,
+      "learning_rate": 8.985126714588739e-06,
+      "loss": 0.4136,
+      "step": 1936
+    },
+    {
+      "epoch": 0.4609983935265068,
+      "grad_norm": 0.38940541074163637,
+      "learning_rate": 8.983962256073457e-06,
+      "loss": 0.4249,
+      "step": 1937
+    },
+    {
+      "epoch": 0.4612363895995716,
+      "grad_norm": 0.3599022107051358,
+      "learning_rate": 8.982797205442823e-06,
+      "loss": 0.3004,
+      "step": 1938
+    },
+    {
+      "epoch": 0.4614743856726364,
+      "grad_norm": 0.3859999702393283,
+      "learning_rate": 8.981631562869997e-06,
+      "loss": 0.3493,
+      "step": 1939
+    },
+    {
+      "epoch": 0.4617123817457012,
+      "grad_norm": 0.38132124586744975,
+      "learning_rate": 8.98046532852822e-06,
+      "loss": 0.4082,
+      "step": 1940
+    },
+    {
+      "epoch": 0.46195037781876597,
+      "grad_norm": 0.40954534803570225,
+      "learning_rate": 8.979298502590821e-06,
+      "loss": 0.3792,
+      "step": 1941
+    },
+    {
+      "epoch": 0.46218837389183076,
+      "grad_norm": 0.3763058419790421,
+      "learning_rate": 8.978131085231223e-06,
+      "loss": 0.314,
+      "step": 1942
+    },
+    {
+      "epoch": 0.46242636996489556,
+      "grad_norm": 0.3801916014322026,
+      "learning_rate": 8.976963076622932e-06,
+      "loss": 0.3608,
+      "step": 1943
+    },
+    {
+      "epoch": 0.46266436603796035,
+      "grad_norm": 0.3754880817697566,
+      "learning_rate": 8.975794476939541e-06,
+      "loss": 0.4133,
+      "step": 1944
+    },
+    {
+      "epoch": 0.46290236211102515,
+      "grad_norm": 0.4458270346361261,
+      "learning_rate": 8.974625286354735e-06,
+      "loss": 0.3359,
+      "step": 1945
+    },
+    {
+      "epoch": 0.46314035818408994,
+      "grad_norm": 0.40596951932507613,
+      "learning_rate": 8.973455505042285e-06,
+      "loss": 0.36,
+      "step": 1946
+    },
+    {
+      "epoch": 0.46337835425715473,
+      "grad_norm": 0.3777591671479931,
+      "learning_rate": 8.972285133176047e-06,
+      "loss": 0.4053,
+      "step": 1947
+    },
+    {
+      "epoch": 0.46361635033021953,
+      "grad_norm": 0.381883104188769,
+      "learning_rate": 8.97111417092997e-06,
+      "loss": 0.3788,
+      "step": 1948
+    },
+    {
+      "epoch": 0.4638543464032843,
+      "grad_norm": 0.41602474134267337,
+      "learning_rate": 8.969942618478085e-06,
+      "loss": 0.3184,
+      "step": 1949
+    },
+    {
+      "epoch": 0.4640923424763491,
+      "grad_norm": 0.3838946466607663,
+      "learning_rate": 8.968770475994514e-06,
+      "loss": 0.3573,
+      "step": 1950
+    },
+    {
+      "epoch": 0.4643303385494139,
+      "grad_norm": 0.3860312821497556,
+      "learning_rate": 8.967597743653471e-06,
+      "loss": 0.4234,
+      "step": 1951
+    },
+    {
+      "epoch": 0.4645683346224787,
+      "grad_norm": 0.36977923888724,
+      "learning_rate": 8.966424421629247e-06,
+      "loss": 0.3242,
+      "step": 1952
+    },
+    {
+      "epoch": 0.4648063306955435,
+      "grad_norm": 0.3937693643527784,
+      "learning_rate": 8.965250510096231e-06,
+      "loss": 0.3117,
+      "step": 1953
+    },
+    {
+      "epoch": 0.4650443267686083,
+      "grad_norm": 0.38742510667851454,
+      "learning_rate": 8.964076009228892e-06,
+      "loss": 0.3899,
+      "step": 1954
+    },
+    {
+      "epoch": 0.4652823228416731,
+      "grad_norm": 0.381510778099069,
+      "learning_rate": 8.962900919201793e-06,
+      "loss": 0.363,
+      "step": 1955
+    },
+    {
+      "epoch": 0.4655203189147379,
+      "grad_norm": 0.4266911809851305,
+      "learning_rate": 8.96172524018958e-06,
+      "loss": 0.3424,
+      "step": 1956
+    },
+    {
+      "epoch": 0.4657583149878027,
+      "grad_norm": 0.39273456552864333,
+      "learning_rate": 8.960548972366987e-06,
+      "loss": 0.3623,
+      "step": 1957
+    },
+    {
+      "epoch": 0.4659963110608675,
+      "grad_norm": 0.39550307752848174,
+      "learning_rate": 8.959372115908838e-06,
+      "loss": 0.4073,
+      "step": 1958
+    },
+    {
+      "epoch": 0.46623430713393227,
+      "grad_norm": 0.3939424995482473,
+      "learning_rate": 8.958194670990043e-06,
+      "loss": 0.3466,
+      "step": 1959
+    },
+    {
+      "epoch": 0.46647230320699706,
+      "grad_norm": 0.4163291327720836,
+      "learning_rate": 8.957016637785599e-06,
+      "loss": 0.3167,
+      "step": 1960
+    },
+    {
+      "epoch": 0.46671029928006186,
+      "grad_norm": 0.3744209884685097,
+      "learning_rate": 8.95583801647059e-06,
+      "loss": 0.3617,
+      "step": 1961
+    },
+    {
+      "epoch": 0.46694829535312665,
+      "grad_norm": 0.36438599515039355,
+      "learning_rate": 8.954658807220189e-06,
+      "loss": 0.3979,
+      "step": 1962
+    },
+    {
+      "epoch": 0.46718629142619145,
+      "grad_norm": 0.36878304664624034,
+      "learning_rate": 8.953479010209655e-06,
+      "loss": 0.3402,
+      "step": 1963
+    },
+    {
+      "epoch": 0.46742428749925624,
+      "grad_norm": 0.38784861609124194,
+      "learning_rate": 8.952298625614335e-06,
+      "loss": 0.3465,
+      "step": 1964
+    },
+    {
+      "epoch": 0.46766228357232104,
+      "grad_norm": 0.36193075171069583,
+      "learning_rate": 8.951117653609666e-06,
+      "loss": 0.4162,
+      "step": 1965
+    },
+    {
+      "epoch": 0.46790027964538583,
+      "grad_norm": 0.4273533447324452,
+      "learning_rate": 8.949936094371168e-06,
+      "loss": 0.357,
+      "step": 1966
+    },
+    {
+      "epoch": 0.4681382757184506,
+      "grad_norm": 0.3748626683477407,
+      "learning_rate": 8.948753948074448e-06,
+      "loss": 0.3286,
+      "step": 1967
+    },
+    {
+      "epoch": 0.4683762717915154,
+      "grad_norm": 0.3615946348473227,
+      "learning_rate": 8.947571214895206e-06,
+      "loss": 0.374,
+      "step": 1968
+    },
+    {
+      "epoch": 0.4686142678645802,
+      "grad_norm": 0.40199712164294776,
+      "learning_rate": 8.946387895009221e-06,
+      "loss": 0.4383,
+      "step": 1969
+    },
+    {
+      "epoch": 0.468852263937645,
+      "grad_norm": 0.40189899947360347,
+      "learning_rate": 8.945203988592368e-06,
+      "loss": 0.3137,
+      "step": 1970
+    },
+    {
+      "epoch": 0.4690902600107098,
+      "grad_norm": 0.42304082101866364,
+      "learning_rate": 8.944019495820602e-06,
+      "loss": 0.3607,
+      "step": 1971
+    },
+    {
+      "epoch": 0.4693282560837746,
+      "grad_norm": 0.3915717145865347,
+      "learning_rate": 8.942834416869967e-06,
+      "loss": 0.4186,
+      "step": 1972
+    },
+    {
+      "epoch": 0.4695662521568394,
+      "grad_norm": 0.3699303309540853,
+      "learning_rate": 8.941648751916598e-06,
+      "loss": 0.3479,
+      "step": 1973
+    },
+    {
+      "epoch": 0.4698042482299042,
+      "grad_norm": 0.3931300817695143,
+      "learning_rate": 8.940462501136712e-06,
+      "loss": 0.3416,
+      "step": 1974
+    },
+    {
+      "epoch": 0.470042244302969,
+      "grad_norm": 0.4174108723423135,
+      "learning_rate": 8.939275664706618e-06,
+      "loss": 0.3809,
+      "step": 1975
+    },
+    {
+      "epoch": 0.4702802403760338,
+      "grad_norm": 0.4017321399357895,
+      "learning_rate": 8.938088242802705e-06,
+      "loss": 0.3999,
+      "step": 1976
+    },
+    {
+      "epoch": 0.47051823644909857,
+      "grad_norm": 0.3796367988353256,
+      "learning_rate": 8.936900235601456e-06,
+      "loss": 0.3103,
+      "step": 1977
+    },
+    {
+      "epoch": 0.47075623252216336,
+      "grad_norm": 0.4341261574035814,
+      "learning_rate": 8.93571164327944e-06,
+      "loss": 0.3357,
+      "step": 1978
+    },
+    {
+      "epoch": 0.47099422859522816,
+      "grad_norm": 0.4070316455404249,
+      "learning_rate": 8.934522466013305e-06,
+      "loss": 0.3898,
+      "step": 1979
+    },
+    {
+      "epoch": 0.47123222466829295,
+      "grad_norm": 0.3974826629955441,
+      "learning_rate": 8.933332703979798e-06,
+      "loss": 0.3888,
+      "step": 1980
+    },
+    {
+      "epoch": 0.47147022074135775,
+      "grad_norm": 0.4184039008610896,
+      "learning_rate": 8.932142357355747e-06,
+      "loss": 0.3078,
+      "step": 1981
+    },
+    {
+      "epoch": 0.47170821681442254,
+      "grad_norm": 0.41469276127770316,
+      "learning_rate": 8.930951426318061e-06,
+      "loss": 0.346,
+      "step": 1982
+    },
+    {
+      "epoch": 0.47194621288748734,
+      "grad_norm": 0.3679140205317907,
+      "learning_rate": 8.929759911043749e-06,
+      "loss": 0.4249,
+      "step": 1983
+    },
+    {
+      "epoch": 0.47218420896055213,
+      "grad_norm": 0.38233800356637293,
+      "learning_rate": 8.928567811709897e-06,
+      "loss": 0.3664,
+      "step": 1984
+    },
+    {
+      "epoch": 0.4724222050336169,
+      "grad_norm": 0.3860957949442366,
+      "learning_rate": 8.927375128493679e-06,
+      "loss": 0.3278,
+      "step": 1985
+    },
+    {
+      "epoch": 0.4726602011066817,
+      "grad_norm": 0.42016968648143477,
+      "learning_rate": 8.92618186157236e-06,
+      "loss": 0.4028,
+      "step": 1986
+    },
+    {
+      "epoch": 0.4728981971797465,
+      "grad_norm": 0.3932672809367911,
+      "learning_rate": 8.924988011123286e-06,
+      "loss": 0.4194,
+      "step": 1987
+    },
+    {
+      "epoch": 0.4731361932528113,
+      "grad_norm": 0.3914761416304697,
+      "learning_rate": 8.923793577323894e-06,
+      "loss": 0.289,
+      "step": 1988
+    },
+    {
+      "epoch": 0.4733741893258761,
+      "grad_norm": 0.41886930971894176,
+      "learning_rate": 8.922598560351705e-06,
+      "loss": 0.3386,
+      "step": 1989
+    },
+    {
+      "epoch": 0.4736121853989409,
+      "grad_norm": 0.40054887442025705,
+      "learning_rate": 8.92140296038433e-06,
+      "loss": 0.4222,
+      "step": 1990
+    },
+    {
+      "epoch": 0.4738501814720057,
+      "grad_norm": 0.37901621734635876,
+      "learning_rate": 8.920206777599467e-06,
+      "loss": 0.3587,
+      "step": 1991
+    },
+    {
+      "epoch": 0.4740881775450705,
+      "grad_norm": 0.37289856531710025,
+      "learning_rate": 8.919010012174894e-06,
+      "loss": 0.3589,
+      "step": 1992
+    },
+    {
+      "epoch": 0.4743261736181353,
+      "grad_norm": 0.37673235179772124,
+      "learning_rate": 8.917812664288481e-06,
+      "loss": 0.3493,
+      "step": 1993
+    },
+    {
+      "epoch": 0.4745641696912001,
+      "grad_norm": 0.40946971737432014,
+      "learning_rate": 8.916614734118184e-06,
+      "loss": 0.4314,
+      "step": 1994
+    },
+    {
+      "epoch": 0.47480216576426487,
+      "grad_norm": 0.3663866235640273,
+      "learning_rate": 8.915416221842045e-06,
+      "loss": 0.3188,
+      "step": 1995
+    },
+    {
+      "epoch": 0.47504016183732967,
+      "grad_norm": 0.3729349793322217,
+      "learning_rate": 8.914217127638194e-06,
+      "loss": 0.33,
+      "step": 1996
+    },
+    {
+      "epoch": 0.47527815791039446,
+      "grad_norm": 0.37819004145885465,
+      "learning_rate": 8.913017451684845e-06,
+      "loss": 0.3864,
+      "step": 1997
+    },
+    {
+      "epoch": 0.47551615398345926,
+      "grad_norm": 0.3816038357588018,
+      "learning_rate": 8.911817194160297e-06,
+      "loss": 0.3707,
+      "step": 1998
+    },
+    {
+      "epoch": 0.47575415005652405,
+      "grad_norm": 0.381594564688525,
+      "learning_rate": 8.910616355242943e-06,
+      "loss": 0.2955,
+      "step": 1999
+    },
+    {
+      "epoch": 0.47599214612958884,
+      "grad_norm": 0.4144362212607731,
+      "learning_rate": 8.909414935111251e-06,
+      "loss": 0.3695,
+      "step": 2000
+    },
+    {
+      "epoch": 0.47623014220265364,
+      "grad_norm": 0.40541513918237165,
+      "learning_rate": 8.908212933943788e-06,
+      "loss": 0.4287,
+      "step": 2001
+    },
+    {
+      "epoch": 0.47646813827571843,
+      "grad_norm": 0.4010091276525552,
+      "learning_rate": 8.907010351919198e-06,
+      "loss": 0.352,
+      "step": 2002
+    },
+    {
+      "epoch": 0.47670613434878323,
+      "grad_norm": 0.37436400202401043,
+      "learning_rate": 8.905807189216216e-06,
+      "loss": 0.3316,
+      "step": 2003
+    },
+    {
+      "epoch": 0.476944130421848,
+      "grad_norm": 0.3610601587615899,
+      "learning_rate": 8.90460344601366e-06,
+      "loss": 0.3977,
+      "step": 2004
+    },
+    {
+      "epoch": 0.4771821264949128,
+      "grad_norm": 0.40390327913148766,
+      "learning_rate": 8.903399122490436e-06,
+      "loss": 0.4071,
+      "step": 2005
+    },
+    {
+      "epoch": 0.4774201225679776,
+      "grad_norm": 0.39707419630426943,
+      "learning_rate": 8.902194218825537e-06,
+      "loss": 0.3261,
+      "step": 2006
+    },
+    {
+      "epoch": 0.4776581186410424,
+      "grad_norm": 0.4064601483285377,
+      "learning_rate": 8.900988735198043e-06,
+      "loss": 0.373,
+      "step": 2007
+    },
+    {
+      "epoch": 0.4778961147141072,
+      "grad_norm": 0.3921311614892764,
+      "learning_rate": 8.899782671787114e-06,
+      "loss": 0.402,
+      "step": 2008
+    },
+    {
+      "epoch": 0.478134110787172,
+      "grad_norm": 0.38829304720202296,
+      "learning_rate": 8.898576028772006e-06,
+      "loss": 0.3733,
+      "step": 2009
+    },
+    {
+      "epoch": 0.4783721068602368,
+      "grad_norm": 0.423689833869032,
+      "learning_rate": 8.897368806332053e-06,
+      "loss": 0.3363,
+      "step": 2010
+    },
+    {
+      "epoch": 0.4786101029333016,
+      "grad_norm": 0.40741104435951725,
+      "learning_rate": 8.896161004646682e-06,
+      "loss": 0.3923,
+      "step": 2011
+    },
+    {
+      "epoch": 0.4788480990063664,
+      "grad_norm": 0.38648441274354406,
+      "learning_rate": 8.894952623895396e-06,
+      "loss": 0.4054,
+      "step": 2012
+    },
+    {
+      "epoch": 0.4790860950794312,
+      "grad_norm": 0.3899353786856898,
+      "learning_rate": 8.893743664257796e-06,
+      "loss": 0.3402,
+      "step": 2013
+    },
+    {
+      "epoch": 0.47932409115249597,
+      "grad_norm": 0.37220325831387135,
+      "learning_rate": 8.892534125913558e-06,
+      "loss": 0.3757,
+      "step": 2014
+    },
+    {
+      "epoch": 0.47956208722556076,
+      "grad_norm": 0.37816661979461397,
+      "learning_rate": 8.891324009042456e-06,
+      "loss": 0.4359,
+      "step": 2015
+    },
+    {
+      "epoch": 0.47980008329862556,
+      "grad_norm": 0.3787006467971716,
+      "learning_rate": 8.890113313824339e-06,
+      "loss": 0.366,
+      "step": 2016
+    },
+    {
+      "epoch": 0.48003807937169035,
+      "grad_norm": 0.36458447262183397,
+      "learning_rate": 8.888902040439145e-06,
+      "loss": 0.3146,
+      "step": 2017
+    },
+    {
+      "epoch": 0.48027607544475515,
+      "grad_norm": 0.39938985399119364,
+      "learning_rate": 8.887690189066899e-06,
+      "loss": 0.3725,
+      "step": 2018
+    },
+    {
+      "epoch": 0.48051407151781994,
+      "grad_norm": 0.3591918893002723,
+      "learning_rate": 8.886477759887717e-06,
+      "loss": 0.3943,
+      "step": 2019
+    },
+    {
+      "epoch": 0.48075206759088474,
+      "grad_norm": 0.36999609920363924,
+      "learning_rate": 8.885264753081794e-06,
+      "loss": 0.3297,
+      "step": 2020
+    },
+    {
+      "epoch": 0.48099006366394953,
+      "grad_norm": 0.407098460962921,
+      "learning_rate": 8.884051168829409e-06,
+      "loss": 0.3605,
+      "step": 2021
+    },
+    {
+      "epoch": 0.4812280597370143,
+      "grad_norm": 0.3847503739932104,
+      "learning_rate": 8.882837007310936e-06,
+      "loss": 0.4087,
+      "step": 2022
+    },
+    {
+      "epoch": 0.4814660558100791,
+      "grad_norm": 0.3953664392850304,
+      "learning_rate": 8.881622268706825e-06,
+      "loss": 0.3865,
+      "step": 2023
+    },
+    {
+      "epoch": 0.4817040518831439,
+      "grad_norm": 0.4139311567463376,
+      "learning_rate": 8.88040695319762e-06,
+      "loss": 0.3024,
+      "step": 2024
+    },
+    {
+      "epoch": 0.4819420479562087,
+      "grad_norm": 0.3915339203989994,
+      "learning_rate": 8.879191060963943e-06,
+      "loss": 0.3784,
+      "step": 2025
+    },
+    {
+      "epoch": 0.4821800440292735,
+      "grad_norm": 0.3600119497807205,
+      "learning_rate": 8.87797459218651e-06,
+      "loss": 0.4668,
+      "step": 2026
+    },
+    {
+      "epoch": 0.4824180401023383,
+      "grad_norm": 0.4115620298847174,
+      "learning_rate": 8.876757547046116e-06,
+      "loss": 0.3102,
+      "step": 2027
+    },
+    {
+      "epoch": 0.4826560361754031,
+      "grad_norm": 0.4194742305434396,
+      "learning_rate": 8.875539925723641e-06,
+      "loss": 0.3474,
+      "step": 2028
+    },
+    {
+      "epoch": 0.4828940322484679,
+      "grad_norm": 0.37165614518858103,
+      "learning_rate": 8.874321728400059e-06,
+      "loss": 0.3922,
+      "step": 2029
+    },
+    {
+      "epoch": 0.4831320283215327,
+      "grad_norm": 0.37833468745165144,
+      "learning_rate": 8.873102955256423e-06,
+      "loss": 0.3875,
+      "step": 2030
+    },
+    {
+      "epoch": 0.4833700243945975,
+      "grad_norm": 0.4131846841190066,
+      "learning_rate": 8.871883606473871e-06,
+      "loss": 0.3283,
+      "step": 2031
+    },
+    {
+      "epoch": 0.48360802046766227,
+      "grad_norm": 0.38823207775132323,
+      "learning_rate": 8.87066368223363e-06,
+      "loss": 0.3778,
+      "step": 2032
+    },
+    {
+      "epoch": 0.48384601654072706,
+      "grad_norm": 0.3691014202938941,
+      "learning_rate": 8.869443182717009e-06,
+      "loss": 0.4339,
+      "step": 2033
+    },
+    {
+      "epoch": 0.48408401261379186,
+      "grad_norm": 0.39215841748503116,
+      "learning_rate": 8.868222108105407e-06,
+      "loss": 0.3159,
+      "step": 2034
+    },
+    {
+      "epoch": 0.48432200868685665,
+      "grad_norm": 0.3882981946824124,
+      "learning_rate": 8.867000458580302e-06,
+      "loss": 0.3129,
+      "step": 2035
+    },
+    {
+      "epoch": 0.48456000475992145,
+      "grad_norm": 0.3918438336109168,
+      "learning_rate": 8.865778234323266e-06,
+      "loss": 0.4061,
+      "step": 2036
+    },
+    {
+      "epoch": 0.48479800083298624,
+      "grad_norm": 0.36841655891359437,
+      "learning_rate": 8.864555435515949e-06,
+      "loss": 0.408,
+      "step": 2037
+    },
+    {
+      "epoch": 0.48503599690605104,
+      "grad_norm": 0.37655472476118107,
+      "learning_rate": 8.863332062340091e-06,
+      "loss": 0.297,
+      "step": 2038
+    },
+    {
+      "epoch": 0.48527399297911583,
+      "grad_norm": 0.3798702816409488,
+      "learning_rate": 8.862108114977512e-06,
+      "loss": 0.3575,
+      "step": 2039
+    },
+    {
+      "epoch": 0.4855119890521806,
+      "grad_norm": 0.36545896885427936,
+      "learning_rate": 8.860883593610126e-06,
+      "loss": 0.4035,
+      "step": 2040
+    },
+    {
+      "epoch": 0.4857499851252454,
+      "grad_norm": 0.35102870721288,
+      "learning_rate": 8.859658498419922e-06,
+      "loss": 0.3564,
+      "step": 2041
+    },
+    {
+      "epoch": 0.4859879811983102,
+      "grad_norm": 0.3576865709579289,
+      "learning_rate": 8.858432829588984e-06,
+      "loss": 0.296,
+      "step": 2042
+    },
+    {
+      "epoch": 0.486225977271375,
+      "grad_norm": 0.36034938518371656,
+      "learning_rate": 8.857206587299471e-06,
+      "loss": 0.3711,
+      "step": 2043
+    },
+    {
+      "epoch": 0.4864639733444398,
+      "grad_norm": 0.4065171435929195,
+      "learning_rate": 8.85597977173364e-06,
+      "loss": 0.4387,
+      "step": 2044
+    },
+    {
+      "epoch": 0.4867019694175046,
+      "grad_norm": 0.3834118830811794,
+      "learning_rate": 8.85475238307382e-06,
+      "loss": 0.2858,
+      "step": 2045
+    },
+    {
+      "epoch": 0.4869399654905694,
+      "grad_norm": 0.3943529386040461,
+      "learning_rate": 8.853524421502436e-06,
+      "loss": 0.3485,
+      "step": 2046
+    },
+    {
+      "epoch": 0.4871779615636342,
+      "grad_norm": 0.4010452585395006,
+      "learning_rate": 8.852295887201988e-06,
+      "loss": 0.4022,
+      "step": 2047
+    },
+    {
+      "epoch": 0.487415957636699,
+      "grad_norm": 0.4092837500931114,
+      "learning_rate": 8.851066780355074e-06,
+      "loss": 0.3798,
+      "step": 2048
+    },
+    {
+      "epoch": 0.4876539537097638,
+      "grad_norm": 0.41343430994676683,
+      "learning_rate": 8.849837101144363e-06,
+      "loss": 0.3158,
+      "step": 2049
+    },
+    {
+      "epoch": 0.48789194978282857,
+      "grad_norm": 0.3737441299895588,
+      "learning_rate": 8.84860684975262e-06,
+      "loss": 0.3551,
+      "step": 2050
+    },
+    {
+      "epoch": 0.48812994585589337,
+      "grad_norm": 0.3745215140069202,
+      "learning_rate": 8.847376026362688e-06,
+      "loss": 0.4366,
+      "step": 2051
+    },
+    {
+      "epoch": 0.48836794192895816,
+      "grad_norm": 0.42064142614172584,
+      "learning_rate": 8.8461446311575e-06,
+      "loss": 0.3181,
+      "step": 2052
+    },
+    {
+      "epoch": 0.48860593800202295,
+      "grad_norm": 0.37945037835962714,
+      "learning_rate": 8.844912664320072e-06,
+      "loss": 0.3606,
+      "step": 2053
+    },
+    {
+      "epoch": 0.48884393407508775,
+      "grad_norm": 0.38167249606508846,
+      "learning_rate": 8.8436801260335e-06,
+      "loss": 0.3845,
+      "step": 2054
+    },
+    {
+      "epoch": 0.48908193014815254,
+      "grad_norm": 0.3746464998144936,
+      "learning_rate": 8.842447016480975e-06,
+      "loss": 0.3902,
+      "step": 2055
+    },
+    {
+      "epoch": 0.48931992622121734,
+      "grad_norm": 0.39703764026174426,
+      "learning_rate": 8.841213335845767e-06,
+      "loss": 0.3346,
+      "step": 2056
+    },
+    {
+      "epoch": 0.48955792229428213,
+      "grad_norm": 0.391301094421723,
+      "learning_rate": 8.839979084311228e-06,
+      "loss": 0.3585,
+      "step": 2057
+    },
+    {
+      "epoch": 0.4897959183673469,
+      "grad_norm": 0.4083430548241745,
+      "learning_rate": 8.8387442620608e-06,
+      "loss": 0.4185,
+      "step": 2058
+    },
+    {
+      "epoch": 0.4900339144404117,
+      "grad_norm": 0.37140532210327554,
+      "learning_rate": 8.837508869278011e-06,
+      "loss": 0.3295,
+      "step": 2059
+    },
+    {
+      "epoch": 0.4902719105134765,
+      "grad_norm": 0.43155618842343063,
+      "learning_rate": 8.836272906146467e-06,
+      "loss": 0.3387,
+      "step": 2060
+    },
+    {
+      "epoch": 0.4905099065865413,
+      "grad_norm": 0.3865957545475916,
+      "learning_rate": 8.835036372849867e-06,
+      "loss": 0.3753,
+      "step": 2061
+    },
+    {
+      "epoch": 0.4907479026596061,
+      "grad_norm": 0.38780511078052327,
+      "learning_rate": 8.833799269571985e-06,
+      "loss": 0.397,
+      "step": 2062
+    },
+    {
+      "epoch": 0.4909858987326709,
+      "grad_norm": 0.4141760815744359,
+      "learning_rate": 8.832561596496689e-06,
+      "loss": 0.3248,
+      "step": 2063
+    },
+    {
+      "epoch": 0.4912238948057357,
+      "grad_norm": 0.4008766368624388,
+      "learning_rate": 8.831323353807928e-06,
+      "loss": 0.342,
+      "step": 2064
+    },
+    {
+      "epoch": 0.4914618908788005,
+      "grad_norm": 0.39406895654952484,
+      "learning_rate": 8.830084541689731e-06,
+      "loss": 0.4475,
+      "step": 2065
+    },
+    {
+      "epoch": 0.4916998869518653,
+      "grad_norm": 0.399495170165796,
+      "learning_rate": 8.828845160326222e-06,
+      "loss": 0.3681,
+      "step": 2066
+    },
+    {
+      "epoch": 0.4919378830249301,
+      "grad_norm": 0.3868671762065268,
+      "learning_rate": 8.827605209901602e-06,
+      "loss": 0.3054,
+      "step": 2067
+    },
+    {
+      "epoch": 0.4921758790979949,
+      "grad_norm": 0.38625669927979583,
+      "learning_rate": 8.826364690600155e-06,
+      "loss": 0.3585,
+      "step": 2068
+    },
+    {
+      "epoch": 0.49241387517105967,
+      "grad_norm": 0.4117055914580924,
+      "learning_rate": 8.825123602606256e-06,
+      "loss": 0.4328,
+      "step": 2069
+    },
+    {
+      "epoch": 0.49265187124412446,
+      "grad_norm": 0.3711886777989222,
+      "learning_rate": 8.82388194610436e-06,
+      "loss": 0.3273,
+      "step": 2070
+    },
+    {
+      "epoch": 0.49288986731718926,
+      "grad_norm": 0.4050709073259558,
+      "learning_rate": 8.82263972127901e-06,
+      "loss": 0.341,
+      "step": 2071
+    },
+    {
+      "epoch": 0.49312786339025405,
+      "grad_norm": 0.39297282044656545,
+      "learning_rate": 8.82139692831483e-06,
+      "loss": 0.3942,
+      "step": 2072
+    },
+    {
+      "epoch": 0.49336585946331885,
+      "grad_norm": 0.4006082152861021,
+      "learning_rate": 8.820153567396528e-06,
+      "loss": 0.3502,
+      "step": 2073
+    },
+    {
+      "epoch": 0.49360385553638364,
+      "grad_norm": 0.4641634293550907,
+      "learning_rate": 8.818909638708901e-06,
+      "loss": 0.3107,
+      "step": 2074
+    },
+    {
+      "epoch": 0.49384185160944843,
+      "grad_norm": 0.3822372933111529,
+      "learning_rate": 8.817665142436826e-06,
+      "loss": 0.3832,
+      "step": 2075
+    },
+    {
+      "epoch": 0.49407984768251323,
+      "grad_norm": 0.36126502781069214,
+      "learning_rate": 8.816420078765267e-06,
+      "loss": 0.4281,
+      "step": 2076
+    },
+    {
+      "epoch": 0.494317843755578,
+      "grad_norm": 0.3805792912065209,
+      "learning_rate": 8.81517444787927e-06,
+      "loss": 0.326,
+      "step": 2077
+    },
+    {
+      "epoch": 0.4945558398286428,
+      "grad_norm": 0.3938968001852603,
+      "learning_rate": 8.813928249963967e-06,
+      "loss": 0.348,
+      "step": 2078
+    },
+    {
+      "epoch": 0.4947938359017076,
+      "grad_norm": 0.38229600062789804,
+      "learning_rate": 8.812681485204575e-06,
+      "loss": 0.4155,
+      "step": 2079
+    },
+    {
+      "epoch": 0.4950318319747724,
+      "grad_norm": 0.4196935982542361,
+      "learning_rate": 8.811434153786392e-06,
+      "loss": 0.377,
+      "step": 2080
+    },
+    {
+      "epoch": 0.4952698280478372,
+      "grad_norm": 0.4143016393191996,
+      "learning_rate": 8.810186255894804e-06,
+      "loss": 0.3329,
+      "step": 2081
+    },
+    {
+      "epoch": 0.495507824120902,
+      "grad_norm": 0.3843063609766728,
+      "learning_rate": 8.808937791715278e-06,
+      "loss": 0.3756,
+      "step": 2082
+    },
+    {
+      "epoch": 0.4957458201939668,
+      "grad_norm": 0.36885161283443524,
+      "learning_rate": 8.807688761433369e-06,
+      "loss": 0.433,
+      "step": 2083
+    },
+    {
+      "epoch": 0.4959838162670316,
+      "grad_norm": 0.36443450880109934,
+      "learning_rate": 8.806439165234711e-06,
+      "loss": 0.3567,
+      "step": 2084
+    },
+    {
+      "epoch": 0.4962218123400964,
+      "grad_norm": 0.38276313847917565,
+      "learning_rate": 8.805189003305026e-06,
+      "loss": 0.3166,
+      "step": 2085
+    },
+    {
+      "epoch": 0.4964598084131612,
+      "grad_norm": 0.38135336290107213,
+      "learning_rate": 8.803938275830122e-06,
+      "loss": 0.3852,
+      "step": 2086
+    },
+    {
+      "epoch": 0.49669780448622597,
+      "grad_norm": 0.3876532972271958,
+      "learning_rate": 8.802686982995882e-06,
+      "loss": 0.394,
+      "step": 2087
+    },
+    {
+      "epoch": 0.49693580055929076,
+      "grad_norm": 0.42237733267854094,
+      "learning_rate": 8.801435124988284e-06,
+      "loss": 0.3224,
+      "step": 2088
+    },
+    {
+      "epoch": 0.49717379663235556,
+      "grad_norm": 0.36357998941592223,
+      "learning_rate": 8.800182701993383e-06,
+      "loss": 0.3489,
+      "step": 2089
+    },
+    {
+      "epoch": 0.49741179270542035,
+      "grad_norm": 0.3776997755369579,
+      "learning_rate": 8.798929714197321e-06,
+      "loss": 0.4065,
+      "step": 2090
+    },
+    {
+      "epoch": 0.49764978877848515,
+      "grad_norm": 0.39974248449207966,
+      "learning_rate": 8.797676161786322e-06,
+      "loss": 0.3716,
+      "step": 2091
+    },
+    {
+      "epoch": 0.49788778485154994,
+      "grad_norm": 0.3512586392644621,
+      "learning_rate": 8.796422044946697e-06,
+      "loss": 0.2985,
+      "step": 2092
+    },
+    {
+      "epoch": 0.49812578092461474,
+      "grad_norm": 0.37530246376072357,
+      "learning_rate": 8.795167363864835e-06,
+      "loss": 0.3765,
+      "step": 2093
+    },
+    {
+      "epoch": 0.49836377699767953,
+      "grad_norm": 0.403359701606089,
+      "learning_rate": 8.793912118727214e-06,
+      "loss": 0.4233,
+      "step": 2094
+    },
+    {
+      "epoch": 0.4986017730707443,
+      "grad_norm": 0.39277782417073076,
+      "learning_rate": 8.792656309720398e-06,
+      "loss": 0.323,
+      "step": 2095
+    },
+    {
+      "epoch": 0.4988397691438091,
+      "grad_norm": 0.3901545968666047,
+      "learning_rate": 8.791399937031027e-06,
+      "loss": 0.3246,
+      "step": 2096
+    },
+    {
+      "epoch": 0.4990777652168739,
+      "grad_norm": 0.4108640260500558,
+      "learning_rate": 8.790143000845832e-06,
+      "loss": 0.4035,
+      "step": 2097
+    },
+    {
+      "epoch": 0.4993157612899387,
+      "grad_norm": 0.43483507105211894,
+      "learning_rate": 8.788885501351622e-06,
+      "loss": 0.3644,
+      "step": 2098
+    },
+    {
+      "epoch": 0.4995537573630035,
+      "grad_norm": 0.3989358061906033,
+      "learning_rate": 8.787627438735295e-06,
+      "loss": 0.3092,
+      "step": 2099
+    },
+    {
+      "epoch": 0.4997917534360683,
+      "grad_norm": 0.39812360364772853,
+      "learning_rate": 8.786368813183829e-06,
+      "loss": 0.3558,
+      "step": 2100
+    },
+    {
+      "epoch": 0.5000297495091331,
+      "grad_norm": 0.404777603959223,
+      "learning_rate": 8.785109624884287e-06,
+      "loss": 0.4303,
+      "step": 2101
+    },
+    {
+      "epoch": 0.5002677455821979,
+      "grad_norm": 0.40005648641359376,
+      "learning_rate": 8.783849874023816e-06,
+      "loss": 0.3249,
+      "step": 2102
+    },
+    {
+      "epoch": 0.5005057416552627,
+      "grad_norm": 0.4387837922924149,
+      "learning_rate": 8.782589560789645e-06,
+      "loss": 0.3341,
+      "step": 2103
+    },
+    {
+      "epoch": 0.5007437377283275,
+      "grad_norm": 0.36939020016413393,
+      "learning_rate": 8.781328685369088e-06,
+      "loss": 0.3658,
+      "step": 2104
+    },
+    {
+      "epoch": 0.5009817338013923,
+      "grad_norm": 0.42034598447699306,
+      "learning_rate": 8.780067247949545e-06,
+      "loss": 0.4267,
+      "step": 2105
+    },
+    {
+      "epoch": 0.5012197298744571,
+      "grad_norm": 0.3744391349615731,
+      "learning_rate": 8.778805248718492e-06,
+      "loss": 0.3169,
+      "step": 2106
+    },
+    {
+      "epoch": 0.5014577259475219,
+      "grad_norm": 0.39455712888275285,
+      "learning_rate": 8.777542687863498e-06,
+      "loss": 0.385,
+      "step": 2107
+    },
+    {
+      "epoch": 0.5016957220205867,
+      "grad_norm": 0.4107559343031647,
+      "learning_rate": 8.776279565572208e-06,
+      "loss": 0.4194,
+      "step": 2108
+    },
+    {
+      "epoch": 0.5019337180936515,
+      "grad_norm": 0.376450072712637,
+      "learning_rate": 8.775015882032355e-06,
+      "loss": 0.3303,
+      "step": 2109
+    },
+    {
+      "epoch": 0.5021717141667162,
+      "grad_norm": 0.39192436430391,
+      "learning_rate": 8.77375163743175e-06,
+      "loss": 0.3186,
+      "step": 2110
+    },
+    {
+      "epoch": 0.5024097102397811,
+      "grad_norm": 0.4103061508772666,
+      "learning_rate": 8.772486831958293e-06,
+      "loss": 0.3535,
+      "step": 2111
+    },
+    {
+      "epoch": 0.5026477063128458,
+      "grad_norm": 0.4061911120479169,
+      "learning_rate": 8.771221465799968e-06,
+      "loss": 0.4365,
+      "step": 2112
+    },
+    {
+      "epoch": 0.5028857023859107,
+      "grad_norm": 0.36093912967050273,
+      "learning_rate": 8.769955539144839e-06,
+      "loss": 0.311,
+      "step": 2113
+    },
+    {
+      "epoch": 0.5031236984589754,
+      "grad_norm": 0.431661016075779,
+      "learning_rate": 8.768689052181051e-06,
+      "loss": 0.3709,
+      "step": 2114
+    },
+    {
+      "epoch": 0.5033616945320403,
+      "grad_norm": 0.37241191111075006,
+      "learning_rate": 8.767422005096838e-06,
+      "loss": 0.4208,
+      "step": 2115
+    },
+    {
+      "epoch": 0.503599690605105,
+      "grad_norm": 0.3900038040150187,
+      "learning_rate": 8.766154398080511e-06,
+      "loss": 0.3656,
+      "step": 2116
+    },
+    {
+      "epoch": 0.5038376866781699,
+      "grad_norm": 0.38734311341707733,
+      "learning_rate": 8.764886231320473e-06,
+      "loss": 0.315,
+      "step": 2117
+    },
+    {
+      "epoch": 0.5040756827512346,
+      "grad_norm": 0.39624067824100434,
+      "learning_rate": 8.7636175050052e-06,
+      "loss": 0.3679,
+      "step": 2118
+    },
+    {
+      "epoch": 0.5043136788242994,
+      "grad_norm": 0.3786762562077862,
+      "learning_rate": 8.76234821932326e-06,
+      "loss": 0.421,
+      "step": 2119
+    },
+    {
+      "epoch": 0.5045516748973642,
+      "grad_norm": 0.3658684215497542,
+      "learning_rate": 8.7610783744633e-06,
+      "loss": 0.3373,
+      "step": 2120
+    },
+    {
+      "epoch": 0.504789670970429,
+      "grad_norm": 0.371242399448406,
+      "learning_rate": 8.759807970614044e-06,
+      "loss": 0.3295,
+      "step": 2121
+    },
+    {
+      "epoch": 0.5050276670434938,
+      "grad_norm": 0.40052728933299747,
+      "learning_rate": 8.758537007964314e-06,
+      "loss": 0.4182,
+      "step": 2122
+    },
+    {
+      "epoch": 0.5052656631165586,
+      "grad_norm": 0.39743565216070303,
+      "learning_rate": 8.757265486703001e-06,
+      "loss": 0.3706,
+      "step": 2123
+    },
+    {
+      "epoch": 0.5055036591896234,
+      "grad_norm": 0.36470933499650765,
+      "learning_rate": 8.75599340701909e-06,
+      "loss": 0.2961,
+      "step": 2124
+    },
+    {
+      "epoch": 0.5057416552626882,
+      "grad_norm": 0.37485254459171946,
+      "learning_rate": 8.754720769101636e-06,
+      "loss": 0.3761,
+      "step": 2125
+    },
+    {
+      "epoch": 0.505979651335753,
+      "grad_norm": 0.37277864735486177,
+      "learning_rate": 8.75344757313979e-06,
+      "loss": 0.4196,
+      "step": 2126
+    },
+    {
+      "epoch": 0.5062176474088178,
+      "grad_norm": 0.39747924989282063,
+      "learning_rate": 8.75217381932278e-06,
+      "loss": 0.3059,
+      "step": 2127
+    },
+    {
+      "epoch": 0.5064556434818825,
+      "grad_norm": 0.3715732881453211,
+      "learning_rate": 8.750899507839913e-06,
+      "loss": 0.322,
+      "step": 2128
+    },
+    {
+      "epoch": 0.5066936395549474,
+      "grad_norm": 0.4053341652611045,
+      "learning_rate": 8.74962463888059e-06,
+      "loss": 0.4358,
+      "step": 2129
+    },
+    {
+      "epoch": 0.5069316356280121,
+      "grad_norm": 0.40118580003340715,
+      "learning_rate": 8.748349212634284e-06,
+      "loss": 0.3766,
+      "step": 2130
+    },
+    {
+      "epoch": 0.507169631701077,
+      "grad_norm": 0.4019091388621105,
+      "learning_rate": 8.747073229290552e-06,
+      "loss": 0.3041,
+      "step": 2131
+    },
+    {
+      "epoch": 0.5074076277741417,
+      "grad_norm": 0.4304176264560313,
+      "learning_rate": 8.745796689039043e-06,
+      "loss": 0.3581,
+      "step": 2132
+    },
+    {
+      "epoch": 0.5076456238472066,
+      "grad_norm": 0.4099748631186,
+      "learning_rate": 8.744519592069479e-06,
+      "loss": 0.4268,
+      "step": 2133
+    },
+    {
+      "epoch": 0.5078836199202713,
+      "grad_norm": 0.40848377251157103,
+      "learning_rate": 8.743241938571667e-06,
+      "loss": 0.3395,
+      "step": 2134
+    },
+    {
+      "epoch": 0.5081216159933362,
+      "grad_norm": 0.38784109641240894,
+      "learning_rate": 8.741963728735502e-06,
+      "loss": 0.3247,
+      "step": 2135
+    },
+    {
+      "epoch": 0.5083596120664009,
+      "grad_norm": 0.387056172080891,
+      "learning_rate": 8.740684962750953e-06,
+      "loss": 0.4079,
+      "step": 2136
+    },
+    {
+      "epoch": 0.5085976081394658,
+      "grad_norm": 0.43968856979286486,
+      "learning_rate": 8.73940564080808e-06,
+      "loss": 0.4165,
+      "step": 2137
+    },
+    {
+      "epoch": 0.5088356042125305,
+      "grad_norm": 0.4069695223863342,
+      "learning_rate": 8.738125763097019e-06,
+      "loss": 0.3109,
+      "step": 2138
+    },
+    {
+      "epoch": 0.5090736002855953,
+      "grad_norm": 0.390822454142334,
+      "learning_rate": 8.736845329807994e-06,
+      "loss": 0.3342,
+      "step": 2139
+    },
+    {
+      "epoch": 0.5093115963586601,
+      "grad_norm": 0.3791678415289624,
+      "learning_rate": 8.735564341131308e-06,
+      "loss": 0.4129,
+      "step": 2140
+    },
+    {
+      "epoch": 0.5095495924317249,
+      "grad_norm": 0.3842175369404934,
+      "learning_rate": 8.734282797257347e-06,
+      "loss": 0.3665,
+      "step": 2141
+    },
+    {
+      "epoch": 0.5097875885047897,
+      "grad_norm": 0.39913401398744197,
+      "learning_rate": 8.733000698376579e-06,
+      "loss": 0.3109,
+      "step": 2142
+    },
+    {
+      "epoch": 0.5100255845778545,
+      "grad_norm": 0.3937291920519653,
+      "learning_rate": 8.73171804467956e-06,
+      "loss": 0.3699,
+      "step": 2143
+    },
+    {
+      "epoch": 0.5102635806509193,
+      "grad_norm": 0.3710499687158456,
+      "learning_rate": 8.73043483635692e-06,
+      "loss": 0.409,
+      "step": 2144
+    },
+    {
+      "epoch": 0.5105015767239841,
+      "grad_norm": 0.4346758984087426,
+      "learning_rate": 8.729151073599376e-06,
+      "loss": 0.3292,
+      "step": 2145
+    },
+    {
+      "epoch": 0.5107395727970488,
+      "grad_norm": 0.40591530956978267,
+      "learning_rate": 8.72786675659773e-06,
+      "loss": 0.3347,
+      "step": 2146
+    },
+    {
+      "epoch": 0.5109775688701137,
+      "grad_norm": 0.37110322190450645,
+      "learning_rate": 8.72658188554286e-06,
+      "loss": 0.431,
+      "step": 2147
+    },
+    {
+      "epoch": 0.5112155649431784,
+      "grad_norm": 0.3762833501337982,
+      "learning_rate": 8.725296460625729e-06,
+      "loss": 0.3465,
+      "step": 2148
+    },
+    {
+      "epoch": 0.5114535610162433,
+      "grad_norm": 0.42166675669869735,
+      "learning_rate": 8.724010482037386e-06,
+      "loss": 0.3255,
+      "step": 2149
+    },
+    {
+      "epoch": 0.511691557089308,
+      "grad_norm": 0.4134995142541755,
+      "learning_rate": 8.722723949968958e-06,
+      "loss": 0.3746,
+      "step": 2150
+    },
+    {
+      "epoch": 0.5119295531623729,
+      "grad_norm": 0.38354990683388107,
+      "learning_rate": 8.721436864611653e-06,
+      "loss": 0.4031,
+      "step": 2151
+    },
+    {
+      "epoch": 0.5121675492354376,
+      "grad_norm": 0.4329904072710047,
+      "learning_rate": 8.720149226156769e-06,
+      "loss": 0.3331,
+      "step": 2152
+    },
+    {
+      "epoch": 0.5124055453085025,
+      "grad_norm": 0.4137030908720267,
+      "learning_rate": 8.718861034795677e-06,
+      "loss": 0.3028,
+      "step": 2153
+    },
+    {
+      "epoch": 0.5126435413815672,
+      "grad_norm": 0.3654824931248918,
+      "learning_rate": 8.717572290719835e-06,
+      "loss": 0.3846,
+      "step": 2154
+    },
+    {
+      "epoch": 0.512881537454632,
+      "grad_norm": 0.422834765499284,
+      "learning_rate": 8.716282994120782e-06,
+      "loss": 0.3844,
+      "step": 2155
+    },
+    {
+      "epoch": 0.5131195335276968,
+      "grad_norm": 0.3922197545354826,
+      "learning_rate": 8.71499314519014e-06,
+      "loss": 0.3464,
+      "step": 2156
+    },
+    {
+      "epoch": 0.5133575296007616,
+      "grad_norm": 0.38918598237414415,
+      "learning_rate": 8.713702744119613e-06,
+      "loss": 0.3361,
+      "step": 2157
+    },
+    {
+      "epoch": 0.5135955256738264,
+      "grad_norm": 0.4107866717596137,
+      "learning_rate": 8.712411791100983e-06,
+      "loss": 0.4021,
+      "step": 2158
+    },
+    {
+      "epoch": 0.5138335217468912,
+      "grad_norm": 0.3901161988399392,
+      "learning_rate": 8.711120286326122e-06,
+      "loss": 0.3225,
+      "step": 2159
+    },
+    {
+      "epoch": 0.514071517819956,
+      "grad_norm": 0.413635276999212,
+      "learning_rate": 8.709828229986978e-06,
+      "loss": 0.3094,
+      "step": 2160
+    },
+    {
+      "epoch": 0.5143095138930208,
+      "grad_norm": 0.37538329913818896,
+      "learning_rate": 8.708535622275581e-06,
+      "loss": 0.4063,
+      "step": 2161
+    },
+    {
+      "epoch": 0.5145475099660856,
+      "grad_norm": 0.37400221595282845,
+      "learning_rate": 8.707242463384046e-06,
+      "loss": 0.4596,
+      "step": 2162
+    },
+    {
+      "epoch": 0.5147855060391504,
+      "grad_norm": 0.40427219364856926,
+      "learning_rate": 8.705948753504569e-06,
+      "loss": 0.3199,
+      "step": 2163
+    },
+    {
+      "epoch": 0.5150235021122151,
+      "grad_norm": 0.3803126661404316,
+      "learning_rate": 8.704654492829428e-06,
+      "loss": 0.3466,
+      "step": 2164
+    },
+    {
+      "epoch": 0.51526149818528,
+      "grad_norm": 0.3755852465058205,
+      "learning_rate": 8.703359681550978e-06,
+      "loss": 0.3883,
+      "step": 2165
+    },
+    {
+      "epoch": 0.5154994942583447,
+      "grad_norm": 0.3908086804440207,
+      "learning_rate": 8.702064319861663e-06,
+      "loss": 0.3674,
+      "step": 2166
+    },
+    {
+      "epoch": 0.5157374903314096,
+      "grad_norm": 0.3771686738573496,
+      "learning_rate": 8.700768407954007e-06,
+      "loss": 0.315,
+      "step": 2167
+    },
+    {
+      "epoch": 0.5159754864044743,
+      "grad_norm": 0.370557512425135,
+      "learning_rate": 8.699471946020612e-06,
+      "loss": 0.3404,
+      "step": 2168
+    },
+    {
+      "epoch": 0.5162134824775391,
+      "grad_norm": 0.3752377056778907,
+      "learning_rate": 8.698174934254164e-06,
+      "loss": 0.4201,
+      "step": 2169
+    },
+    {
+      "epoch": 0.5164514785506039,
+      "grad_norm": 0.4184085294213279,
+      "learning_rate": 8.696877372847434e-06,
+      "loss": 0.3294,
+      "step": 2170
+    },
+    {
+      "epoch": 0.5166894746236687,
+      "grad_norm": 0.388912569698524,
+      "learning_rate": 8.69557926199327e-06,
+      "loss": 0.359,
+      "step": 2171
+    },
+    {
+      "epoch": 0.5169274706967335,
+      "grad_norm": 0.365614463397713,
+      "learning_rate": 8.694280601884603e-06,
+      "loss": 0.4009,
+      "step": 2172
+    },
+    {
+      "epoch": 0.5171654667697982,
+      "grad_norm": 0.42621501649673044,
+      "learning_rate": 8.692981392714445e-06,
+      "loss": 0.36,
+      "step": 2173
+    },
+    {
+      "epoch": 0.5174034628428631,
+      "grad_norm": 0.41391982376423003,
+      "learning_rate": 8.691681634675895e-06,
+      "loss": 0.3199,
+      "step": 2174
+    },
+    {
+      "epoch": 0.5176414589159278,
+      "grad_norm": 0.41365783101678866,
+      "learning_rate": 8.690381327962125e-06,
+      "loss": 0.3919,
+      "step": 2175
+    },
+    {
+      "epoch": 0.5178794549889927,
+      "grad_norm": 0.37354876476127546,
+      "learning_rate": 8.689080472766393e-06,
+      "loss": 0.4031,
+      "step": 2176
+    },
+    {
+      "epoch": 0.5181174510620574,
+      "grad_norm": 0.4203856959561625,
+      "learning_rate": 8.687779069282041e-06,
+      "loss": 0.3208,
+      "step": 2177
+    },
+    {
+      "epoch": 0.5183554471351223,
+      "grad_norm": 0.40423490700726816,
+      "learning_rate": 8.686477117702488e-06,
+      "loss": 0.3334,
+      "step": 2178
+    },
+    {
+      "epoch": 0.518593443208187,
+      "grad_norm": 0.3509157300766899,
+      "learning_rate": 8.685174618221235e-06,
+      "loss": 0.4177,
+      "step": 2179
+    },
+    {
+      "epoch": 0.5188314392812519,
+      "grad_norm": 0.40079035375275884,
+      "learning_rate": 8.683871571031867e-06,
+      "loss": 0.4139,
+      "step": 2180
+    },
+    {
+      "epoch": 0.5190694353543166,
+      "grad_norm": 0.37163196418771477,
+      "learning_rate": 8.68256797632805e-06,
+      "loss": 0.3169,
+      "step": 2181
+    },
+    {
+      "epoch": 0.5193074314273814,
+      "grad_norm": 0.4054280532520997,
+      "learning_rate": 8.681263834303528e-06,
+      "loss": 0.4119,
+      "step": 2182
+    },
+    {
+      "epoch": 0.5195454275004462,
+      "grad_norm": 0.4273379334199248,
+      "learning_rate": 8.67995914515213e-06,
+      "loss": 0.4099,
+      "step": 2183
+    },
+    {
+      "epoch": 0.519783423573511,
+      "grad_norm": 0.38570960886267935,
+      "learning_rate": 8.678653909067767e-06,
+      "loss": 0.3451,
+      "step": 2184
+    },
+    {
+      "epoch": 0.5200214196465758,
+      "grad_norm": 0.39683790687415654,
+      "learning_rate": 8.677348126244427e-06,
+      "loss": 0.316,
+      "step": 2185
+    },
+    {
+      "epoch": 0.5202594157196406,
+      "grad_norm": 0.38461925483513904,
+      "learning_rate": 8.676041796876183e-06,
+      "loss": 0.3881,
+      "step": 2186
+    },
+    {
+      "epoch": 0.5204974117927054,
+      "grad_norm": 0.4254609757724267,
+      "learning_rate": 8.674734921157185e-06,
+      "loss": 0.4075,
+      "step": 2187
+    },
+    {
+      "epoch": 0.5207354078657702,
+      "grad_norm": 0.36619327488390596,
+      "learning_rate": 8.67342749928167e-06,
+      "loss": 0.3159,
+      "step": 2188
+    },
+    {
+      "epoch": 0.520973403938835,
+      "grad_norm": 0.41853879431263746,
+      "learning_rate": 8.672119531443951e-06,
+      "loss": 0.3249,
+      "step": 2189
+    },
+    {
+      "epoch": 0.5212114000118998,
+      "grad_norm": 0.37412197275330067,
+      "learning_rate": 8.67081101783843e-06,
+      "loss": 0.4339,
+      "step": 2190
+    },
+    {
+      "epoch": 0.5214493960849645,
+      "grad_norm": 0.3840297484734256,
+      "learning_rate": 8.669501958659576e-06,
+      "loss": 0.3418,
+      "step": 2191
+    },
+    {
+      "epoch": 0.5216873921580294,
+      "grad_norm": 0.39214557075111384,
+      "learning_rate": 8.668192354101953e-06,
+      "loss": 0.3065,
+      "step": 2192
+    },
+    {
+      "epoch": 0.5219253882310941,
+      "grad_norm": 0.3979212333040952,
+      "learning_rate": 8.666882204360201e-06,
+      "loss": 0.3639,
+      "step": 2193
+    },
+    {
+      "epoch": 0.522163384304159,
+      "grad_norm": 0.43111388602684464,
+      "learning_rate": 8.665571509629038e-06,
+      "loss": 0.3999,
+      "step": 2194
+    },
+    {
+      "epoch": 0.5224013803772237,
+      "grad_norm": 0.39713249026912617,
+      "learning_rate": 8.664260270103265e-06,
+      "loss": 0.329,
+      "step": 2195
+    },
+    {
+      "epoch": 0.5226393764502886,
+      "grad_norm": 0.3899241343272046,
+      "learning_rate": 8.662948485977768e-06,
+      "loss": 0.3436,
+      "step": 2196
+    },
+    {
+      "epoch": 0.5228773725233533,
+      "grad_norm": 0.3939001732463367,
+      "learning_rate": 8.661636157447511e-06,
+      "loss": 0.4149,
+      "step": 2197
+    },
+    {
+      "epoch": 0.5231153685964182,
+      "grad_norm": 0.3957340016317853,
+      "learning_rate": 8.660323284707535e-06,
+      "loss": 0.3787,
+      "step": 2198
+    },
+    {
+      "epoch": 0.5233533646694829,
+      "grad_norm": 0.44601803382908783,
+      "learning_rate": 8.659009867952966e-06,
+      "loss": 0.2919,
+      "step": 2199
+    },
+    {
+      "epoch": 0.5235913607425478,
+      "grad_norm": 0.3932173337224158,
+      "learning_rate": 8.657695907379011e-06,
+      "loss": 0.3767,
+      "step": 2200
+    },
+    {
+      "epoch": 0.5238293568156125,
+      "grad_norm": 0.408961622615314,
+      "learning_rate": 8.65638140318096e-06,
+      "loss": 0.4167,
+      "step": 2201
+    },
+    {
+      "epoch": 0.5240673528886773,
+      "grad_norm": 0.3895863266832368,
+      "learning_rate": 8.655066355554175e-06,
+      "loss": 0.3162,
+      "step": 2202
+    },
+    {
+      "epoch": 0.5243053489617421,
+      "grad_norm": 0.385176418198331,
+      "learning_rate": 8.65375076469411e-06,
+      "loss": 0.323,
+      "step": 2203
+    },
+    {
+      "epoch": 0.5245433450348069,
+      "grad_norm": 0.36711047121100115,
+      "learning_rate": 8.652434630796288e-06,
+      "loss": 0.4063,
+      "step": 2204
+    },
+    {
+      "epoch": 0.5247813411078717,
+      "grad_norm": 0.3996610302929268,
+      "learning_rate": 8.651117954056325e-06,
+      "loss": 0.4035,
+      "step": 2205
+    },
+    {
+      "epoch": 0.5250193371809365,
+      "grad_norm": 0.38507884013978433,
+      "learning_rate": 8.649800734669912e-06,
+      "loss": 0.3275,
+      "step": 2206
+    },
+    {
+      "epoch": 0.5252573332540013,
+      "grad_norm": 0.3649799910820221,
+      "learning_rate": 8.648482972832815e-06,
+      "loss": 0.3651,
+      "step": 2207
+    },
+    {
+      "epoch": 0.5254953293270661,
+      "grad_norm": 0.37617511815362664,
+      "learning_rate": 8.647164668740891e-06,
+      "loss": 0.4303,
+      "step": 2208
+    },
+    {
+      "epoch": 0.5257333254001308,
+      "grad_norm": 0.38432072478719503,
+      "learning_rate": 8.64584582259007e-06,
+      "loss": 0.3724,
+      "step": 2209
+    },
+    {
+      "epoch": 0.5259713214731957,
+      "grad_norm": 0.42322209539590827,
+      "learning_rate": 8.644526434576365e-06,
+      "loss": 0.323,
+      "step": 2210
+    },
+    {
+      "epoch": 0.5262093175462604,
+      "grad_norm": 0.40391854416882916,
+      "learning_rate": 8.64320650489587e-06,
+      "loss": 0.3575,
+      "step": 2211
+    },
+    {
+      "epoch": 0.5264473136193253,
+      "grad_norm": 0.3835559240154108,
+      "learning_rate": 8.641886033744762e-06,
+      "loss": 0.426,
+      "step": 2212
+    },
+    {
+      "epoch": 0.52668530969239,
+      "grad_norm": 0.3936562166696423,
+      "learning_rate": 8.640565021319293e-06,
+      "loss": 0.3247,
+      "step": 2213
+    },
+    {
+      "epoch": 0.5269233057654549,
+      "grad_norm": 0.39202674206549953,
+      "learning_rate": 8.639243467815798e-06,
+      "loss": 0.3597,
+      "step": 2214
+    },
+    {
+      "epoch": 0.5271613018385196,
+      "grad_norm": 0.3783225902955775,
+      "learning_rate": 8.637921373430694e-06,
+      "loss": 0.4475,
+      "step": 2215
+    },
+    {
+      "epoch": 0.5273992979115845,
+      "grad_norm": 0.3600937320401172,
+      "learning_rate": 8.636598738360476e-06,
+      "loss": 0.3617,
+      "step": 2216
+    },
+    {
+      "epoch": 0.5276372939846492,
+      "grad_norm": 0.4160704153133727,
+      "learning_rate": 8.63527556280172e-06,
+      "loss": 0.3446,
+      "step": 2217
+    },
+    {
+      "epoch": 0.527875290057714,
+      "grad_norm": 0.3913242338353506,
+      "learning_rate": 8.633951846951081e-06,
+      "loss": 0.3822,
+      "step": 2218
+    },
+    {
+      "epoch": 0.5281132861307788,
+      "grad_norm": 0.4035152853862847,
+      "learning_rate": 8.6326275910053e-06,
+      "loss": 0.4297,
+      "step": 2219
+    },
+    {
+      "epoch": 0.5283512822038436,
+      "grad_norm": 0.409203322526621,
+      "learning_rate": 8.631302795161192e-06,
+      "loss": 0.3351,
+      "step": 2220
+    },
+    {
+      "epoch": 0.5285892782769084,
+      "grad_norm": 0.39436563517242,
+      "learning_rate": 8.629977459615655e-06,
+      "loss": 0.3404,
+      "step": 2221
+    },
+    {
+      "epoch": 0.5288272743499732,
+      "grad_norm": 0.3590194240170449,
+      "learning_rate": 8.628651584565665e-06,
+      "loss": 0.3969,
+      "step": 2222
+    },
+    {
+      "epoch": 0.529065270423038,
+      "grad_norm": 0.4019072389485261,
+      "learning_rate": 8.627325170208282e-06,
+      "loss": 0.3476,
+      "step": 2223
+    },
+    {
+      "epoch": 0.5293032664961028,
+      "grad_norm": 0.37619114031005646,
+      "learning_rate": 8.625998216740643e-06,
+      "loss": 0.3222,
+      "step": 2224
+    },
+    {
+      "epoch": 0.5295412625691676,
+      "grad_norm": 0.4034865597690215,
+      "learning_rate": 8.624670724359964e-06,
+      "loss": 0.3545,
+      "step": 2225
+    },
+    {
+      "epoch": 0.5297792586422324,
+      "grad_norm": 0.408006482551879,
+      "learning_rate": 8.623342693263549e-06,
+      "loss": 0.4144,
+      "step": 2226
+    },
+    {
+      "epoch": 0.5300172547152971,
+      "grad_norm": 0.38382326688221563,
+      "learning_rate": 8.62201412364877e-06,
+      "loss": 0.3342,
+      "step": 2227
+    },
+    {
+      "epoch": 0.530255250788362,
+      "grad_norm": 0.41258702064324476,
+      "learning_rate": 8.620685015713089e-06,
+      "loss": 0.3147,
+      "step": 2228
+    },
+    {
+      "epoch": 0.5304932468614267,
+      "grad_norm": 0.38204434060333653,
+      "learning_rate": 8.619355369654043e-06,
+      "loss": 0.3818,
+      "step": 2229
+    },
+    {
+      "epoch": 0.5307312429344916,
+      "grad_norm": 0.38216547595409855,
+      "learning_rate": 8.61802518566925e-06,
+      "loss": 0.383,
+      "step": 2230
+    },
+    {
+      "epoch": 0.5309692390075563,
+      "grad_norm": 0.4029134683361265,
+      "learning_rate": 8.616694463956409e-06,
+      "loss": 0.2774,
+      "step": 2231
+    },
+    {
+      "epoch": 0.5312072350806212,
+      "grad_norm": 0.37301434376960896,
+      "learning_rate": 8.615363204713299e-06,
+      "loss": 0.3619,
+      "step": 2232
+    },
+    {
+      "epoch": 0.5314452311536859,
+      "grad_norm": 0.38496277632142917,
+      "learning_rate": 8.614031408137775e-06,
+      "loss": 0.4241,
+      "step": 2233
+    },
+    {
+      "epoch": 0.5316832272267508,
+      "grad_norm": 0.33800484116848534,
+      "learning_rate": 8.612699074427777e-06,
+      "loss": 0.3205,
+      "step": 2234
+    },
+    {
+      "epoch": 0.5319212232998155,
+      "grad_norm": 0.38034462571684113,
+      "learning_rate": 8.611366203781323e-06,
+      "loss": 0.3176,
+      "step": 2235
+    },
+    {
+      "epoch": 0.5321592193728804,
+      "grad_norm": 0.41034018337162753,
+      "learning_rate": 8.610032796396513e-06,
+      "loss": 0.3932,
+      "step": 2236
+    },
+    {
+      "epoch": 0.5323972154459451,
+      "grad_norm": 0.3624140867849283,
+      "learning_rate": 8.60869885247152e-06,
+      "loss": 0.3799,
+      "step": 2237
+    },
+    {
+      "epoch": 0.5326352115190099,
+      "grad_norm": 0.39211384701169555,
+      "learning_rate": 8.607364372204602e-06,
+      "loss": 0.3236,
+      "step": 2238
+    },
+    {
+      "epoch": 0.5328732075920747,
+      "grad_norm": 0.3631336275865908,
+      "learning_rate": 8.606029355794095e-06,
+      "loss": 0.3626,
+      "step": 2239
+    },
+    {
+      "epoch": 0.5331112036651395,
+      "grad_norm": 0.3499747372288337,
+      "learning_rate": 8.604693803438418e-06,
+      "loss": 0.3962,
+      "step": 2240
+    },
+    {
+      "epoch": 0.5333491997382043,
+      "grad_norm": 0.3833522235637884,
+      "learning_rate": 8.603357715336067e-06,
+      "loss": 0.3912,
+      "step": 2241
+    },
+    {
+      "epoch": 0.5335871958112691,
+      "grad_norm": 0.40874050154372005,
+      "learning_rate": 8.602021091685615e-06,
+      "loss": 0.3315,
+      "step": 2242
+    },
+    {
+      "epoch": 0.5338251918843339,
+      "grad_norm": 0.40195329644632777,
+      "learning_rate": 8.600683932685721e-06,
+      "loss": 0.3904,
+      "step": 2243
+    },
+    {
+      "epoch": 0.5340631879573987,
+      "grad_norm": 0.3846575731760141,
+      "learning_rate": 8.599346238535118e-06,
+      "loss": 0.4735,
+      "step": 2244
+    },
+    {
+      "epoch": 0.5343011840304634,
+      "grad_norm": 0.36643018837011293,
+      "learning_rate": 8.59800800943262e-06,
+      "loss": 0.3197,
+      "step": 2245
+    },
+    {
+      "epoch": 0.5345391801035283,
+      "grad_norm": 0.40617951682169934,
+      "learning_rate": 8.596669245577119e-06,
+      "loss": 0.3588,
+      "step": 2246
+    },
+    {
+      "epoch": 0.534777176176593,
+      "grad_norm": 0.3698247677981035,
+      "learning_rate": 8.595329947167593e-06,
+      "loss": 0.4227,
+      "step": 2247
+    },
+    {
+      "epoch": 0.5350151722496579,
+      "grad_norm": 0.3755104238478636,
+      "learning_rate": 8.593990114403093e-06,
+      "loss": 0.3951,
+      "step": 2248
+    },
+    {
+      "epoch": 0.5352531683227226,
+      "grad_norm": 0.37805284252091315,
+      "learning_rate": 8.59264974748275e-06,
+      "loss": 0.2891,
+      "step": 2249
+    },
+    {
+      "epoch": 0.5354911643957875,
+      "grad_norm": 0.37337307896849703,
+      "learning_rate": 8.591308846605777e-06,
+      "loss": 0.3513,
+      "step": 2250
+    },
+    {
+      "epoch": 0.5357291604688522,
+      "grad_norm": 0.3460371741397967,
+      "learning_rate": 8.589967411971464e-06,
+      "loss": 0.4251,
+      "step": 2251
+    },
+    {
+      "epoch": 0.5359671565419171,
+      "grad_norm": 0.37459542754801345,
+      "learning_rate": 8.588625443779183e-06,
+      "loss": 0.3355,
+      "step": 2252
+    },
+    {
+      "epoch": 0.5362051526149818,
+      "grad_norm": 0.40230507957953726,
+      "learning_rate": 8.587282942228382e-06,
+      "loss": 0.34,
+      "step": 2253
+    },
+    {
+      "epoch": 0.5364431486880467,
+      "grad_norm": 0.372308451561018,
+      "learning_rate": 8.585939907518591e-06,
+      "loss": 0.3885,
+      "step": 2254
+    },
+    {
+      "epoch": 0.5366811447611114,
+      "grad_norm": 0.3689804359627597,
+      "learning_rate": 8.584596339849419e-06,
+      "loss": 0.3928,
+      "step": 2255
+    },
+    {
+      "epoch": 0.5369191408341762,
+      "grad_norm": 0.37560524421780966,
+      "learning_rate": 8.583252239420549e-06,
+      "loss": 0.3303,
+      "step": 2256
+    },
+    {
+      "epoch": 0.537157136907241,
+      "grad_norm": 0.3842271683641431,
+      "learning_rate": 8.581907606431754e-06,
+      "loss": 0.3536,
+      "step": 2257
+    },
+    {
+      "epoch": 0.5373951329803058,
+      "grad_norm": 0.4002665238506151,
+      "learning_rate": 8.580562441082876e-06,
+      "loss": 0.4165,
+      "step": 2258
+    },
+    {
+      "epoch": 0.5376331290533706,
+      "grad_norm": 0.38907378275117815,
+      "learning_rate": 8.579216743573839e-06,
+      "loss": 0.2978,
+      "step": 2259
+    },
+    {
+      "epoch": 0.5378711251264354,
+      "grad_norm": 0.3956313105170694,
+      "learning_rate": 8.577870514104651e-06,
+      "loss": 0.3106,
+      "step": 2260
+    },
+    {
+      "epoch": 0.5381091211995002,
+      "grad_norm": 0.39047993680689397,
+      "learning_rate": 8.57652375287539e-06,
+      "loss": 0.3575,
+      "step": 2261
+    },
+    {
+      "epoch": 0.538347117272565,
+      "grad_norm": 0.41768920949511756,
+      "learning_rate": 8.575176460086221e-06,
+      "loss": 0.4167,
+      "step": 2262
+    },
+    {
+      "epoch": 0.5385851133456298,
+      "grad_norm": 0.3870552235120542,
+      "learning_rate": 8.573828635937384e-06,
+      "loss": 0.2996,
+      "step": 2263
+    },
+    {
+      "epoch": 0.5388231094186946,
+      "grad_norm": 0.4596709809606415,
+      "learning_rate": 8.5724802806292e-06,
+      "loss": 0.373,
+      "step": 2264
+    },
+    {
+      "epoch": 0.5390611054917593,
+      "grad_norm": 0.3706967842183925,
+      "learning_rate": 8.571131394362069e-06,
+      "loss": 0.404,
+      "step": 2265
+    },
+    {
+      "epoch": 0.5392991015648242,
+      "grad_norm": 0.40181886987790577,
+      "learning_rate": 8.569781977336464e-06,
+      "loss": 0.3562,
+      "step": 2266
+    },
+    {
+      "epoch": 0.5395370976378889,
+      "grad_norm": 0.4298167654016625,
+      "learning_rate": 8.568432029752947e-06,
+      "loss": 0.3202,
+      "step": 2267
+    },
+    {
+      "epoch": 0.5397750937109538,
+      "grad_norm": 0.39862560057903823,
+      "learning_rate": 8.56708155181215e-06,
+      "loss": 0.3843,
+      "step": 2268
+    },
+    {
+      "epoch": 0.5400130897840185,
+      "grad_norm": 0.41266735819620026,
+      "learning_rate": 8.565730543714791e-06,
+      "loss": 0.4155,
+      "step": 2269
+    },
+    {
+      "epoch": 0.5402510858570834,
+      "grad_norm": 0.3974166536006753,
+      "learning_rate": 8.564379005661661e-06,
+      "loss": 0.3293,
+      "step": 2270
+    },
+    {
+      "epoch": 0.5404890819301481,
+      "grad_norm": 0.35669423871612993,
+      "learning_rate": 8.563026937853633e-06,
+      "loss": 0.3605,
+      "step": 2271
+    },
+    {
+      "epoch": 0.540727078003213,
+      "grad_norm": 0.4127273022491468,
+      "learning_rate": 8.561674340491656e-06,
+      "loss": 0.409,
+      "step": 2272
+    },
+    {
+      "epoch": 0.5409650740762777,
+      "grad_norm": 0.4680110594563534,
+      "learning_rate": 8.56032121377676e-06,
+      "loss": 0.3709,
+      "step": 2273
+    },
+    {
+      "epoch": 0.5412030701493425,
+      "grad_norm": 0.4124449173785681,
+      "learning_rate": 8.558967557910054e-06,
+      "loss": 0.2934,
+      "step": 2274
+    },
+    {
+      "epoch": 0.5414410662224073,
+      "grad_norm": 0.4639894069096805,
+      "learning_rate": 8.557613373092724e-06,
+      "loss": 0.3645,
+      "step": 2275
+    },
+    {
+      "epoch": 0.5416790622954721,
+      "grad_norm": 0.3845035156200861,
+      "learning_rate": 8.556258659526036e-06,
+      "loss": 0.4354,
+      "step": 2276
+    },
+    {
+      "epoch": 0.5419170583685369,
+      "grad_norm": 0.4096705605468606,
+      "learning_rate": 8.554903417411333e-06,
+      "loss": 0.3055,
+      "step": 2277
+    },
+    {
+      "epoch": 0.5421550544416017,
+      "grad_norm": 0.3894517106688162,
+      "learning_rate": 8.553547646950037e-06,
+      "loss": 0.3147,
+      "step": 2278
+    },
+    {
+      "epoch": 0.5423930505146665,
+      "grad_norm": 0.41144516694641703,
+      "learning_rate": 8.552191348343653e-06,
+      "loss": 0.4009,
+      "step": 2279
+    },
+    {
+      "epoch": 0.5426310465877313,
+      "grad_norm": 0.40545518632374106,
+      "learning_rate": 8.550834521793757e-06,
+      "loss": 0.3549,
+      "step": 2280
+    },
+    {
+      "epoch": 0.542869042660796,
+      "grad_norm": 0.4536869892584681,
+      "learning_rate": 8.549477167502006e-06,
+      "loss": 0.2982,
+      "step": 2281
+    },
+    {
+      "epoch": 0.5431070387338609,
+      "grad_norm": 0.36914461803860876,
+      "learning_rate": 8.54811928567014e-06,
+      "loss": 0.3515,
+      "step": 2282
+    },
+    {
+      "epoch": 0.5433450348069256,
+      "grad_norm": 0.38659861108474836,
+      "learning_rate": 8.546760876499968e-06,
+      "loss": 0.4102,
+      "step": 2283
+    },
+    {
+      "epoch": 0.5435830308799905,
+      "grad_norm": 0.40982842576957934,
+      "learning_rate": 8.545401940193392e-06,
+      "loss": 0.3362,
+      "step": 2284
+    },
+    {
+      "epoch": 0.5438210269530552,
+      "grad_norm": 0.3991434411628737,
+      "learning_rate": 8.544042476952377e-06,
+      "loss": 0.3119,
+      "step": 2285
+    },
+    {
+      "epoch": 0.5440590230261201,
+      "grad_norm": 0.3929815486547205,
+      "learning_rate": 8.542682486978973e-06,
+      "loss": 0.3739,
+      "step": 2286
+    },
+    {
+      "epoch": 0.5442970190991848,
+      "grad_norm": 0.42239476535343035,
+      "learning_rate": 8.541321970475312e-06,
+      "loss": 0.3884,
+      "step": 2287
+    },
+    {
+      "epoch": 0.5445350151722497,
+      "grad_norm": 0.4368412726721034,
+      "learning_rate": 8.539960927643596e-06,
+      "loss": 0.3054,
+      "step": 2288
+    },
+    {
+      "epoch": 0.5447730112453144,
+      "grad_norm": 0.39557679784627,
+      "learning_rate": 8.538599358686112e-06,
+      "loss": 0.353,
+      "step": 2289
+    },
+    {
+      "epoch": 0.5450110073183793,
+      "grad_norm": 0.39922321971295105,
+      "learning_rate": 8.537237263805225e-06,
+      "loss": 0.4309,
+      "step": 2290
+    },
+    {
+      "epoch": 0.545249003391444,
+      "grad_norm": 0.39924479565216386,
+      "learning_rate": 8.53587464320337e-06,
+      "loss": 0.3412,
+      "step": 2291
+    },
+    {
+      "epoch": 0.5454869994645088,
+      "grad_norm": 0.39342949827033796,
+      "learning_rate": 8.534511497083073e-06,
+      "loss": 0.2964,
+      "step": 2292
+    },
+    {
+      "epoch": 0.5457249955375736,
+      "grad_norm": 0.37825134436273367,
+      "learning_rate": 8.533147825646925e-06,
+      "loss": 0.3704,
+      "step": 2293
+    },
+    {
+      "epoch": 0.5459629916106384,
+      "grad_norm": 0.38022616425475375,
+      "learning_rate": 8.531783629097608e-06,
+      "loss": 0.405,
+      "step": 2294
+    },
+    {
+      "epoch": 0.5462009876837032,
+      "grad_norm": 0.36260140531944307,
+      "learning_rate": 8.530418907637868e-06,
+      "loss": 0.3217,
+      "step": 2295
+    },
+    {
+      "epoch": 0.546438983756768,
+      "grad_norm": 0.37353441118295333,
+      "learning_rate": 8.529053661470542e-06,
+      "loss": 0.3432,
+      "step": 2296
+    },
+    {
+      "epoch": 0.5466769798298328,
+      "grad_norm": 0.40946792741351434,
+      "learning_rate": 8.527687890798537e-06,
+      "loss": 0.4199,
+      "step": 2297
+    },
+    {
+      "epoch": 0.5469149759028976,
+      "grad_norm": 0.3726039053557827,
+      "learning_rate": 8.52632159582484e-06,
+      "loss": 0.3707,
+      "step": 2298
+    },
+    {
+      "epoch": 0.5471529719759624,
+      "grad_norm": 0.3659302138094137,
+      "learning_rate": 8.524954776752516e-06,
+      "loss": 0.338,
+      "step": 2299
+    },
+    {
+      "epoch": 0.5473909680490272,
+      "grad_norm": 0.3829909878854894,
+      "learning_rate": 8.52358743378471e-06,
+      "loss": 0.3499,
+      "step": 2300
+    },
+    {
+      "epoch": 0.5476289641220919,
+      "grad_norm": 0.3890459899862676,
+      "learning_rate": 8.522219567124643e-06,
+      "loss": 0.4132,
+      "step": 2301
+    },
+    {
+      "epoch": 0.5478669601951568,
+      "grad_norm": 0.36286174283101486,
+      "learning_rate": 8.520851176975612e-06,
+      "loss": 0.35,
+      "step": 2302
+    },
+    {
+      "epoch": 0.5481049562682215,
+      "grad_norm": 0.38768718004776836,
+      "learning_rate": 8.519482263540994e-06,
+      "loss": 0.3365,
+      "step": 2303
+    },
+    {
+      "epoch": 0.5483429523412864,
+      "grad_norm": 0.4160688543785431,
+      "learning_rate": 8.518112827024245e-06,
+      "loss": 0.4039,
+      "step": 2304
+    },
+    {
+      "epoch": 0.5485809484143511,
+      "grad_norm": 0.36929499946058547,
+      "learning_rate": 8.516742867628895e-06,
+      "loss": 0.4113,
+      "step": 2305
+    },
+    {
+      "epoch": 0.548818944487416,
+      "grad_norm": 0.4434563141145236,
+      "learning_rate": 8.515372385558554e-06,
+      "loss": 0.3342,
+      "step": 2306
+    },
+    {
+      "epoch": 0.5490569405604807,
+      "grad_norm": 0.37489035852359814,
+      "learning_rate": 8.514001381016912e-06,
+      "loss": 0.3657,
+      "step": 2307
+    },
+    {
+      "epoch": 0.5492949366335456,
+      "grad_norm": 0.35701192424068157,
+      "learning_rate": 8.512629854207733e-06,
+      "loss": 0.4158,
+      "step": 2308
+    },
+    {
+      "epoch": 0.5495329327066103,
+      "grad_norm": 0.4554656989283448,
+      "learning_rate": 8.511257805334859e-06,
+      "loss": 0.3355,
+      "step": 2309
+    },
+    {
+      "epoch": 0.5497709287796751,
+      "grad_norm": 0.3669973654578771,
+      "learning_rate": 8.509885234602209e-06,
+      "loss": 0.3158,
+      "step": 2310
+    },
+    {
+      "epoch": 0.5500089248527399,
+      "grad_norm": 0.391116918799618,
+      "learning_rate": 8.508512142213784e-06,
+      "loss": 0.3917,
+      "step": 2311
+    },
+    {
+      "epoch": 0.5502469209258047,
+      "grad_norm": 0.40840152784809935,
+      "learning_rate": 8.507138528373658e-06,
+      "loss": 0.4089,
+      "step": 2312
+    },
+    {
+      "epoch": 0.5504849169988695,
+      "grad_norm": 0.39490420844357255,
+      "learning_rate": 8.505764393285985e-06,
+      "loss": 0.3205,
+      "step": 2313
+    },
+    {
+      "epoch": 0.5507229130719343,
+      "grad_norm": 0.3845395660774653,
+      "learning_rate": 8.504389737154994e-06,
+      "loss": 0.391,
+      "step": 2314
+    },
+    {
+      "epoch": 0.5509609091449991,
+      "grad_norm": 0.3909360241094245,
+      "learning_rate": 8.503014560184994e-06,
+      "loss": 0.4145,
+      "step": 2315
+    },
+    {
+      "epoch": 0.5511989052180639,
+      "grad_norm": 0.38762183020768226,
+      "learning_rate": 8.50163886258037e-06,
+      "loss": 0.3754,
+      "step": 2316
+    },
+    {
+      "epoch": 0.5514369012911287,
+      "grad_norm": 0.3912728587608663,
+      "learning_rate": 8.500262644545584e-06,
+      "loss": 0.3224,
+      "step": 2317
+    },
+    {
+      "epoch": 0.5516748973641935,
+      "grad_norm": 0.37973815027865854,
+      "learning_rate": 8.498885906285177e-06,
+      "loss": 0.3811,
+      "step": 2318
+    },
+    {
+      "epoch": 0.5519128934372582,
+      "grad_norm": 0.3579851250106683,
+      "learning_rate": 8.497508648003765e-06,
+      "loss": 0.4085,
+      "step": 2319
+    },
+    {
+      "epoch": 0.5521508895103231,
+      "grad_norm": 0.39660478319056564,
+      "learning_rate": 8.496130869906046e-06,
+      "loss": 0.3407,
+      "step": 2320
+    },
+    {
+      "epoch": 0.5523888855833878,
+      "grad_norm": 0.47401560455711944,
+      "learning_rate": 8.49475257219679e-06,
+      "loss": 0.3387,
+      "step": 2321
+    },
+    {
+      "epoch": 0.5526268816564527,
+      "grad_norm": 0.37354092809224154,
+      "learning_rate": 8.493373755080843e-06,
+      "loss": 0.3914,
+      "step": 2322
+    },
+    {
+      "epoch": 0.5528648777295174,
+      "grad_norm": 0.35292148040523397,
+      "learning_rate": 8.491994418763136e-06,
+      "loss": 0.3585,
+      "step": 2323
+    },
+    {
+      "epoch": 0.5531028738025823,
+      "grad_norm": 0.5613791127606462,
+      "learning_rate": 8.49061456344867e-06,
+      "loss": 0.2989,
+      "step": 2324
+    },
+    {
+      "epoch": 0.553340869875647,
+      "grad_norm": 0.41661056361625665,
+      "learning_rate": 8.489234189342526e-06,
+      "loss": 0.3642,
+      "step": 2325
+    },
+    {
+      "epoch": 0.5535788659487119,
+      "grad_norm": 0.3734315892480904,
+      "learning_rate": 8.487853296649861e-06,
+      "loss": 0.4333,
+      "step": 2326
+    },
+    {
+      "epoch": 0.5538168620217766,
+      "grad_norm": 0.3674453219548101,
+      "learning_rate": 8.486471885575912e-06,
+      "loss": 0.3408,
+      "step": 2327
+    },
+    {
+      "epoch": 0.5540548580948415,
+      "grad_norm": 0.3673572223097309,
+      "learning_rate": 8.48508995632599e-06,
+      "loss": 0.3572,
+      "step": 2328
+    },
+    {
+      "epoch": 0.5542928541679062,
+      "grad_norm": 0.4046956426382339,
+      "learning_rate": 8.483707509105483e-06,
+      "loss": 0.3804,
+      "step": 2329
+    },
+    {
+      "epoch": 0.554530850240971,
+      "grad_norm": 0.35220241891020276,
+      "learning_rate": 8.482324544119858e-06,
+      "loss": 0.3549,
+      "step": 2330
+    },
+    {
+      "epoch": 0.5547688463140358,
+      "grad_norm": 0.4301526345521306,
+      "learning_rate": 8.480941061574656e-06,
+      "loss": 0.3218,
+      "step": 2331
+    },
+    {
+      "epoch": 0.5550068423871006,
+      "grad_norm": 0.38794458883354294,
+      "learning_rate": 8.479557061675498e-06,
+      "loss": 0.3864,
+      "step": 2332
+    },
+    {
+      "epoch": 0.5552448384601654,
+      "grad_norm": 0.37899037555503806,
+      "learning_rate": 8.478172544628082e-06,
+      "loss": 0.4119,
+      "step": 2333
+    },
+    {
+      "epoch": 0.5554828345332302,
+      "grad_norm": 0.4193850656712972,
+      "learning_rate": 8.476787510638179e-06,
+      "loss": 0.362,
+      "step": 2334
+    },
+    {
+      "epoch": 0.555720830606295,
+      "grad_norm": 0.37309432031469925,
+      "learning_rate": 8.47540195991164e-06,
+      "loss": 0.3088,
+      "step": 2335
+    },
+    {
+      "epoch": 0.5559588266793598,
+      "grad_norm": 0.3870361410415046,
+      "learning_rate": 8.474015892654394e-06,
+      "loss": 0.4003,
+      "step": 2336
+    },
+    {
+      "epoch": 0.5561968227524245,
+      "grad_norm": 0.4108399130708448,
+      "learning_rate": 8.472629309072443e-06,
+      "loss": 0.4116,
+      "step": 2337
+    },
+    {
+      "epoch": 0.5564348188254894,
+      "grad_norm": 0.38847211210505317,
+      "learning_rate": 8.471242209371867e-06,
+      "loss": 0.3489,
+      "step": 2338
+    },
+    {
+      "epoch": 0.5566728148985541,
+      "grad_norm": 0.3632605487642382,
+      "learning_rate": 8.469854593758825e-06,
+      "loss": 0.3579,
+      "step": 2339
+    },
+    {
+      "epoch": 0.556910810971619,
+      "grad_norm": 0.36587914178181785,
+      "learning_rate": 8.468466462439549e-06,
+      "loss": 0.4237,
+      "step": 2340
+    },
+    {
+      "epoch": 0.5571488070446837,
+      "grad_norm": 0.3955575662234394,
+      "learning_rate": 8.467077815620352e-06,
+      "loss": 0.3263,
+      "step": 2341
+    },
+    {
+      "epoch": 0.5573868031177486,
+      "grad_norm": 0.4044906304039566,
+      "learning_rate": 8.46568865350762e-06,
+      "loss": 0.3015,
+      "step": 2342
+    },
+    {
+      "epoch": 0.5576247991908133,
+      "grad_norm": 0.3719462975550843,
+      "learning_rate": 8.464298976307816e-06,
+      "loss": 0.3709,
+      "step": 2343
+    },
+    {
+      "epoch": 0.5578627952638782,
+      "grad_norm": 0.41922112217020985,
+      "learning_rate": 8.462908784227484e-06,
+      "loss": 0.4194,
+      "step": 2344
+    },
+    {
+      "epoch": 0.5581007913369429,
+      "grad_norm": 0.4271019026096489,
+      "learning_rate": 8.461518077473236e-06,
+      "loss": 0.3211,
+      "step": 2345
+    },
+    {
+      "epoch": 0.5583387874100078,
+      "grad_norm": 0.401362760699648,
+      "learning_rate": 8.46012685625177e-06,
+      "loss": 0.3347,
+      "step": 2346
+    },
+    {
+      "epoch": 0.5585767834830725,
+      "grad_norm": 0.3788197761561941,
+      "learning_rate": 8.458735120769853e-06,
+      "loss": 0.3966,
+      "step": 2347
+    },
+    {
+      "epoch": 0.5588147795561373,
+      "grad_norm": 0.37058769198361474,
+      "learning_rate": 8.457342871234331e-06,
+      "loss": 0.3761,
+      "step": 2348
+    },
+    {
+      "epoch": 0.5590527756292021,
+      "grad_norm": 0.3701385464501336,
+      "learning_rate": 8.455950107852127e-06,
+      "loss": 0.296,
+      "step": 2349
+    },
+    {
+      "epoch": 0.5592907717022669,
+      "grad_norm": 0.39037984018929806,
+      "learning_rate": 8.454556830830242e-06,
+      "loss": 0.3621,
+      "step": 2350
+    },
+    {
+      "epoch": 0.5595287677753317,
+      "grad_norm": 0.3678062640318312,
+      "learning_rate": 8.453163040375751e-06,
+      "loss": 0.4257,
+      "step": 2351
+    },
+    {
+      "epoch": 0.5597667638483965,
+      "grad_norm": 0.3905948837382555,
+      "learning_rate": 8.451768736695806e-06,
+      "loss": 0.3342,
+      "step": 2352
+    },
+    {
+      "epoch": 0.5600047599214613,
+      "grad_norm": 0.4279691685882921,
+      "learning_rate": 8.450373919997633e-06,
+      "loss": 0.3265,
+      "step": 2353
+    },
+    {
+      "epoch": 0.5602427559945261,
+      "grad_norm": 0.39184884481824495,
+      "learning_rate": 8.448978590488538e-06,
+      "loss": 0.398,
+      "step": 2354
+    },
+    {
+      "epoch": 0.5604807520675908,
+      "grad_norm": 0.3667348181679106,
+      "learning_rate": 8.447582748375899e-06,
+      "loss": 0.3963,
+      "step": 2355
+    },
+    {
+      "epoch": 0.5607187481406557,
+      "grad_norm": 0.4044466646021283,
+      "learning_rate": 8.446186393867175e-06,
+      "loss": 0.3226,
+      "step": 2356
+    },
+    {
+      "epoch": 0.5609567442137204,
+      "grad_norm": 0.42695375000264973,
+      "learning_rate": 8.444789527169899e-06,
+      "loss": 0.345,
+      "step": 2357
+    },
+    {
+      "epoch": 0.5611947402867853,
+      "grad_norm": 0.38004783330852054,
+      "learning_rate": 8.44339214849168e-06,
+      "loss": 0.3978,
+      "step": 2358
+    },
+    {
+      "epoch": 0.56143273635985,
+      "grad_norm": 0.3996997964059031,
+      "learning_rate": 8.441994258040202e-06,
+      "loss": 0.3303,
+      "step": 2359
+    },
+    {
+      "epoch": 0.5616707324329149,
+      "grad_norm": 0.39510081986039514,
+      "learning_rate": 8.440595856023226e-06,
+      "loss": 0.3358,
+      "step": 2360
+    },
+    {
+      "epoch": 0.5619087285059796,
+      "grad_norm": 0.4191208470001434,
+      "learning_rate": 8.439196942648589e-06,
+      "loss": 0.4067,
+      "step": 2361
+    },
+    {
+      "epoch": 0.5621467245790445,
+      "grad_norm": 0.4303815244296002,
+      "learning_rate": 8.437797518124205e-06,
+      "loss": 0.4148,
+      "step": 2362
+    },
+    {
+      "epoch": 0.5623847206521092,
+      "grad_norm": 0.3989720793075796,
+      "learning_rate": 8.436397582658062e-06,
+      "loss": 0.3123,
+      "step": 2363
+    },
+    {
+      "epoch": 0.562622716725174,
+      "grad_norm": 0.3880785671355131,
+      "learning_rate": 8.434997136458227e-06,
+      "loss": 0.3449,
+      "step": 2364
+    },
+    {
+      "epoch": 0.5628607127982388,
+      "grad_norm": 0.35392253974214943,
+      "learning_rate": 8.433596179732838e-06,
+      "loss": 0.4147,
+      "step": 2365
+    },
+    {
+      "epoch": 0.5630987088713036,
+      "grad_norm": 0.35980860715044094,
+      "learning_rate": 8.432194712690117e-06,
+      "loss": 0.344,
+      "step": 2366
+    },
+    {
+      "epoch": 0.5633367049443684,
+      "grad_norm": 0.3585696578371747,
+      "learning_rate": 8.430792735538352e-06,
+      "loss": 0.3067,
+      "step": 2367
+    },
+    {
+      "epoch": 0.5635747010174332,
+      "grad_norm": 0.3443163729830074,
+      "learning_rate": 8.429390248485911e-06,
+      "loss": 0.3627,
+      "step": 2368
+    },
+    {
+      "epoch": 0.563812697090498,
+      "grad_norm": 0.3649074447807937,
+      "learning_rate": 8.42798725174124e-06,
+      "loss": 0.4325,
+      "step": 2369
+    },
+    {
+      "epoch": 0.5640506931635628,
+      "grad_norm": 0.3798587786258275,
+      "learning_rate": 8.426583745512862e-06,
+      "loss": 0.3246,
+      "step": 2370
+    },
+    {
+      "epoch": 0.5642886892366276,
+      "grad_norm": 0.40308412350983575,
+      "learning_rate": 8.425179730009368e-06,
+      "loss": 0.3384,
+      "step": 2371
+    },
+    {
+      "epoch": 0.5645266853096924,
+      "grad_norm": 0.4049157079324643,
+      "learning_rate": 8.423775205439433e-06,
+      "loss": 0.4236,
+      "step": 2372
+    },
+    {
+      "epoch": 0.5647646813827571,
+      "grad_norm": 0.3975137139248059,
+      "learning_rate": 8.4223701720118e-06,
+      "loss": 0.3573,
+      "step": 2373
+    },
+    {
+      "epoch": 0.565002677455822,
+      "grad_norm": 0.40160322607563,
+      "learning_rate": 8.420964629935294e-06,
+      "loss": 0.306,
+      "step": 2374
+    },
+    {
+      "epoch": 0.5652406735288867,
+      "grad_norm": 0.37308395646020814,
+      "learning_rate": 8.419558579418813e-06,
+      "loss": 0.3918,
+      "step": 2375
+    },
+    {
+      "epoch": 0.5654786696019516,
+      "grad_norm": 0.3514852696457898,
+      "learning_rate": 8.418152020671335e-06,
+      "loss": 0.4363,
+      "step": 2376
+    },
+    {
+      "epoch": 0.5657166656750163,
+      "grad_norm": 0.6416951229578869,
+      "learning_rate": 8.416744953901904e-06,
+      "loss": 0.3119,
+      "step": 2377
+    },
+    {
+      "epoch": 0.5659546617480812,
+      "grad_norm": 0.3436478077704659,
+      "learning_rate": 8.415337379319645e-06,
+      "loss": 0.3405,
+      "step": 2378
+    },
+    {
+      "epoch": 0.5661926578211459,
+      "grad_norm": 0.38888307187061655,
+      "learning_rate": 8.41392929713376e-06,
+      "loss": 0.3862,
+      "step": 2379
+    },
+    {
+      "epoch": 0.5664306538942108,
+      "grad_norm": 0.4044080349860496,
+      "learning_rate": 8.412520707553527e-06,
+      "loss": 0.4068,
+      "step": 2380
+    },
+    {
+      "epoch": 0.5666686499672755,
+      "grad_norm": 0.3738358783673891,
+      "learning_rate": 8.411111610788294e-06,
+      "loss": 0.3103,
+      "step": 2381
+    },
+    {
+      "epoch": 0.5669066460403404,
+      "grad_norm": 0.3894619985619245,
+      "learning_rate": 8.40970200704749e-06,
+      "loss": 0.3569,
+      "step": 2382
+    },
+    {
+      "epoch": 0.5671446421134051,
+      "grad_norm": 0.37956208132491615,
+      "learning_rate": 8.408291896540613e-06,
+      "loss": 0.4404,
+      "step": 2383
+    },
+    {
+      "epoch": 0.56738263818647,
+      "grad_norm": 0.4385341684832731,
+      "learning_rate": 8.406881279477244e-06,
+      "loss": 0.3509,
+      "step": 2384
+    },
+    {
+      "epoch": 0.5676206342595347,
+      "grad_norm": 0.43016413455724534,
+      "learning_rate": 8.405470156067038e-06,
+      "loss": 0.3038,
+      "step": 2385
+    },
+    {
+      "epoch": 0.5678586303325995,
+      "grad_norm": 0.41401646258681746,
+      "learning_rate": 8.404058526519717e-06,
+      "loss": 0.3866,
+      "step": 2386
+    },
+    {
+      "epoch": 0.5680966264056643,
+      "grad_norm": 0.3977676509749422,
+      "learning_rate": 8.402646391045085e-06,
+      "loss": 0.4203,
+      "step": 2387
+    },
+    {
+      "epoch": 0.5683346224787291,
+      "grad_norm": 0.47631691258225883,
+      "learning_rate": 8.401233749853024e-06,
+      "loss": 0.3177,
+      "step": 2388
+    },
+    {
+      "epoch": 0.5685726185517939,
+      "grad_norm": 0.3878321830383998,
+      "learning_rate": 8.399820603153483e-06,
+      "loss": 0.3613,
+      "step": 2389
+    },
+    {
+      "epoch": 0.5688106146248587,
+      "grad_norm": 0.3714565178735775,
+      "learning_rate": 8.398406951156496e-06,
+      "loss": 0.426,
+      "step": 2390
+    },
+    {
+      "epoch": 0.5690486106979235,
+      "grad_norm": 0.39492294312602905,
+      "learning_rate": 8.396992794072162e-06,
+      "loss": 0.3286,
+      "step": 2391
+    },
+    {
+      "epoch": 0.5692866067709883,
+      "grad_norm": 0.4043641921314243,
+      "learning_rate": 8.395578132110663e-06,
+      "loss": 0.3344,
+      "step": 2392
+    },
+    {
+      "epoch": 0.569524602844053,
+      "grad_norm": 0.39639917678624054,
+      "learning_rate": 8.394162965482249e-06,
+      "loss": 0.3757,
+      "step": 2393
+    },
+    {
+      "epoch": 0.5697625989171179,
+      "grad_norm": 0.37351452297347937,
+      "learning_rate": 8.39274729439725e-06,
+      "loss": 0.3834,
+      "step": 2394
+    },
+    {
+      "epoch": 0.5700005949901826,
+      "grad_norm": 0.4062182778345128,
+      "learning_rate": 8.391331119066071e-06,
+      "loss": 0.304,
+      "step": 2395
+    },
+    {
+      "epoch": 0.5702385910632475,
+      "grad_norm": 0.36745005612369885,
+      "learning_rate": 8.389914439699191e-06,
+      "loss": 0.3195,
+      "step": 2396
+    },
+    {
+      "epoch": 0.5704765871363122,
+      "grad_norm": 0.3687137278697919,
+      "learning_rate": 8.388497256507163e-06,
+      "loss": 0.3726,
+      "step": 2397
+    },
+    {
+      "epoch": 0.5707145832093771,
+      "grad_norm": 0.3792259871749328,
+      "learning_rate": 8.387079569700615e-06,
+      "loss": 0.3311,
+      "step": 2398
+    },
+    {
+      "epoch": 0.5709525792824418,
+      "grad_norm": 0.40696921581904394,
+      "learning_rate": 8.38566137949025e-06,
+      "loss": 0.2977,
+      "step": 2399
+    },
+    {
+      "epoch": 0.5711905753555067,
+      "grad_norm": 0.3699322815190473,
+      "learning_rate": 8.384242686086848e-06,
+      "loss": 0.3794,
+      "step": 2400
+    },
+    {
+      "epoch": 0.5714285714285714,
+      "grad_norm": 0.3826635436528603,
+      "learning_rate": 8.382823489701262e-06,
+      "loss": 0.4304,
+      "step": 2401
+    },
+    {
+      "epoch": 0.5716665675016362,
+      "grad_norm": 0.38070938472969035,
+      "learning_rate": 8.381403790544416e-06,
+      "loss": 0.3305,
+      "step": 2402
+    },
+    {
+      "epoch": 0.571904563574701,
+      "grad_norm": 0.3868544933608521,
+      "learning_rate": 8.379983588827314e-06,
+      "loss": 0.3067,
+      "step": 2403
+    },
+    {
+      "epoch": 0.5721425596477658,
+      "grad_norm": 0.39043334928429774,
+      "learning_rate": 8.378562884761037e-06,
+      "loss": 0.385,
+      "step": 2404
+    },
+    {
+      "epoch": 0.5723805557208306,
+      "grad_norm": 0.3953822016500048,
+      "learning_rate": 8.37714167855673e-06,
+      "loss": 0.3968,
+      "step": 2405
+    },
+    {
+      "epoch": 0.5726185517938954,
+      "grad_norm": 0.36014258721015646,
+      "learning_rate": 8.375719970425626e-06,
+      "loss": 0.3058,
+      "step": 2406
+    },
+    {
+      "epoch": 0.5728565478669602,
+      "grad_norm": 0.38242727883148314,
+      "learning_rate": 8.374297760579024e-06,
+      "loss": 0.3667,
+      "step": 2407
+    },
+    {
+      "epoch": 0.573094543940025,
+      "grad_norm": 0.3583190778095351,
+      "learning_rate": 8.372875049228295e-06,
+      "loss": 0.448,
+      "step": 2408
+    },
+    {
+      "epoch": 0.5733325400130898,
+      "grad_norm": 0.43781235930710827,
+      "learning_rate": 8.371451836584894e-06,
+      "loss": 0.3132,
+      "step": 2409
+    },
+    {
+      "epoch": 0.5735705360861546,
+      "grad_norm": 0.3688130777579906,
+      "learning_rate": 8.370028122860346e-06,
+      "loss": 0.3088,
+      "step": 2410
+    },
+    {
+      "epoch": 0.5738085321592193,
+      "grad_norm": 0.398460180753574,
+      "learning_rate": 8.368603908266248e-06,
+      "loss": 0.3805,
+      "step": 2411
+    },
+    {
+      "epoch": 0.5740465282322842,
+      "grad_norm": 0.39559802571213665,
+      "learning_rate": 8.367179193014275e-06,
+      "loss": 0.4192,
+      "step": 2412
+    },
+    {
+      "epoch": 0.5742845243053489,
+      "grad_norm": 0.41593476553175096,
+      "learning_rate": 8.365753977316171e-06,
+      "loss": 0.3284,
+      "step": 2413
+    },
+    {
+      "epoch": 0.5745225203784138,
+      "grad_norm": 0.3718359537773827,
+      "learning_rate": 8.364328261383763e-06,
+      "loss": 0.34,
+      "step": 2414
+    },
+    {
+      "epoch": 0.5747605164514785,
+      "grad_norm": 0.3557624684742045,
+      "learning_rate": 8.362902045428945e-06,
+      "loss": 0.4296,
+      "step": 2415
+    },
+    {
+      "epoch": 0.5749985125245434,
+      "grad_norm": 0.402514906677194,
+      "learning_rate": 8.36147532966369e-06,
+      "loss": 0.3406,
+      "step": 2416
+    },
+    {
+      "epoch": 0.5752365085976081,
+      "grad_norm": 0.39462407875443245,
+      "learning_rate": 8.360048114300041e-06,
+      "loss": 0.3383,
+      "step": 2417
+    },
+    {
+      "epoch": 0.575474504670673,
+      "grad_norm": 0.41311640572509384,
+      "learning_rate": 8.35862039955012e-06,
+      "loss": 0.3611,
+      "step": 2418
+    },
+    {
+      "epoch": 0.5757125007437377,
+      "grad_norm": 0.3917287001092687,
+      "learning_rate": 8.357192185626118e-06,
+      "loss": 0.4186,
+      "step": 2419
+    },
+    {
+      "epoch": 0.5759504968168025,
+      "grad_norm": 0.3467390799072633,
+      "learning_rate": 8.355763472740305e-06,
+      "loss": 0.3304,
+      "step": 2420
+    },
+    {
+      "epoch": 0.5761884928898673,
+      "grad_norm": 0.38769794783433975,
+      "learning_rate": 8.354334261105023e-06,
+      "loss": 0.3565,
+      "step": 2421
+    },
+    {
+      "epoch": 0.5764264889629321,
+      "grad_norm": 0.4041528840850423,
+      "learning_rate": 8.352904550932687e-06,
+      "loss": 0.4234,
+      "step": 2422
+    },
+    {
+      "epoch": 0.5766644850359969,
+      "grad_norm": 0.4100086929644872,
+      "learning_rate": 8.351474342435786e-06,
+      "loss": 0.356,
+      "step": 2423
+    },
+    {
+      "epoch": 0.5769024811090617,
+      "grad_norm": 0.36794983188883607,
+      "learning_rate": 8.350043635826888e-06,
+      "loss": 0.3224,
+      "step": 2424
+    },
+    {
+      "epoch": 0.5771404771821265,
+      "grad_norm": 0.40729170540747445,
+      "learning_rate": 8.34861243131863e-06,
+      "loss": 0.3484,
+      "step": 2425
+    },
+    {
+      "epoch": 0.5773784732551913,
+      "grad_norm": 0.40251376202597505,
+      "learning_rate": 8.347180729123724e-06,
+      "loss": 0.4518,
+      "step": 2426
+    },
+    {
+      "epoch": 0.577616469328256,
+      "grad_norm": 0.3852352526845826,
+      "learning_rate": 8.345748529454956e-06,
+      "loss": 0.3455,
+      "step": 2427
+    },
+    {
+      "epoch": 0.5778544654013209,
+      "grad_norm": 0.39160730175824005,
+      "learning_rate": 8.344315832525187e-06,
+      "loss": 0.3527,
+      "step": 2428
+    },
+    {
+      "epoch": 0.5780924614743856,
+      "grad_norm": 0.38108553823066177,
+      "learning_rate": 8.342882638547351e-06,
+      "loss": 0.4142,
+      "step": 2429
+    },
+    {
+      "epoch": 0.5783304575474505,
+      "grad_norm": 0.4092273445218704,
+      "learning_rate": 8.341448947734454e-06,
+      "loss": 0.3947,
+      "step": 2430
+    },
+    {
+      "epoch": 0.5785684536205152,
+      "grad_norm": 0.40333249898205964,
+      "learning_rate": 8.340014760299582e-06,
+      "loss": 0.3326,
+      "step": 2431
+    },
+    {
+      "epoch": 0.5788064496935801,
+      "grad_norm": 0.3729044861254844,
+      "learning_rate": 8.33858007645589e-06,
+      "loss": 0.3735,
+      "step": 2432
+    },
+    {
+      "epoch": 0.5790444457666448,
+      "grad_norm": 0.3671984616197141,
+      "learning_rate": 8.337144896416602e-06,
+      "loss": 0.3966,
+      "step": 2433
+    },
+    {
+      "epoch": 0.5792824418397097,
+      "grad_norm": 0.3836501393931116,
+      "learning_rate": 8.335709220395029e-06,
+      "loss": 0.3285,
+      "step": 2434
+    },
+    {
+      "epoch": 0.5795204379127744,
+      "grad_norm": 0.35340054131259374,
+      "learning_rate": 8.334273048604541e-06,
+      "loss": 0.3307,
+      "step": 2435
+    },
+    {
+      "epoch": 0.5797584339858393,
+      "grad_norm": 0.3439612755818031,
+      "learning_rate": 8.332836381258596e-06,
+      "loss": 0.3899,
+      "step": 2436
+    },
+    {
+      "epoch": 0.579996430058904,
+      "grad_norm": 0.3569322611942811,
+      "learning_rate": 8.331399218570711e-06,
+      "loss": 0.4009,
+      "step": 2437
+    },
+    {
+      "epoch": 0.5802344261319689,
+      "grad_norm": 0.6662367868543072,
+      "learning_rate": 8.32996156075449e-06,
+      "loss": 0.3147,
+      "step": 2438
+    },
+    {
+      "epoch": 0.5804724222050336,
+      "grad_norm": 0.4227515779891762,
+      "learning_rate": 8.328523408023599e-06,
+      "loss": 0.3299,
+      "step": 2439
+    },
+    {
+      "epoch": 0.5807104182780984,
+      "grad_norm": 0.37706784919884295,
+      "learning_rate": 8.327084760591788e-06,
+      "loss": 0.4154,
+      "step": 2440
+    },
+    {
+      "epoch": 0.5809484143511632,
+      "grad_norm": 0.36959641829833456,
+      "learning_rate": 8.325645618672873e-06,
+      "loss": 0.3566,
+      "step": 2441
+    },
+    {
+      "epoch": 0.581186410424228,
+      "grad_norm": 0.38877783129131926,
+      "learning_rate": 8.324205982480747e-06,
+      "loss": 0.3094,
+      "step": 2442
+    },
+    {
+      "epoch": 0.5814244064972928,
+      "grad_norm": 0.4093291177503294,
+      "learning_rate": 8.322765852229373e-06,
+      "loss": 0.3788,
+      "step": 2443
+    },
+    {
+      "epoch": 0.5816624025703576,
+      "grad_norm": 0.35993564376286113,
+      "learning_rate": 8.321325228132793e-06,
+      "loss": 0.3971,
+      "step": 2444
+    },
+    {
+      "epoch": 0.5819003986434224,
+      "grad_norm": 0.37858365267915045,
+      "learning_rate": 8.31988411040512e-06,
+      "loss": 0.3312,
+      "step": 2445
+    },
+    {
+      "epoch": 0.5821383947164872,
+      "grad_norm": 0.36232519560188203,
+      "learning_rate": 8.318442499260538e-06,
+      "loss": 0.3209,
+      "step": 2446
+    },
+    {
+      "epoch": 0.582376390789552,
+      "grad_norm": 0.37803394847163874,
+      "learning_rate": 8.317000394913304e-06,
+      "loss": 0.4168,
+      "step": 2447
+    },
+    {
+      "epoch": 0.5826143868626168,
+      "grad_norm": 0.39433181440916604,
+      "learning_rate": 8.315557797577754e-06,
+      "loss": 0.3812,
+      "step": 2448
+    },
+    {
+      "epoch": 0.5828523829356815,
+      "grad_norm": 0.403795746000504,
+      "learning_rate": 8.314114707468293e-06,
+      "loss": 0.2969,
+      "step": 2449
+    },
+    {
+      "epoch": 0.5830903790087464,
+      "grad_norm": 0.3611167532562934,
+      "learning_rate": 8.312671124799398e-06,
+      "loss": 0.3553,
+      "step": 2450
+    },
+    {
+      "epoch": 0.5833283750818111,
+      "grad_norm": 0.38084432296705717,
+      "learning_rate": 8.311227049785623e-06,
+      "loss": 0.4011,
+      "step": 2451
+    },
+    {
+      "epoch": 0.583566371154876,
+      "grad_norm": 0.3655837949285881,
+      "learning_rate": 8.309782482641595e-06,
+      "loss": 0.3379,
+      "step": 2452
+    },
+    {
+      "epoch": 0.5838043672279407,
+      "grad_norm": 0.3913423734099842,
+      "learning_rate": 8.308337423582006e-06,
+      "loss": 0.3441,
+      "step": 2453
+    },
+    {
+      "epoch": 0.5840423633010056,
+      "grad_norm": 0.39397581704554785,
+      "learning_rate": 8.306891872821635e-06,
+      "loss": 0.3829,
+      "step": 2454
+    },
+    {
+      "epoch": 0.5842803593740703,
+      "grad_norm": 0.3843675181792697,
+      "learning_rate": 8.30544583057532e-06,
+      "loss": 0.3929,
+      "step": 2455
+    },
+    {
+      "epoch": 0.5845183554471352,
+      "grad_norm": 0.3832844659583936,
+      "learning_rate": 8.303999297057983e-06,
+      "loss": 0.3161,
+      "step": 2456
+    },
+    {
+      "epoch": 0.5847563515201999,
+      "grad_norm": 0.39763166006214656,
+      "learning_rate": 8.302552272484613e-06,
+      "loss": 0.3667,
+      "step": 2457
+    },
+    {
+      "epoch": 0.5849943475932647,
+      "grad_norm": 0.3585575011269157,
+      "learning_rate": 8.301104757070276e-06,
+      "loss": 0.4027,
+      "step": 2458
+    },
+    {
+      "epoch": 0.5852323436663295,
+      "grad_norm": 0.38820380899321066,
+      "learning_rate": 8.299656751030105e-06,
+      "loss": 0.3341,
+      "step": 2459
+    },
+    {
+      "epoch": 0.5854703397393943,
+      "grad_norm": 0.3836385113109399,
+      "learning_rate": 8.298208254579311e-06,
+      "loss": 0.3173,
+      "step": 2460
+    },
+    {
+      "epoch": 0.5857083358124591,
+      "grad_norm": 0.38574649888799817,
+      "learning_rate": 8.296759267933178e-06,
+      "loss": 0.3862,
+      "step": 2461
+    },
+    {
+      "epoch": 0.5859463318855239,
+      "grad_norm": 0.3684542994428625,
+      "learning_rate": 8.29530979130706e-06,
+      "loss": 0.3995,
+      "step": 2462
+    },
+    {
+      "epoch": 0.5861843279585887,
+      "grad_norm": 0.39106683323722646,
+      "learning_rate": 8.293859824916383e-06,
+      "loss": 0.3097,
+      "step": 2463
+    },
+    {
+      "epoch": 0.5864223240316535,
+      "grad_norm": 0.4653396505467396,
+      "learning_rate": 8.29240936897665e-06,
+      "loss": 0.3342,
+      "step": 2464
+    },
+    {
+      "epoch": 0.5866603201047182,
+      "grad_norm": 0.3895719288501651,
+      "learning_rate": 8.290958423703437e-06,
+      "loss": 0.4107,
+      "step": 2465
+    },
+    {
+      "epoch": 0.5868983161777831,
+      "grad_norm": 0.4018410614287424,
+      "learning_rate": 8.289506989312386e-06,
+      "loss": 0.3257,
+      "step": 2466
+    },
+    {
+      "epoch": 0.5871363122508478,
+      "grad_norm": 0.3944410384951886,
+      "learning_rate": 8.288055066019218e-06,
+      "loss": 0.3238,
+      "step": 2467
+    },
+    {
+      "epoch": 0.5873743083239127,
+      "grad_norm": 0.38792256065200875,
+      "learning_rate": 8.286602654039724e-06,
+      "loss": 0.3826,
+      "step": 2468
+    },
+    {
+      "epoch": 0.5876123043969774,
+      "grad_norm": 0.38502496740303777,
+      "learning_rate": 8.28514975358977e-06,
+      "loss": 0.3997,
+      "step": 2469
+    },
+    {
+      "epoch": 0.5878503004700423,
+      "grad_norm": 0.37703883930285037,
+      "learning_rate": 8.283696364885293e-06,
+      "loss": 0.3356,
+      "step": 2470
+    },
+    {
+      "epoch": 0.588088296543107,
+      "grad_norm": 0.3861610449357546,
+      "learning_rate": 8.282242488142299e-06,
+      "loss": 0.3348,
+      "step": 2471
+    },
+    {
+      "epoch": 0.5883262926161719,
+      "grad_norm": 0.3780090152003883,
+      "learning_rate": 8.280788123576873e-06,
+      "loss": 0.4045,
+      "step": 2472
+    },
+    {
+      "epoch": 0.5885642886892366,
+      "grad_norm": 0.3801772021397056,
+      "learning_rate": 8.279333271405171e-06,
+      "loss": 0.3941,
+      "step": 2473
+    },
+    {
+      "epoch": 0.5888022847623015,
+      "grad_norm": 0.3903145537175543,
+      "learning_rate": 8.277877931843417e-06,
+      "loss": 0.32,
+      "step": 2474
+    },
+    {
+      "epoch": 0.5890402808353662,
+      "grad_norm": 0.3762448521852173,
+      "learning_rate": 8.276422105107911e-06,
+      "loss": 0.3519,
+      "step": 2475
+    },
+    {
+      "epoch": 0.589278276908431,
+      "grad_norm": 0.4024863557944298,
+      "learning_rate": 8.274965791415026e-06,
+      "loss": 0.4126,
+      "step": 2476
+    },
+    {
+      "epoch": 0.5895162729814958,
+      "grad_norm": 0.3810042080891356,
+      "learning_rate": 8.273508990981206e-06,
+      "loss": 0.3461,
+      "step": 2477
+    },
+    {
+      "epoch": 0.5897542690545606,
+      "grad_norm": 0.39863438603953905,
+      "learning_rate": 8.272051704022965e-06,
+      "loss": 0.3328,
+      "step": 2478
+    },
+    {
+      "epoch": 0.5899922651276254,
+      "grad_norm": 0.3867478091098192,
+      "learning_rate": 8.270593930756897e-06,
+      "loss": 0.377,
+      "step": 2479
+    },
+    {
+      "epoch": 0.5902302612006902,
+      "grad_norm": 0.3879898329337162,
+      "learning_rate": 8.269135671399659e-06,
+      "loss": 0.3646,
+      "step": 2480
+    },
+    {
+      "epoch": 0.590468257273755,
+      "grad_norm": 0.36772292163364306,
+      "learning_rate": 8.267676926167986e-06,
+      "loss": 0.3086,
+      "step": 2481
+    },
+    {
+      "epoch": 0.5907062533468198,
+      "grad_norm": 0.3565818145381248,
+      "learning_rate": 8.266217695278682e-06,
+      "loss": 0.355,
+      "step": 2482
+    },
+    {
+      "epoch": 0.5909442494198845,
+      "grad_norm": 0.37046723786837815,
+      "learning_rate": 8.264757978948627e-06,
+      "loss": 0.4076,
+      "step": 2483
+    },
+    {
+      "epoch": 0.5911822454929494,
+      "grad_norm": 0.4249706612993194,
+      "learning_rate": 8.263297777394772e-06,
+      "loss": 0.3327,
+      "step": 2484
+    },
+    {
+      "epoch": 0.5914202415660141,
+      "grad_norm": 0.3951802875165859,
+      "learning_rate": 8.261837090834135e-06,
+      "loss": 0.3376,
+      "step": 2485
+    },
+    {
+      "epoch": 0.591658237639079,
+      "grad_norm": 0.3673787092456175,
+      "learning_rate": 8.260375919483812e-06,
+      "loss": 0.3509,
+      "step": 2486
+    },
+    {
+      "epoch": 0.5918962337121437,
+      "grad_norm": 0.3896876553874139,
+      "learning_rate": 8.258914263560971e-06,
+      "loss": 0.4089,
+      "step": 2487
+    },
+    {
+      "epoch": 0.5921342297852086,
+      "grad_norm": 0.37425201217841664,
+      "learning_rate": 8.257452123282847e-06,
+      "loss": 0.3259,
+      "step": 2488
+    },
+    {
+      "epoch": 0.5923722258582733,
+      "grad_norm": 0.3859687694225435,
+      "learning_rate": 8.255989498866754e-06,
+      "loss": 0.3243,
+      "step": 2489
+    },
+    {
+      "epoch": 0.5926102219313382,
+      "grad_norm": 0.3708026305032604,
+      "learning_rate": 8.254526390530071e-06,
+      "loss": 0.3954,
+      "step": 2490
+    },
+    {
+      "epoch": 0.5928482180044029,
+      "grad_norm": 0.3794998880457793,
+      "learning_rate": 8.253062798490255e-06,
+      "loss": 0.3535,
+      "step": 2491
+    },
+    {
+      "epoch": 0.5930862140774678,
+      "grad_norm": 0.4489229665961381,
+      "learning_rate": 8.251598722964828e-06,
+      "loss": 0.3371,
+      "step": 2492
+    },
+    {
+      "epoch": 0.5933242101505325,
+      "grad_norm": 0.3631137000424927,
+      "learning_rate": 8.250134164171391e-06,
+      "loss": 0.3718,
+      "step": 2493
+    },
+    {
+      "epoch": 0.5935622062235973,
+      "grad_norm": 0.3527479506637244,
+      "learning_rate": 8.248669122327612e-06,
+      "loss": 0.3933,
+      "step": 2494
+    },
+    {
+      "epoch": 0.5938002022966621,
+      "grad_norm": 0.3742555835211287,
+      "learning_rate": 8.247203597651234e-06,
+      "loss": 0.3274,
+      "step": 2495
+    },
+    {
+      "epoch": 0.5940381983697269,
+      "grad_norm": 0.4308062065966235,
+      "learning_rate": 8.24573759036007e-06,
+      "loss": 0.3666,
+      "step": 2496
+    },
+    {
+      "epoch": 0.5942761944427917,
+      "grad_norm": 0.3966932502889181,
+      "learning_rate": 8.244271100672004e-06,
+      "loss": 0.4066,
+      "step": 2497
+    },
+    {
+      "epoch": 0.5945141905158565,
+      "grad_norm": 0.35946849534322295,
+      "learning_rate": 8.242804128804993e-06,
+      "loss": 0.3555,
+      "step": 2498
+    },
+    {
+      "epoch": 0.5947521865889213,
+      "grad_norm": 0.37828665852870985,
+      "learning_rate": 8.241336674977064e-06,
+      "loss": 0.3126,
+      "step": 2499
+    },
+    {
+      "epoch": 0.5949901826619861,
+      "grad_norm": 0.38560613627125817,
+      "learning_rate": 8.23986873940632e-06,
+      "loss": 0.3595,
+      "step": 2500
+    },
+    {
+      "epoch": 0.5952281787350509,
+      "grad_norm": 0.3933069848593732,
+      "learning_rate": 8.238400322310931e-06,
+      "loss": 0.4599,
+      "step": 2501
+    },
+    {
+      "epoch": 0.5954661748081157,
+      "grad_norm": 0.37906655075786916,
+      "learning_rate": 8.23693142390914e-06,
+      "loss": 0.3189,
+      "step": 2502
+    },
+    {
+      "epoch": 0.5957041708811804,
+      "grad_norm": 0.3597221601036792,
+      "learning_rate": 8.23546204441926e-06,
+      "loss": 0.3001,
+      "step": 2503
+    },
+    {
+      "epoch": 0.5959421669542453,
+      "grad_norm": 0.36715814207073877,
+      "learning_rate": 8.233992184059681e-06,
+      "loss": 0.3793,
+      "step": 2504
+    },
+    {
+      "epoch": 0.59618016302731,
+      "grad_norm": 0.37986642282050925,
+      "learning_rate": 8.232521843048855e-06,
+      "loss": 0.378,
+      "step": 2505
+    },
+    {
+      "epoch": 0.5964181591003749,
+      "grad_norm": 0.38641730071730823,
+      "learning_rate": 8.231051021605316e-06,
+      "loss": 0.3321,
+      "step": 2506
+    },
+    {
+      "epoch": 0.5966561551734396,
+      "grad_norm": 0.4161340087618939,
+      "learning_rate": 8.229579719947664e-06,
+      "loss": 0.3855,
+      "step": 2507
+    },
+    {
+      "epoch": 0.5968941512465045,
+      "grad_norm": 0.3557911482780271,
+      "learning_rate": 8.228107938294568e-06,
+      "loss": 0.3944,
+      "step": 2508
+    },
+    {
+      "epoch": 0.5971321473195692,
+      "grad_norm": 0.37982110625589854,
+      "learning_rate": 8.226635676864774e-06,
+      "loss": 0.3175,
+      "step": 2509
+    },
+    {
+      "epoch": 0.5973701433926341,
+      "grad_norm": 0.37294853078019546,
+      "learning_rate": 8.225162935877096e-06,
+      "loss": 0.3159,
+      "step": 2510
+    },
+    {
+      "epoch": 0.5976081394656988,
+      "grad_norm": 0.3732432601966647,
+      "learning_rate": 8.223689715550417e-06,
+      "loss": 0.4083,
+      "step": 2511
+    },
+    {
+      "epoch": 0.5978461355387636,
+      "grad_norm": 0.37372480614911496,
+      "learning_rate": 8.222216016103697e-06,
+      "loss": 0.4262,
+      "step": 2512
+    },
+    {
+      "epoch": 0.5980841316118284,
+      "grad_norm": 0.35343278143642487,
+      "learning_rate": 8.220741837755964e-06,
+      "loss": 0.3131,
+      "step": 2513
+    },
+    {
+      "epoch": 0.5983221276848932,
+      "grad_norm": 0.36855420142012446,
+      "learning_rate": 8.219267180726315e-06,
+      "loss": 0.3692,
+      "step": 2514
+    },
+    {
+      "epoch": 0.598560123757958,
+      "grad_norm": 0.3994967531361665,
+      "learning_rate": 8.217792045233924e-06,
+      "loss": 0.3973,
+      "step": 2515
+    },
+    {
+      "epoch": 0.5987981198310228,
+      "grad_norm": 0.4108525006798571,
+      "learning_rate": 8.216316431498028e-06,
+      "loss": 0.3718,
+      "step": 2516
+    },
+    {
+      "epoch": 0.5990361159040876,
+      "grad_norm": 0.40662138405492826,
+      "learning_rate": 8.214840339737943e-06,
+      "loss": 0.2975,
+      "step": 2517
+    },
+    {
+      "epoch": 0.5992741119771524,
+      "grad_norm": 0.3728366285892858,
+      "learning_rate": 8.213363770173054e-06,
+      "loss": 0.3559,
+      "step": 2518
+    },
+    {
+      "epoch": 0.5995121080502172,
+      "grad_norm": 0.37944723614507436,
+      "learning_rate": 8.211886723022814e-06,
+      "loss": 0.4264,
+      "step": 2519
+    },
+    {
+      "epoch": 0.599750104123282,
+      "grad_norm": 0.3611408857103594,
+      "learning_rate": 8.210409198506748e-06,
+      "loss": 0.3163,
+      "step": 2520
+    },
+    {
+      "epoch": 0.5999881001963467,
+      "grad_norm": 0.3729043334589071,
+      "learning_rate": 8.208931196844453e-06,
+      "loss": 0.3315,
+      "step": 2521
+    },
+    {
+      "epoch": 0.6002260962694116,
+      "grad_norm": 0.39851502121087623,
+      "learning_rate": 8.207452718255597e-06,
+      "loss": 0.4049,
+      "step": 2522
+    },
+    {
+      "epoch": 0.6004640923424763,
+      "grad_norm": 0.402522005091446,
+      "learning_rate": 8.20597376295992e-06,
+      "loss": 0.3749,
+      "step": 2523
+    },
+    {
+      "epoch": 0.6007020884155412,
+      "grad_norm": 0.4137714780671095,
+      "learning_rate": 8.204494331177229e-06,
+      "loss": 0.3014,
+      "step": 2524
+    },
+    {
+      "epoch": 0.6009400844886059,
+      "grad_norm": 0.3901604913593569,
+      "learning_rate": 8.203014423127405e-06,
+      "loss": 0.3693,
+      "step": 2525
+    },
+    {
+      "epoch": 0.6011780805616708,
+      "grad_norm": 0.3839384353345892,
+      "learning_rate": 8.201534039030398e-06,
+      "loss": 0.4382,
+      "step": 2526
+    },
+    {
+      "epoch": 0.6014160766347355,
+      "grad_norm": 0.39209431032725645,
+      "learning_rate": 8.20005317910623e-06,
+      "loss": 0.3409,
+      "step": 2527
+    },
+    {
+      "epoch": 0.6016540727078004,
+      "grad_norm": 0.39203793684317206,
+      "learning_rate": 8.198571843574997e-06,
+      "loss": 0.3048,
+      "step": 2528
+    },
+    {
+      "epoch": 0.6018920687808651,
+      "grad_norm": 0.3696671599005409,
+      "learning_rate": 8.197090032656858e-06,
+      "loss": 0.4085,
+      "step": 2529
+    },
+    {
+      "epoch": 0.60213006485393,
+      "grad_norm": 0.36909421514297636,
+      "learning_rate": 8.195607746572047e-06,
+      "loss": 0.3601,
+      "step": 2530
+    },
+    {
+      "epoch": 0.6023680609269947,
+      "grad_norm": 0.4061715297820262,
+      "learning_rate": 8.19412498554087e-06,
+      "loss": 0.3074,
+      "step": 2531
+    },
+    {
+      "epoch": 0.6026060570000595,
+      "grad_norm": 0.42169562952568157,
+      "learning_rate": 8.192641749783703e-06,
+      "loss": 0.3869,
+      "step": 2532
+    },
+    {
+      "epoch": 0.6028440530731243,
+      "grad_norm": 0.37904313486242136,
+      "learning_rate": 8.191158039520986e-06,
+      "loss": 0.4187,
+      "step": 2533
+    },
+    {
+      "epoch": 0.6030820491461891,
+      "grad_norm": 0.3933415457883819,
+      "learning_rate": 8.18967385497324e-06,
+      "loss": 0.3349,
+      "step": 2534
+    },
+    {
+      "epoch": 0.6033200452192539,
+      "grad_norm": 0.37746188889384785,
+      "learning_rate": 8.188189196361052e-06,
+      "loss": 0.3126,
+      "step": 2535
+    },
+    {
+      "epoch": 0.6035580412923187,
+      "grad_norm": 0.37077728868779936,
+      "learning_rate": 8.186704063905078e-06,
+      "loss": 0.3702,
+      "step": 2536
+    },
+    {
+      "epoch": 0.6037960373653835,
+      "grad_norm": 0.384234124457375,
+      "learning_rate": 8.185218457826043e-06,
+      "loss": 0.4062,
+      "step": 2537
+    },
+    {
+      "epoch": 0.6040340334384483,
+      "grad_norm": 8.292846660801779,
+      "learning_rate": 8.183732378344747e-06,
+      "loss": 0.3263,
+      "step": 2538
+    },
+    {
+      "epoch": 0.604272029511513,
+      "grad_norm": 0.3985849899773034,
+      "learning_rate": 8.18224582568206e-06,
+      "loss": 0.3693,
+      "step": 2539
+    },
+    {
+      "epoch": 0.6045100255845779,
+      "grad_norm": 0.36445215405013764,
+      "learning_rate": 8.180758800058914e-06,
+      "loss": 0.3774,
+      "step": 2540
+    },
+    {
+      "epoch": 0.6047480216576426,
+      "grad_norm": 0.38537810047531207,
+      "learning_rate": 8.179271301696326e-06,
+      "loss": 0.3585,
+      "step": 2541
+    },
+    {
+      "epoch": 0.6049860177307075,
+      "grad_norm": 0.3722062094603994,
+      "learning_rate": 8.17778333081537e-06,
+      "loss": 0.3261,
+      "step": 2542
+    },
+    {
+      "epoch": 0.6052240138037722,
+      "grad_norm": 0.43085688766278635,
+      "learning_rate": 8.176294887637195e-06,
+      "loss": 0.3491,
+      "step": 2543
+    },
+    {
+      "epoch": 0.6054620098768371,
+      "grad_norm": 0.4080559076822167,
+      "learning_rate": 8.174805972383024e-06,
+      "loss": 0.3951,
+      "step": 2544
+    },
+    {
+      "epoch": 0.6057000059499018,
+      "grad_norm": 0.37312331547160477,
+      "learning_rate": 8.173316585274144e-06,
+      "loss": 0.3025,
+      "step": 2545
+    },
+    {
+      "epoch": 0.6059380020229667,
+      "grad_norm": 0.3667533329998851,
+      "learning_rate": 8.171826726531916e-06,
+      "loss": 0.3169,
+      "step": 2546
+    },
+    {
+      "epoch": 0.6061759980960314,
+      "grad_norm": 0.3893058682249162,
+      "learning_rate": 8.170336396377767e-06,
+      "loss": 0.4053,
+      "step": 2547
+    },
+    {
+      "epoch": 0.6064139941690962,
+      "grad_norm": 0.36637391718899653,
+      "learning_rate": 8.168845595033202e-06,
+      "loss": 0.3559,
+      "step": 2548
+    },
+    {
+      "epoch": 0.606651990242161,
+      "grad_norm": 0.41529994785116214,
+      "learning_rate": 8.167354322719785e-06,
+      "loss": 0.3117,
+      "step": 2549
+    },
+    {
+      "epoch": 0.6068899863152258,
+      "grad_norm": 0.38610454842175557,
+      "learning_rate": 8.165862579659161e-06,
+      "loss": 0.3633,
+      "step": 2550
+    },
+    {
+      "epoch": 0.6071279823882906,
+      "grad_norm": 0.3753416629984086,
+      "learning_rate": 8.164370366073038e-06,
+      "loss": 0.4248,
+      "step": 2551
+    },
+    {
+      "epoch": 0.6073659784613554,
+      "grad_norm": 0.394779197165333,
+      "learning_rate": 8.162877682183197e-06,
+      "loss": 0.3117,
+      "step": 2552
+    },
+    {
+      "epoch": 0.6076039745344202,
+      "grad_norm": 0.4086616702369126,
+      "learning_rate": 8.161384528211485e-06,
+      "loss": 0.3191,
+      "step": 2553
+    },
+    {
+      "epoch": 0.607841970607485,
+      "grad_norm": 0.3788225625439413,
+      "learning_rate": 8.159890904379823e-06,
+      "loss": 0.3859,
+      "step": 2554
+    },
+    {
+      "epoch": 0.6080799666805498,
+      "grad_norm": 0.38335294969113787,
+      "learning_rate": 8.158396810910201e-06,
+      "loss": 0.3708,
+      "step": 2555
+    },
+    {
+      "epoch": 0.6083179627536146,
+      "grad_norm": 0.38664017085300634,
+      "learning_rate": 8.156902248024678e-06,
+      "loss": 0.3318,
+      "step": 2556
+    },
+    {
+      "epoch": 0.6085559588266793,
+      "grad_norm": 0.38453529856278706,
+      "learning_rate": 8.155407215945382e-06,
+      "loss": 0.3584,
+      "step": 2557
+    },
+    {
+      "epoch": 0.6087939548997442,
+      "grad_norm": 0.42462216171053496,
+      "learning_rate": 8.153911714894513e-06,
+      "loss": 0.4105,
+      "step": 2558
+    },
+    {
+      "epoch": 0.6090319509728089,
+      "grad_norm": 0.39687180601172,
+      "learning_rate": 8.152415745094342e-06,
+      "loss": 0.3442,
+      "step": 2559
+    },
+    {
+      "epoch": 0.6092699470458738,
+      "grad_norm": 0.4136807644278921,
+      "learning_rate": 8.150919306767202e-06,
+      "loss": 0.3311,
+      "step": 2560
+    },
+    {
+      "epoch": 0.6095079431189385,
+      "grad_norm": 0.35777343605244677,
+      "learning_rate": 8.149422400135503e-06,
+      "loss": 0.3723,
+      "step": 2561
+    },
+    {
+      "epoch": 0.6097459391920034,
+      "grad_norm": 0.4469538872661931,
+      "learning_rate": 8.14792502542172e-06,
+      "loss": 0.3759,
+      "step": 2562
+    },
+    {
+      "epoch": 0.6099839352650681,
+      "grad_norm": 0.39114145249451315,
+      "learning_rate": 8.146427182848407e-06,
+      "loss": 0.2963,
+      "step": 2563
+    },
+    {
+      "epoch": 0.610221931338133,
+      "grad_norm": 0.39846643949056665,
+      "learning_rate": 8.144928872638174e-06,
+      "loss": 0.3149,
+      "step": 2564
+    },
+    {
+      "epoch": 0.6104599274111977,
+      "grad_norm": 0.39040261177798785,
+      "learning_rate": 8.143430095013706e-06,
+      "loss": 0.3969,
+      "step": 2565
+    },
+    {
+      "epoch": 0.6106979234842626,
+      "grad_norm": 0.3735836629849792,
+      "learning_rate": 8.141930850197765e-06,
+      "loss": 0.3382,
+      "step": 2566
+    },
+    {
+      "epoch": 0.6109359195573273,
+      "grad_norm": 0.3736061629320796,
+      "learning_rate": 8.14043113841317e-06,
+      "loss": 0.307,
+      "step": 2567
+    },
+    {
+      "epoch": 0.6111739156303921,
+      "grad_norm": 0.35064240017216614,
+      "learning_rate": 8.138930959882818e-06,
+      "loss": 0.3508,
+      "step": 2568
+    },
+    {
+      "epoch": 0.6114119117034569,
+      "grad_norm": 0.39486353842354627,
+      "learning_rate": 8.137430314829671e-06,
+      "loss": 0.4483,
+      "step": 2569
+    },
+    {
+      "epoch": 0.6116499077765217,
+      "grad_norm": 0.3759570099382165,
+      "learning_rate": 8.135929203476764e-06,
+      "loss": 0.343,
+      "step": 2570
+    },
+    {
+      "epoch": 0.6118879038495865,
+      "grad_norm": 0.376912992715495,
+      "learning_rate": 8.134427626047198e-06,
+      "loss": 0.3342,
+      "step": 2571
+    },
+    {
+      "epoch": 0.6121258999226513,
+      "grad_norm": 0.3675782252622046,
+      "learning_rate": 8.132925582764144e-06,
+      "loss": 0.4009,
+      "step": 2572
+    },
+    {
+      "epoch": 0.6123638959957161,
+      "grad_norm": 0.37894873028788706,
+      "learning_rate": 8.131423073850845e-06,
+      "loss": 0.3797,
+      "step": 2573
+    },
+    {
+      "epoch": 0.6126018920687809,
+      "grad_norm": 0.41501690842186384,
+      "learning_rate": 8.129920099530608e-06,
+      "loss": 0.3149,
+      "step": 2574
+    },
+    {
+      "epoch": 0.6128398881418456,
+      "grad_norm": 0.3901236279329649,
+      "learning_rate": 8.128416660026816e-06,
+      "loss": 0.3559,
+      "step": 2575
+    },
+    {
+      "epoch": 0.6130778842149105,
+      "grad_norm": 0.36410270538740547,
+      "learning_rate": 8.126912755562913e-06,
+      "loss": 0.4105,
+      "step": 2576
+    },
+    {
+      "epoch": 0.6133158802879752,
+      "grad_norm": 0.40445769613053906,
+      "learning_rate": 8.125408386362419e-06,
+      "loss": 0.3565,
+      "step": 2577
+    },
+    {
+      "epoch": 0.6135538763610401,
+      "grad_norm": 0.37585605678511325,
+      "learning_rate": 8.12390355264892e-06,
+      "loss": 0.299,
+      "step": 2578
+    },
+    {
+      "epoch": 0.6137918724341048,
+      "grad_norm": 0.35091890724717334,
+      "learning_rate": 8.122398254646071e-06,
+      "loss": 0.3952,
+      "step": 2579
+    },
+    {
+      "epoch": 0.6140298685071697,
+      "grad_norm": 0.3858838022367494,
+      "learning_rate": 8.120892492577598e-06,
+      "loss": 0.4277,
+      "step": 2580
+    },
+    {
+      "epoch": 0.6142678645802344,
+      "grad_norm": 0.39895986019373647,
+      "learning_rate": 8.119386266667292e-06,
+      "loss": 0.3109,
+      "step": 2581
+    },
+    {
+      "epoch": 0.6145058606532993,
+      "grad_norm": 0.4111713762728162,
+      "learning_rate": 8.117879577139019e-06,
+      "loss": 0.3601,
+      "step": 2582
+    },
+    {
+      "epoch": 0.614743856726364,
+      "grad_norm": 0.3873631664853874,
+      "learning_rate": 8.116372424216705e-06,
+      "loss": 0.4458,
+      "step": 2583
+    },
+    {
+      "epoch": 0.6149818527994289,
+      "grad_norm": 0.3810771917479411,
+      "learning_rate": 8.114864808124356e-06,
+      "loss": 0.296,
+      "step": 2584
+    },
+    {
+      "epoch": 0.6152198488724936,
+      "grad_norm": 0.39306780621427023,
+      "learning_rate": 8.113356729086038e-06,
+      "loss": 0.3077,
+      "step": 2585
+    },
+    {
+      "epoch": 0.6154578449455584,
+      "grad_norm": 0.3857981143058693,
+      "learning_rate": 8.111848187325889e-06,
+      "loss": 0.3732,
+      "step": 2586
+    },
+    {
+      "epoch": 0.6156958410186232,
+      "grad_norm": 0.37904552046137224,
+      "learning_rate": 8.110339183068117e-06,
+      "loss": 0.3811,
+      "step": 2587
+    },
+    {
+      "epoch": 0.615933837091688,
+      "grad_norm": 0.35684929770087614,
+      "learning_rate": 8.108829716536993e-06,
+      "loss": 0.3357,
+      "step": 2588
+    },
+    {
+      "epoch": 0.6161718331647528,
+      "grad_norm": 0.37821384127046404,
+      "learning_rate": 8.107319787956866e-06,
+      "loss": 0.3508,
+      "step": 2589
+    },
+    {
+      "epoch": 0.6164098292378176,
+      "grad_norm": 0.3574697327965667,
+      "learning_rate": 8.105809397552148e-06,
+      "loss": 0.4393,
+      "step": 2590
+    },
+    {
+      "epoch": 0.6166478253108824,
+      "grad_norm": 0.39258448327977785,
+      "learning_rate": 8.10429854554732e-06,
+      "loss": 0.3101,
+      "step": 2591
+    },
+    {
+      "epoch": 0.6168858213839472,
+      "grad_norm": 0.3888467385192024,
+      "learning_rate": 8.10278723216693e-06,
+      "loss": 0.3177,
+      "step": 2592
+    },
+    {
+      "epoch": 0.617123817457012,
+      "grad_norm": 0.3852959961633021,
+      "learning_rate": 8.101275457635601e-06,
+      "loss": 0.3663,
+      "step": 2593
+    },
+    {
+      "epoch": 0.6173618135300768,
+      "grad_norm": 0.40121539043640597,
+      "learning_rate": 8.099763222178015e-06,
+      "loss": 0.405,
+      "step": 2594
+    },
+    {
+      "epoch": 0.6175998096031415,
+      "grad_norm": 0.4081613317939108,
+      "learning_rate": 8.098250526018927e-06,
+      "loss": 0.3019,
+      "step": 2595
+    },
+    {
+      "epoch": 0.6178378056762064,
+      "grad_norm": 0.45909803930292775,
+      "learning_rate": 8.096737369383167e-06,
+      "loss": 0.339,
+      "step": 2596
+    },
+    {
+      "epoch": 0.6180758017492711,
+      "grad_norm": 0.39288919774336073,
+      "learning_rate": 8.095223752495625e-06,
+      "loss": 0.3849,
+      "step": 2597
+    },
+    {
+      "epoch": 0.618313797822336,
+      "grad_norm": 0.40645028794870114,
+      "learning_rate": 8.09370967558126e-06,
+      "loss": 0.3484,
+      "step": 2598
+    },
+    {
+      "epoch": 0.6185517938954007,
+      "grad_norm": 0.36773838171647716,
+      "learning_rate": 8.092195138865102e-06,
+      "loss": 0.3205,
+      "step": 2599
+    },
+    {
+      "epoch": 0.6187897899684656,
+      "grad_norm": 0.39094679473606364,
+      "learning_rate": 8.090680142572251e-06,
+      "loss": 0.3719,
+      "step": 2600
+    },
+    {
+      "epoch": 0.6190277860415303,
+      "grad_norm": 0.37593142152302145,
+      "learning_rate": 8.089164686927869e-06,
+      "loss": 0.4178,
+      "step": 2601
+    },
+    {
+      "epoch": 0.6192657821145952,
+      "grad_norm": 0.3875234390518961,
+      "learning_rate": 8.087648772157193e-06,
+      "loss": 0.3594,
+      "step": 2602
+    },
+    {
+      "epoch": 0.6195037781876599,
+      "grad_norm": 0.3937548041329991,
+      "learning_rate": 8.086132398485525e-06,
+      "loss": 0.3272,
+      "step": 2603
+    },
+    {
+      "epoch": 0.6197417742607247,
+      "grad_norm": 0.4882360341404345,
+      "learning_rate": 8.084615566138234e-06,
+      "loss": 0.3972,
+      "step": 2604
+    },
+    {
+      "epoch": 0.6199797703337895,
+      "grad_norm": 0.3644617243184282,
+      "learning_rate": 8.083098275340762e-06,
+      "loss": 0.3796,
+      "step": 2605
+    },
+    {
+      "epoch": 0.6202177664068543,
+      "grad_norm": 0.3464988959256686,
+      "learning_rate": 8.081580526318614e-06,
+      "loss": 0.3145,
+      "step": 2606
+    },
+    {
+      "epoch": 0.6204557624799191,
+      "grad_norm": 0.37402621710256423,
+      "learning_rate": 8.080062319297364e-06,
+      "loss": 0.3315,
+      "step": 2607
+    },
+    {
+      "epoch": 0.6206937585529839,
+      "grad_norm": 0.40711958296792816,
+      "learning_rate": 8.078543654502656e-06,
+      "loss": 0.4,
+      "step": 2608
+    },
+    {
+      "epoch": 0.6209317546260487,
+      "grad_norm": 0.36492939402646146,
+      "learning_rate": 8.077024532160202e-06,
+      "loss": 0.3082,
+      "step": 2609
+    },
+    {
+      "epoch": 0.6211697506991135,
+      "grad_norm": 0.37941276134920326,
+      "learning_rate": 8.075504952495781e-06,
+      "loss": 0.3377,
+      "step": 2610
+    },
+    {
+      "epoch": 0.6214077467721782,
+      "grad_norm": 0.4020215489655943,
+      "learning_rate": 8.07398491573524e-06,
+      "loss": 0.3639,
+      "step": 2611
+    },
+    {
+      "epoch": 0.6216457428452431,
+      "grad_norm": 0.4020370198246242,
+      "learning_rate": 8.072464422104493e-06,
+      "loss": 0.4031,
+      "step": 2612
+    },
+    {
+      "epoch": 0.6218837389183078,
+      "grad_norm": 0.4204655833624215,
+      "learning_rate": 8.070943471829524e-06,
+      "loss": 0.2973,
+      "step": 2613
+    },
+    {
+      "epoch": 0.6221217349913727,
+      "grad_norm": 0.4006137318758786,
+      "learning_rate": 8.069422065136386e-06,
+      "loss": 0.3611,
+      "step": 2614
+    },
+    {
+      "epoch": 0.6223597310644374,
+      "grad_norm": 0.3662514094870652,
+      "learning_rate": 8.067900202251191e-06,
+      "loss": 0.397,
+      "step": 2615
+    },
+    {
+      "epoch": 0.6225977271375023,
+      "grad_norm": 0.44999481000949004,
+      "learning_rate": 8.066377883400132e-06,
+      "loss": 0.3546,
+      "step": 2616
+    },
+    {
+      "epoch": 0.622835723210567,
+      "grad_norm": 0.364463764085942,
+      "learning_rate": 8.064855108809461e-06,
+      "loss": 0.2948,
+      "step": 2617
+    },
+    {
+      "epoch": 0.6230737192836319,
+      "grad_norm": 0.3964833253922973,
+      "learning_rate": 8.063331878705499e-06,
+      "loss": 0.3764,
+      "step": 2618
+    },
+    {
+      "epoch": 0.6233117153566966,
+      "grad_norm": 0.3514192516820972,
+      "learning_rate": 8.061808193314638e-06,
+      "loss": 0.4172,
+      "step": 2619
+    },
+    {
+      "epoch": 0.6235497114297615,
+      "grad_norm": 0.3548064383030754,
+      "learning_rate": 8.060284052863334e-06,
+      "loss": 0.3202,
+      "step": 2620
+    },
+    {
+      "epoch": 0.6237877075028262,
+      "grad_norm": 0.38064126138563015,
+      "learning_rate": 8.05875945757811e-06,
+      "loss": 0.3158,
+      "step": 2621
+    },
+    {
+      "epoch": 0.624025703575891,
+      "grad_norm": 0.3849786777328987,
+      "learning_rate": 8.057234407685563e-06,
+      "loss": 0.3875,
+      "step": 2622
+    },
+    {
+      "epoch": 0.6242636996489558,
+      "grad_norm": 0.3737306119830634,
+      "learning_rate": 8.05570890341235e-06,
+      "loss": 0.3512,
+      "step": 2623
+    },
+    {
+      "epoch": 0.6245016957220206,
+      "grad_norm": 0.4064964030178299,
+      "learning_rate": 8.054182944985198e-06,
+      "loss": 0.3145,
+      "step": 2624
+    },
+    {
+      "epoch": 0.6247396917950854,
+      "grad_norm": 0.38133945827871857,
+      "learning_rate": 8.052656532630905e-06,
+      "loss": 0.36,
+      "step": 2625
+    },
+    {
+      "epoch": 0.6249776878681502,
+      "grad_norm": 0.34831792324190414,
+      "learning_rate": 8.051129666576331e-06,
+      "loss": 0.4264,
+      "step": 2626
+    },
+    {
+      "epoch": 0.625215683941215,
+      "grad_norm": 0.38814660085579894,
+      "learning_rate": 8.049602347048408e-06,
+      "loss": 0.3214,
+      "step": 2627
+    },
+    {
+      "epoch": 0.6254536800142798,
+      "grad_norm": 0.4292810547509034,
+      "learning_rate": 8.048074574274132e-06,
+      "loss": 0.3469,
+      "step": 2628
+    },
+    {
+      "epoch": 0.6256916760873446,
+      "grad_norm": 0.36341426198547644,
+      "learning_rate": 8.04654634848057e-06,
+      "loss": 0.3922,
+      "step": 2629
+    },
+    {
+      "epoch": 0.6259296721604094,
+      "grad_norm": 0.3821109148057445,
+      "learning_rate": 8.045017669894851e-06,
+      "loss": 0.3777,
+      "step": 2630
+    },
+    {
+      "epoch": 0.6261676682334741,
+      "grad_norm": 0.37289488804653775,
+      "learning_rate": 8.043488538744177e-06,
+      "loss": 0.2943,
+      "step": 2631
+    },
+    {
+      "epoch": 0.626405664306539,
+      "grad_norm": 0.37258093559025074,
+      "learning_rate": 8.041958955255815e-06,
+      "loss": 0.342,
+      "step": 2632
+    },
+    {
+      "epoch": 0.6266436603796037,
+      "grad_norm": 0.3486184579114556,
+      "learning_rate": 8.040428919657095e-06,
+      "loss": 0.4144,
+      "step": 2633
+    },
+    {
+      "epoch": 0.6268816564526686,
+      "grad_norm": 0.3741035400050582,
+      "learning_rate": 8.038898432175424e-06,
+      "loss": 0.3246,
+      "step": 2634
+    },
+    {
+      "epoch": 0.6271196525257333,
+      "grad_norm": 0.40892825888298057,
+      "learning_rate": 8.037367493038265e-06,
+      "loss": 0.3332,
+      "step": 2635
+    },
+    {
+      "epoch": 0.6273576485987982,
+      "grad_norm": 0.36522383480533593,
+      "learning_rate": 8.035836102473155e-06,
+      "loss": 0.3566,
+      "step": 2636
+    },
+    {
+      "epoch": 0.6275956446718629,
+      "grad_norm": 0.3652591745540261,
+      "learning_rate": 8.0343042607077e-06,
+      "loss": 0.3765,
+      "step": 2637
+    },
+    {
+      "epoch": 0.6278336407449278,
+      "grad_norm": 0.35844676767075323,
+      "learning_rate": 8.032771967969566e-06,
+      "loss": 0.3158,
+      "step": 2638
+    },
+    {
+      "epoch": 0.6280716368179925,
+      "grad_norm": 0.3931558454911546,
+      "learning_rate": 8.03123922448649e-06,
+      "loss": 0.3568,
+      "step": 2639
+    },
+    {
+      "epoch": 0.6283096328910573,
+      "grad_norm": 0.36411863899036223,
+      "learning_rate": 8.029706030486274e-06,
+      "loss": 0.4586,
+      "step": 2640
+    },
+    {
+      "epoch": 0.6285476289641221,
+      "grad_norm": 0.3796092362122607,
+      "learning_rate": 8.028172386196794e-06,
+      "loss": 0.372,
+      "step": 2641
+    },
+    {
+      "epoch": 0.6287856250371869,
+      "grad_norm": 0.3761030511051456,
+      "learning_rate": 8.026638291845982e-06,
+      "loss": 0.3092,
+      "step": 2642
+    },
+    {
+      "epoch": 0.6290236211102517,
+      "grad_norm": 0.3699761519507502,
+      "learning_rate": 8.025103747661844e-06,
+      "loss": 0.3805,
+      "step": 2643
+    },
+    {
+      "epoch": 0.6292616171833165,
+      "grad_norm": 0.373140231509292,
+      "learning_rate": 8.023568753872453e-06,
+      "loss": 0.388,
+      "step": 2644
+    },
+    {
+      "epoch": 0.6294996132563813,
+      "grad_norm": 0.3637236022266386,
+      "learning_rate": 8.022033310705946e-06,
+      "loss": 0.3184,
+      "step": 2645
+    },
+    {
+      "epoch": 0.6297376093294461,
+      "grad_norm": 0.3915776794451119,
+      "learning_rate": 8.020497418390527e-06,
+      "loss": 0.3349,
+      "step": 2646
+    },
+    {
+      "epoch": 0.6299756054025109,
+      "grad_norm": 0.36532602213156506,
+      "learning_rate": 8.018961077154468e-06,
+      "loss": 0.3957,
+      "step": 2647
+    },
+    {
+      "epoch": 0.6302136014755757,
+      "grad_norm": 0.4140563457038236,
+      "learning_rate": 8.017424287226107e-06,
+      "loss": 0.415,
+      "step": 2648
+    },
+    {
+      "epoch": 0.6304515975486404,
+      "grad_norm": 0.3769113334994522,
+      "learning_rate": 8.01588704883385e-06,
+      "loss": 0.3145,
+      "step": 2649
+    },
+    {
+      "epoch": 0.6306895936217053,
+      "grad_norm": 1.5368218754503258,
+      "learning_rate": 8.014349362206167e-06,
+      "loss": 0.374,
+      "step": 2650
+    },
+    {
+      "epoch": 0.63092758969477,
+      "grad_norm": 0.38350013956172013,
+      "learning_rate": 8.012811227571597e-06,
+      "loss": 0.4003,
+      "step": 2651
+    },
+    {
+      "epoch": 0.6311655857678349,
+      "grad_norm": 0.3866132408928916,
+      "learning_rate": 8.011272645158747e-06,
+      "loss": 0.3235,
+      "step": 2652
+    },
+    {
+      "epoch": 0.6314035818408996,
+      "grad_norm": 0.40608371668229537,
+      "learning_rate": 8.009733615196287e-06,
+      "loss": 0.3244,
+      "step": 2653
+    },
+    {
+      "epoch": 0.6316415779139645,
+      "grad_norm": 0.38096062005900233,
+      "learning_rate": 8.008194137912955e-06,
+      "loss": 0.412,
+      "step": 2654
+    },
+    {
+      "epoch": 0.6318795739870292,
+      "grad_norm": 0.3862436056336853,
+      "learning_rate": 8.006654213537553e-06,
+      "loss": 0.3634,
+      "step": 2655
+    },
+    {
+      "epoch": 0.6321175700600941,
+      "grad_norm": 0.3976502335701635,
+      "learning_rate": 8.005113842298954e-06,
+      "loss": 0.3234,
+      "step": 2656
+    },
+    {
+      "epoch": 0.6323555661331588,
+      "grad_norm": 0.39127711057264086,
+      "learning_rate": 8.003573024426094e-06,
+      "loss": 0.3673,
+      "step": 2657
+    },
+    {
+      "epoch": 0.6325935622062236,
+      "grad_norm": 0.3896004400805801,
+      "learning_rate": 8.00203176014798e-06,
+      "loss": 0.4458,
+      "step": 2658
+    },
+    {
+      "epoch": 0.6328315582792884,
+      "grad_norm": 0.46398616972299095,
+      "learning_rate": 8.000490049693678e-06,
+      "loss": 0.3429,
+      "step": 2659
+    },
+    {
+      "epoch": 0.6330695543523532,
+      "grad_norm": 0.46831349058115457,
+      "learning_rate": 7.998947893292328e-06,
+      "loss": 0.3135,
+      "step": 2660
+    },
+    {
+      "epoch": 0.633307550425418,
+      "grad_norm": 0.4258334551178332,
+      "learning_rate": 7.99740529117313e-06,
+      "loss": 0.381,
+      "step": 2661
+    },
+    {
+      "epoch": 0.6335455464984828,
+      "grad_norm": 0.413976046137531,
+      "learning_rate": 7.995862243565352e-06,
+      "loss": 0.4138,
+      "step": 2662
+    },
+    {
+      "epoch": 0.6337835425715476,
+      "grad_norm": 0.43240881559747,
+      "learning_rate": 7.994318750698333e-06,
+      "loss": 0.3083,
+      "step": 2663
+    },
+    {
+      "epoch": 0.6340215386446124,
+      "grad_norm": 0.4246524529086203,
+      "learning_rate": 7.99277481280147e-06,
+      "loss": 0.3523,
+      "step": 2664
+    },
+    {
+      "epoch": 0.6342595347176772,
+      "grad_norm": 0.38578881152202554,
+      "learning_rate": 7.991230430104233e-06,
+      "loss": 0.3935,
+      "step": 2665
+    },
+    {
+      "epoch": 0.634497530790742,
+      "grad_norm": 0.7982507745714851,
+      "learning_rate": 7.989685602836155e-06,
+      "loss": 0.3685,
+      "step": 2666
+    },
+    {
+      "epoch": 0.6347355268638067,
+      "grad_norm": 0.39291885711093316,
+      "learning_rate": 7.988140331226835e-06,
+      "loss": 0.302,
+      "step": 2667
+    },
+    {
+      "epoch": 0.6349735229368716,
+      "grad_norm": 0.3774641923130378,
+      "learning_rate": 7.986594615505938e-06,
+      "loss": 0.3455,
+      "step": 2668
+    },
+    {
+      "epoch": 0.6352115190099363,
+      "grad_norm": 0.3849751604819079,
+      "learning_rate": 7.985048455903195e-06,
+      "loss": 0.3941,
+      "step": 2669
+    },
+    {
+      "epoch": 0.6354495150830012,
+      "grad_norm": 0.4028547987540656,
+      "learning_rate": 7.983501852648408e-06,
+      "loss": 0.3105,
+      "step": 2670
+    },
+    {
+      "epoch": 0.6356875111560659,
+      "grad_norm": 0.4175776893284019,
+      "learning_rate": 7.981954805971434e-06,
+      "loss": 0.3202,
+      "step": 2671
+    },
+    {
+      "epoch": 0.6359255072291308,
+      "grad_norm": 0.3948825669750941,
+      "learning_rate": 7.98040731610221e-06,
+      "loss": 0.386,
+      "step": 2672
+    },
+    {
+      "epoch": 0.6361635033021955,
+      "grad_norm": 0.37485861058760167,
+      "learning_rate": 7.978859383270723e-06,
+      "loss": 0.3553,
+      "step": 2673
+    },
+    {
+      "epoch": 0.6364014993752604,
+      "grad_norm": 0.38429792606989394,
+      "learning_rate": 7.97731100770704e-06,
+      "loss": 0.305,
+      "step": 2674
+    },
+    {
+      "epoch": 0.6366394954483251,
+      "grad_norm": 0.40243135178828765,
+      "learning_rate": 7.975762189641287e-06,
+      "loss": 0.3782,
+      "step": 2675
+    },
+    {
+      "epoch": 0.63687749152139,
+      "grad_norm": 0.3537177295193898,
+      "learning_rate": 7.974212929303655e-06,
+      "loss": 0.4115,
+      "step": 2676
+    },
+    {
+      "epoch": 0.6371154875944547,
+      "grad_norm": 0.4063244285595607,
+      "learning_rate": 7.972663226924404e-06,
+      "loss": 0.3373,
+      "step": 2677
+    },
+    {
+      "epoch": 0.6373534836675195,
+      "grad_norm": 0.7174554198990031,
+      "learning_rate": 7.971113082733855e-06,
+      "loss": 0.319,
+      "step": 2678
+    },
+    {
+      "epoch": 0.6375914797405843,
+      "grad_norm": 0.38291427234154246,
+      "learning_rate": 7.969562496962402e-06,
+      "loss": 0.4063,
+      "step": 2679
+    },
+    {
+      "epoch": 0.6378294758136491,
+      "grad_norm": 0.37205693582330984,
+      "learning_rate": 7.968011469840498e-06,
+      "loss": 0.3852,
+      "step": 2680
+    },
+    {
+      "epoch": 0.6380674718867139,
+      "grad_norm": 0.37289319443279956,
+      "learning_rate": 7.966460001598666e-06,
+      "loss": 0.3108,
+      "step": 2681
+    },
+    {
+      "epoch": 0.6383054679597787,
+      "grad_norm": 0.36566089153086934,
+      "learning_rate": 7.96490809246749e-06,
+      "loss": 0.3616,
+      "step": 2682
+    },
+    {
+      "epoch": 0.6385434640328435,
+      "grad_norm": 0.3792165993672128,
+      "learning_rate": 7.963355742677622e-06,
+      "loss": 0.3968,
+      "step": 2683
+    },
+    {
+      "epoch": 0.6387814601059083,
+      "grad_norm": 0.3798446258828693,
+      "learning_rate": 7.961802952459782e-06,
+      "loss": 0.3367,
+      "step": 2684
+    },
+    {
+      "epoch": 0.639019456178973,
+      "grad_norm": 0.3490183798631681,
+      "learning_rate": 7.96024972204475e-06,
+      "loss": 0.295,
+      "step": 2685
+    },
+    {
+      "epoch": 0.6392574522520379,
+      "grad_norm": 0.3964584193841697,
+      "learning_rate": 7.958696051663378e-06,
+      "loss": 0.3989,
+      "step": 2686
+    },
+    {
+      "epoch": 0.6394954483251026,
+      "grad_norm": 0.41876375542959604,
+      "learning_rate": 7.957141941546579e-06,
+      "loss": 0.4196,
+      "step": 2687
+    },
+    {
+      "epoch": 0.6397334443981675,
+      "grad_norm": 0.3704391565272174,
+      "learning_rate": 7.95558739192533e-06,
+      "loss": 0.3395,
+      "step": 2688
+    },
+    {
+      "epoch": 0.6399714404712322,
+      "grad_norm": 0.38705193099976026,
+      "learning_rate": 7.954032403030676e-06,
+      "loss": 0.3671,
+      "step": 2689
+    },
+    {
+      "epoch": 0.6402094365442971,
+      "grad_norm": 0.3439386120652561,
+      "learning_rate": 7.952476975093729e-06,
+      "loss": 0.3779,
+      "step": 2690
+    },
+    {
+      "epoch": 0.6404474326173618,
+      "grad_norm": 0.40968077374028966,
+      "learning_rate": 7.950921108345663e-06,
+      "loss": 0.3443,
+      "step": 2691
+    },
+    {
+      "epoch": 0.6406854286904267,
+      "grad_norm": 0.40943399377681217,
+      "learning_rate": 7.949364803017716e-06,
+      "loss": 0.3285,
+      "step": 2692
+    },
+    {
+      "epoch": 0.6409234247634914,
+      "grad_norm": 0.35009359805640566,
+      "learning_rate": 7.947808059341198e-06,
+      "loss": 0.3749,
+      "step": 2693
+    },
+    {
+      "epoch": 0.6411614208365563,
+      "grad_norm": 0.41593770811256764,
+      "learning_rate": 7.946250877547477e-06,
+      "loss": 0.4014,
+      "step": 2694
+    },
+    {
+      "epoch": 0.641399416909621,
+      "grad_norm": 0.37165753141343205,
+      "learning_rate": 7.944693257867988e-06,
+      "loss": 0.3249,
+      "step": 2695
+    },
+    {
+      "epoch": 0.6416374129826858,
+      "grad_norm": 0.36739798172367116,
+      "learning_rate": 7.943135200534231e-06,
+      "loss": 0.3509,
+      "step": 2696
+    },
+    {
+      "epoch": 0.6418754090557506,
+      "grad_norm": 0.3724656741195384,
+      "learning_rate": 7.941576705777775e-06,
+      "loss": 0.4075,
+      "step": 2697
+    },
+    {
+      "epoch": 0.6421134051288154,
+      "grad_norm": 0.36653430005871285,
+      "learning_rate": 7.940017773830251e-06,
+      "loss": 0.3887,
+      "step": 2698
+    },
+    {
+      "epoch": 0.6423514012018802,
+      "grad_norm": 0.3954242148262836,
+      "learning_rate": 7.93845840492335e-06,
+      "loss": 0.2837,
+      "step": 2699
+    },
+    {
+      "epoch": 0.642589397274945,
+      "grad_norm": 0.37217288826320916,
+      "learning_rate": 7.936898599288837e-06,
+      "loss": 0.367,
+      "step": 2700
+    },
+    {
+      "epoch": 0.6428273933480098,
+      "grad_norm": 0.3433351247909528,
+      "learning_rate": 7.93533835715854e-06,
+      "loss": 0.413,
+      "step": 2701
+    },
+    {
+      "epoch": 0.6430653894210746,
+      "grad_norm": 0.3710719613065313,
+      "learning_rate": 7.933777678764342e-06,
+      "loss": 0.3481,
+      "step": 2702
+    },
+    {
+      "epoch": 0.6433033854941393,
+      "grad_norm": 0.4051379668160816,
+      "learning_rate": 7.932216564338207e-06,
+      "loss": 0.3487,
+      "step": 2703
+    },
+    {
+      "epoch": 0.6435413815672042,
+      "grad_norm": 0.35261923047482774,
+      "learning_rate": 7.930655014112149e-06,
+      "loss": 0.3673,
+      "step": 2704
+    },
+    {
+      "epoch": 0.6437793776402689,
+      "grad_norm": 0.37561470708818495,
+      "learning_rate": 7.929093028318254e-06,
+      "loss": 0.3952,
+      "step": 2705
+    },
+    {
+      "epoch": 0.6440173737133338,
+      "grad_norm": 0.38085667134847434,
+      "learning_rate": 7.927530607188674e-06,
+      "loss": 0.2997,
+      "step": 2706
+    },
+    {
+      "epoch": 0.6442553697863985,
+      "grad_norm": 0.38309188860484494,
+      "learning_rate": 7.925967750955621e-06,
+      "loss": 0.3318,
+      "step": 2707
+    },
+    {
+      "epoch": 0.6444933658594634,
+      "grad_norm": 0.35539044141417847,
+      "learning_rate": 7.924404459851376e-06,
+      "loss": 0.4216,
+      "step": 2708
+    },
+    {
+      "epoch": 0.6447313619325281,
+      "grad_norm": 0.36545999682638586,
+      "learning_rate": 7.92284073410828e-06,
+      "loss": 0.3565,
+      "step": 2709
+    },
+    {
+      "epoch": 0.644969358005593,
+      "grad_norm": 0.3853826020640622,
+      "learning_rate": 7.921276573958747e-06,
+      "loss": 0.3193,
+      "step": 2710
+    },
+    {
+      "epoch": 0.6452073540786577,
+      "grad_norm": 0.370265598492799,
+      "learning_rate": 7.919711979635245e-06,
+      "loss": 0.3624,
+      "step": 2711
+    },
+    {
+      "epoch": 0.6454453501517224,
+      "grad_norm": 0.38428982530476763,
+      "learning_rate": 7.918146951370312e-06,
+      "loss": 0.368,
+      "step": 2712
+    },
+    {
+      "epoch": 0.6456833462247873,
+      "grad_norm": 0.3863324681379504,
+      "learning_rate": 7.916581489396551e-06,
+      "loss": 0.3382,
+      "step": 2713
+    },
+    {
+      "epoch": 0.645921342297852,
+      "grad_norm": 0.40476758870600216,
+      "learning_rate": 7.915015593946627e-06,
+      "loss": 0.3452,
+      "step": 2714
+    },
+    {
+      "epoch": 0.6461593383709169,
+      "grad_norm": 0.350755584302665,
+      "learning_rate": 7.913449265253272e-06,
+      "loss": 0.3541,
+      "step": 2715
+    },
+    {
+      "epoch": 0.6463973344439816,
+      "grad_norm": 0.3917385422279502,
+      "learning_rate": 7.911882503549282e-06,
+      "loss": 0.3615,
+      "step": 2716
+    },
+    {
+      "epoch": 0.6466353305170465,
+      "grad_norm": 0.4148508130598785,
+      "learning_rate": 7.910315309067515e-06,
+      "loss": 0.3237,
+      "step": 2717
+    },
+    {
+      "epoch": 0.6468733265901112,
+      "grad_norm": 0.3655588510807828,
+      "learning_rate": 7.908747682040893e-06,
+      "loss": 0.3717,
+      "step": 2718
+    },
+    {
+      "epoch": 0.6471113226631761,
+      "grad_norm": 0.3786662704253251,
+      "learning_rate": 7.907179622702409e-06,
+      "loss": 0.3995,
+      "step": 2719
+    },
+    {
+      "epoch": 0.6473493187362408,
+      "grad_norm": 0.41077148959741994,
+      "learning_rate": 7.905611131285114e-06,
+      "loss": 0.3225,
+      "step": 2720
+    },
+    {
+      "epoch": 0.6475873148093056,
+      "grad_norm": 0.3845293510123214,
+      "learning_rate": 7.904042208022121e-06,
+      "loss": 0.3184,
+      "step": 2721
+    },
+    {
+      "epoch": 0.6478253108823704,
+      "grad_norm": 0.38026816465993274,
+      "learning_rate": 7.902472853146614e-06,
+      "loss": 0.4047,
+      "step": 2722
+    },
+    {
+      "epoch": 0.6480633069554352,
+      "grad_norm": 0.3831104022471865,
+      "learning_rate": 7.90090306689184e-06,
+      "loss": 0.3919,
+      "step": 2723
+    },
+    {
+      "epoch": 0.6483013030285,
+      "grad_norm": 0.39343951127809484,
+      "learning_rate": 7.899332849491101e-06,
+      "loss": 0.2942,
+      "step": 2724
+    },
+    {
+      "epoch": 0.6485392991015648,
+      "grad_norm": 0.37517985298715045,
+      "learning_rate": 7.897762201177777e-06,
+      "loss": 0.3449,
+      "step": 2725
+    },
+    {
+      "epoch": 0.6487772951746296,
+      "grad_norm": 0.36054421682014914,
+      "learning_rate": 7.896191122185302e-06,
+      "loss": 0.3968,
+      "step": 2726
+    },
+    {
+      "epoch": 0.6490152912476944,
+      "grad_norm": 0.38545956806454257,
+      "learning_rate": 7.894619612747177e-06,
+      "loss": 0.3261,
+      "step": 2727
+    },
+    {
+      "epoch": 0.6492532873207592,
+      "grad_norm": 0.4049278658685888,
+      "learning_rate": 7.89304767309697e-06,
+      "loss": 0.3517,
+      "step": 2728
+    },
+    {
+      "epoch": 0.649491283393824,
+      "grad_norm": 0.377732909086466,
+      "learning_rate": 7.891475303468307e-06,
+      "loss": 0.377,
+      "step": 2729
+    },
+    {
+      "epoch": 0.6497292794668887,
+      "grad_norm": 0.35831707426820597,
+      "learning_rate": 7.889902504094883e-06,
+      "loss": 0.386,
+      "step": 2730
+    },
+    {
+      "epoch": 0.6499672755399536,
+      "grad_norm": 0.38647507461352604,
+      "learning_rate": 7.888329275210454e-06,
+      "loss": 0.3116,
+      "step": 2731
+    },
+    {
+      "epoch": 0.6502052716130183,
+      "grad_norm": 0.41222519034063054,
+      "learning_rate": 7.88675561704884e-06,
+      "loss": 0.3485,
+      "step": 2732
+    },
+    {
+      "epoch": 0.6504432676860832,
+      "grad_norm": 0.3540005364042782,
+      "learning_rate": 7.885181529843928e-06,
+      "loss": 0.3811,
+      "step": 2733
+    },
+    {
+      "epoch": 0.6506812637591479,
+      "grad_norm": 0.3889803639057776,
+      "learning_rate": 7.883607013829664e-06,
+      "loss": 0.3164,
+      "step": 2734
+    },
+    {
+      "epoch": 0.6509192598322128,
+      "grad_norm": 0.3825132936787645,
+      "learning_rate": 7.88203206924006e-06,
+      "loss": 0.3355,
+      "step": 2735
+    },
+    {
+      "epoch": 0.6511572559052775,
+      "grad_norm": 0.3600797185755466,
+      "learning_rate": 7.880456696309194e-06,
+      "loss": 0.3822,
+      "step": 2736
+    },
+    {
+      "epoch": 0.6513952519783424,
+      "grad_norm": 0.36788983076651144,
+      "learning_rate": 7.878880895271203e-06,
+      "loss": 0.4035,
+      "step": 2737
+    },
+    {
+      "epoch": 0.6516332480514071,
+      "grad_norm": 0.36838811546682065,
+      "learning_rate": 7.87730466636029e-06,
+      "loss": 0.3201,
+      "step": 2738
+    },
+    {
+      "epoch": 0.651871244124472,
+      "grad_norm": 0.4199311085900462,
+      "learning_rate": 7.875728009810723e-06,
+      "loss": 0.3354,
+      "step": 2739
+    },
+    {
+      "epoch": 0.6521092401975367,
+      "grad_norm": 0.3700461680317869,
+      "learning_rate": 7.874150925856832e-06,
+      "loss": 0.3907,
+      "step": 2740
+    },
+    {
+      "epoch": 0.6523472362706015,
+      "grad_norm": 0.3658208922433521,
+      "learning_rate": 7.87257341473301e-06,
+      "loss": 0.3436,
+      "step": 2741
+    },
+    {
+      "epoch": 0.6525852323436663,
+      "grad_norm": 0.3673071582783027,
+      "learning_rate": 7.870995476673716e-06,
+      "loss": 0.3026,
+      "step": 2742
+    },
+    {
+      "epoch": 0.6528232284167311,
+      "grad_norm": 0.36996570740087614,
+      "learning_rate": 7.869417111913469e-06,
+      "loss": 0.3494,
+      "step": 2743
+    },
+    {
+      "epoch": 0.6530612244897959,
+      "grad_norm": 0.387133905688407,
+      "learning_rate": 7.867838320686852e-06,
+      "loss": 0.4002,
+      "step": 2744
+    },
+    {
+      "epoch": 0.6532992205628607,
+      "grad_norm": 0.4136666892086636,
+      "learning_rate": 7.866259103228513e-06,
+      "loss": 0.3187,
+      "step": 2745
+    },
+    {
+      "epoch": 0.6535372166359255,
+      "grad_norm": 0.39398446442473684,
+      "learning_rate": 7.864679459773165e-06,
+      "loss": 0.3197,
+      "step": 2746
+    },
+    {
+      "epoch": 0.6537752127089903,
+      "grad_norm": 0.3856438918664348,
+      "learning_rate": 7.863099390555579e-06,
+      "loss": 0.3823,
+      "step": 2747
+    },
+    {
+      "epoch": 0.654013208782055,
+      "grad_norm": 0.3908108331472674,
+      "learning_rate": 7.861518895810597e-06,
+      "loss": 0.36,
+      "step": 2748
+    },
+    {
+      "epoch": 0.6542512048551199,
+      "grad_norm": 0.3906604552323348,
+      "learning_rate": 7.859937975773113e-06,
+      "loss": 0.3083,
+      "step": 2749
+    },
+    {
+      "epoch": 0.6544892009281846,
+      "grad_norm": 0.3644262557255445,
+      "learning_rate": 7.858356630678095e-06,
+      "loss": 0.3834,
+      "step": 2750
+    },
+    {
+      "epoch": 0.6547271970012495,
+      "grad_norm": 0.3859873524132499,
+      "learning_rate": 7.85677486076057e-06,
+      "loss": 0.4009,
+      "step": 2751
+    },
+    {
+      "epoch": 0.6549651930743142,
+      "grad_norm": 0.4012225445856221,
+      "learning_rate": 7.855192666255627e-06,
+      "loss": 0.3174,
+      "step": 2752
+    },
+    {
+      "epoch": 0.6552031891473791,
+      "grad_norm": 0.3894743297706602,
+      "learning_rate": 7.853610047398422e-06,
+      "loss": 0.3163,
+      "step": 2753
+    },
+    {
+      "epoch": 0.6554411852204438,
+      "grad_norm": 0.36865693864488674,
+      "learning_rate": 7.852027004424166e-06,
+      "loss": 0.4053,
+      "step": 2754
+    },
+    {
+      "epoch": 0.6556791812935087,
+      "grad_norm": 0.3871705949523487,
+      "learning_rate": 7.850443537568142e-06,
+      "loss": 0.3511,
+      "step": 2755
+    },
+    {
+      "epoch": 0.6559171773665734,
+      "grad_norm": 0.40039522901635954,
+      "learning_rate": 7.848859647065692e-06,
+      "loss": 0.3183,
+      "step": 2756
+    },
+    {
+      "epoch": 0.6561551734396383,
+      "grad_norm": 0.3899094558661615,
+      "learning_rate": 7.847275333152222e-06,
+      "loss": 0.3536,
+      "step": 2757
+    },
+    {
+      "epoch": 0.656393169512703,
+      "grad_norm": 0.38854298448856595,
+      "learning_rate": 7.845690596063198e-06,
+      "loss": 0.3992,
+      "step": 2758
+    },
+    {
+      "epoch": 0.6566311655857678,
+      "grad_norm": 0.37350144359617254,
+      "learning_rate": 7.844105436034156e-06,
+      "loss": 0.3449,
+      "step": 2759
+    },
+    {
+      "epoch": 0.6568691616588326,
+      "grad_norm": 0.44114084013334837,
+      "learning_rate": 7.842519853300683e-06,
+      "loss": 0.2977,
+      "step": 2760
+    },
+    {
+      "epoch": 0.6571071577318974,
+      "grad_norm": 0.35999935925172966,
+      "learning_rate": 7.840933848098437e-06,
+      "loss": 0.3916,
+      "step": 2761
+    },
+    {
+      "epoch": 0.6573451538049622,
+      "grad_norm": 0.39899431523147383,
+      "learning_rate": 7.839347420663143e-06,
+      "loss": 0.4253,
+      "step": 2762
+    },
+    {
+      "epoch": 0.657583149878027,
+      "grad_norm": 0.3511330110093876,
+      "learning_rate": 7.837760571230582e-06,
+      "loss": 0.3323,
+      "step": 2763
+    },
+    {
+      "epoch": 0.6578211459510918,
+      "grad_norm": 0.4461907551327487,
+      "learning_rate": 7.836173300036594e-06,
+      "loss": 0.3239,
+      "step": 2764
+    },
+    {
+      "epoch": 0.6580591420241566,
+      "grad_norm": 0.39054055621536493,
+      "learning_rate": 7.83458560731709e-06,
+      "loss": 0.4205,
+      "step": 2765
+    },
+    {
+      "epoch": 0.6582971380972213,
+      "grad_norm": 0.3769385946466861,
+      "learning_rate": 7.832997493308043e-06,
+      "loss": 0.3511,
+      "step": 2766
+    },
+    {
+      "epoch": 0.6585351341702862,
+      "grad_norm": 0.3986011878362815,
+      "learning_rate": 7.831408958245483e-06,
+      "loss": 0.296,
+      "step": 2767
+    },
+    {
+      "epoch": 0.6587731302433509,
+      "grad_norm": 0.3880140645019253,
+      "learning_rate": 7.829820002365504e-06,
+      "loss": 0.3771,
+      "step": 2768
+    },
+    {
+      "epoch": 0.6590111263164158,
+      "grad_norm": 0.38212257026061813,
+      "learning_rate": 7.828230625904269e-06,
+      "loss": 0.423,
+      "step": 2769
+    },
+    {
+      "epoch": 0.6592491223894805,
+      "grad_norm": 0.38852972753947673,
+      "learning_rate": 7.826640829097994e-06,
+      "loss": 0.3055,
+      "step": 2770
+    },
+    {
+      "epoch": 0.6594871184625454,
+      "grad_norm": 0.392631438114691,
+      "learning_rate": 7.825050612182965e-06,
+      "loss": 0.3385,
+      "step": 2771
+    },
+    {
+      "epoch": 0.6597251145356101,
+      "grad_norm": 0.37085742233686686,
+      "learning_rate": 7.823459975395527e-06,
+      "loss": 0.3848,
+      "step": 2772
+    },
+    {
+      "epoch": 0.659963110608675,
+      "grad_norm": 0.3834859007180493,
+      "learning_rate": 7.821868918972087e-06,
+      "loss": 0.3313,
+      "step": 2773
+    },
+    {
+      "epoch": 0.6602011066817397,
+      "grad_norm": 0.3822005143101434,
+      "learning_rate": 7.820277443149114e-06,
+      "loss": 0.3139,
+      "step": 2774
+    },
+    {
+      "epoch": 0.6604391027548046,
+      "grad_norm": 0.3751827860273947,
+      "learning_rate": 7.818685548163144e-06,
+      "loss": 0.3772,
+      "step": 2775
+    },
+    {
+      "epoch": 0.6606770988278693,
+      "grad_norm": 0.36208369988094957,
+      "learning_rate": 7.817093234250772e-06,
+      "loss": 0.4038,
+      "step": 2776
+    },
+    {
+      "epoch": 0.6609150949009341,
+      "grad_norm": 0.394839034258573,
+      "learning_rate": 7.815500501648654e-06,
+      "loss": 0.3098,
+      "step": 2777
+    },
+    {
+      "epoch": 0.6611530909739989,
+      "grad_norm": 0.37324421123913026,
+      "learning_rate": 7.813907350593509e-06,
+      "loss": 0.2987,
+      "step": 2778
+    },
+    {
+      "epoch": 0.6613910870470637,
+      "grad_norm": 0.3522307791159532,
+      "learning_rate": 7.812313781322119e-06,
+      "loss": 0.3708,
+      "step": 2779
+    },
+    {
+      "epoch": 0.6616290831201285,
+      "grad_norm": 0.38147797701046887,
+      "learning_rate": 7.810719794071326e-06,
+      "loss": 0.3826,
+      "step": 2780
+    },
+    {
+      "epoch": 0.6618670791931933,
+      "grad_norm": 0.3818587837039667,
+      "learning_rate": 7.809125389078038e-06,
+      "loss": 0.306,
+      "step": 2781
+    },
+    {
+      "epoch": 0.6621050752662581,
+      "grad_norm": 0.4383170403521546,
+      "learning_rate": 7.807530566579225e-06,
+      "loss": 0.3566,
+      "step": 2782
+    },
+    {
+      "epoch": 0.6623430713393229,
+      "grad_norm": 0.35732395462798827,
+      "learning_rate": 7.805935326811913e-06,
+      "loss": 0.4335,
+      "step": 2783
+    },
+    {
+      "epoch": 0.6625810674123876,
+      "grad_norm": 0.3504216615398166,
+      "learning_rate": 7.804339670013196e-06,
+      "loss": 0.3143,
+      "step": 2784
+    },
+    {
+      "epoch": 0.6628190634854525,
+      "grad_norm": 0.3661223386576208,
+      "learning_rate": 7.802743596420228e-06,
+      "loss": 0.3529,
+      "step": 2785
+    },
+    {
+      "epoch": 0.6630570595585172,
+      "grad_norm": 0.3723251622104599,
+      "learning_rate": 7.801147106270227e-06,
+      "loss": 0.3883,
+      "step": 2786
+    },
+    {
+      "epoch": 0.6632950556315821,
+      "grad_norm": 0.36878452917502985,
+      "learning_rate": 7.799550199800468e-06,
+      "loss": 0.421,
+      "step": 2787
+    },
+    {
+      "epoch": 0.6635330517046468,
+      "grad_norm": 0.3618774249811299,
+      "learning_rate": 7.797952877248289e-06,
+      "loss": 0.3128,
+      "step": 2788
+    },
+    {
+      "epoch": 0.6637710477777117,
+      "grad_norm": 0.4123556625672267,
+      "learning_rate": 7.796355138851098e-06,
+      "loss": 0.3594,
+      "step": 2789
+    },
+    {
+      "epoch": 0.6640090438507764,
+      "grad_norm": 0.335938751220196,
+      "learning_rate": 7.794756984846353e-06,
+      "loss": 0.4006,
+      "step": 2790
+    },
+    {
+      "epoch": 0.6642470399238413,
+      "grad_norm": 0.38401019782464624,
+      "learning_rate": 7.793158415471582e-06,
+      "loss": 0.3682,
+      "step": 2791
+    },
+    {
+      "epoch": 0.664485035996906,
+      "grad_norm": 0.35133531035355053,
+      "learning_rate": 7.791559430964371e-06,
+      "loss": 0.2978,
+      "step": 2792
+    },
+    {
+      "epoch": 0.6647230320699709,
+      "grad_norm": 0.39741836114549955,
+      "learning_rate": 7.789960031562368e-06,
+      "loss": 0.3708,
+      "step": 2793
+    },
+    {
+      "epoch": 0.6649610281430356,
+      "grad_norm": 0.3886500686427725,
+      "learning_rate": 7.788360217503284e-06,
+      "loss": 0.4344,
+      "step": 2794
+    },
+    {
+      "epoch": 0.6651990242161004,
+      "grad_norm": 0.38181964678908115,
+      "learning_rate": 7.786759989024891e-06,
+      "loss": 0.3091,
+      "step": 2795
+    },
+    {
+      "epoch": 0.6654370202891652,
+      "grad_norm": 0.3575252771082044,
+      "learning_rate": 7.785159346365024e-06,
+      "loss": 0.3533,
+      "step": 2796
+    },
+    {
+      "epoch": 0.66567501636223,
+      "grad_norm": 0.35735170640215147,
+      "learning_rate": 7.783558289761575e-06,
+      "loss": 0.4113,
+      "step": 2797
+    },
+    {
+      "epoch": 0.6659130124352948,
+      "grad_norm": 0.3802244001409177,
+      "learning_rate": 7.781956819452503e-06,
+      "loss": 0.3463,
+      "step": 2798
+    },
+    {
+      "epoch": 0.6661510085083596,
+      "grad_norm": 0.3589197801986663,
+      "learning_rate": 7.780354935675824e-06,
+      "loss": 0.3054,
+      "step": 2799
+    },
+    {
+      "epoch": 0.6663890045814244,
+      "grad_norm": 0.37241774516966863,
+      "learning_rate": 7.778752638669621e-06,
+      "loss": 0.344,
+      "step": 2800
+    },
+    {
+      "epoch": 0.6666270006544892,
+      "grad_norm": 0.38446901953737084,
+      "learning_rate": 7.777149928672032e-06,
+      "loss": 0.389,
+      "step": 2801
+    },
+    {
+      "epoch": 0.666864996727554,
+      "grad_norm": 0.3748374371731264,
+      "learning_rate": 7.775546805921259e-06,
+      "loss": 0.2864,
+      "step": 2802
+    },
+    {
+      "epoch": 0.6671029928006188,
+      "grad_norm": 0.4043541069673673,
+      "learning_rate": 7.773943270655568e-06,
+      "loss": 0.3045,
+      "step": 2803
+    },
+    {
+      "epoch": 0.6673409888736835,
+      "grad_norm": 0.4320616556631244,
+      "learning_rate": 7.772339323113283e-06,
+      "loss": 0.3828,
+      "step": 2804
+    },
+    {
+      "epoch": 0.6675789849467484,
+      "grad_norm": 0.3884296375705101,
+      "learning_rate": 7.770734963532791e-06,
+      "loss": 0.3935,
+      "step": 2805
+    },
+    {
+      "epoch": 0.6678169810198131,
+      "grad_norm": 0.3947257784010564,
+      "learning_rate": 7.769130192152538e-06,
+      "loss": 0.3206,
+      "step": 2806
+    },
+    {
+      "epoch": 0.668054977092878,
+      "grad_norm": 0.4031463873354111,
+      "learning_rate": 7.767525009211032e-06,
+      "loss": 0.3593,
+      "step": 2807
+    },
+    {
+      "epoch": 0.6682929731659427,
+      "grad_norm": 0.33474899996590485,
+      "learning_rate": 7.765919414946846e-06,
+      "loss": 0.387,
+      "step": 2808
+    },
+    {
+      "epoch": 0.6685309692390076,
+      "grad_norm": 0.3959331136357287,
+      "learning_rate": 7.76431340959861e-06,
+      "loss": 0.3433,
+      "step": 2809
+    },
+    {
+      "epoch": 0.6687689653120723,
+      "grad_norm": 0.3963425292776564,
+      "learning_rate": 7.762706993405014e-06,
+      "loss": 0.2888,
+      "step": 2810
+    },
+    {
+      "epoch": 0.6690069613851372,
+      "grad_norm": 0.37455109669088693,
+      "learning_rate": 7.761100166604814e-06,
+      "loss": 0.3744,
+      "step": 2811
+    },
+    {
+      "epoch": 0.6692449574582019,
+      "grad_norm": 0.37164146720882896,
+      "learning_rate": 7.759492929436821e-06,
+      "loss": 0.3976,
+      "step": 2812
+    },
+    {
+      "epoch": 0.6694829535312667,
+      "grad_norm": 0.37019035117029214,
+      "learning_rate": 7.757885282139913e-06,
+      "loss": 0.3273,
+      "step": 2813
+    },
+    {
+      "epoch": 0.6697209496043315,
+      "grad_norm": 0.3867936235753999,
+      "learning_rate": 7.756277224953027e-06,
+      "loss": 0.3476,
+      "step": 2814
+    },
+    {
+      "epoch": 0.6699589456773963,
+      "grad_norm": 0.37145512873107056,
+      "learning_rate": 7.754668758115157e-06,
+      "loss": 0.3941,
+      "step": 2815
+    },
+    {
+      "epoch": 0.6701969417504611,
+      "grad_norm": 0.36270472677005344,
+      "learning_rate": 7.753059881865361e-06,
+      "loss": 0.3415,
+      "step": 2816
+    },
+    {
+      "epoch": 0.6704349378235259,
+      "grad_norm": 0.4044012125764732,
+      "learning_rate": 7.751450596442761e-06,
+      "loss": 0.3126,
+      "step": 2817
+    },
+    {
+      "epoch": 0.6706729338965907,
+      "grad_norm": 0.4117053779057893,
+      "learning_rate": 7.749840902086534e-06,
+      "loss": 0.3767,
+      "step": 2818
+    },
+    {
+      "epoch": 0.6709109299696555,
+      "grad_norm": 0.3709103066934568,
+      "learning_rate": 7.748230799035922e-06,
+      "loss": 0.4183,
+      "step": 2819
+    },
+    {
+      "epoch": 0.6711489260427203,
+      "grad_norm": 0.36765736344691075,
+      "learning_rate": 7.746620287530224e-06,
+      "loss": 0.3289,
+      "step": 2820
+    },
+    {
+      "epoch": 0.6713869221157851,
+      "grad_norm": 0.39213657203388363,
+      "learning_rate": 7.745009367808805e-06,
+      "loss": 0.3612,
+      "step": 2821
+    },
+    {
+      "epoch": 0.6716249181888498,
+      "grad_norm": 0.3694609673330917,
+      "learning_rate": 7.743398040111085e-06,
+      "loss": 0.4092,
+      "step": 2822
+    },
+    {
+      "epoch": 0.6718629142619147,
+      "grad_norm": 0.36370204232238,
+      "learning_rate": 7.741786304676546e-06,
+      "loss": 0.3647,
+      "step": 2823
+    },
+    {
+      "epoch": 0.6721009103349794,
+      "grad_norm": 0.39103037281531944,
+      "learning_rate": 7.740174161744734e-06,
+      "loss": 0.3049,
+      "step": 2824
+    },
+    {
+      "epoch": 0.6723389064080443,
+      "grad_norm": 0.4294014674026649,
+      "learning_rate": 7.738561611555256e-06,
+      "loss": 0.3702,
+      "step": 2825
+    },
+    {
+      "epoch": 0.672576902481109,
+      "grad_norm": 0.378252161646541,
+      "learning_rate": 7.736948654347771e-06,
+      "loss": 0.429,
+      "step": 2826
+    },
+    {
+      "epoch": 0.6728148985541739,
+      "grad_norm": 0.3733849472381206,
+      "learning_rate": 7.735335290362008e-06,
+      "loss": 0.3083,
+      "step": 2827
+    },
+    {
+      "epoch": 0.6730528946272386,
+      "grad_norm": 0.3875790384170856,
+      "learning_rate": 7.733721519837751e-06,
+      "loss": 0.3125,
+      "step": 2828
+    },
+    {
+      "epoch": 0.6732908907003035,
+      "grad_norm": 0.357812704637255,
+      "learning_rate": 7.732107343014848e-06,
+      "loss": 0.4005,
+      "step": 2829
+    },
+    {
+      "epoch": 0.6735288867733682,
+      "grad_norm": 0.3644020205564977,
+      "learning_rate": 7.730492760133204e-06,
+      "loss": 0.4116,
+      "step": 2830
+    },
+    {
+      "epoch": 0.673766882846433,
+      "grad_norm": 0.39330628647179383,
+      "learning_rate": 7.728877771432787e-06,
+      "loss": 0.322,
+      "step": 2831
+    },
+    {
+      "epoch": 0.6740048789194978,
+      "grad_norm": 0.39515618942825725,
+      "learning_rate": 7.727262377153625e-06,
+      "loss": 0.3311,
+      "step": 2832
+    },
+    {
+      "epoch": 0.6742428749925626,
+      "grad_norm": 0.3826839966849655,
+      "learning_rate": 7.725646577535803e-06,
+      "loss": 0.4147,
+      "step": 2833
+    },
+    {
+      "epoch": 0.6744808710656274,
+      "grad_norm": 0.40336985165060796,
+      "learning_rate": 7.724030372819473e-06,
+      "loss": 0.3492,
+      "step": 2834
+    },
+    {
+      "epoch": 0.6747188671386922,
+      "grad_norm": 0.40161595680714757,
+      "learning_rate": 7.722413763244837e-06,
+      "loss": 0.3058,
+      "step": 2835
+    },
+    {
+      "epoch": 0.674956863211757,
+      "grad_norm": 0.37111456991939507,
+      "learning_rate": 7.720796749052169e-06,
+      "loss": 0.3988,
+      "step": 2836
+    },
+    {
+      "epoch": 0.6751948592848218,
+      "grad_norm": 0.4093260967053505,
+      "learning_rate": 7.719179330481791e-06,
+      "loss": 0.3985,
+      "step": 2837
+    },
+    {
+      "epoch": 0.6754328553578866,
+      "grad_norm": 0.37469388622058286,
+      "learning_rate": 7.7175615077741e-06,
+      "loss": 0.3206,
+      "step": 2838
+    },
+    {
+      "epoch": 0.6756708514309514,
+      "grad_norm": 0.36305043278624066,
+      "learning_rate": 7.715943281169539e-06,
+      "loss": 0.3477,
+      "step": 2839
+    },
+    {
+      "epoch": 0.6759088475040161,
+      "grad_norm": 0.36935236818452855,
+      "learning_rate": 7.714324650908615e-06,
+      "loss": 0.3998,
+      "step": 2840
+    },
+    {
+      "epoch": 0.676146843577081,
+      "grad_norm": 0.3948705853817137,
+      "learning_rate": 7.7127056172319e-06,
+      "loss": 0.3422,
+      "step": 2841
+    },
+    {
+      "epoch": 0.6763848396501457,
+      "grad_norm": 0.3796382485482362,
+      "learning_rate": 7.711086180380021e-06,
+      "loss": 0.3149,
+      "step": 2842
+    },
+    {
+      "epoch": 0.6766228357232106,
+      "grad_norm": 0.3616110674940236,
+      "learning_rate": 7.709466340593666e-06,
+      "loss": 0.364,
+      "step": 2843
+    },
+    {
+      "epoch": 0.6768608317962753,
+      "grad_norm": 0.3662500931977446,
+      "learning_rate": 7.707846098113583e-06,
+      "loss": 0.3967,
+      "step": 2844
+    },
+    {
+      "epoch": 0.6770988278693402,
+      "grad_norm": 0.37281181697106686,
+      "learning_rate": 7.706225453180583e-06,
+      "loss": 0.3279,
+      "step": 2845
+    },
+    {
+      "epoch": 0.6773368239424049,
+      "grad_norm": 0.3855893672663316,
+      "learning_rate": 7.704604406035531e-06,
+      "loss": 0.3496,
+      "step": 2846
+    },
+    {
+      "epoch": 0.6775748200154698,
+      "grad_norm": 0.3889507681326313,
+      "learning_rate": 7.702982956919356e-06,
+      "loss": 0.4173,
+      "step": 2847
+    },
+    {
+      "epoch": 0.6778128160885345,
+      "grad_norm": 0.3971306511498377,
+      "learning_rate": 7.701361106073044e-06,
+      "loss": 0.3444,
+      "step": 2848
+    },
+    {
+      "epoch": 0.6780508121615993,
+      "grad_norm": 0.36005635998666913,
+      "learning_rate": 7.699738853737646e-06,
+      "loss": 0.2921,
+      "step": 2849
+    },
+    {
+      "epoch": 0.6782888082346641,
+      "grad_norm": 0.38741602947105674,
+      "learning_rate": 7.698116200154262e-06,
+      "loss": 0.3202,
+      "step": 2850
+    },
+    {
+      "epoch": 0.6785268043077289,
+      "grad_norm": 0.3676779858892718,
+      "learning_rate": 7.696493145564065e-06,
+      "loss": 0.4059,
+      "step": 2851
+    },
+    {
+      "epoch": 0.6787648003807937,
+      "grad_norm": 0.3417732883784283,
+      "learning_rate": 7.694869690208278e-06,
+      "loss": 0.3071,
+      "step": 2852
+    },
+    {
+      "epoch": 0.6790027964538585,
+      "grad_norm": 0.3806094615926385,
+      "learning_rate": 7.693245834328186e-06,
+      "loss": 0.3005,
+      "step": 2853
+    },
+    {
+      "epoch": 0.6792407925269233,
+      "grad_norm": 0.37415443493343387,
+      "learning_rate": 7.691621578165135e-06,
+      "loss": 0.3775,
+      "step": 2854
+    },
+    {
+      "epoch": 0.6794787885999881,
+      "grad_norm": 0.3588357321765243,
+      "learning_rate": 7.689996921960533e-06,
+      "loss": 0.3828,
+      "step": 2855
+    },
+    {
+      "epoch": 0.6797167846730529,
+      "grad_norm": 0.3928642880638878,
+      "learning_rate": 7.68837186595584e-06,
+      "loss": 0.3388,
+      "step": 2856
+    },
+    {
+      "epoch": 0.6799547807461177,
+      "grad_norm": 0.39713877557243726,
+      "learning_rate": 7.686746410392579e-06,
+      "loss": 0.3547,
+      "step": 2857
+    },
+    {
+      "epoch": 0.6801927768191824,
+      "grad_norm": 0.37028184933204206,
+      "learning_rate": 7.685120555512335e-06,
+      "loss": 0.3991,
+      "step": 2858
+    },
+    {
+      "epoch": 0.6804307728922473,
+      "grad_norm": 0.44868723676569244,
+      "learning_rate": 7.68349430155675e-06,
+      "loss": 0.2975,
+      "step": 2859
+    },
+    {
+      "epoch": 0.680668768965312,
+      "grad_norm": 0.38969868882705583,
+      "learning_rate": 7.681867648767527e-06,
+      "loss": 0.3064,
+      "step": 2860
+    },
+    {
+      "epoch": 0.6809067650383769,
+      "grad_norm": 0.38048777908327175,
+      "learning_rate": 7.680240597386423e-06,
+      "loss": 0.3544,
+      "step": 2861
+    },
+    {
+      "epoch": 0.6811447611114416,
+      "grad_norm": 0.37492467378771765,
+      "learning_rate": 7.678613147655263e-06,
+      "loss": 0.4007,
+      "step": 2862
+    },
+    {
+      "epoch": 0.6813827571845065,
+      "grad_norm": 0.3689709476215601,
+      "learning_rate": 7.676985299815921e-06,
+      "loss": 0.3204,
+      "step": 2863
+    },
+    {
+      "epoch": 0.6816207532575712,
+      "grad_norm": 0.4036262585926096,
+      "learning_rate": 7.675357054110337e-06,
+      "loss": 0.3487,
+      "step": 2864
+    },
+    {
+      "epoch": 0.6818587493306361,
+      "grad_norm": 0.41388190325234303,
+      "learning_rate": 7.673728410780512e-06,
+      "loss": 0.3926,
+      "step": 2865
+    },
+    {
+      "epoch": 0.6820967454037008,
+      "grad_norm": 0.3896246709459209,
+      "learning_rate": 7.6720993700685e-06,
+      "loss": 0.3495,
+      "step": 2866
+    },
+    {
+      "epoch": 0.6823347414767656,
+      "grad_norm": 0.38993044742482114,
+      "learning_rate": 7.670469932216416e-06,
+      "loss": 0.2993,
+      "step": 2867
+    },
+    {
+      "epoch": 0.6825727375498304,
+      "grad_norm": 0.3709106695995631,
+      "learning_rate": 7.668840097466438e-06,
+      "loss": 0.4017,
+      "step": 2868
+    },
+    {
+      "epoch": 0.6828107336228952,
+      "grad_norm": 0.39844192153673497,
+      "learning_rate": 7.667209866060795e-06,
+      "loss": 0.4329,
+      "step": 2869
+    },
+    {
+      "epoch": 0.68304872969596,
+      "grad_norm": 0.394396694213486,
+      "learning_rate": 7.665579238241783e-06,
+      "loss": 0.3091,
+      "step": 2870
+    },
+    {
+      "epoch": 0.6832867257690248,
+      "grad_norm": 0.39853662948827007,
+      "learning_rate": 7.663948214251754e-06,
+      "loss": 0.336,
+      "step": 2871
+    },
+    {
+      "epoch": 0.6835247218420896,
+      "grad_norm": 0.3630956355542585,
+      "learning_rate": 7.662316794333115e-06,
+      "loss": 0.4052,
+      "step": 2872
+    },
+    {
+      "epoch": 0.6837627179151544,
+      "grad_norm": 0.3963082232165837,
+      "learning_rate": 7.660684978728341e-06,
+      "loss": 0.3736,
+      "step": 2873
+    },
+    {
+      "epoch": 0.6840007139882192,
+      "grad_norm": 0.3977137944295455,
+      "learning_rate": 7.659052767679956e-06,
+      "loss": 0.3113,
+      "step": 2874
+    },
+    {
+      "epoch": 0.684238710061284,
+      "grad_norm": 0.4015812355864788,
+      "learning_rate": 7.657420161430548e-06,
+      "loss": 0.3585,
+      "step": 2875
+    },
+    {
+      "epoch": 0.6844767061343487,
+      "grad_norm": 0.39210402493400554,
+      "learning_rate": 7.655787160222762e-06,
+      "loss": 0.4295,
+      "step": 2876
+    },
+    {
+      "epoch": 0.6847147022074136,
+      "grad_norm": 0.39681491199767854,
+      "learning_rate": 7.654153764299304e-06,
+      "loss": 0.2991,
+      "step": 2877
+    },
+    {
+      "epoch": 0.6849526982804783,
+      "grad_norm": 0.4163108651945542,
+      "learning_rate": 7.652519973902935e-06,
+      "loss": 0.3332,
+      "step": 2878
+    },
+    {
+      "epoch": 0.6851906943535432,
+      "grad_norm": 0.3748847281909277,
+      "learning_rate": 7.650885789276477e-06,
+      "loss": 0.3468,
+      "step": 2879
+    },
+    {
+      "epoch": 0.6854286904266079,
+      "grad_norm": 0.421568928920472,
+      "learning_rate": 7.649251210662812e-06,
+      "loss": 0.373,
+      "step": 2880
+    },
+    {
+      "epoch": 0.6856666864996728,
+      "grad_norm": 0.37882377572325376,
+      "learning_rate": 7.647616238304876e-06,
+      "loss": 0.3311,
+      "step": 2881
+    },
+    {
+      "epoch": 0.6859046825727375,
+      "grad_norm": 0.37714593364667043,
+      "learning_rate": 7.64598087244567e-06,
+      "loss": 0.3475,
+      "step": 2882
+    },
+    {
+      "epoch": 0.6861426786458024,
+      "grad_norm": 0.3855958441067824,
+      "learning_rate": 7.644345113328248e-06,
+      "loss": 0.4108,
+      "step": 2883
+    },
+    {
+      "epoch": 0.6863806747188671,
+      "grad_norm": 0.3927803834584747,
+      "learning_rate": 7.642708961195723e-06,
+      "loss": 0.3284,
+      "step": 2884
+    },
+    {
+      "epoch": 0.686618670791932,
+      "grad_norm": 0.39284438670389804,
+      "learning_rate": 7.641072416291271e-06,
+      "loss": 0.3087,
+      "step": 2885
+    },
+    {
+      "epoch": 0.6868566668649967,
+      "grad_norm": 0.38222011385116755,
+      "learning_rate": 7.639435478858119e-06,
+      "loss": 0.3611,
+      "step": 2886
+    },
+    {
+      "epoch": 0.6870946629380615,
+      "grad_norm": 0.35343748065798736,
+      "learning_rate": 7.637798149139559e-06,
+      "loss": 0.4157,
+      "step": 2887
+    },
+    {
+      "epoch": 0.6873326590111263,
+      "grad_norm": 0.4761170077236074,
+      "learning_rate": 7.636160427378938e-06,
+      "loss": 0.3342,
+      "step": 2888
+    },
+    {
+      "epoch": 0.6875706550841911,
+      "grad_norm": 0.3653390110713056,
+      "learning_rate": 7.634522313819664e-06,
+      "loss": 0.332,
+      "step": 2889
+    },
+    {
+      "epoch": 0.6878086511572559,
+      "grad_norm": 0.357610536904034,
+      "learning_rate": 7.632883808705196e-06,
+      "loss": 0.4209,
+      "step": 2890
+    },
+    {
+      "epoch": 0.6880466472303207,
+      "grad_norm": 0.35884223240749147,
+      "learning_rate": 7.631244912279061e-06,
+      "loss": 0.3103,
+      "step": 2891
+    },
+    {
+      "epoch": 0.6882846433033855,
+      "grad_norm": 0.3932088139549123,
+      "learning_rate": 7.629605624784839e-06,
+      "loss": 0.3257,
+      "step": 2892
+    },
+    {
+      "epoch": 0.6885226393764503,
+      "grad_norm": 0.3756997656933469,
+      "learning_rate": 7.627965946466167e-06,
+      "loss": 0.3667,
+      "step": 2893
+    },
+    {
+      "epoch": 0.688760635449515,
+      "grad_norm": 0.41363244284700523,
+      "learning_rate": 7.626325877566741e-06,
+      "loss": 0.421,
+      "step": 2894
+    },
+    {
+      "epoch": 0.6889986315225799,
+      "grad_norm": 0.38949498606977967,
+      "learning_rate": 7.624685418330319e-06,
+      "loss": 0.3179,
+      "step": 2895
+    },
+    {
+      "epoch": 0.6892366275956446,
+      "grad_norm": 0.37724094080660947,
+      "learning_rate": 7.623044569000712e-06,
+      "loss": 0.3282,
+      "step": 2896
+    },
+    {
+      "epoch": 0.6894746236687095,
+      "grad_norm": 0.5285214493849093,
+      "learning_rate": 7.621403329821792e-06,
+      "loss": 0.393,
+      "step": 2897
+    },
+    {
+      "epoch": 0.6897126197417742,
+      "grad_norm": 0.3726869379004517,
+      "learning_rate": 7.619761701037486e-06,
+      "loss": 0.3831,
+      "step": 2898
+    },
+    {
+      "epoch": 0.6899506158148391,
+      "grad_norm": 0.39521137312351157,
+      "learning_rate": 7.618119682891782e-06,
+      "loss": 0.3289,
+      "step": 2899
+    },
+    {
+      "epoch": 0.6901886118879038,
+      "grad_norm": 0.4132662165624947,
+      "learning_rate": 7.6164772756287234e-06,
+      "loss": 0.3614,
+      "step": 2900
+    },
+    {
+      "epoch": 0.6904266079609687,
+      "grad_norm": 0.3713002154182572,
+      "learning_rate": 7.614834479492413e-06,
+      "loss": 0.3873,
+      "step": 2901
+    },
+    {
+      "epoch": 0.6906646040340334,
+      "grad_norm": 0.4066260605116146,
+      "learning_rate": 7.613191294727011e-06,
+      "loss": 0.3409,
+      "step": 2902
+    },
+    {
+      "epoch": 0.6909026001070983,
+      "grad_norm": 0.4241202111497487,
+      "learning_rate": 7.611547721576738e-06,
+      "loss": 0.3538,
+      "step": 2903
+    },
+    {
+      "epoch": 0.691140596180163,
+      "grad_norm": 0.4686662314847007,
+      "learning_rate": 7.609903760285864e-06,
+      "loss": 0.4259,
+      "step": 2904
+    },
+    {
+      "epoch": 0.6913785922532278,
+      "grad_norm": 0.362668095436962,
+      "learning_rate": 7.608259411098725e-06,
+      "loss": 0.3968,
+      "step": 2905
+    },
+    {
+      "epoch": 0.6916165883262926,
+      "grad_norm": 0.40773309376570577,
+      "learning_rate": 7.606614674259714e-06,
+      "loss": 0.3047,
+      "step": 2906
+    },
+    {
+      "epoch": 0.6918545843993574,
+      "grad_norm": 0.35828275597855175,
+      "learning_rate": 7.6049695500132754e-06,
+      "loss": 0.3743,
+      "step": 2907
+    },
+    {
+      "epoch": 0.6920925804724222,
+      "grad_norm": 0.3603447573938342,
+      "learning_rate": 7.603324038603921e-06,
+      "loss": 0.3842,
+      "step": 2908
+    },
+    {
+      "epoch": 0.692330576545487,
+      "grad_norm": 0.40350976963719254,
+      "learning_rate": 7.601678140276209e-06,
+      "loss": 0.3228,
+      "step": 2909
+    },
+    {
+      "epoch": 0.6925685726185518,
+      "grad_norm": 0.38318140197317035,
+      "learning_rate": 7.600031855274764e-06,
+      "loss": 0.3013,
+      "step": 2910
+    },
+    {
+      "epoch": 0.6928065686916166,
+      "grad_norm": 0.3810631038611205,
+      "learning_rate": 7.598385183844263e-06,
+      "loss": 0.3572,
+      "step": 2911
+    },
+    {
+      "epoch": 0.6930445647646813,
+      "grad_norm": 0.3784857718340065,
+      "learning_rate": 7.5967381262294435e-06,
+      "loss": 0.3771,
+      "step": 2912
+    },
+    {
+      "epoch": 0.6932825608377462,
+      "grad_norm": 0.37559330533248253,
+      "learning_rate": 7.595090682675098e-06,
+      "loss": 0.3135,
+      "step": 2913
+    },
+    {
+      "epoch": 0.6935205569108109,
+      "grad_norm": 0.365340499769047,
+      "learning_rate": 7.593442853426077e-06,
+      "loss": 0.337,
+      "step": 2914
+    },
+    {
+      "epoch": 0.6937585529838758,
+      "grad_norm": 0.3680511519981457,
+      "learning_rate": 7.59179463872729e-06,
+      "loss": 0.4097,
+      "step": 2915
+    },
+    {
+      "epoch": 0.6939965490569405,
+      "grad_norm": 0.3766039547521425,
+      "learning_rate": 7.590146038823702e-06,
+      "loss": 0.332,
+      "step": 2916
+    },
+    {
+      "epoch": 0.6942345451300054,
+      "grad_norm": 0.3889666332240844,
+      "learning_rate": 7.588497053960335e-06,
+      "loss": 0.3063,
+      "step": 2917
+    },
+    {
+      "epoch": 0.6944725412030701,
+      "grad_norm": 0.377287587936367,
+      "learning_rate": 7.586847684382269e-06,
+      "loss": 0.3576,
+      "step": 2918
+    },
+    {
+      "epoch": 0.694710537276135,
+      "grad_norm": 0.3803646716074299,
+      "learning_rate": 7.585197930334642e-06,
+      "loss": 0.3992,
+      "step": 2919
+    },
+    {
+      "epoch": 0.6949485333491997,
+      "grad_norm": 0.3630382010768588,
+      "learning_rate": 7.5835477920626474e-06,
+      "loss": 0.3336,
+      "step": 2920
+    },
+    {
+      "epoch": 0.6951865294222646,
+      "grad_norm": 0.4145598312089378,
+      "learning_rate": 7.5818972698115375e-06,
+      "loss": 0.3337,
+      "step": 2921
+    },
+    {
+      "epoch": 0.6954245254953293,
+      "grad_norm": 0.40683667940888457,
+      "learning_rate": 7.580246363826621e-06,
+      "loss": 0.4267,
+      "step": 2922
+    },
+    {
+      "epoch": 0.6956625215683941,
+      "grad_norm": 0.4208382587332909,
+      "learning_rate": 7.578595074353262e-06,
+      "loss": 0.3757,
+      "step": 2923
+    },
+    {
+      "epoch": 0.6959005176414589,
+      "grad_norm": 0.35944681590578803,
+      "learning_rate": 7.5769434016368845e-06,
+      "loss": 0.3203,
+      "step": 2924
+    },
+    {
+      "epoch": 0.6961385137145237,
+      "grad_norm": 0.3599434098310551,
+      "learning_rate": 7.575291345922966e-06,
+      "loss": 0.3583,
+      "step": 2925
+    },
+    {
+      "epoch": 0.6963765097875885,
+      "grad_norm": 0.3756154404949079,
+      "learning_rate": 7.5736389074570425e-06,
+      "loss": 0.408,
+      "step": 2926
+    },
+    {
+      "epoch": 0.6966145058606533,
+      "grad_norm": 0.367210127037896,
+      "learning_rate": 7.571986086484711e-06,
+      "loss": 0.3372,
+      "step": 2927
+    },
+    {
+      "epoch": 0.6968525019337181,
+      "grad_norm": 0.3875760302017146,
+      "learning_rate": 7.570332883251618e-06,
+      "loss": 0.3293,
+      "step": 2928
+    },
+    {
+      "epoch": 0.6970904980067829,
+      "grad_norm": 0.39687946684154773,
+      "learning_rate": 7.568679298003472e-06,
+      "loss": 0.3794,
+      "step": 2929
+    },
+    {
+      "epoch": 0.6973284940798476,
+      "grad_norm": 0.3774932208465836,
+      "learning_rate": 7.567025330986035e-06,
+      "loss": 0.3996,
+      "step": 2930
+    },
+    {
+      "epoch": 0.6975664901529125,
+      "grad_norm": 0.37825723085597085,
+      "learning_rate": 7.565370982445131e-06,
+      "loss": 0.3063,
+      "step": 2931
+    },
+    {
+      "epoch": 0.6978044862259772,
+      "grad_norm": 0.3701480630486983,
+      "learning_rate": 7.563716252626632e-06,
+      "loss": 0.3392,
+      "step": 2932
+    },
+    {
+      "epoch": 0.6980424822990421,
+      "grad_norm": 0.363483092582883,
+      "learning_rate": 7.562061141776476e-06,
+      "loss": 0.4165,
+      "step": 2933
+    },
+    {
+      "epoch": 0.6982804783721068,
+      "grad_norm": 0.40095465937550545,
+      "learning_rate": 7.560405650140652e-06,
+      "loss": 0.3086,
+      "step": 2934
+    },
+    {
+      "epoch": 0.6985184744451717,
+      "grad_norm": 0.5381288739496879,
+      "learning_rate": 7.5587497779652065e-06,
+      "loss": 0.2924,
+      "step": 2935
+    },
+    {
+      "epoch": 0.6987564705182364,
+      "grad_norm": 0.3515417031170579,
+      "learning_rate": 7.557093525496245e-06,
+      "loss": 0.389,
+      "step": 2936
+    },
+    {
+      "epoch": 0.6989944665913013,
+      "grad_norm": 0.37096216086334566,
+      "learning_rate": 7.555436892979926e-06,
+      "loss": 0.421,
+      "step": 2937
+    },
+    {
+      "epoch": 0.699232462664366,
+      "grad_norm": 0.3482586299052201,
+      "learning_rate": 7.553779880662465e-06,
+      "loss": 0.2726,
+      "step": 2938
+    },
+    {
+      "epoch": 0.6994704587374309,
+      "grad_norm": 0.4102693276938636,
+      "learning_rate": 7.552122488790136e-06,
+      "loss": 0.3548,
+      "step": 2939
+    },
+    {
+      "epoch": 0.6997084548104956,
+      "grad_norm": 0.3581143681257741,
+      "learning_rate": 7.55046471760927e-06,
+      "loss": 0.379,
+      "step": 2940
+    },
+    {
+      "epoch": 0.6999464508835604,
+      "grad_norm": 0.4034626868783178,
+      "learning_rate": 7.548806567366251e-06,
+      "loss": 0.376,
+      "step": 2941
+    },
+    {
+      "epoch": 0.7001844469566252,
+      "grad_norm": 0.37940003906957404,
+      "learning_rate": 7.547148038307521e-06,
+      "loss": 0.3032,
+      "step": 2942
+    },
+    {
+      "epoch": 0.70042244302969,
+      "grad_norm": 0.37237400358199735,
+      "learning_rate": 7.545489130679581e-06,
+      "loss": 0.3705,
+      "step": 2943
+    },
+    {
+      "epoch": 0.7006604391027548,
+      "grad_norm": 0.3714730152118366,
+      "learning_rate": 7.543829844728983e-06,
+      "loss": 0.4293,
+      "step": 2944
+    },
+    {
+      "epoch": 0.7008984351758196,
+      "grad_norm": 0.35794214306124905,
+      "learning_rate": 7.542170180702337e-06,
+      "loss": 0.2971,
+      "step": 2945
+    },
+    {
+      "epoch": 0.7011364312488844,
+      "grad_norm": 0.43530263361830346,
+      "learning_rate": 7.540510138846313e-06,
+      "loss": 0.324,
+      "step": 2946
+    },
+    {
+      "epoch": 0.7013744273219492,
+      "grad_norm": 0.38306179336745,
+      "learning_rate": 7.538849719407632e-06,
+      "loss": 0.3798,
+      "step": 2947
+    },
+    {
+      "epoch": 0.701612423395014,
+      "grad_norm": 0.39934257512599863,
+      "learning_rate": 7.5371889226330765e-06,
+      "loss": 0.3281,
+      "step": 2948
+    },
+    {
+      "epoch": 0.7018504194680788,
+      "grad_norm": 0.36426837248809085,
+      "learning_rate": 7.53552774876948e-06,
+      "loss": 0.3066,
+      "step": 2949
+    },
+    {
+      "epoch": 0.7020884155411435,
+      "grad_norm": 0.4139378891941689,
+      "learning_rate": 7.533866198063734e-06,
+      "loss": 0.3819,
+      "step": 2950
+    },
+    {
+      "epoch": 0.7023264116142084,
+      "grad_norm": 0.3662560224348692,
+      "learning_rate": 7.532204270762786e-06,
+      "loss": 0.4475,
+      "step": 2951
+    },
+    {
+      "epoch": 0.7025644076872731,
+      "grad_norm": 0.3990867976425942,
+      "learning_rate": 7.530541967113639e-06,
+      "loss": 0.3468,
+      "step": 2952
+    },
+    {
+      "epoch": 0.702802403760338,
+      "grad_norm": 0.3856725854345983,
+      "learning_rate": 7.528879287363354e-06,
+      "loss": 0.3259,
+      "step": 2953
+    },
+    {
+      "epoch": 0.7030403998334027,
+      "grad_norm": 0.37688679209097037,
+      "learning_rate": 7.527216231759045e-06,
+      "loss": 0.3637,
+      "step": 2954
+    },
+    {
+      "epoch": 0.7032783959064676,
+      "grad_norm": 0.3667821119467153,
+      "learning_rate": 7.525552800547883e-06,
+      "loss": 0.3878,
+      "step": 2955
+    },
+    {
+      "epoch": 0.7035163919795323,
+      "grad_norm": 0.3753572526176831,
+      "learning_rate": 7.523888993977097e-06,
+      "loss": 0.3284,
+      "step": 2956
+    },
+    {
+      "epoch": 0.7037543880525972,
+      "grad_norm": 0.38058502140052836,
+      "learning_rate": 7.522224812293968e-06,
+      "loss": 0.3636,
+      "step": 2957
+    },
+    {
+      "epoch": 0.7039923841256619,
+      "grad_norm": 0.38834168887671905,
+      "learning_rate": 7.5205602557458345e-06,
+      "loss": 0.4213,
+      "step": 2958
+    },
+    {
+      "epoch": 0.7042303801987267,
+      "grad_norm": 0.37141054900070025,
+      "learning_rate": 7.518895324580091e-06,
+      "loss": 0.3236,
+      "step": 2959
+    },
+    {
+      "epoch": 0.7044683762717915,
+      "grad_norm": 0.39065100626003746,
+      "learning_rate": 7.517230019044188e-06,
+      "loss": 0.3049,
+      "step": 2960
+    },
+    {
+      "epoch": 0.7047063723448563,
+      "grad_norm": 0.3666524852773353,
+      "learning_rate": 7.51556433938563e-06,
+      "loss": 0.3719,
+      "step": 2961
+    },
+    {
+      "epoch": 0.7049443684179211,
+      "grad_norm": 0.3817964753531332,
+      "learning_rate": 7.513898285851982e-06,
+      "loss": 0.4357,
+      "step": 2962
+    },
+    {
+      "epoch": 0.7051823644909859,
+      "grad_norm": 0.4198232988530167,
+      "learning_rate": 7.512231858690856e-06,
+      "loss": 0.3348,
+      "step": 2963
+    },
+    {
+      "epoch": 0.7054203605640507,
+      "grad_norm": 0.3964784823438321,
+      "learning_rate": 7.510565058149927e-06,
+      "loss": 0.3519,
+      "step": 2964
+    },
+    {
+      "epoch": 0.7056583566371155,
+      "grad_norm": 0.34675732231223394,
+      "learning_rate": 7.508897884476921e-06,
+      "loss": 0.3907,
+      "step": 2965
+    },
+    {
+      "epoch": 0.7058963527101803,
+      "grad_norm": 0.3578070224969587,
+      "learning_rate": 7.507230337919623e-06,
+      "loss": 0.335,
+      "step": 2966
+    },
+    {
+      "epoch": 0.7061343487832451,
+      "grad_norm": 0.42191824768221664,
+      "learning_rate": 7.505562418725869e-06,
+      "loss": 0.296,
+      "step": 2967
+    },
+    {
+      "epoch": 0.7063723448563098,
+      "grad_norm": 0.3651458752052437,
+      "learning_rate": 7.503894127143558e-06,
+      "loss": 0.3755,
+      "step": 2968
+    },
+    {
+      "epoch": 0.7066103409293747,
+      "grad_norm": 0.4570811159042686,
+      "learning_rate": 7.5022254634206345e-06,
+      "loss": 0.3917,
+      "step": 2969
+    },
+    {
+      "epoch": 0.7068483370024394,
+      "grad_norm": 0.34715154210823396,
+      "learning_rate": 7.500556427805106e-06,
+      "loss": 0.3153,
+      "step": 2970
+    },
+    {
+      "epoch": 0.7070863330755043,
+      "grad_norm": 0.3947211815602096,
+      "learning_rate": 7.498887020545031e-06,
+      "loss": 0.3417,
+      "step": 2971
+    },
+    {
+      "epoch": 0.707324329148569,
+      "grad_norm": 0.37334961366076164,
+      "learning_rate": 7.497217241888525e-06,
+      "loss": 0.4165,
+      "step": 2972
+    },
+    {
+      "epoch": 0.7075623252216339,
+      "grad_norm": 0.3509931391175888,
+      "learning_rate": 7.495547092083758e-06,
+      "loss": 0.413,
+      "step": 2973
+    },
+    {
+      "epoch": 0.7078003212946986,
+      "grad_norm": 0.4205974782723252,
+      "learning_rate": 7.493876571378958e-06,
+      "loss": 0.2983,
+      "step": 2974
+    },
+    {
+      "epoch": 0.7080383173677635,
+      "grad_norm": 0.40308557662169947,
+      "learning_rate": 7.492205680022402e-06,
+      "loss": 0.3522,
+      "step": 2975
+    },
+    {
+      "epoch": 0.7082763134408282,
+      "grad_norm": 0.3511659283555987,
+      "learning_rate": 7.490534418262429e-06,
+      "loss": 0.4201,
+      "step": 2976
+    },
+    {
+      "epoch": 0.708514309513893,
+      "grad_norm": 0.38333575891336563,
+      "learning_rate": 7.488862786347428e-06,
+      "loss": 0.3414,
+      "step": 2977
+    },
+    {
+      "epoch": 0.7087523055869578,
+      "grad_norm": 0.38546643322989155,
+      "learning_rate": 7.487190784525847e-06,
+      "loss": 0.3206,
+      "step": 2978
+    },
+    {
+      "epoch": 0.7089903016600226,
+      "grad_norm": 0.36230466619537316,
+      "learning_rate": 7.485518413046185e-06,
+      "loss": 0.3884,
+      "step": 2979
+    },
+    {
+      "epoch": 0.7092282977330874,
+      "grad_norm": 0.36360795125731105,
+      "learning_rate": 7.4838456721569975e-06,
+      "loss": 0.3583,
+      "step": 2980
+    },
+    {
+      "epoch": 0.7094662938061522,
+      "grad_norm": 0.3836337184101506,
+      "learning_rate": 7.482172562106894e-06,
+      "loss": 0.3268,
+      "step": 2981
+    },
+    {
+      "epoch": 0.709704289879217,
+      "grad_norm": 0.4106360537544092,
+      "learning_rate": 7.480499083144544e-06,
+      "loss": 0.3347,
+      "step": 2982
+    },
+    {
+      "epoch": 0.7099422859522818,
+      "grad_norm": 0.3748286533906386,
+      "learning_rate": 7.478825235518665e-06,
+      "loss": 0.3911,
+      "step": 2983
+    },
+    {
+      "epoch": 0.7101802820253466,
+      "grad_norm": 0.42427243151949334,
+      "learning_rate": 7.477151019478033e-06,
+      "loss": 0.3345,
+      "step": 2984
+    },
+    {
+      "epoch": 0.7104182780984114,
+      "grad_norm": 0.3954736604508801,
+      "learning_rate": 7.4754764352714775e-06,
+      "loss": 0.3117,
+      "step": 2985
+    },
+    {
+      "epoch": 0.7106562741714761,
+      "grad_norm": 0.35115884372592393,
+      "learning_rate": 7.4738014831478825e-06,
+      "loss": 0.3906,
+      "step": 2986
+    },
+    {
+      "epoch": 0.710894270244541,
+      "grad_norm": 0.3945125666106555,
+      "learning_rate": 7.472126163356189e-06,
+      "loss": 0.4087,
+      "step": 2987
+    },
+    {
+      "epoch": 0.7111322663176057,
+      "grad_norm": 0.36333273981672193,
+      "learning_rate": 7.47045047614539e-06,
+      "loss": 0.3242,
+      "step": 2988
+    },
+    {
+      "epoch": 0.7113702623906706,
+      "grad_norm": 0.3914705133621617,
+      "learning_rate": 7.468774421764534e-06,
+      "loss": 0.3465,
+      "step": 2989
+    },
+    {
+      "epoch": 0.7116082584637353,
+      "grad_norm": 0.3823529571498665,
+      "learning_rate": 7.467098000462726e-06,
+      "loss": 0.3816,
+      "step": 2990
+    },
+    {
+      "epoch": 0.7118462545368002,
+      "grad_norm": 0.3548613233280991,
+      "learning_rate": 7.465421212489121e-06,
+      "loss": 0.3438,
+      "step": 2991
+    },
+    {
+      "epoch": 0.7120842506098649,
+      "grad_norm": 0.39262400367886485,
+      "learning_rate": 7.463744058092932e-06,
+      "loss": 0.352,
+      "step": 2992
+    },
+    {
+      "epoch": 0.7123222466829298,
+      "grad_norm": 0.367798301154,
+      "learning_rate": 7.462066537523427e-06,
+      "loss": 0.3347,
+      "step": 2993
+    },
+    {
+      "epoch": 0.7125602427559945,
+      "grad_norm": 0.38423044064973816,
+      "learning_rate": 7.460388651029925e-06,
+      "loss": 0.4565,
+      "step": 2994
+    },
+    {
+      "epoch": 0.7127982388290593,
+      "grad_norm": 0.3640635090647599,
+      "learning_rate": 7.458710398861802e-06,
+      "loss": 0.3073,
+      "step": 2995
+    },
+    {
+      "epoch": 0.7130362349021241,
+      "grad_norm": 0.4006113452209261,
+      "learning_rate": 7.457031781268488e-06,
+      "loss": 0.3733,
+      "step": 2996
+    },
+    {
+      "epoch": 0.7132742309751889,
+      "grad_norm": 0.3807254297958794,
+      "learning_rate": 7.455352798499468e-06,
+      "loss": 0.4349,
+      "step": 2997
+    },
+    {
+      "epoch": 0.7135122270482537,
+      "grad_norm": 0.39126757394311634,
+      "learning_rate": 7.453673450804279e-06,
+      "loss": 0.3575,
+      "step": 2998
+    },
+    {
+      "epoch": 0.7137502231213185,
+      "grad_norm": 0.4052451533875098,
+      "learning_rate": 7.451993738432514e-06,
+      "loss": 0.2873,
+      "step": 2999
+    },
+    {
+      "epoch": 0.7139882191943833,
+      "grad_norm": 0.39364111352601927,
+      "learning_rate": 7.450313661633821e-06,
+      "loss": 0.3606,
+      "step": 3000
+    },
+    {
+      "epoch": 0.7142262152674481,
+      "grad_norm": 0.40280636612509435,
+      "learning_rate": 7.448633220657901e-06,
+      "loss": 0.383,
+      "step": 3001
+    },
+    {
+      "epoch": 0.7144642113405129,
+      "grad_norm": 0.4339937070314304,
+      "learning_rate": 7.4469524157545055e-06,
+      "loss": 0.336,
+      "step": 3002
+    },
+    {
+      "epoch": 0.7147022074135777,
+      "grad_norm": 0.40489755454617965,
+      "learning_rate": 7.445271247173449e-06,
+      "loss": 0.3121,
+      "step": 3003
+    },
+    {
+      "epoch": 0.7149402034866424,
+      "grad_norm": 0.4041470391859187,
+      "learning_rate": 7.44358971516459e-06,
+      "loss": 0.396,
+      "step": 3004
+    },
+    {
+      "epoch": 0.7151781995597073,
+      "grad_norm": 0.39500293731975633,
+      "learning_rate": 7.441907819977849e-06,
+      "loss": 0.4046,
+      "step": 3005
+    },
+    {
+      "epoch": 0.715416195632772,
+      "grad_norm": 0.41370890008241495,
+      "learning_rate": 7.440225561863197e-06,
+      "loss": 0.3536,
+      "step": 3006
+    },
+    {
+      "epoch": 0.7156541917058369,
+      "grad_norm": 0.3939257241455852,
+      "learning_rate": 7.438542941070657e-06,
+      "loss": 0.3451,
+      "step": 3007
+    },
+    {
+      "epoch": 0.7158921877789016,
+      "grad_norm": 0.35452897061585553,
+      "learning_rate": 7.436859957850309e-06,
+      "loss": 0.3992,
+      "step": 3008
+    },
+    {
+      "epoch": 0.7161301838519665,
+      "grad_norm": 0.3816527448874161,
+      "learning_rate": 7.435176612452286e-06,
+      "loss": 0.3353,
+      "step": 3009
+    },
+    {
+      "epoch": 0.7163681799250312,
+      "grad_norm": 0.41454236361169766,
+      "learning_rate": 7.4334929051267755e-06,
+      "loss": 0.3166,
+      "step": 3010
+    },
+    {
+      "epoch": 0.7166061759980961,
+      "grad_norm": 0.3602376065948697,
+      "learning_rate": 7.431808836124018e-06,
+      "loss": 0.3797,
+      "step": 3011
+    },
+    {
+      "epoch": 0.7168441720711608,
+      "grad_norm": 0.39512210096179917,
+      "learning_rate": 7.4301244056943075e-06,
+      "loss": 0.4029,
+      "step": 3012
+    },
+    {
+      "epoch": 0.7170821681442257,
+      "grad_norm": 0.3827942527690594,
+      "learning_rate": 7.42843961408799e-06,
+      "loss": 0.3173,
+      "step": 3013
+    },
+    {
+      "epoch": 0.7173201642172904,
+      "grad_norm": 0.39446387386662557,
+      "learning_rate": 7.426754461555471e-06,
+      "loss": 0.3544,
+      "step": 3014
+    },
+    {
+      "epoch": 0.7175581602903552,
+      "grad_norm": 0.35472318916385837,
+      "learning_rate": 7.425068948347204e-06,
+      "loss": 0.4169,
+      "step": 3015
+    },
+    {
+      "epoch": 0.71779615636342,
+      "grad_norm": 0.38263262613927496,
+      "learning_rate": 7.423383074713697e-06,
+      "loss": 0.3714,
+      "step": 3016
+    },
+    {
+      "epoch": 0.7180341524364848,
+      "grad_norm": 0.3808280361027985,
+      "learning_rate": 7.421696840905515e-06,
+      "loss": 0.2928,
+      "step": 3017
+    },
+    {
+      "epoch": 0.7182721485095496,
+      "grad_norm": 0.36752434940666945,
+      "learning_rate": 7.4200102471732704e-06,
+      "loss": 0.3481,
+      "step": 3018
+    },
+    {
+      "epoch": 0.7185101445826144,
+      "grad_norm": 0.38919742791659145,
+      "learning_rate": 7.4183232937676375e-06,
+      "loss": 0.391,
+      "step": 3019
+    },
+    {
+      "epoch": 0.7187481406556792,
+      "grad_norm": 0.37374333756044126,
+      "learning_rate": 7.416635980939335e-06,
+      "loss": 0.348,
+      "step": 3020
+    },
+    {
+      "epoch": 0.718986136728744,
+      "grad_norm": 0.36920174596848016,
+      "learning_rate": 7.414948308939141e-06,
+      "loss": 0.3441,
+      "step": 3021
+    },
+    {
+      "epoch": 0.7192241328018087,
+      "grad_norm": 0.3899921201145649,
+      "learning_rate": 7.413260278017887e-06,
+      "loss": 0.3838,
+      "step": 3022
+    },
+    {
+      "epoch": 0.7194621288748736,
+      "grad_norm": 0.43144843112558834,
+      "learning_rate": 7.411571888426452e-06,
+      "loss": 0.3486,
+      "step": 3023
+    },
+    {
+      "epoch": 0.7197001249479383,
+      "grad_norm": 0.3911155710827359,
+      "learning_rate": 7.4098831404157765e-06,
+      "loss": 0.3315,
+      "step": 3024
+    },
+    {
+      "epoch": 0.7199381210210032,
+      "grad_norm": 0.42440357958109975,
+      "learning_rate": 7.408194034236849e-06,
+      "loss": 0.3736,
+      "step": 3025
+    },
+    {
+      "epoch": 0.7201761170940679,
+      "grad_norm": 0.3680061754320904,
+      "learning_rate": 7.40650457014071e-06,
+      "loss": 0.3912,
+      "step": 3026
+    },
+    {
+      "epoch": 0.7204141131671328,
+      "grad_norm": 0.3830697542132779,
+      "learning_rate": 7.404814748378461e-06,
+      "loss": 0.3287,
+      "step": 3027
+    },
+    {
+      "epoch": 0.7206521092401975,
+      "grad_norm": 0.37264948361322325,
+      "learning_rate": 7.403124569201246e-06,
+      "loss": 0.3231,
+      "step": 3028
+    },
+    {
+      "epoch": 0.7208901053132624,
+      "grad_norm": 0.42134581649859754,
+      "learning_rate": 7.4014340328602685e-06,
+      "loss": 0.3565,
+      "step": 3029
+    },
+    {
+      "epoch": 0.7211281013863271,
+      "grad_norm": 0.41130909439310925,
+      "learning_rate": 7.399743139606788e-06,
+      "loss": 0.4445,
+      "step": 3030
+    },
+    {
+      "epoch": 0.721366097459392,
+      "grad_norm": 0.4343180622590514,
+      "learning_rate": 7.398051889692108e-06,
+      "loss": 0.3357,
+      "step": 3031
+    },
+    {
+      "epoch": 0.7216040935324567,
+      "grad_norm": 0.39112408012774896,
+      "learning_rate": 7.396360283367594e-06,
+      "loss": 0.3501,
+      "step": 3032
+    },
+    {
+      "epoch": 0.7218420896055215,
+      "grad_norm": 0.3354333543079656,
+      "learning_rate": 7.394668320884658e-06,
+      "loss": 0.3829,
+      "step": 3033
+    },
+    {
+      "epoch": 0.7220800856785863,
+      "grad_norm": 0.34658725353941033,
+      "learning_rate": 7.392976002494768e-06,
+      "loss": 0.3088,
+      "step": 3034
+    },
+    {
+      "epoch": 0.7223180817516511,
+      "grad_norm": 0.42267425856173196,
+      "learning_rate": 7.391283328449445e-06,
+      "loss": 0.3284,
+      "step": 3035
+    },
+    {
+      "epoch": 0.7225560778247159,
+      "grad_norm": 0.3696528618085508,
+      "learning_rate": 7.389590299000262e-06,
+      "loss": 0.3727,
+      "step": 3036
+    },
+    {
+      "epoch": 0.7227940738977807,
+      "grad_norm": 0.37189620964899023,
+      "learning_rate": 7.387896914398845e-06,
+      "loss": 0.4148,
+      "step": 3037
+    },
+    {
+      "epoch": 0.7230320699708455,
+      "grad_norm": 0.4135419215724609,
+      "learning_rate": 7.386203174896872e-06,
+      "loss": 0.3078,
+      "step": 3038
+    },
+    {
+      "epoch": 0.7232700660439103,
+      "grad_norm": 0.3992204444214033,
+      "learning_rate": 7.384509080746076e-06,
+      "loss": 0.3575,
+      "step": 3039
+    },
+    {
+      "epoch": 0.723508062116975,
+      "grad_norm": 0.35922624214255505,
+      "learning_rate": 7.382814632198241e-06,
+      "loss": 0.4262,
+      "step": 3040
+    },
+    {
+      "epoch": 0.7237460581900399,
+      "grad_norm": 0.3793102673645999,
+      "learning_rate": 7.381119829505204e-06,
+      "loss": 0.365,
+      "step": 3041
+    },
+    {
+      "epoch": 0.7239840542631046,
+      "grad_norm": 0.3944308333847217,
+      "learning_rate": 7.379424672918853e-06,
+      "loss": 0.2977,
+      "step": 3042
+    },
+    {
+      "epoch": 0.7242220503361695,
+      "grad_norm": 0.36373682832023985,
+      "learning_rate": 7.377729162691131e-06,
+      "loss": 0.3509,
+      "step": 3043
+    },
+    {
+      "epoch": 0.7244600464092342,
+      "grad_norm": 0.3627646085804695,
+      "learning_rate": 7.376033299074035e-06,
+      "loss": 0.4184,
+      "step": 3044
+    },
+    {
+      "epoch": 0.7246980424822991,
+      "grad_norm": 0.3871761600279167,
+      "learning_rate": 7.3743370823196096e-06,
+      "loss": 0.3216,
+      "step": 3045
+    },
+    {
+      "epoch": 0.7249360385553638,
+      "grad_norm": 0.3798739905567255,
+      "learning_rate": 7.372640512679955e-06,
+      "loss": 0.3456,
+      "step": 3046
+    },
+    {
+      "epoch": 0.7251740346284287,
+      "grad_norm": 0.4259458917487435,
+      "learning_rate": 7.370943590407225e-06,
+      "loss": 0.4103,
+      "step": 3047
+    },
+    {
+      "epoch": 0.7254120307014934,
+      "grad_norm": 0.3626974255553493,
+      "learning_rate": 7.369246315753623e-06,
+      "loss": 0.3433,
+      "step": 3048
+    },
+    {
+      "epoch": 0.7256500267745583,
+      "grad_norm": 0.4435787230045141,
+      "learning_rate": 7.367548688971407e-06,
+      "loss": 0.2972,
+      "step": 3049
+    },
+    {
+      "epoch": 0.725888022847623,
+      "grad_norm": 0.3554006057479478,
+      "learning_rate": 7.365850710312883e-06,
+      "loss": 0.3688,
+      "step": 3050
+    },
+    {
+      "epoch": 0.7261260189206878,
+      "grad_norm": 0.36442397072863186,
+      "learning_rate": 7.364152380030416e-06,
+      "loss": 0.4098,
+      "step": 3051
+    },
+    {
+      "epoch": 0.7263640149937526,
+      "grad_norm": 0.4410087294940679,
+      "learning_rate": 7.3624536983764195e-06,
+      "loss": 0.3272,
+      "step": 3052
+    },
+    {
+      "epoch": 0.7266020110668174,
+      "grad_norm": 0.4282754810487593,
+      "learning_rate": 7.3607546656033594e-06,
+      "loss": 0.3241,
+      "step": 3053
+    },
+    {
+      "epoch": 0.7268400071398822,
+      "grad_norm": 0.4083398243189906,
+      "learning_rate": 7.359055281963753e-06,
+      "loss": 0.3797,
+      "step": 3054
+    },
+    {
+      "epoch": 0.727078003212947,
+      "grad_norm": 0.3965480234646783,
+      "learning_rate": 7.357355547710172e-06,
+      "loss": 0.3904,
+      "step": 3055
+    },
+    {
+      "epoch": 0.7273159992860118,
+      "grad_norm": 0.42026409912961626,
+      "learning_rate": 7.355655463095239e-06,
+      "loss": 0.3146,
+      "step": 3056
+    },
+    {
+      "epoch": 0.7275539953590766,
+      "grad_norm": 0.3771142416353585,
+      "learning_rate": 7.3539550283716265e-06,
+      "loss": 0.3445,
+      "step": 3057
+    },
+    {
+      "epoch": 0.7277919914321413,
+      "grad_norm": 0.3551162990197736,
+      "learning_rate": 7.352254243792064e-06,
+      "loss": 0.4232,
+      "step": 3058
+    },
+    {
+      "epoch": 0.7280299875052062,
+      "grad_norm": 0.4085035997854665,
+      "learning_rate": 7.350553109609329e-06,
+      "loss": 0.3106,
+      "step": 3059
+    },
+    {
+      "epoch": 0.7282679835782709,
+      "grad_norm": 0.3847757510606681,
+      "learning_rate": 7.348851626076252e-06,
+      "loss": 0.3063,
+      "step": 3060
+    },
+    {
+      "epoch": 0.7285059796513358,
+      "grad_norm": 0.36865435960657855,
+      "learning_rate": 7.347149793445715e-06,
+      "loss": 0.3544,
+      "step": 3061
+    },
+    {
+      "epoch": 0.7287439757244005,
+      "grad_norm": 0.3922363467897953,
+      "learning_rate": 7.345447611970653e-06,
+      "loss": 0.3822,
+      "step": 3062
+    },
+    {
+      "epoch": 0.7289819717974654,
+      "grad_norm": 0.34369944175304884,
+      "learning_rate": 7.3437450819040536e-06,
+      "loss": 0.3206,
+      "step": 3063
+    },
+    {
+      "epoch": 0.7292199678705301,
+      "grad_norm": 0.35824405322282227,
+      "learning_rate": 7.342042203498952e-06,
+      "loss": 0.3554,
+      "step": 3064
+    },
+    {
+      "epoch": 0.729457963943595,
+      "grad_norm": 0.39512318030878885,
+      "learning_rate": 7.34033897700844e-06,
+      "loss": 0.3839,
+      "step": 3065
+    },
+    {
+      "epoch": 0.7296959600166597,
+      "grad_norm": 0.38854090842538846,
+      "learning_rate": 7.338635402685659e-06,
+      "loss": 0.3378,
+      "step": 3066
+    },
+    {
+      "epoch": 0.7299339560897246,
+      "grad_norm": 0.38073840280648125,
+      "learning_rate": 7.336931480783801e-06,
+      "loss": 0.3343,
+      "step": 3067
+    },
+    {
+      "epoch": 0.7301719521627893,
+      "grad_norm": 0.3869439775508733,
+      "learning_rate": 7.335227211556113e-06,
+      "loss": 0.3474,
+      "step": 3068
+    },
+    {
+      "epoch": 0.7304099482358541,
+      "grad_norm": 0.36331918337163177,
+      "learning_rate": 7.3335225952558904e-06,
+      "loss": 0.4093,
+      "step": 3069
+    },
+    {
+      "epoch": 0.7306479443089189,
+      "grad_norm": 0.3676892295018712,
+      "learning_rate": 7.3318176321364835e-06,
+      "loss": 0.3082,
+      "step": 3070
+    },
+    {
+      "epoch": 0.7308859403819837,
+      "grad_norm": 0.37227813295867007,
+      "learning_rate": 7.330112322451287e-06,
+      "loss": 0.3266,
+      "step": 3071
+    },
+    {
+      "epoch": 0.7311239364550485,
+      "grad_norm": 0.37682897636401325,
+      "learning_rate": 7.328406666453757e-06,
+      "loss": 0.4061,
+      "step": 3072
+    },
+    {
+      "epoch": 0.7313619325281133,
+      "grad_norm": 0.43099604922414647,
+      "learning_rate": 7.326700664397395e-06,
+      "loss": 0.3787,
+      "step": 3073
+    },
+    {
+      "epoch": 0.7315999286011781,
+      "grad_norm": 0.38646511617724494,
+      "learning_rate": 7.324994316535753e-06,
+      "loss": 0.3184,
+      "step": 3074
+    },
+    {
+      "epoch": 0.7318379246742429,
+      "grad_norm": 0.3745081198422119,
+      "learning_rate": 7.323287623122439e-06,
+      "loss": 0.3465,
+      "step": 3075
+    },
+    {
+      "epoch": 0.7320759207473077,
+      "grad_norm": 0.3868492244329305,
+      "learning_rate": 7.321580584411108e-06,
+      "loss": 0.4354,
+      "step": 3076
+    },
+    {
+      "epoch": 0.7323139168203725,
+      "grad_norm": 0.3746511813825389,
+      "learning_rate": 7.31987320065547e-06,
+      "loss": 0.3515,
+      "step": 3077
+    },
+    {
+      "epoch": 0.7325519128934372,
+      "grad_norm": 0.46299638156809153,
+      "learning_rate": 7.318165472109282e-06,
+      "loss": 0.3055,
+      "step": 3078
+    },
+    {
+      "epoch": 0.7327899089665021,
+      "grad_norm": 0.4037577857446162,
+      "learning_rate": 7.3164573990263574e-06,
+      "loss": 0.3781,
+      "step": 3079
+    },
+    {
+      "epoch": 0.7330279050395668,
+      "grad_norm": 0.3996195774993303,
+      "learning_rate": 7.314748981660555e-06,
+      "loss": 0.3987,
+      "step": 3080
+    },
+    {
+      "epoch": 0.7332659011126317,
+      "grad_norm": 0.3936618750557137,
+      "learning_rate": 7.313040220265792e-06,
+      "loss": 0.3083,
+      "step": 3081
+    },
+    {
+      "epoch": 0.7335038971856964,
+      "grad_norm": 0.38121369149338524,
+      "learning_rate": 7.31133111509603e-06,
+      "loss": 0.3229,
+      "step": 3082
+    },
+    {
+      "epoch": 0.7337418932587613,
+      "grad_norm": 0.388825782352307,
+      "learning_rate": 7.309621666405284e-06,
+      "loss": 0.4128,
+      "step": 3083
+    },
+    {
+      "epoch": 0.733979889331826,
+      "grad_norm": 0.3715499437212395,
+      "learning_rate": 7.307911874447622e-06,
+      "loss": 0.3383,
+      "step": 3084
+    },
+    {
+      "epoch": 0.7342178854048909,
+      "grad_norm": 0.4232766232253306,
+      "learning_rate": 7.306201739477159e-06,
+      "loss": 0.3137,
+      "step": 3085
+    },
+    {
+      "epoch": 0.7344558814779556,
+      "grad_norm": 0.345889539039532,
+      "learning_rate": 7.304491261748067e-06,
+      "loss": 0.3484,
+      "step": 3086
+    },
+    {
+      "epoch": 0.7346938775510204,
+      "grad_norm": 0.37961353019617056,
+      "learning_rate": 7.302780441514561e-06,
+      "loss": 0.3989,
+      "step": 3087
+    },
+    {
+      "epoch": 0.7349318736240852,
+      "grad_norm": 0.3772100128276546,
+      "learning_rate": 7.3010692790309145e-06,
+      "loss": 0.3109,
+      "step": 3088
+    },
+    {
+      "epoch": 0.73516986969715,
+      "grad_norm": 0.4148043405953893,
+      "learning_rate": 7.2993577745514475e-06,
+      "loss": 0.339,
+      "step": 3089
+    },
+    {
+      "epoch": 0.7354078657702148,
+      "grad_norm": 0.4041915618136653,
+      "learning_rate": 7.2976459283305326e-06,
+      "loss": 0.3987,
+      "step": 3090
+    },
+    {
+      "epoch": 0.7356458618432796,
+      "grad_norm": 0.37070530009594416,
+      "learning_rate": 7.2959337406225894e-06,
+      "loss": 0.3478,
+      "step": 3091
+    },
+    {
+      "epoch": 0.7358838579163444,
+      "grad_norm": 0.3405354595093119,
+      "learning_rate": 7.294221211682096e-06,
+      "loss": 0.3366,
+      "step": 3092
+    },
+    {
+      "epoch": 0.7361218539894092,
+      "grad_norm": 0.36518838895931155,
+      "learning_rate": 7.292508341763574e-06,
+      "loss": 0.3426,
+      "step": 3093
+    },
+    {
+      "epoch": 0.736359850062474,
+      "grad_norm": 0.37822344257306156,
+      "learning_rate": 7.290795131121595e-06,
+      "loss": 0.3879,
+      "step": 3094
+    },
+    {
+      "epoch": 0.7365978461355388,
+      "grad_norm": 0.3730174664708697,
+      "learning_rate": 7.289081580010792e-06,
+      "loss": 0.3241,
+      "step": 3095
+    },
+    {
+      "epoch": 0.7368358422086035,
+      "grad_norm": 0.360220342804819,
+      "learning_rate": 7.287367688685835e-06,
+      "loss": 0.3325,
+      "step": 3096
+    },
+    {
+      "epoch": 0.7370738382816684,
+      "grad_norm": 0.36585884890518433,
+      "learning_rate": 7.285653457401453e-06,
+      "loss": 0.3872,
+      "step": 3097
+    },
+    {
+      "epoch": 0.7373118343547331,
+      "grad_norm": 0.5429383224292799,
+      "learning_rate": 7.283938886412424e-06,
+      "loss": 0.3578,
+      "step": 3098
+    },
+    {
+      "epoch": 0.737549830427798,
+      "grad_norm": 0.41494686957593174,
+      "learning_rate": 7.2822239759735735e-06,
+      "loss": 0.3477,
+      "step": 3099
+    },
+    {
+      "epoch": 0.7377878265008627,
+      "grad_norm": 0.3816939297486696,
+      "learning_rate": 7.280508726339781e-06,
+      "loss": 0.3684,
+      "step": 3100
+    },
+    {
+      "epoch": 0.7380258225739276,
+      "grad_norm": 0.37029991676249097,
+      "learning_rate": 7.278793137765976e-06,
+      "loss": 0.4007,
+      "step": 3101
+    },
+    {
+      "epoch": 0.7382638186469923,
+      "grad_norm": 0.4339643261053269,
+      "learning_rate": 7.277077210507135e-06,
+      "loss": 0.3341,
+      "step": 3102
+    },
+    {
+      "epoch": 0.7385018147200572,
+      "grad_norm": 0.37177092531956185,
+      "learning_rate": 7.2753609448182885e-06,
+      "loss": 0.3178,
+      "step": 3103
+    },
+    {
+      "epoch": 0.7387398107931219,
+      "grad_norm": 0.36083182065635616,
+      "learning_rate": 7.273644340954515e-06,
+      "loss": 0.3974,
+      "step": 3104
+    },
+    {
+      "epoch": 0.7389778068661867,
+      "grad_norm": 0.3742211097385228,
+      "learning_rate": 7.271927399170946e-06,
+      "loss": 0.3764,
+      "step": 3105
+    },
+    {
+      "epoch": 0.7392158029392515,
+      "grad_norm": 0.39314424003897525,
+      "learning_rate": 7.270210119722761e-06,
+      "loss": 0.3178,
+      "step": 3106
+    },
+    {
+      "epoch": 0.7394537990123163,
+      "grad_norm": 0.387246764459355,
+      "learning_rate": 7.2684925028651875e-06,
+      "loss": 0.3463,
+      "step": 3107
+    },
+    {
+      "epoch": 0.7396917950853811,
+      "grad_norm": 0.3455422305424231,
+      "learning_rate": 7.26677454885351e-06,
+      "loss": 0.4018,
+      "step": 3108
+    },
+    {
+      "epoch": 0.7399297911584459,
+      "grad_norm": 0.3680423220780498,
+      "learning_rate": 7.265056257943059e-06,
+      "loss": 0.3318,
+      "step": 3109
+    },
+    {
+      "epoch": 0.7401677872315107,
+      "grad_norm": 0.4218016718701506,
+      "learning_rate": 7.2633376303892115e-06,
+      "loss": 0.3327,
+      "step": 3110
+    },
+    {
+      "epoch": 0.7404057833045755,
+      "grad_norm": 0.4797971091280992,
+      "learning_rate": 7.2616186664474e-06,
+      "loss": 0.3781,
+      "step": 3111
+    },
+    {
+      "epoch": 0.7406437793776403,
+      "grad_norm": 0.3656564240573133,
+      "learning_rate": 7.259899366373105e-06,
+      "loss": 0.3589,
+      "step": 3112
+    },
+    {
+      "epoch": 0.7408817754507051,
+      "grad_norm": 0.39727650704719514,
+      "learning_rate": 7.258179730421856e-06,
+      "loss": 0.302,
+      "step": 3113
+    },
+    {
+      "epoch": 0.7411197715237698,
+      "grad_norm": 0.3854240175463639,
+      "learning_rate": 7.256459758849236e-06,
+      "loss": 0.3257,
+      "step": 3114
+    },
+    {
+      "epoch": 0.7413577675968347,
+      "grad_norm": 0.35574522570705447,
+      "learning_rate": 7.254739451910872e-06,
+      "loss": 0.3877,
+      "step": 3115
+    },
+    {
+      "epoch": 0.7415957636698994,
+      "grad_norm": 0.4630875140371236,
+      "learning_rate": 7.253018809862448e-06,
+      "loss": 0.3405,
+      "step": 3116
+    },
+    {
+      "epoch": 0.7418337597429643,
+      "grad_norm": 0.41595067569227445,
+      "learning_rate": 7.251297832959691e-06,
+      "loss": 0.3055,
+      "step": 3117
+    },
+    {
+      "epoch": 0.742071755816029,
+      "grad_norm": 0.35751446775150947,
+      "learning_rate": 7.249576521458381e-06,
+      "loss": 0.3468,
+      "step": 3118
+    },
+    {
+      "epoch": 0.7423097518890939,
+      "grad_norm": 0.3782939975585627,
+      "learning_rate": 7.247854875614348e-06,
+      "loss": 0.4108,
+      "step": 3119
+    },
+    {
+      "epoch": 0.7425477479621586,
+      "grad_norm": 0.36443090923466487,
+      "learning_rate": 7.246132895683472e-06,
+      "loss": 0.3252,
+      "step": 3120
+    },
+    {
+      "epoch": 0.7427857440352235,
+      "grad_norm": 0.3521241114001235,
+      "learning_rate": 7.244410581921679e-06,
+      "loss": 0.3367,
+      "step": 3121
+    },
+    {
+      "epoch": 0.7430237401082882,
+      "grad_norm": 0.40051085069078274,
+      "learning_rate": 7.242687934584952e-06,
+      "loss": 0.3951,
+      "step": 3122
+    },
+    {
+      "epoch": 0.743261736181353,
+      "grad_norm": 0.44680675909155915,
+      "learning_rate": 7.2409649539293155e-06,
+      "loss": 0.3604,
+      "step": 3123
+    },
+    {
+      "epoch": 0.7434997322544178,
+      "grad_norm": 0.41922316103507395,
+      "learning_rate": 7.239241640210849e-06,
+      "loss": 0.3254,
+      "step": 3124
+    },
+    {
+      "epoch": 0.7437377283274826,
+      "grad_norm": 0.3777661176299726,
+      "learning_rate": 7.2375179936856775e-06,
+      "loss": 0.371,
+      "step": 3125
+    },
+    {
+      "epoch": 0.7439757244005474,
+      "grad_norm": 0.37648805520020007,
+      "learning_rate": 7.235794014609978e-06,
+      "loss": 0.4046,
+      "step": 3126
+    },
+    {
+      "epoch": 0.7442137204736122,
+      "grad_norm": 0.410705623333718,
+      "learning_rate": 7.234069703239979e-06,
+      "loss": 0.3264,
+      "step": 3127
+    },
+    {
+      "epoch": 0.744451716546677,
+      "grad_norm": 0.36702681685700017,
+      "learning_rate": 7.2323450598319535e-06,
+      "loss": 0.3487,
+      "step": 3128
+    },
+    {
+      "epoch": 0.7446897126197418,
+      "grad_norm": 0.3670128177868576,
+      "learning_rate": 7.230620084642226e-06,
+      "loss": 0.3881,
+      "step": 3129
+    },
+    {
+      "epoch": 0.7449277086928066,
+      "grad_norm": 0.40207445243333073,
+      "learning_rate": 7.228894777927171e-06,
+      "loss": 0.3942,
+      "step": 3130
+    },
+    {
+      "epoch": 0.7451657047658714,
+      "grad_norm": 0.40080124947818796,
+      "learning_rate": 7.227169139943211e-06,
+      "loss": 0.3075,
+      "step": 3131
+    },
+    {
+      "epoch": 0.7454037008389361,
+      "grad_norm": 0.39628761333956425,
+      "learning_rate": 7.22544317094682e-06,
+      "loss": 0.3387,
+      "step": 3132
+    },
+    {
+      "epoch": 0.745641696912001,
+      "grad_norm": 0.37788202720786135,
+      "learning_rate": 7.223716871194519e-06,
+      "loss": 0.4307,
+      "step": 3133
+    },
+    {
+      "epoch": 0.7458796929850657,
+      "grad_norm": 0.37196320757312157,
+      "learning_rate": 7.221990240942878e-06,
+      "loss": 0.3459,
+      "step": 3134
+    },
+    {
+      "epoch": 0.7461176890581306,
+      "grad_norm": 0.37988100060665997,
+      "learning_rate": 7.220263280448518e-06,
+      "loss": 0.2937,
+      "step": 3135
+    },
+    {
+      "epoch": 0.7463556851311953,
+      "grad_norm": 0.37254514931619725,
+      "learning_rate": 7.21853598996811e-06,
+      "loss": 0.4057,
+      "step": 3136
+    },
+    {
+      "epoch": 0.7465936812042602,
+      "grad_norm": 0.3900480382097283,
+      "learning_rate": 7.216808369758368e-06,
+      "loss": 0.381,
+      "step": 3137
+    },
+    {
+      "epoch": 0.7468316772773249,
+      "grad_norm": 0.3796183272267863,
+      "learning_rate": 7.215080420076061e-06,
+      "loss": 0.3289,
+      "step": 3138
+    },
+    {
+      "epoch": 0.7470696733503898,
+      "grad_norm": 0.3806521308902775,
+      "learning_rate": 7.2133521411780075e-06,
+      "loss": 0.3483,
+      "step": 3139
+    },
+    {
+      "epoch": 0.7473076694234545,
+      "grad_norm": 0.38838395980321355,
+      "learning_rate": 7.211623533321067e-06,
+      "loss": 0.3974,
+      "step": 3140
+    },
+    {
+      "epoch": 0.7475456654965194,
+      "grad_norm": 0.41947045054774285,
+      "learning_rate": 7.209894596762158e-06,
+      "loss": 0.3447,
+      "step": 3141
+    },
+    {
+      "epoch": 0.7477836615695841,
+      "grad_norm": 0.3540371480756213,
+      "learning_rate": 7.2081653317582414e-06,
+      "loss": 0.3406,
+      "step": 3142
+    },
+    {
+      "epoch": 0.7480216576426489,
+      "grad_norm": 0.3826391985417264,
+      "learning_rate": 7.20643573856633e-06,
+      "loss": 0.3343,
+      "step": 3143
+    },
+    {
+      "epoch": 0.7482596537157137,
+      "grad_norm": 0.3538807128710463,
+      "learning_rate": 7.204705817443483e-06,
+      "loss": 0.4207,
+      "step": 3144
+    },
+    {
+      "epoch": 0.7484976497887785,
+      "grad_norm": 0.352020290864496,
+      "learning_rate": 7.202975568646809e-06,
+      "loss": 0.3251,
+      "step": 3145
+    },
+    {
+      "epoch": 0.7487356458618433,
+      "grad_norm": 0.39868157169625523,
+      "learning_rate": 7.201244992433466e-06,
+      "loss": 0.3339,
+      "step": 3146
+    },
+    {
+      "epoch": 0.7489736419349081,
+      "grad_norm": 0.4023017435952385,
+      "learning_rate": 7.199514089060662e-06,
+      "loss": 0.3981,
+      "step": 3147
+    },
+    {
+      "epoch": 0.7492116380079729,
+      "grad_norm": 0.3741576715187835,
+      "learning_rate": 7.19778285878565e-06,
+      "loss": 0.3561,
+      "step": 3148
+    },
+    {
+      "epoch": 0.7494496340810377,
+      "grad_norm": 0.4709729162057381,
+      "learning_rate": 7.196051301865736e-06,
+      "loss": 0.3399,
+      "step": 3149
+    },
+    {
+      "epoch": 0.7496876301541024,
+      "grad_norm": 0.41862196677750924,
+      "learning_rate": 7.19431941855827e-06,
+      "loss": 0.3235,
+      "step": 3150
+    },
+    {
+      "epoch": 0.7499256262271673,
+      "grad_norm": 0.36071859439792425,
+      "learning_rate": 7.192587209120654e-06,
+      "loss": 0.3893,
+      "step": 3151
+    },
+    {
+      "epoch": 0.750163622300232,
+      "grad_norm": 0.3699965349983255,
+      "learning_rate": 7.190854673810337e-06,
+      "loss": 0.3109,
+      "step": 3152
+    },
+    {
+      "epoch": 0.7504016183732969,
+      "grad_norm": 0.42095108252199576,
+      "learning_rate": 7.189121812884816e-06,
+      "loss": 0.3121,
+      "step": 3153
+    },
+    {
+      "epoch": 0.7506396144463616,
+      "grad_norm": 0.37196059546980936,
+      "learning_rate": 7.1873886266016365e-06,
+      "loss": 0.3859,
+      "step": 3154
+    },
+    {
+      "epoch": 0.7508776105194265,
+      "grad_norm": 0.3728280714448741,
+      "learning_rate": 7.185655115218395e-06,
+      "loss": 0.3766,
+      "step": 3155
+    },
+    {
+      "epoch": 0.7511156065924912,
+      "grad_norm": 0.37533698264848414,
+      "learning_rate": 7.183921278992731e-06,
+      "loss": 0.2943,
+      "step": 3156
+    },
+    {
+      "epoch": 0.7513536026655561,
+      "grad_norm": 0.3918689428403605,
+      "learning_rate": 7.18218711818234e-06,
+      "loss": 0.3503,
+      "step": 3157
+    },
+    {
+      "epoch": 0.7515915987386208,
+      "grad_norm": 0.42511294665911836,
+      "learning_rate": 7.180452633044958e-06,
+      "loss": 0.4205,
+      "step": 3158
+    },
+    {
+      "epoch": 0.7518295948116857,
+      "grad_norm": 0.39558443408700916,
+      "learning_rate": 7.178717823838371e-06,
+      "loss": 0.3445,
+      "step": 3159
+    },
+    {
+      "epoch": 0.7520675908847504,
+      "grad_norm": 0.3861088188015497,
+      "learning_rate": 7.176982690820418e-06,
+      "loss": 0.3184,
+      "step": 3160
+    },
+    {
+      "epoch": 0.7523055869578152,
+      "grad_norm": 0.3598250325709718,
+      "learning_rate": 7.175247234248979e-06,
+      "loss": 0.3543,
+      "step": 3161
+    },
+    {
+      "epoch": 0.75254358303088,
+      "grad_norm": 0.38572071033062194,
+      "learning_rate": 7.173511454381991e-06,
+      "loss": 0.3887,
+      "step": 3162
+    },
+    {
+      "epoch": 0.7527815791039448,
+      "grad_norm": 0.3597558768595933,
+      "learning_rate": 7.171775351477429e-06,
+      "loss": 0.3176,
+      "step": 3163
+    },
+    {
+      "epoch": 0.7530195751770096,
+      "grad_norm": 0.3519928195184308,
+      "learning_rate": 7.170038925793323e-06,
+      "loss": 0.3598,
+      "step": 3164
+    },
+    {
+      "epoch": 0.7532575712500744,
+      "grad_norm": 0.40143875905094994,
+      "learning_rate": 7.16830217758775e-06,
+      "loss": 0.405,
+      "step": 3165
+    },
+    {
+      "epoch": 0.7534955673231392,
+      "grad_norm": 0.39193156061116613,
+      "learning_rate": 7.16656510711883e-06,
+      "loss": 0.3403,
+      "step": 3166
+    },
+    {
+      "epoch": 0.753733563396204,
+      "grad_norm": 0.38720279747566255,
+      "learning_rate": 7.164827714644738e-06,
+      "loss": 0.292,
+      "step": 3167
+    },
+    {
+      "epoch": 0.7539715594692687,
+      "grad_norm": 0.36467759519856513,
+      "learning_rate": 7.163090000423691e-06,
+      "loss": 0.3451,
+      "step": 3168
+    },
+    {
+      "epoch": 0.7542095555423336,
+      "grad_norm": 0.41427804995589623,
+      "learning_rate": 7.161351964713959e-06,
+      "loss": 0.4082,
+      "step": 3169
+    },
+    {
+      "epoch": 0.7544475516153983,
+      "grad_norm": 0.38192749026353107,
+      "learning_rate": 7.159613607773857e-06,
+      "loss": 0.3061,
+      "step": 3170
+    },
+    {
+      "epoch": 0.7546855476884632,
+      "grad_norm": 0.43361143142318026,
+      "learning_rate": 7.157874929861745e-06,
+      "loss": 0.3096,
+      "step": 3171
+    },
+    {
+      "epoch": 0.7549235437615279,
+      "grad_norm": 0.3554010401349325,
+      "learning_rate": 7.156135931236034e-06,
+      "loss": 0.4144,
+      "step": 3172
+    },
+    {
+      "epoch": 0.7551615398345928,
+      "grad_norm": 0.39011150823679097,
+      "learning_rate": 7.1543966121551845e-06,
+      "loss": 0.3801,
+      "step": 3173
+    },
+    {
+      "epoch": 0.7553995359076575,
+      "grad_norm": 0.3817077245262734,
+      "learning_rate": 7.152656972877702e-06,
+      "loss": 0.2873,
+      "step": 3174
+    },
+    {
+      "epoch": 0.7556375319807224,
+      "grad_norm": 0.3720557616381678,
+      "learning_rate": 7.150917013662138e-06,
+      "loss": 0.3617,
+      "step": 3175
+    },
+    {
+      "epoch": 0.7558755280537871,
+      "grad_norm": 0.34769187060464707,
+      "learning_rate": 7.149176734767095e-06,
+      "loss": 0.4133,
+      "step": 3176
+    },
+    {
+      "epoch": 0.756113524126852,
+      "grad_norm": 0.3835724087378952,
+      "learning_rate": 7.147436136451221e-06,
+      "loss": 0.325,
+      "step": 3177
+    },
+    {
+      "epoch": 0.7563515201999167,
+      "grad_norm": 0.37749841830040526,
+      "learning_rate": 7.145695218973213e-06,
+      "loss": 0.3102,
+      "step": 3178
+    },
+    {
+      "epoch": 0.7565895162729815,
+      "grad_norm": 0.374520542733197,
+      "learning_rate": 7.143953982591813e-06,
+      "loss": 0.3978,
+      "step": 3179
+    },
+    {
+      "epoch": 0.7568275123460463,
+      "grad_norm": 0.4023541368854894,
+      "learning_rate": 7.142212427565812e-06,
+      "loss": 0.3807,
+      "step": 3180
+    },
+    {
+      "epoch": 0.7570655084191111,
+      "grad_norm": 0.35873098883174603,
+      "learning_rate": 7.140470554154048e-06,
+      "loss": 0.3559,
+      "step": 3181
+    },
+    {
+      "epoch": 0.7573035044921759,
+      "grad_norm": 0.3867236526972635,
+      "learning_rate": 7.138728362615408e-06,
+      "loss": 0.3331,
+      "step": 3182
+    },
+    {
+      "epoch": 0.7575415005652407,
+      "grad_norm": 0.34735983378932106,
+      "learning_rate": 7.136985853208824e-06,
+      "loss": 0.398,
+      "step": 3183
+    },
+    {
+      "epoch": 0.7577794966383055,
+      "grad_norm": 0.38443189061398053,
+      "learning_rate": 7.135243026193275e-06,
+      "loss": 0.3334,
+      "step": 3184
+    },
+    {
+      "epoch": 0.7580174927113703,
+      "grad_norm": 0.3821702158023499,
+      "learning_rate": 7.13349988182779e-06,
+      "loss": 0.3129,
+      "step": 3185
+    },
+    {
+      "epoch": 0.758255488784435,
+      "grad_norm": 0.3569641423841968,
+      "learning_rate": 7.131756420371441e-06,
+      "loss": 0.38,
+      "step": 3186
+    },
+    {
+      "epoch": 0.7584934848574999,
+      "grad_norm": 0.40219483194623007,
+      "learning_rate": 7.130012642083351e-06,
+      "loss": 0.4149,
+      "step": 3187
+    },
+    {
+      "epoch": 0.7587314809305646,
+      "grad_norm": 0.3708101360867531,
+      "learning_rate": 7.128268547222688e-06,
+      "loss": 0.3625,
+      "step": 3188
+    },
+    {
+      "epoch": 0.7589694770036295,
+      "grad_norm": 0.37603123822726897,
+      "learning_rate": 7.126524136048669e-06,
+      "loss": 0.33,
+      "step": 3189
+    },
+    {
+      "epoch": 0.7592074730766942,
+      "grad_norm": 0.3625022362919271,
+      "learning_rate": 7.124779408820555e-06,
+      "loss": 0.4318,
+      "step": 3190
+    },
+    {
+      "epoch": 0.7594454691497591,
+      "grad_norm": 0.39739967957247313,
+      "learning_rate": 7.123034365797657e-06,
+      "loss": 0.3272,
+      "step": 3191
+    },
+    {
+      "epoch": 0.7596834652228238,
+      "grad_norm": 0.3702784091648422,
+      "learning_rate": 7.121289007239331e-06,
+      "loss": 0.3268,
+      "step": 3192
+    },
+    {
+      "epoch": 0.7599214612958887,
+      "grad_norm": 0.3516224416164428,
+      "learning_rate": 7.119543333404981e-06,
+      "loss": 0.3504,
+      "step": 3193
+    },
+    {
+      "epoch": 0.7601594573689534,
+      "grad_norm": 0.3559370032742599,
+      "learning_rate": 7.117797344554056e-06,
+      "loss": 0.4026,
+      "step": 3194
+    },
+    {
+      "epoch": 0.7603974534420183,
+      "grad_norm": 0.3734187380388612,
+      "learning_rate": 7.116051040946053e-06,
+      "loss": 0.3204,
+      "step": 3195
+    },
+    {
+      "epoch": 0.760635449515083,
+      "grad_norm": 0.3579819166208651,
+      "learning_rate": 7.114304422840517e-06,
+      "loss": 0.3277,
+      "step": 3196
+    },
+    {
+      "epoch": 0.7608734455881478,
+      "grad_norm": 0.3752870062603591,
+      "learning_rate": 7.112557490497038e-06,
+      "loss": 0.3955,
+      "step": 3197
+    },
+    {
+      "epoch": 0.7611114416612126,
+      "grad_norm": 0.3707029313698444,
+      "learning_rate": 7.1108102441752546e-06,
+      "loss": 0.3666,
+      "step": 3198
+    },
+    {
+      "epoch": 0.7613494377342774,
+      "grad_norm": 0.3867018314993227,
+      "learning_rate": 7.109062684134851e-06,
+      "loss": 0.3063,
+      "step": 3199
+    },
+    {
+      "epoch": 0.7615874338073422,
+      "grad_norm": 0.40876052113414646,
+      "learning_rate": 7.107314810635555e-06,
+      "loss": 0.3568,
+      "step": 3200
+    },
+    {
+      "epoch": 0.761825429880407,
+      "grad_norm": 0.4030334876778494,
+      "learning_rate": 7.105566623937145e-06,
+      "loss": 0.4414,
+      "step": 3201
+    },
+    {
+      "epoch": 0.7620634259534718,
+      "grad_norm": 0.36214917368846855,
+      "learning_rate": 7.103818124299446e-06,
+      "loss": 0.3108,
+      "step": 3202
+    },
+    {
+      "epoch": 0.7623014220265366,
+      "grad_norm": 0.36698016620497026,
+      "learning_rate": 7.102069311982329e-06,
+      "loss": 0.3015,
+      "step": 3203
+    },
+    {
+      "epoch": 0.7625394180996014,
+      "grad_norm": 0.368871180085732,
+      "learning_rate": 7.100320187245711e-06,
+      "loss": 0.3934,
+      "step": 3204
+    },
+    {
+      "epoch": 0.7627774141726662,
+      "grad_norm": 0.39902520639387057,
+      "learning_rate": 7.098570750349552e-06,
+      "loss": 0.3871,
+      "step": 3205
+    },
+    {
+      "epoch": 0.7630154102457309,
+      "grad_norm": 0.37446900240784753,
+      "learning_rate": 7.096821001553863e-06,
+      "loss": 0.3109,
+      "step": 3206
+    },
+    {
+      "epoch": 0.7632534063187958,
+      "grad_norm": 0.33881592006301775,
+      "learning_rate": 7.0950709411187e-06,
+      "loss": 0.3471,
+      "step": 3207
+    },
+    {
+      "epoch": 0.7634914023918605,
+      "grad_norm": 0.372761489856163,
+      "learning_rate": 7.093320569304168e-06,
+      "loss": 0.4049,
+      "step": 3208
+    },
+    {
+      "epoch": 0.7637293984649254,
+      "grad_norm": 0.4017927278263893,
+      "learning_rate": 7.0915698863704094e-06,
+      "loss": 0.3293,
+      "step": 3209
+    },
+    {
+      "epoch": 0.7639673945379901,
+      "grad_norm": 0.4054778790918442,
+      "learning_rate": 7.089818892577625e-06,
+      "loss": 0.335,
+      "step": 3210
+    },
+    {
+      "epoch": 0.764205390611055,
+      "grad_norm": 0.3850085404080853,
+      "learning_rate": 7.088067588186053e-06,
+      "loss": 0.3492,
+      "step": 3211
+    },
+    {
+      "epoch": 0.7644433866841197,
+      "grad_norm": 0.37942848234718374,
+      "learning_rate": 7.086315973455982e-06,
+      "loss": 0.4199,
+      "step": 3212
+    },
+    {
+      "epoch": 0.7646813827571846,
+      "grad_norm": 0.4002793707382145,
+      "learning_rate": 7.084564048647742e-06,
+      "loss": 0.3223,
+      "step": 3213
+    },
+    {
+      "epoch": 0.7649193788302493,
+      "grad_norm": 0.3758467906510097,
+      "learning_rate": 7.082811814021717e-06,
+      "loss": 0.3411,
+      "step": 3214
+    },
+    {
+      "epoch": 0.7651573749033141,
+      "grad_norm": 0.38353128527707686,
+      "learning_rate": 7.08105926983833e-06,
+      "loss": 0.4176,
+      "step": 3215
+    },
+    {
+      "epoch": 0.7653953709763789,
+      "grad_norm": 0.41062374283228403,
+      "learning_rate": 7.0793064163580515e-06,
+      "loss": 0.3443,
+      "step": 3216
+    },
+    {
+      "epoch": 0.7656333670494437,
+      "grad_norm": 0.38600351854755544,
+      "learning_rate": 7.0775532538414005e-06,
+      "loss": 0.3275,
+      "step": 3217
+    },
+    {
+      "epoch": 0.7658713631225085,
+      "grad_norm": 0.3588188716781081,
+      "learning_rate": 7.0757997825489395e-06,
+      "loss": 0.368,
+      "step": 3218
+    },
+    {
+      "epoch": 0.7661093591955733,
+      "grad_norm": 0.3582749191881199,
+      "learning_rate": 7.074046002741279e-06,
+      "loss": 0.3938,
+      "step": 3219
+    },
+    {
+      "epoch": 0.7663473552686381,
+      "grad_norm": 0.3597466282489841,
+      "learning_rate": 7.072291914679072e-06,
+      "loss": 0.3072,
+      "step": 3220
+    },
+    {
+      "epoch": 0.7665853513417029,
+      "grad_norm": 0.3920888152123116,
+      "learning_rate": 7.070537518623022e-06,
+      "loss": 0.2984,
+      "step": 3221
+    },
+    {
+      "epoch": 0.7668233474147677,
+      "grad_norm": 0.3641437351355874,
+      "learning_rate": 7.068782814833872e-06,
+      "loss": 0.3744,
+      "step": 3222
+    },
+    {
+      "epoch": 0.7670613434878325,
+      "grad_norm": 0.37937338820809674,
+      "learning_rate": 7.067027803572417e-06,
+      "loss": 0.3666,
+      "step": 3223
+    },
+    {
+      "epoch": 0.7672993395608972,
+      "grad_norm": 0.40998865963471826,
+      "learning_rate": 7.065272485099496e-06,
+      "loss": 0.2939,
+      "step": 3224
+    },
+    {
+      "epoch": 0.7675373356339621,
+      "grad_norm": 0.4023210745015225,
+      "learning_rate": 7.06351685967599e-06,
+      "loss": 0.3818,
+      "step": 3225
+    },
+    {
+      "epoch": 0.7677753317070268,
+      "grad_norm": 0.37345454403274114,
+      "learning_rate": 7.061760927562831e-06,
+      "loss": 0.4636,
+      "step": 3226
+    },
+    {
+      "epoch": 0.7680133277800917,
+      "grad_norm": 0.4223720881715161,
+      "learning_rate": 7.060004689020991e-06,
+      "loss": 0.3236,
+      "step": 3227
+    },
+    {
+      "epoch": 0.7682513238531564,
+      "grad_norm": 0.40982795591831794,
+      "learning_rate": 7.058248144311493e-06,
+      "loss": 0.3324,
+      "step": 3228
+    },
+    {
+      "epoch": 0.7684893199262213,
+      "grad_norm": 0.3557070641164775,
+      "learning_rate": 7.056491293695401e-06,
+      "loss": 0.3833,
+      "step": 3229
+    },
+    {
+      "epoch": 0.768727315999286,
+      "grad_norm": 0.3797321346430915,
+      "learning_rate": 7.05473413743383e-06,
+      "loss": 0.3552,
+      "step": 3230
+    },
+    {
+      "epoch": 0.7689653120723509,
+      "grad_norm": 0.4115569949091365,
+      "learning_rate": 7.052976675787932e-06,
+      "loss": 0.3028,
+      "step": 3231
+    },
+    {
+      "epoch": 0.7692033081454156,
+      "grad_norm": 0.42313222429594644,
+      "learning_rate": 7.051218909018913e-06,
+      "loss": 0.36,
+      "step": 3232
+    },
+    {
+      "epoch": 0.7694413042184804,
+      "grad_norm": 0.3576077007683352,
+      "learning_rate": 7.04946083738802e-06,
+      "loss": 0.3817,
+      "step": 3233
+    },
+    {
+      "epoch": 0.7696793002915452,
+      "grad_norm": 0.4065000621443697,
+      "learning_rate": 7.047702461156545e-06,
+      "loss": 0.3447,
+      "step": 3234
+    },
+    {
+      "epoch": 0.76991729636461,
+      "grad_norm": 0.38063058594261784,
+      "learning_rate": 7.045943780585826e-06,
+      "loss": 0.2826,
+      "step": 3235
+    },
+    {
+      "epoch": 0.7701552924376748,
+      "grad_norm": 0.410395519650393,
+      "learning_rate": 7.044184795937248e-06,
+      "loss": 0.3782,
+      "step": 3236
+    },
+    {
+      "epoch": 0.7703932885107396,
+      "grad_norm": 0.38149692275265434,
+      "learning_rate": 7.042425507472237e-06,
+      "loss": 0.4149,
+      "step": 3237
+    },
+    {
+      "epoch": 0.7706312845838044,
+      "grad_norm": 0.3809955756687497,
+      "learning_rate": 7.040665915452269e-06,
+      "loss": 0.3169,
+      "step": 3238
+    },
+    {
+      "epoch": 0.7708692806568692,
+      "grad_norm": 0.3697765244878929,
+      "learning_rate": 7.038906020138863e-06,
+      "loss": 0.3496,
+      "step": 3239
+    },
+    {
+      "epoch": 0.771107276729934,
+      "grad_norm": 0.4309372907682824,
+      "learning_rate": 7.037145821793582e-06,
+      "loss": 0.3958,
+      "step": 3240
+    },
+    {
+      "epoch": 0.7713452728029988,
+      "grad_norm": 0.38099748409784767,
+      "learning_rate": 7.035385320678035e-06,
+      "loss": 0.3601,
+      "step": 3241
+    },
+    {
+      "epoch": 0.7715832688760635,
+      "grad_norm": 0.4276643539057699,
+      "learning_rate": 7.033624517053878e-06,
+      "loss": 0.3132,
+      "step": 3242
+    },
+    {
+      "epoch": 0.7718212649491284,
+      "grad_norm": 0.3518664176009511,
+      "learning_rate": 7.031863411182806e-06,
+      "loss": 0.3615,
+      "step": 3243
+    },
+    {
+      "epoch": 0.7720592610221931,
+      "grad_norm": 0.362211191847796,
+      "learning_rate": 7.0301020033265655e-06,
+      "loss": 0.4409,
+      "step": 3244
+    },
+    {
+      "epoch": 0.772297257095258,
+      "grad_norm": 0.41910572530406326,
+      "learning_rate": 7.0283402937469455e-06,
+      "loss": 0.3131,
+      "step": 3245
+    },
+    {
+      "epoch": 0.7725352531683227,
+      "grad_norm": 0.4314331090536776,
+      "learning_rate": 7.0265782827057804e-06,
+      "loss": 0.3372,
+      "step": 3246
+    },
+    {
+      "epoch": 0.7727732492413876,
+      "grad_norm": 0.3693815905880875,
+      "learning_rate": 7.024815970464947e-06,
+      "loss": 0.4088,
+      "step": 3247
+    },
+    {
+      "epoch": 0.7730112453144523,
+      "grad_norm": 0.37981415522062073,
+      "learning_rate": 7.023053357286366e-06,
+      "loss": 0.3587,
+      "step": 3248
+    },
+    {
+      "epoch": 0.7732492413875172,
+      "grad_norm": 0.408493661115404,
+      "learning_rate": 7.0212904434320115e-06,
+      "loss": 0.307,
+      "step": 3249
+    },
+    {
+      "epoch": 0.7734872374605819,
+      "grad_norm": 0.37233119228531636,
+      "learning_rate": 7.019527229163891e-06,
+      "loss": 0.3307,
+      "step": 3250
+    },
+    {
+      "epoch": 0.7737252335336468,
+      "grad_norm": 0.355235494120951,
+      "learning_rate": 7.0177637147440645e-06,
+      "loss": 0.4161,
+      "step": 3251
+    },
+    {
+      "epoch": 0.7739632296067115,
+      "grad_norm": 0.41229891884255343,
+      "learning_rate": 7.015999900434632e-06,
+      "loss": 0.3474,
+      "step": 3252
+    },
+    {
+      "epoch": 0.7742012256797762,
+      "grad_norm": 0.37602579982370005,
+      "learning_rate": 7.0142357864977425e-06,
+      "loss": 0.3039,
+      "step": 3253
+    },
+    {
+      "epoch": 0.7744392217528411,
+      "grad_norm": 0.3958312928917149,
+      "learning_rate": 7.012471373195584e-06,
+      "loss": 0.3848,
+      "step": 3254
+    },
+    {
+      "epoch": 0.7746772178259058,
+      "grad_norm": 0.4121169912150407,
+      "learning_rate": 7.010706660790393e-06,
+      "loss": 0.3912,
+      "step": 3255
+    },
+    {
+      "epoch": 0.7749152138989707,
+      "grad_norm": 0.3824825922233828,
+      "learning_rate": 7.0089416495444505e-06,
+      "loss": 0.3104,
+      "step": 3256
+    },
+    {
+      "epoch": 0.7751532099720354,
+      "grad_norm": 0.3844237240799921,
+      "learning_rate": 7.007176339720079e-06,
+      "loss": 0.3511,
+      "step": 3257
+    },
+    {
+      "epoch": 0.7753912060451003,
+      "grad_norm": 0.4050734698257971,
+      "learning_rate": 7.005410731579649e-06,
+      "loss": 0.3956,
+      "step": 3258
+    },
+    {
+      "epoch": 0.775629202118165,
+      "grad_norm": 0.38094079518404117,
+      "learning_rate": 7.003644825385574e-06,
+      "loss": 0.3168,
+      "step": 3259
+    },
+    {
+      "epoch": 0.7758671981912298,
+      "grad_norm": 0.3756921808647231,
+      "learning_rate": 7.001878621400309e-06,
+      "loss": 0.3121,
+      "step": 3260
+    },
+    {
+      "epoch": 0.7761051942642946,
+      "grad_norm": 0.3795064413693575,
+      "learning_rate": 7.000112119886356e-06,
+      "loss": 0.3722,
+      "step": 3261
+    },
+    {
+      "epoch": 0.7763431903373594,
+      "grad_norm": 0.38072119281394334,
+      "learning_rate": 6.998345321106264e-06,
+      "loss": 0.3769,
+      "step": 3262
+    },
+    {
+      "epoch": 0.7765811864104242,
+      "grad_norm": 0.3521206764121899,
+      "learning_rate": 6.996578225322619e-06,
+      "loss": 0.3074,
+      "step": 3263
+    },
+    {
+      "epoch": 0.776819182483489,
+      "grad_norm": 0.3934398383104075,
+      "learning_rate": 6.994810832798056e-06,
+      "loss": 0.329,
+      "step": 3264
+    },
+    {
+      "epoch": 0.7770571785565538,
+      "grad_norm": 0.3832785826014656,
+      "learning_rate": 6.993043143795255e-06,
+      "loss": 0.405,
+      "step": 3265
+    },
+    {
+      "epoch": 0.7772951746296186,
+      "grad_norm": 0.3642849945216878,
+      "learning_rate": 6.991275158576936e-06,
+      "loss": 0.3423,
+      "step": 3266
+    },
+    {
+      "epoch": 0.7775331707026834,
+      "grad_norm": 0.40701200067939586,
+      "learning_rate": 6.989506877405867e-06,
+      "loss": 0.3073,
+      "step": 3267
+    },
+    {
+      "epoch": 0.7777711667757482,
+      "grad_norm": 0.38472233590714217,
+      "learning_rate": 6.9877383005448595e-06,
+      "loss": 0.3395,
+      "step": 3268
+    },
+    {
+      "epoch": 0.7780091628488129,
+      "grad_norm": 0.3839123853902827,
+      "learning_rate": 6.9859694282567655e-06,
+      "loss": 0.3998,
+      "step": 3269
+    },
+    {
+      "epoch": 0.7782471589218778,
+      "grad_norm": 0.35916227359386177,
+      "learning_rate": 6.9842002608044844e-06,
+      "loss": 0.3028,
+      "step": 3270
+    },
+    {
+      "epoch": 0.7784851549949425,
+      "grad_norm": 0.37216903957857644,
+      "learning_rate": 6.9824307984509565e-06,
+      "loss": 0.3197,
+      "step": 3271
+    },
+    {
+      "epoch": 0.7787231510680074,
+      "grad_norm": 0.401474993410537,
+      "learning_rate": 6.98066104145917e-06,
+      "loss": 0.4028,
+      "step": 3272
+    },
+    {
+      "epoch": 0.7789611471410721,
+      "grad_norm": 0.4764009593683862,
+      "learning_rate": 6.9788909900921546e-06,
+      "loss": 0.3624,
+      "step": 3273
+    },
+    {
+      "epoch": 0.779199143214137,
+      "grad_norm": 0.35941136835596654,
+      "learning_rate": 6.977120644612981e-06,
+      "loss": 0.3168,
+      "step": 3274
+    },
+    {
+      "epoch": 0.7794371392872017,
+      "grad_norm": 0.47474946631582,
+      "learning_rate": 6.975350005284769e-06,
+      "loss": 0.3574,
+      "step": 3275
+    },
+    {
+      "epoch": 0.7796751353602666,
+      "grad_norm": 0.3664417092614242,
+      "learning_rate": 6.973579072370678e-06,
+      "loss": 0.4299,
+      "step": 3276
+    },
+    {
+      "epoch": 0.7799131314333313,
+      "grad_norm": 0.3583726867946201,
+      "learning_rate": 6.971807846133912e-06,
+      "loss": 0.3236,
+      "step": 3277
+    },
+    {
+      "epoch": 0.7801511275063961,
+      "grad_norm": 0.4022285999884666,
+      "learning_rate": 6.97003632683772e-06,
+      "loss": 0.3138,
+      "step": 3278
+    },
+    {
+      "epoch": 0.7803891235794609,
+      "grad_norm": 0.35817090327334505,
+      "learning_rate": 6.9682645147453954e-06,
+      "loss": 0.3843,
+      "step": 3279
+    },
+    {
+      "epoch": 0.7806271196525257,
+      "grad_norm": 0.3846970980247595,
+      "learning_rate": 6.966492410120269e-06,
+      "loss": 0.3713,
+      "step": 3280
+    },
+    {
+      "epoch": 0.7808651157255905,
+      "grad_norm": 0.366789717423642,
+      "learning_rate": 6.964720013225723e-06,
+      "loss": 0.2951,
+      "step": 3281
+    },
+    {
+      "epoch": 0.7811031117986553,
+      "grad_norm": 0.38839627264622095,
+      "learning_rate": 6.962947324325178e-06,
+      "loss": 0.3634,
+      "step": 3282
+    },
+    {
+      "epoch": 0.7813411078717201,
+      "grad_norm": 0.36282780184060254,
+      "learning_rate": 6.9611743436821e-06,
+      "loss": 0.3993,
+      "step": 3283
+    },
+    {
+      "epoch": 0.7815791039447849,
+      "grad_norm": 0.35508176540562836,
+      "learning_rate": 6.959401071559997e-06,
+      "loss": 0.3159,
+      "step": 3284
+    },
+    {
+      "epoch": 0.7818171000178497,
+      "grad_norm": 0.4152503707635735,
+      "learning_rate": 6.957627508222421e-06,
+      "loss": 0.2873,
+      "step": 3285
+    },
+    {
+      "epoch": 0.7820550960909145,
+      "grad_norm": 0.39876662297919807,
+      "learning_rate": 6.955853653932969e-06,
+      "loss": 0.3802,
+      "step": 3286
+    },
+    {
+      "epoch": 0.7822930921639792,
+      "grad_norm": 0.36696139093226415,
+      "learning_rate": 6.9540795089552785e-06,
+      "loss": 0.3693,
+      "step": 3287
+    },
+    {
+      "epoch": 0.7825310882370441,
+      "grad_norm": 0.3600324867306773,
+      "learning_rate": 6.952305073553031e-06,
+      "loss": 0.3051,
+      "step": 3288
+    },
+    {
+      "epoch": 0.7827690843101088,
+      "grad_norm": 0.3489749012167579,
+      "learning_rate": 6.950530347989952e-06,
+      "loss": 0.3532,
+      "step": 3289
+    },
+    {
+      "epoch": 0.7830070803831737,
+      "grad_norm": 0.3620983180455999,
+      "learning_rate": 6.9487553325298086e-06,
+      "loss": 0.3963,
+      "step": 3290
+    },
+    {
+      "epoch": 0.7832450764562384,
+      "grad_norm": 0.36785158602869183,
+      "learning_rate": 6.946980027436413e-06,
+      "loss": 0.343,
+      "step": 3291
+    },
+    {
+      "epoch": 0.7834830725293033,
+      "grad_norm": 0.3599934529541224,
+      "learning_rate": 6.94520443297362e-06,
+      "loss": 0.2774,
+      "step": 3292
+    },
+    {
+      "epoch": 0.783721068602368,
+      "grad_norm": 0.38715941357436445,
+      "learning_rate": 6.943428549405327e-06,
+      "loss": 0.3519,
+      "step": 3293
+    },
+    {
+      "epoch": 0.7839590646754329,
+      "grad_norm": 0.35349331453822797,
+      "learning_rate": 6.941652376995471e-06,
+      "loss": 0.3988,
+      "step": 3294
+    },
+    {
+      "epoch": 0.7841970607484976,
+      "grad_norm": 0.3955985233260755,
+      "learning_rate": 6.93987591600804e-06,
+      "loss": 0.3389,
+      "step": 3295
+    },
+    {
+      "epoch": 0.7844350568215624,
+      "grad_norm": 0.34663056196232733,
+      "learning_rate": 6.938099166707058e-06,
+      "loss": 0.314,
+      "step": 3296
+    },
+    {
+      "epoch": 0.7846730528946272,
+      "grad_norm": 0.38906902208324257,
+      "learning_rate": 6.936322129356592e-06,
+      "loss": 0.3764,
+      "step": 3297
+    },
+    {
+      "epoch": 0.784911048967692,
+      "grad_norm": 0.41156216306072313,
+      "learning_rate": 6.934544804220755e-06,
+      "loss": 0.3957,
+      "step": 3298
+    },
+    {
+      "epoch": 0.7851490450407568,
+      "grad_norm": 0.35518958539778267,
+      "learning_rate": 6.932767191563703e-06,
+      "loss": 0.282,
+      "step": 3299
+    },
+    {
+      "epoch": 0.7853870411138216,
+      "grad_norm": 0.47033631889554306,
+      "learning_rate": 6.9309892916496315e-06,
+      "loss": 0.3442,
+      "step": 3300
+    },
+    {
+      "epoch": 0.7856250371868864,
+      "grad_norm": 0.3666911616164395,
+      "learning_rate": 6.929211104742781e-06,
+      "loss": 0.4094,
+      "step": 3301
+    },
+    {
+      "epoch": 0.7858630332599512,
+      "grad_norm": 0.3577310928807679,
+      "learning_rate": 6.927432631107434e-06,
+      "loss": 0.3025,
+      "step": 3302
+    },
+    {
+      "epoch": 0.786101029333016,
+      "grad_norm": 0.37429247897055223,
+      "learning_rate": 6.925653871007916e-06,
+      "loss": 0.3226,
+      "step": 3303
+    },
+    {
+      "epoch": 0.7863390254060808,
+      "grad_norm": 0.3880037905135578,
+      "learning_rate": 6.923874824708594e-06,
+      "loss": 0.354,
+      "step": 3304
+    },
+    {
+      "epoch": 0.7865770214791455,
+      "grad_norm": 0.4433660243451845,
+      "learning_rate": 6.922095492473877e-06,
+      "loss": 0.3544,
+      "step": 3305
+    },
+    {
+      "epoch": 0.7868150175522104,
+      "grad_norm": 0.3759346004041654,
+      "learning_rate": 6.920315874568222e-06,
+      "loss": 0.2971,
+      "step": 3306
+    },
+    {
+      "epoch": 0.7870530136252751,
+      "grad_norm": 0.3785964570018848,
+      "learning_rate": 6.918535971256121e-06,
+      "loss": 0.3874,
+      "step": 3307
+    },
+    {
+      "epoch": 0.78729100969834,
+      "grad_norm": 0.3558284632586464,
+      "learning_rate": 6.91675578280211e-06,
+      "loss": 0.422,
+      "step": 3308
+    },
+    {
+      "epoch": 0.7875290057714047,
+      "grad_norm": 0.37377127129597065,
+      "learning_rate": 6.914975309470775e-06,
+      "loss": 0.3282,
+      "step": 3309
+    },
+    {
+      "epoch": 0.7877670018444696,
+      "grad_norm": 0.40012751504223365,
+      "learning_rate": 6.913194551526733e-06,
+      "loss": 0.3119,
+      "step": 3310
+    },
+    {
+      "epoch": 0.7880049979175343,
+      "grad_norm": 0.3754850496234726,
+      "learning_rate": 6.911413509234651e-06,
+      "loss": 0.3494,
+      "step": 3311
+    },
+    {
+      "epoch": 0.7882429939905992,
+      "grad_norm": 0.40489835887499337,
+      "learning_rate": 6.9096321828592336e-06,
+      "loss": 0.4102,
+      "step": 3312
+    },
+    {
+      "epoch": 0.7884809900636639,
+      "grad_norm": 0.43391003763332675,
+      "learning_rate": 6.9078505726652345e-06,
+      "loss": 0.3132,
+      "step": 3313
+    },
+    {
+      "epoch": 0.7887189861367288,
+      "grad_norm": 0.3897042525840326,
+      "learning_rate": 6.906068678917442e-06,
+      "loss": 0.3263,
+      "step": 3314
+    },
+    {
+      "epoch": 0.7889569822097935,
+      "grad_norm": 0.3983465394931579,
+      "learning_rate": 6.904286501880688e-06,
+      "loss": 0.4121,
+      "step": 3315
+    },
+    {
+      "epoch": 0.7891949782828583,
+      "grad_norm": 0.3960830865161725,
+      "learning_rate": 6.902504041819853e-06,
+      "loss": 0.3513,
+      "step": 3316
+    },
+    {
+      "epoch": 0.7894329743559231,
+      "grad_norm": 0.4092338028148718,
+      "learning_rate": 6.900721298999849e-06,
+      "loss": 0.3118,
+      "step": 3317
+    },
+    {
+      "epoch": 0.7896709704289879,
+      "grad_norm": 0.37481354646128145,
+      "learning_rate": 6.8989382736856405e-06,
+      "loss": 0.3556,
+      "step": 3318
+    },
+    {
+      "epoch": 0.7899089665020527,
+      "grad_norm": 0.4041714681573852,
+      "learning_rate": 6.897154966142225e-06,
+      "loss": 0.3973,
+      "step": 3319
+    },
+    {
+      "epoch": 0.7901469625751175,
+      "grad_norm": 0.3961472202444824,
+      "learning_rate": 6.89537137663465e-06,
+      "loss": 0.2757,
+      "step": 3320
+    },
+    {
+      "epoch": 0.7903849586481823,
+      "grad_norm": 0.3476476966030962,
+      "learning_rate": 6.893587505427997e-06,
+      "loss": 0.3343,
+      "step": 3321
+    },
+    {
+      "epoch": 0.7906229547212471,
+      "grad_norm": 0.4234915791131859,
+      "learning_rate": 6.891803352787396e-06,
+      "loss": 0.3884,
+      "step": 3322
+    },
+    {
+      "epoch": 0.7908609507943118,
+      "grad_norm": 0.3688518168774175,
+      "learning_rate": 6.890018918978018e-06,
+      "loss": 0.3443,
+      "step": 3323
+    },
+    {
+      "epoch": 0.7910989468673767,
+      "grad_norm": 0.44308279354142094,
+      "learning_rate": 6.888234204265071e-06,
+      "loss": 0.2956,
+      "step": 3324
+    },
+    {
+      "epoch": 0.7913369429404414,
+      "grad_norm": 0.3728873788461928,
+      "learning_rate": 6.8864492089138076e-06,
+      "loss": 0.3573,
+      "step": 3325
+    },
+    {
+      "epoch": 0.7915749390135063,
+      "grad_norm": 0.3691109038727998,
+      "learning_rate": 6.8846639331895235e-06,
+      "loss": 0.415,
+      "step": 3326
+    },
+    {
+      "epoch": 0.791812935086571,
+      "grad_norm": 0.40429091720127247,
+      "learning_rate": 6.882878377357555e-06,
+      "loss": 0.3342,
+      "step": 3327
+    },
+    {
+      "epoch": 0.7920509311596359,
+      "grad_norm": 0.35173869530824603,
+      "learning_rate": 6.881092541683279e-06,
+      "loss": 0.3036,
+      "step": 3328
+    },
+    {
+      "epoch": 0.7922889272327006,
+      "grad_norm": 0.3691282683210561,
+      "learning_rate": 6.879306426432116e-06,
+      "loss": 0.3957,
+      "step": 3329
+    },
+    {
+      "epoch": 0.7925269233057655,
+      "grad_norm": 0.3547986919135188,
+      "learning_rate": 6.877520031869527e-06,
+      "loss": 0.3853,
+      "step": 3330
+    },
+    {
+      "epoch": 0.7927649193788302,
+      "grad_norm": 0.41969998025380606,
+      "learning_rate": 6.875733358261012e-06,
+      "loss": 0.312,
+      "step": 3331
+    },
+    {
+      "epoch": 0.793002915451895,
+      "grad_norm": 0.3666600692656378,
+      "learning_rate": 6.873946405872116e-06,
+      "loss": 0.3293,
+      "step": 3332
+    },
+    {
+      "epoch": 0.7932409115249598,
+      "grad_norm": 0.3793195334251969,
+      "learning_rate": 6.872159174968427e-06,
+      "loss": 0.3999,
+      "step": 3333
+    },
+    {
+      "epoch": 0.7934789075980246,
+      "grad_norm": 0.3653083899549032,
+      "learning_rate": 6.870371665815567e-06,
+      "loss": 0.347,
+      "step": 3334
+    },
+    {
+      "epoch": 0.7937169036710894,
+      "grad_norm": 0.42000680796497764,
+      "learning_rate": 6.868583878679209e-06,
+      "loss": 0.3133,
+      "step": 3335
+    },
+    {
+      "epoch": 0.7939548997441542,
+      "grad_norm": 0.3798826072515094,
+      "learning_rate": 6.866795813825059e-06,
+      "loss": 0.3855,
+      "step": 3336
+    },
+    {
+      "epoch": 0.794192895817219,
+      "grad_norm": 0.4204127342603903,
+      "learning_rate": 6.8650074715188695e-06,
+      "loss": 0.4339,
+      "step": 3337
+    },
+    {
+      "epoch": 0.7944308918902838,
+      "grad_norm": 0.3987640353825363,
+      "learning_rate": 6.863218852026432e-06,
+      "loss": 0.3097,
+      "step": 3338
+    },
+    {
+      "epoch": 0.7946688879633486,
+      "grad_norm": 0.3758268701259083,
+      "learning_rate": 6.861429955613579e-06,
+      "loss": 0.3308,
+      "step": 3339
+    },
+    {
+      "epoch": 0.7949068840364134,
+      "grad_norm": 0.3465046232838342,
+      "learning_rate": 6.859640782546183e-06,
+      "loss": 0.3863,
+      "step": 3340
+    },
+    {
+      "epoch": 0.7951448801094781,
+      "grad_norm": 0.3592668570581204,
+      "learning_rate": 6.8578513330901645e-06,
+      "loss": 0.3441,
+      "step": 3341
+    },
+    {
+      "epoch": 0.795382876182543,
+      "grad_norm": 0.3714878302081704,
+      "learning_rate": 6.856061607511475e-06,
+      "loss": 0.3282,
+      "step": 3342
+    },
+    {
+      "epoch": 0.7956208722556077,
+      "grad_norm": 0.3717879012940753,
+      "learning_rate": 6.854271606076114e-06,
+      "loss": 0.3704,
+      "step": 3343
+    },
+    {
+      "epoch": 0.7958588683286726,
+      "grad_norm": 0.36524627168544693,
+      "learning_rate": 6.85248132905012e-06,
+      "loss": 0.4022,
+      "step": 3344
+    },
+    {
+      "epoch": 0.7960968644017373,
+      "grad_norm": 0.3880622674807223,
+      "learning_rate": 6.850690776699574e-06,
+      "loss": 0.3121,
+      "step": 3345
+    },
+    {
+      "epoch": 0.7963348604748022,
+      "grad_norm": 0.38708118353283516,
+      "learning_rate": 6.848899949290592e-06,
+      "loss": 0.3444,
+      "step": 3346
+    },
+    {
+      "epoch": 0.7965728565478669,
+      "grad_norm": 0.3721719168783397,
+      "learning_rate": 6.847108847089339e-06,
+      "loss": 0.3891,
+      "step": 3347
+    },
+    {
+      "epoch": 0.7968108526209318,
+      "grad_norm": 0.4005020642472224,
+      "learning_rate": 6.8453174703620155e-06,
+      "loss": 0.3959,
+      "step": 3348
+    },
+    {
+      "epoch": 0.7970488486939965,
+      "grad_norm": 0.3522265566325754,
+      "learning_rate": 6.843525819374866e-06,
+      "loss": 0.3006,
+      "step": 3349
+    },
+    {
+      "epoch": 0.7972868447670614,
+      "grad_norm": 0.359814924986342,
+      "learning_rate": 6.841733894394172e-06,
+      "loss": 0.3521,
+      "step": 3350
+    },
+    {
+      "epoch": 0.7975248408401261,
+      "grad_norm": 0.3908892005179815,
+      "learning_rate": 6.839941695686261e-06,
+      "loss": 0.4427,
+      "step": 3351
+    },
+    {
+      "epoch": 0.7977628369131909,
+      "grad_norm": 0.39295135554595395,
+      "learning_rate": 6.838149223517495e-06,
+      "loss": 0.3039,
+      "step": 3352
+    },
+    {
+      "epoch": 0.7980008329862557,
+      "grad_norm": 0.3999452447353942,
+      "learning_rate": 6.836356478154279e-06,
+      "loss": 0.3105,
+      "step": 3353
+    },
+    {
+      "epoch": 0.7982388290593205,
+      "grad_norm": 0.36113133945632986,
+      "learning_rate": 6.834563459863064e-06,
+      "loss": 0.4003,
+      "step": 3354
+    },
+    {
+      "epoch": 0.7984768251323853,
+      "grad_norm": 0.3993728396402891,
+      "learning_rate": 6.832770168910332e-06,
+      "loss": 0.3429,
+      "step": 3355
+    },
+    {
+      "epoch": 0.7987148212054501,
+      "grad_norm": 0.36265914054354226,
+      "learning_rate": 6.830976605562614e-06,
+      "loss": 0.3019,
+      "step": 3356
+    },
+    {
+      "epoch": 0.7989528172785149,
+      "grad_norm": 0.3760662642357464,
+      "learning_rate": 6.829182770086474e-06,
+      "loss": 0.4016,
+      "step": 3357
+    },
+    {
+      "epoch": 0.7991908133515797,
+      "grad_norm": 0.3558139104694944,
+      "learning_rate": 6.8273886627485245e-06,
+      "loss": 0.4189,
+      "step": 3358
+    },
+    {
+      "epoch": 0.7994288094246444,
+      "grad_norm": 0.42200701223038184,
+      "learning_rate": 6.825594283815411e-06,
+      "loss": 0.3216,
+      "step": 3359
+    },
+    {
+      "epoch": 0.7996668054977093,
+      "grad_norm": 0.3628739848145437,
+      "learning_rate": 6.8237996335538245e-06,
+      "loss": 0.3099,
+      "step": 3360
+    },
+    {
+      "epoch": 0.799904801570774,
+      "grad_norm": 0.3622850476056826,
+      "learning_rate": 6.822004712230493e-06,
+      "loss": 0.3894,
+      "step": 3361
+    },
+    {
+      "epoch": 0.8001427976438389,
+      "grad_norm": 0.37608877349141273,
+      "learning_rate": 6.820209520112188e-06,
+      "loss": 0.3764,
+      "step": 3362
+    },
+    {
+      "epoch": 0.8003807937169036,
+      "grad_norm": 0.3921518941166787,
+      "learning_rate": 6.8184140574657185e-06,
+      "loss": 0.2933,
+      "step": 3363
+    },
+    {
+      "epoch": 0.8006187897899685,
+      "grad_norm": 0.3709070132502542,
+      "learning_rate": 6.816618324557934e-06,
+      "loss": 0.3279,
+      "step": 3364
+    },
+    {
+      "epoch": 0.8008567858630332,
+      "grad_norm": 0.3457649897592504,
+      "learning_rate": 6.8148223216557275e-06,
+      "loss": 0.421,
+      "step": 3365
+    },
+    {
+      "epoch": 0.8010947819360981,
+      "grad_norm": 0.363710583757097,
+      "learning_rate": 6.813026049026026e-06,
+      "loss": 0.3431,
+      "step": 3366
+    },
+    {
+      "epoch": 0.8013327780091628,
+      "grad_norm": 0.44025195568280423,
+      "learning_rate": 6.8112295069358005e-06,
+      "loss": 0.2966,
+      "step": 3367
+    },
+    {
+      "epoch": 0.8015707740822277,
+      "grad_norm": 0.44684338780747807,
+      "learning_rate": 6.809432695652063e-06,
+      "loss": 0.3689,
+      "step": 3368
+    },
+    {
+      "epoch": 0.8018087701552924,
+      "grad_norm": 0.3866828600799754,
+      "learning_rate": 6.807635615441866e-06,
+      "loss": 0.4114,
+      "step": 3369
+    },
+    {
+      "epoch": 0.8020467662283572,
+      "grad_norm": 0.3608818486936804,
+      "learning_rate": 6.805838266572296e-06,
+      "loss": 0.3066,
+      "step": 3370
+    },
+    {
+      "epoch": 0.802284762301422,
+      "grad_norm": 0.3903835514254373,
+      "learning_rate": 6.804040649310485e-06,
+      "loss": 0.3154,
+      "step": 3371
+    },
+    {
+      "epoch": 0.8025227583744868,
+      "grad_norm": 0.4174028450269264,
+      "learning_rate": 6.802242763923603e-06,
+      "loss": 0.3932,
+      "step": 3372
+    },
+    {
+      "epoch": 0.8027607544475516,
+      "grad_norm": 0.3902998602460606,
+      "learning_rate": 6.800444610678862e-06,
+      "loss": 0.3686,
+      "step": 3373
+    },
+    {
+      "epoch": 0.8029987505206164,
+      "grad_norm": 0.37542329189902135,
+      "learning_rate": 6.798646189843512e-06,
+      "loss": 0.2835,
+      "step": 3374
+    },
+    {
+      "epoch": 0.8032367465936812,
+      "grad_norm": 0.5826616163085041,
+      "learning_rate": 6.796847501684839e-06,
+      "loss": 0.3604,
+      "step": 3375
+    },
+    {
+      "epoch": 0.803474742666746,
+      "grad_norm": 0.36773753481276283,
+      "learning_rate": 6.795048546470178e-06,
+      "loss": 0.3924,
+      "step": 3376
+    },
+    {
+      "epoch": 0.8037127387398108,
+      "grad_norm": 0.3653653462075791,
+      "learning_rate": 6.793249324466895e-06,
+      "loss": 0.3302,
+      "step": 3377
+    },
+    {
+      "epoch": 0.8039507348128756,
+      "grad_norm": 0.3674665028573087,
+      "learning_rate": 6.7914498359424e-06,
+      "loss": 0.2912,
+      "step": 3378
+    },
+    {
+      "epoch": 0.8041887308859403,
+      "grad_norm": 0.3535016420982183,
+      "learning_rate": 6.78965008116414e-06,
+      "loss": 0.3777,
+      "step": 3379
+    },
+    {
+      "epoch": 0.8044267269590052,
+      "grad_norm": 0.3809175328016834,
+      "learning_rate": 6.787850060399604e-06,
+      "loss": 0.4065,
+      "step": 3380
+    },
+    {
+      "epoch": 0.8046647230320699,
+      "grad_norm": 0.38371526792218086,
+      "learning_rate": 6.78604977391632e-06,
+      "loss": 0.301,
+      "step": 3381
+    },
+    {
+      "epoch": 0.8049027191051348,
+      "grad_norm": 0.37885740937735535,
+      "learning_rate": 6.784249221981856e-06,
+      "loss": 0.3888,
+      "step": 3382
+    },
+    {
+      "epoch": 0.8051407151781995,
+      "grad_norm": 0.37193003169198097,
+      "learning_rate": 6.782448404863816e-06,
+      "loss": 0.3764,
+      "step": 3383
+    },
+    {
+      "epoch": 0.8053787112512644,
+      "grad_norm": 0.40081137867530753,
+      "learning_rate": 6.780647322829849e-06,
+      "loss": 0.3487,
+      "step": 3384
+    },
+    {
+      "epoch": 0.8056167073243291,
+      "grad_norm": 0.38584530133036227,
+      "learning_rate": 6.778845976147638e-06,
+      "loss": 0.3073,
+      "step": 3385
+    },
+    {
+      "epoch": 0.805854703397394,
+      "grad_norm": 0.3737571701313767,
+      "learning_rate": 6.777044365084907e-06,
+      "loss": 0.389,
+      "step": 3386
+    },
+    {
+      "epoch": 0.8060926994704587,
+      "grad_norm": 0.38914805596699603,
+      "learning_rate": 6.775242489909423e-06,
+      "loss": 0.3707,
+      "step": 3387
+    },
+    {
+      "epoch": 0.8063306955435235,
+      "grad_norm": 0.3908919574280197,
+      "learning_rate": 6.773440350888986e-06,
+      "loss": 0.3293,
+      "step": 3388
+    },
+    {
+      "epoch": 0.8065686916165883,
+      "grad_norm": 0.3732810073093854,
+      "learning_rate": 6.771637948291441e-06,
+      "loss": 0.3214,
+      "step": 3389
+    },
+    {
+      "epoch": 0.8068066876896531,
+      "grad_norm": 0.40490720807254776,
+      "learning_rate": 6.769835282384669e-06,
+      "loss": 0.3995,
+      "step": 3390
+    },
+    {
+      "epoch": 0.8070446837627179,
+      "grad_norm": 0.39783899196176287,
+      "learning_rate": 6.768032353436591e-06,
+      "loss": 0.364,
+      "step": 3391
+    },
+    {
+      "epoch": 0.8072826798357827,
+      "grad_norm": 0.36646595976846763,
+      "learning_rate": 6.766229161715165e-06,
+      "loss": 0.2909,
+      "step": 3392
+    },
+    {
+      "epoch": 0.8075206759088475,
+      "grad_norm": 0.4596377496880723,
+      "learning_rate": 6.764425707488393e-06,
+      "loss": 0.3568,
+      "step": 3393
+    },
+    {
+      "epoch": 0.8077586719819123,
+      "grad_norm": 0.38919838139408713,
+      "learning_rate": 6.76262199102431e-06,
+      "loss": 0.4245,
+      "step": 3394
+    },
+    {
+      "epoch": 0.807996668054977,
+      "grad_norm": 0.3756696902040613,
+      "learning_rate": 6.760818012590993e-06,
+      "loss": 0.3169,
+      "step": 3395
+    },
+    {
+      "epoch": 0.8082346641280419,
+      "grad_norm": 0.38466137056243654,
+      "learning_rate": 6.75901377245656e-06,
+      "loss": 0.3322,
+      "step": 3396
+    },
+    {
+      "epoch": 0.8084726602011066,
+      "grad_norm": 0.372823608083243,
+      "learning_rate": 6.757209270889164e-06,
+      "loss": 0.3765,
+      "step": 3397
+    },
+    {
+      "epoch": 0.8087106562741715,
+      "grad_norm": 0.40319286820868333,
+      "learning_rate": 6.755404508156999e-06,
+      "loss": 0.3428,
+      "step": 3398
+    },
+    {
+      "epoch": 0.8089486523472362,
+      "grad_norm": 0.37211593039177365,
+      "learning_rate": 6.753599484528297e-06,
+      "loss": 0.3013,
+      "step": 3399
+    },
+    {
+      "epoch": 0.8091866484203011,
+      "grad_norm": 0.36450494947046486,
+      "learning_rate": 6.75179420027133e-06,
+      "loss": 0.3378,
+      "step": 3400
+    },
+    {
+      "epoch": 0.8094246444933658,
+      "grad_norm": 0.35084818549241026,
+      "learning_rate": 6.749988655654408e-06,
+      "loss": 0.4222,
+      "step": 3401
+    },
+    {
+      "epoch": 0.8096626405664307,
+      "grad_norm": 0.38771334806435903,
+      "learning_rate": 6.748182850945878e-06,
+      "loss": 0.3184,
+      "step": 3402
+    },
+    {
+      "epoch": 0.8099006366394954,
+      "grad_norm": 0.40751175987989763,
+      "learning_rate": 6.746376786414129e-06,
+      "loss": 0.3106,
+      "step": 3403
+    },
+    {
+      "epoch": 0.8101386327125603,
+      "grad_norm": 0.3763443373885834,
+      "learning_rate": 6.744570462327588e-06,
+      "loss": 0.4034,
+      "step": 3404
+    },
+    {
+      "epoch": 0.810376628785625,
+      "grad_norm": 0.3912057352411011,
+      "learning_rate": 6.742763878954716e-06,
+      "loss": 0.3267,
+      "step": 3405
+    },
+    {
+      "epoch": 0.8106146248586898,
+      "grad_norm": 0.38899642397811124,
+      "learning_rate": 6.740957036564018e-06,
+      "loss": 0.2985,
+      "step": 3406
+    },
+    {
+      "epoch": 0.8108526209317546,
+      "grad_norm": 0.38659168634889995,
+      "learning_rate": 6.739149935424036e-06,
+      "loss": 0.3583,
+      "step": 3407
+    },
+    {
+      "epoch": 0.8110906170048194,
+      "grad_norm": 0.38015341694039984,
+      "learning_rate": 6.737342575803347e-06,
+      "loss": 0.4034,
+      "step": 3408
+    },
+    {
+      "epoch": 0.8113286130778842,
+      "grad_norm": 0.3784853530949804,
+      "learning_rate": 6.735534957970573e-06,
+      "loss": 0.3077,
+      "step": 3409
+    },
+    {
+      "epoch": 0.811566609150949,
+      "grad_norm": 0.3715243825768867,
+      "learning_rate": 6.733727082194369e-06,
+      "loss": 0.3189,
+      "step": 3410
+    },
+    {
+      "epoch": 0.8118046052240138,
+      "grad_norm": 0.3920251893951503,
+      "learning_rate": 6.73191894874343e-06,
+      "loss": 0.3666,
+      "step": 3411
+    },
+    {
+      "epoch": 0.8120426012970786,
+      "grad_norm": 0.4047784901748257,
+      "learning_rate": 6.73011055788649e-06,
+      "loss": 0.412,
+      "step": 3412
+    },
+    {
+      "epoch": 0.8122805973701434,
+      "grad_norm": 0.34961388235995544,
+      "learning_rate": 6.728301909892318e-06,
+      "loss": 0.3028,
+      "step": 3413
+    },
+    {
+      "epoch": 0.8125185934432082,
+      "grad_norm": 0.4964751847494611,
+      "learning_rate": 6.72649300502973e-06,
+      "loss": 0.3354,
+      "step": 3414
+    },
+    {
+      "epoch": 0.8127565895162729,
+      "grad_norm": 0.3737712857277735,
+      "learning_rate": 6.724683843567567e-06,
+      "loss": 0.3834,
+      "step": 3415
+    },
+    {
+      "epoch": 0.8129945855893378,
+      "grad_norm": 0.38533893171844813,
+      "learning_rate": 6.7228744257747195e-06,
+      "loss": 0.338,
+      "step": 3416
+    },
+    {
+      "epoch": 0.8132325816624025,
+      "grad_norm": 0.39656870787598,
+      "learning_rate": 6.72106475192011e-06,
+      "loss": 0.2903,
+      "step": 3417
+    },
+    {
+      "epoch": 0.8134705777354674,
+      "grad_norm": 0.35585546765967224,
+      "learning_rate": 6.719254822272701e-06,
+      "loss": 0.3375,
+      "step": 3418
+    },
+    {
+      "epoch": 0.8137085738085321,
+      "grad_norm": 0.37372380792440973,
+      "learning_rate": 6.717444637101494e-06,
+      "loss": 0.4215,
+      "step": 3419
+    },
+    {
+      "epoch": 0.813946569881597,
+      "grad_norm": 0.38397186576664505,
+      "learning_rate": 6.715634196675527e-06,
+      "loss": 0.3081,
+      "step": 3420
+    },
+    {
+      "epoch": 0.8141845659546617,
+      "grad_norm": 0.42512101870564933,
+      "learning_rate": 6.713823501263874e-06,
+      "loss": 0.3501,
+      "step": 3421
+    },
+    {
+      "epoch": 0.8144225620277266,
+      "grad_norm": 0.39007776244652564,
+      "learning_rate": 6.712012551135651e-06,
+      "loss": 0.3927,
+      "step": 3422
+    },
+    {
+      "epoch": 0.8146605581007913,
+      "grad_norm": 0.37533224466831305,
+      "learning_rate": 6.71020134656001e-06,
+      "loss": 0.3562,
+      "step": 3423
+    },
+    {
+      "epoch": 0.8148985541738561,
+      "grad_norm": 0.3691992345183985,
+      "learning_rate": 6.708389887806142e-06,
+      "loss": 0.3062,
+      "step": 3424
+    },
+    {
+      "epoch": 0.8151365502469209,
+      "grad_norm": 0.3836542262713381,
+      "learning_rate": 6.706578175143271e-06,
+      "loss": 0.3577,
+      "step": 3425
+    },
+    {
+      "epoch": 0.8153745463199857,
+      "grad_norm": 0.34274121993433526,
+      "learning_rate": 6.704766208840666e-06,
+      "loss": 0.3955,
+      "step": 3426
+    },
+    {
+      "epoch": 0.8156125423930505,
+      "grad_norm": 0.3785837148211837,
+      "learning_rate": 6.702953989167627e-06,
+      "loss": 0.328,
+      "step": 3427
+    },
+    {
+      "epoch": 0.8158505384661153,
+      "grad_norm": 0.367311727838402,
+      "learning_rate": 6.701141516393497e-06,
+      "loss": 0.3133,
+      "step": 3428
+    },
+    {
+      "epoch": 0.8160885345391801,
+      "grad_norm": 0.38174012941670704,
+      "learning_rate": 6.6993287907876526e-06,
+      "loss": 0.3642,
+      "step": 3429
+    },
+    {
+      "epoch": 0.8163265306122449,
+      "grad_norm": 0.36620752112375166,
+      "learning_rate": 6.6975158126195114e-06,
+      "loss": 0.4239,
+      "step": 3430
+    },
+    {
+      "epoch": 0.8165645266853097,
+      "grad_norm": 0.388324045604444,
+      "learning_rate": 6.695702582158527e-06,
+      "loss": 0.3357,
+      "step": 3431
+    },
+    {
+      "epoch": 0.8168025227583745,
+      "grad_norm": 0.3707836017064938,
+      "learning_rate": 6.693889099674188e-06,
+      "loss": 0.3521,
+      "step": 3432
+    },
+    {
+      "epoch": 0.8170405188314392,
+      "grad_norm": 0.3610380265483462,
+      "learning_rate": 6.692075365436024e-06,
+      "loss": 0.3975,
+      "step": 3433
+    },
+    {
+      "epoch": 0.8172785149045041,
+      "grad_norm": 0.3848858140680098,
+      "learning_rate": 6.690261379713601e-06,
+      "loss": 0.3366,
+      "step": 3434
+    },
+    {
+      "epoch": 0.8175165109775688,
+      "grad_norm": 0.41597504909165217,
+      "learning_rate": 6.688447142776522e-06,
+      "loss": 0.3135,
+      "step": 3435
+    },
+    {
+      "epoch": 0.8177545070506337,
+      "grad_norm": 0.3631226395036403,
+      "learning_rate": 6.6866326548944276e-06,
+      "loss": 0.3642,
+      "step": 3436
+    },
+    {
+      "epoch": 0.8179925031236984,
+      "grad_norm": 0.3821455357196431,
+      "learning_rate": 6.684817916336994e-06,
+      "loss": 0.3862,
+      "step": 3437
+    },
+    {
+      "epoch": 0.8182304991967633,
+      "grad_norm": 0.3585464421983625,
+      "learning_rate": 6.683002927373938e-06,
+      "loss": 0.3127,
+      "step": 3438
+    },
+    {
+      "epoch": 0.818468495269828,
+      "grad_norm": 0.37218554816765576,
+      "learning_rate": 6.681187688275013e-06,
+      "loss": 0.3131,
+      "step": 3439
+    },
+    {
+      "epoch": 0.8187064913428929,
+      "grad_norm": 0.3627840886800982,
+      "learning_rate": 6.679372199310006e-06,
+      "loss": 0.4235,
+      "step": 3440
+    },
+    {
+      "epoch": 0.8189444874159576,
+      "grad_norm": 0.4463956845389517,
+      "learning_rate": 6.677556460748744e-06,
+      "loss": 0.3329,
+      "step": 3441
+    },
+    {
+      "epoch": 0.8191824834890225,
+      "grad_norm": 0.41940300133110525,
+      "learning_rate": 6.675740472861092e-06,
+      "loss": 0.2945,
+      "step": 3442
+    },
+    {
+      "epoch": 0.8194204795620872,
+      "grad_norm": 0.3861526446671477,
+      "learning_rate": 6.673924235916948e-06,
+      "loss": 0.3535,
+      "step": 3443
+    },
+    {
+      "epoch": 0.819658475635152,
+      "grad_norm": 0.44522995704309204,
+      "learning_rate": 6.672107750186255e-06,
+      "loss": 0.3907,
+      "step": 3444
+    },
+    {
+      "epoch": 0.8198964717082168,
+      "grad_norm": 0.43143411864941417,
+      "learning_rate": 6.670291015938983e-06,
+      "loss": 0.2998,
+      "step": 3445
+    },
+    {
+      "epoch": 0.8201344677812816,
+      "grad_norm": 0.4500423391786268,
+      "learning_rate": 6.6684740334451445e-06,
+      "loss": 0.3279,
+      "step": 3446
+    },
+    {
+      "epoch": 0.8203724638543464,
+      "grad_norm": 0.3777270892774342,
+      "learning_rate": 6.666656802974789e-06,
+      "loss": 0.361,
+      "step": 3447
+    },
+    {
+      "epoch": 0.8206104599274112,
+      "grad_norm": 0.43511204091192573,
+      "learning_rate": 6.664839324798002e-06,
+      "loss": 0.394,
+      "step": 3448
+    },
+    {
+      "epoch": 0.820848456000476,
+      "grad_norm": 0.37499179792352805,
+      "learning_rate": 6.663021599184904e-06,
+      "loss": 0.3216,
+      "step": 3449
+    },
+    {
+      "epoch": 0.8210864520735408,
+      "grad_norm": 0.42216483258775034,
+      "learning_rate": 6.661203626405656e-06,
+      "loss": 0.3902,
+      "step": 3450
+    },
+    {
+      "epoch": 0.8213244481466055,
+      "grad_norm": 0.39104996064612785,
+      "learning_rate": 6.659385406730452e-06,
+      "loss": 0.4236,
+      "step": 3451
+    },
+    {
+      "epoch": 0.8215624442196704,
+      "grad_norm": 0.36054918099443395,
+      "learning_rate": 6.6575669404295265e-06,
+      "loss": 0.3189,
+      "step": 3452
+    },
+    {
+      "epoch": 0.8218004402927351,
+      "grad_norm": 0.40811287467124596,
+      "learning_rate": 6.6557482277731465e-06,
+      "loss": 0.3119,
+      "step": 3453
+    },
+    {
+      "epoch": 0.8220384363658,
+      "grad_norm": 0.36440658348004856,
+      "learning_rate": 6.653929269031618e-06,
+      "loss": 0.3753,
+      "step": 3454
+    },
+    {
+      "epoch": 0.8222764324388647,
+      "grad_norm": 0.4096319668844619,
+      "learning_rate": 6.652110064475286e-06,
+      "loss": 0.3708,
+      "step": 3455
+    },
+    {
+      "epoch": 0.8225144285119296,
+      "grad_norm": 0.4138762726742505,
+      "learning_rate": 6.650290614374526e-06,
+      "loss": 0.2696,
+      "step": 3456
+    },
+    {
+      "epoch": 0.8227524245849943,
+      "grad_norm": 0.38274207266784216,
+      "learning_rate": 6.648470918999754e-06,
+      "loss": 0.3701,
+      "step": 3457
+    },
+    {
+      "epoch": 0.8229904206580592,
+      "grad_norm": 0.38596079001363315,
+      "learning_rate": 6.646650978621422e-06,
+      "loss": 0.3932,
+      "step": 3458
+    },
+    {
+      "epoch": 0.8232284167311239,
+      "grad_norm": 0.4033700687920439,
+      "learning_rate": 6.644830793510019e-06,
+      "loss": 0.3134,
+      "step": 3459
+    },
+    {
+      "epoch": 0.8234664128041888,
+      "grad_norm": 0.3957395882295056,
+      "learning_rate": 6.64301036393607e-06,
+      "loss": 0.2913,
+      "step": 3460
+    },
+    {
+      "epoch": 0.8237044088772535,
+      "grad_norm": 0.37332886096760914,
+      "learning_rate": 6.641189690170135e-06,
+      "loss": 0.3702,
+      "step": 3461
+    },
+    {
+      "epoch": 0.8239424049503183,
+      "grad_norm": 0.382271014403154,
+      "learning_rate": 6.639368772482809e-06,
+      "loss": 0.3848,
+      "step": 3462
+    },
+    {
+      "epoch": 0.8241804010233831,
+      "grad_norm": 0.42293254876321,
+      "learning_rate": 6.637547611144729e-06,
+      "loss": 0.3168,
+      "step": 3463
+    },
+    {
+      "epoch": 0.8244183970964479,
+      "grad_norm": 0.5064018854163713,
+      "learning_rate": 6.635726206426562e-06,
+      "loss": 0.3296,
+      "step": 3464
+    },
+    {
+      "epoch": 0.8246563931695127,
+      "grad_norm": 0.3699399246485159,
+      "learning_rate": 6.633904558599015e-06,
+      "loss": 0.379,
+      "step": 3465
+    },
+    {
+      "epoch": 0.8248943892425775,
+      "grad_norm": 0.36822069349704023,
+      "learning_rate": 6.63208266793283e-06,
+      "loss": 0.3622,
+      "step": 3466
+    },
+    {
+      "epoch": 0.8251323853156423,
+      "grad_norm": 0.37564917131168607,
+      "learning_rate": 6.630260534698784e-06,
+      "loss": 0.2927,
+      "step": 3467
+    },
+    {
+      "epoch": 0.8253703813887071,
+      "grad_norm": 0.359534156281574,
+      "learning_rate": 6.628438159167691e-06,
+      "loss": 0.3703,
+      "step": 3468
+    },
+    {
+      "epoch": 0.8256083774617718,
+      "grad_norm": 0.41724063858650523,
+      "learning_rate": 6.626615541610404e-06,
+      "loss": 0.4272,
+      "step": 3469
+    },
+    {
+      "epoch": 0.8258463735348367,
+      "grad_norm": 0.35435885568070713,
+      "learning_rate": 6.624792682297807e-06,
+      "loss": 0.2995,
+      "step": 3470
+    },
+    {
+      "epoch": 0.8260843696079014,
+      "grad_norm": 0.3924531439546138,
+      "learning_rate": 6.62296958150082e-06,
+      "loss": 0.3106,
+      "step": 3471
+    },
+    {
+      "epoch": 0.8263223656809663,
+      "grad_norm": 0.3793607220180146,
+      "learning_rate": 6.621146239490405e-06,
+      "loss": 0.3642,
+      "step": 3472
+    },
+    {
+      "epoch": 0.826560361754031,
+      "grad_norm": 0.3757765538144861,
+      "learning_rate": 6.619322656537552e-06,
+      "loss": 0.3751,
+      "step": 3473
+    },
+    {
+      "epoch": 0.8267983578270959,
+      "grad_norm": 0.379936104170262,
+      "learning_rate": 6.6174988329132935e-06,
+      "loss": 0.3125,
+      "step": 3474
+    },
+    {
+      "epoch": 0.8270363539001606,
+      "grad_norm": 0.36029253203578504,
+      "learning_rate": 6.615674768888693e-06,
+      "loss": 0.3439,
+      "step": 3475
+    },
+    {
+      "epoch": 0.8272743499732255,
+      "grad_norm": 0.400916363212535,
+      "learning_rate": 6.613850464734852e-06,
+      "loss": 0.3963,
+      "step": 3476
+    },
+    {
+      "epoch": 0.8275123460462902,
+      "grad_norm": 0.39903964241453344,
+      "learning_rate": 6.6120259207229074e-06,
+      "loss": 0.3212,
+      "step": 3477
+    },
+    {
+      "epoch": 0.827750342119355,
+      "grad_norm": 0.4140249986684947,
+      "learning_rate": 6.61020113712403e-06,
+      "loss": 0.3088,
+      "step": 3478
+    },
+    {
+      "epoch": 0.8279883381924198,
+      "grad_norm": 0.3755700342561937,
+      "learning_rate": 6.60837611420943e-06,
+      "loss": 0.3936,
+      "step": 3479
+    },
+    {
+      "epoch": 0.8282263342654846,
+      "grad_norm": 0.3669468151714262,
+      "learning_rate": 6.606550852250351e-06,
+      "loss": 0.3653,
+      "step": 3480
+    },
+    {
+      "epoch": 0.8284643303385494,
+      "grad_norm": 0.3630825299629299,
+      "learning_rate": 6.60472535151807e-06,
+      "loss": 0.2969,
+      "step": 3481
+    },
+    {
+      "epoch": 0.8287023264116142,
+      "grad_norm": 0.48144687843420203,
+      "learning_rate": 6.602899612283903e-06,
+      "loss": 0.3609,
+      "step": 3482
+    },
+    {
+      "epoch": 0.828940322484679,
+      "grad_norm": 0.3375193982835446,
+      "learning_rate": 6.6010736348192e-06,
+      "loss": 0.4184,
+      "step": 3483
+    },
+    {
+      "epoch": 0.8291783185577438,
+      "grad_norm": 0.40206568719269975,
+      "learning_rate": 6.599247419395346e-06,
+      "loss": 0.3326,
+      "step": 3484
+    },
+    {
+      "epoch": 0.8294163146308086,
+      "grad_norm": 0.35032011037790545,
+      "learning_rate": 6.597420966283762e-06,
+      "loss": 0.3043,
+      "step": 3485
+    },
+    {
+      "epoch": 0.8296543107038734,
+      "grad_norm": 0.36114199055488116,
+      "learning_rate": 6.595594275755905e-06,
+      "loss": 0.3566,
+      "step": 3486
+    },
+    {
+      "epoch": 0.8298923067769381,
+      "grad_norm": 0.40221894140390374,
+      "learning_rate": 6.593767348083264e-06,
+      "loss": 0.4178,
+      "step": 3487
+    },
+    {
+      "epoch": 0.830130302850003,
+      "grad_norm": 0.4219667697893653,
+      "learning_rate": 6.591940183537369e-06,
+      "loss": 0.301,
+      "step": 3488
+    },
+    {
+      "epoch": 0.8303682989230677,
+      "grad_norm": 0.4089831045897673,
+      "learning_rate": 6.590112782389779e-06,
+      "loss": 0.3545,
+      "step": 3489
+    },
+    {
+      "epoch": 0.8306062949961326,
+      "grad_norm": 0.3568585115767541,
+      "learning_rate": 6.588285144912092e-06,
+      "loss": 0.4073,
+      "step": 3490
+    },
+    {
+      "epoch": 0.8308442910691973,
+      "grad_norm": 0.39444918515542654,
+      "learning_rate": 6.58645727137594e-06,
+      "loss": 0.3751,
+      "step": 3491
+    },
+    {
+      "epoch": 0.8310822871422622,
+      "grad_norm": 0.40709353343588783,
+      "learning_rate": 6.584629162052991e-06,
+      "loss": 0.3463,
+      "step": 3492
+    },
+    {
+      "epoch": 0.8313202832153269,
+      "grad_norm": 0.3926505024184683,
+      "learning_rate": 6.582800817214947e-06,
+      "loss": 0.372,
+      "step": 3493
+    },
+    {
+      "epoch": 0.8315582792883918,
+      "grad_norm": 0.39799166748115355,
+      "learning_rate": 6.5809722371335425e-06,
+      "loss": 0.4018,
+      "step": 3494
+    },
+    {
+      "epoch": 0.8317962753614565,
+      "grad_norm": 0.38694896220045694,
+      "learning_rate": 6.579143422080555e-06,
+      "loss": 0.3137,
+      "step": 3495
+    },
+    {
+      "epoch": 0.8320342714345214,
+      "grad_norm": 0.3643527025654742,
+      "learning_rate": 6.577314372327788e-06,
+      "loss": 0.3131,
+      "step": 3496
+    },
+    {
+      "epoch": 0.8322722675075861,
+      "grad_norm": 0.3818835372248485,
+      "learning_rate": 6.575485088147085e-06,
+      "loss": 0.3707,
+      "step": 3497
+    },
+    {
+      "epoch": 0.8325102635806509,
+      "grad_norm": 0.40425205370807665,
+      "learning_rate": 6.57365556981032e-06,
+      "loss": 0.3499,
+      "step": 3498
+    },
+    {
+      "epoch": 0.8327482596537157,
+      "grad_norm": 0.4772831093586169,
+      "learning_rate": 6.571825817589409e-06,
+      "loss": 0.3433,
+      "step": 3499
+    },
+    {
+      "epoch": 0.8329862557267805,
+      "grad_norm": 0.37500312469382296,
+      "learning_rate": 6.569995831756296e-06,
+      "loss": 0.3384,
+      "step": 3500
+    },
+    {
+      "epoch": 0.8332242517998453,
+      "grad_norm": 0.3878483799536773,
+      "learning_rate": 6.568165612582963e-06,
+      "loss": 0.4041,
+      "step": 3501
+    },
+    {
+      "epoch": 0.8334622478729101,
+      "grad_norm": 0.39314546980554216,
+      "learning_rate": 6.566335160341425e-06,
+      "loss": 0.3057,
+      "step": 3502
+    },
+    {
+      "epoch": 0.8337002439459749,
+      "grad_norm": 0.39459765976708316,
+      "learning_rate": 6.564504475303732e-06,
+      "loss": 0.3598,
+      "step": 3503
+    },
+    {
+      "epoch": 0.8339382400190397,
+      "grad_norm": 0.37557187665586705,
+      "learning_rate": 6.562673557741972e-06,
+      "loss": 0.3976,
+      "step": 3504
+    },
+    {
+      "epoch": 0.8341762360921045,
+      "grad_norm": 0.3666761239770583,
+      "learning_rate": 6.560842407928261e-06,
+      "loss": 0.3936,
+      "step": 3505
+    },
+    {
+      "epoch": 0.8344142321651693,
+      "grad_norm": 0.43334042433611564,
+      "learning_rate": 6.559011026134755e-06,
+      "loss": 0.309,
+      "step": 3506
+    },
+    {
+      "epoch": 0.834652228238234,
+      "grad_norm": 0.3639925280228689,
+      "learning_rate": 6.557179412633643e-06,
+      "loss": 0.3225,
+      "step": 3507
+    },
+    {
+      "epoch": 0.8348902243112989,
+      "grad_norm": 0.368213342084047,
+      "learning_rate": 6.555347567697147e-06,
+      "loss": 0.4027,
+      "step": 3508
+    },
+    {
+      "epoch": 0.8351282203843636,
+      "grad_norm": 0.40280786786827166,
+      "learning_rate": 6.553515491597525e-06,
+      "loss": 0.2956,
+      "step": 3509
+    },
+    {
+      "epoch": 0.8353662164574285,
+      "grad_norm": 0.3901118662358053,
+      "learning_rate": 6.55168318460707e-06,
+      "loss": 0.3167,
+      "step": 3510
+    },
+    {
+      "epoch": 0.8356042125304932,
+      "grad_norm": 0.36744551369501394,
+      "learning_rate": 6.549850646998106e-06,
+      "loss": 0.3774,
+      "step": 3511
+    },
+    {
+      "epoch": 0.8358422086035581,
+      "grad_norm": 0.42446857731031934,
+      "learning_rate": 6.548017879042993e-06,
+      "loss": 0.3903,
+      "step": 3512
+    },
+    {
+      "epoch": 0.8360802046766228,
+      "grad_norm": 0.4358337216970375,
+      "learning_rate": 6.546184881014128e-06,
+      "loss": 0.316,
+      "step": 3513
+    },
+    {
+      "epoch": 0.8363182007496877,
+      "grad_norm": 0.37095075714460507,
+      "learning_rate": 6.54435165318394e-06,
+      "loss": 0.3294,
+      "step": 3514
+    },
+    {
+      "epoch": 0.8365561968227524,
+      "grad_norm": 0.3735370617178222,
+      "learning_rate": 6.54251819582489e-06,
+      "loss": 0.4038,
+      "step": 3515
+    },
+    {
+      "epoch": 0.8367941928958172,
+      "grad_norm": 0.3754617184639044,
+      "learning_rate": 6.5406845092094775e-06,
+      "loss": 0.3109,
+      "step": 3516
+    },
+    {
+      "epoch": 0.837032188968882,
+      "grad_norm": 0.4256438483351794,
+      "learning_rate": 6.5388505936102305e-06,
+      "loss": 0.3236,
+      "step": 3517
+    },
+    {
+      "epoch": 0.8372701850419468,
+      "grad_norm": 0.36228819168311865,
+      "learning_rate": 6.537016449299718e-06,
+      "loss": 0.3672,
+      "step": 3518
+    },
+    {
+      "epoch": 0.8375081811150116,
+      "grad_norm": 0.3567201288288663,
+      "learning_rate": 6.5351820765505345e-06,
+      "loss": 0.3876,
+      "step": 3519
+    },
+    {
+      "epoch": 0.8377461771880764,
+      "grad_norm": 0.36599561427777705,
+      "learning_rate": 6.533347475635316e-06,
+      "loss": 0.3043,
+      "step": 3520
+    },
+    {
+      "epoch": 0.8379841732611412,
+      "grad_norm": 0.4053564651011694,
+      "learning_rate": 6.531512646826731e-06,
+      "loss": 0.3253,
+      "step": 3521
+    },
+    {
+      "epoch": 0.838222169334206,
+      "grad_norm": 0.37600903870469027,
+      "learning_rate": 6.529677590397478e-06,
+      "loss": 0.4036,
+      "step": 3522
+    },
+    {
+      "epoch": 0.8384601654072708,
+      "grad_norm": 0.361124659195214,
+      "learning_rate": 6.527842306620294e-06,
+      "loss": 0.3696,
+      "step": 3523
+    },
+    {
+      "epoch": 0.8386981614803356,
+      "grad_norm": 0.3908512268752686,
+      "learning_rate": 6.5260067957679455e-06,
+      "loss": 0.2881,
+      "step": 3524
+    },
+    {
+      "epoch": 0.8389361575534003,
+      "grad_norm": 0.391451223833398,
+      "learning_rate": 6.524171058113236e-06,
+      "loss": 0.3734,
+      "step": 3525
+    },
+    {
+      "epoch": 0.8391741536264652,
+      "grad_norm": 0.34623716604454546,
+      "learning_rate": 6.522335093928999e-06,
+      "loss": 0.4447,
+      "step": 3526
+    },
+    {
+      "epoch": 0.8394121496995299,
+      "grad_norm": 0.35943956265557137,
+      "learning_rate": 6.520498903488108e-06,
+      "loss": 0.3042,
+      "step": 3527
+    },
+    {
+      "epoch": 0.8396501457725948,
+      "grad_norm": 0.39161434146649104,
+      "learning_rate": 6.518662487063464e-06,
+      "loss": 0.291,
+      "step": 3528
+    },
+    {
+      "epoch": 0.8398881418456595,
+      "grad_norm": 0.37421840622968505,
+      "learning_rate": 6.516825844928005e-06,
+      "loss": 0.3847,
+      "step": 3529
+    },
+    {
+      "epoch": 0.8401261379187244,
+      "grad_norm": 0.37611154296694504,
+      "learning_rate": 6.514988977354701e-06,
+      "loss": 0.3662,
+      "step": 3530
+    },
+    {
+      "epoch": 0.8403641339917891,
+      "grad_norm": 0.3647185189803361,
+      "learning_rate": 6.513151884616556e-06,
+      "loss": 0.3128,
+      "step": 3531
+    },
+    {
+      "epoch": 0.840602130064854,
+      "grad_norm": 0.3821277967922997,
+      "learning_rate": 6.511314566986608e-06,
+      "loss": 0.3503,
+      "step": 3532
+    },
+    {
+      "epoch": 0.8408401261379187,
+      "grad_norm": 0.35310735826505824,
+      "learning_rate": 6.5094770247379256e-06,
+      "loss": 0.3922,
+      "step": 3533
+    },
+    {
+      "epoch": 0.8410781222109835,
+      "grad_norm": 0.39211042496153803,
+      "learning_rate": 6.507639258143615e-06,
+      "loss": 0.3153,
+      "step": 3534
+    },
+    {
+      "epoch": 0.8413161182840483,
+      "grad_norm": 0.37374980390462503,
+      "learning_rate": 6.5058012674768136e-06,
+      "loss": 0.2926,
+      "step": 3535
+    },
+    {
+      "epoch": 0.8415541143571131,
+      "grad_norm": 0.3480206219661194,
+      "learning_rate": 6.5039630530106925e-06,
+      "loss": 0.3595,
+      "step": 3536
+    },
+    {
+      "epoch": 0.8417921104301779,
+      "grad_norm": 0.37436697934855795,
+      "learning_rate": 6.502124615018456e-06,
+      "loss": 0.3896,
+      "step": 3537
+    },
+    {
+      "epoch": 0.8420301065032427,
+      "grad_norm": 0.3783010027033652,
+      "learning_rate": 6.50028595377334e-06,
+      "loss": 0.3203,
+      "step": 3538
+    },
+    {
+      "epoch": 0.8422681025763075,
+      "grad_norm": 0.4307862323394335,
+      "learning_rate": 6.498447069548617e-06,
+      "loss": 0.3373,
+      "step": 3539
+    },
+    {
+      "epoch": 0.8425060986493723,
+      "grad_norm": 0.4059325152116869,
+      "learning_rate": 6.496607962617588e-06,
+      "loss": 0.3827,
+      "step": 3540
+    },
+    {
+      "epoch": 0.842744094722437,
+      "grad_norm": 0.4279651224236223,
+      "learning_rate": 6.494768633253593e-06,
+      "loss": 0.3371,
+      "step": 3541
+    },
+    {
+      "epoch": 0.8429820907955019,
+      "grad_norm": 0.3780915639053451,
+      "learning_rate": 6.492929081729999e-06,
+      "loss": 0.3081,
+      "step": 3542
+    },
+    {
+      "epoch": 0.8432200868685666,
+      "grad_norm": 0.42345332107805106,
+      "learning_rate": 6.491089308320212e-06,
+      "loss": 0.3686,
+      "step": 3543
+    },
+    {
+      "epoch": 0.8434580829416315,
+      "grad_norm": 0.3479273768259802,
+      "learning_rate": 6.489249313297665e-06,
+      "loss": 0.4429,
+      "step": 3544
+    },
+    {
+      "epoch": 0.8436960790146962,
+      "grad_norm": 0.41111841314637215,
+      "learning_rate": 6.487409096935828e-06,
+      "loss": 0.3142,
+      "step": 3545
+    },
+    {
+      "epoch": 0.8439340750877611,
+      "grad_norm": 0.41453492108831,
+      "learning_rate": 6.485568659508201e-06,
+      "loss": 0.3128,
+      "step": 3546
+    },
+    {
+      "epoch": 0.8441720711608258,
+      "grad_norm": 0.3412459944917207,
+      "learning_rate": 6.483728001288322e-06,
+      "loss": 0.3878,
+      "step": 3547
+    },
+    {
+      "epoch": 0.8444100672338907,
+      "grad_norm": 0.3580627139101501,
+      "learning_rate": 6.481887122549755e-06,
+      "loss": 0.3663,
+      "step": 3548
+    },
+    {
+      "epoch": 0.8446480633069554,
+      "grad_norm": 0.3881281457560659,
+      "learning_rate": 6.480046023566101e-06,
+      "loss": 0.3162,
+      "step": 3549
+    },
+    {
+      "epoch": 0.8448860593800203,
+      "grad_norm": 0.38842732635621374,
+      "learning_rate": 6.4782047046109956e-06,
+      "loss": 0.3745,
+      "step": 3550
+    },
+    {
+      "epoch": 0.845124055453085,
+      "grad_norm": 0.368464384136309,
+      "learning_rate": 6.476363165958101e-06,
+      "loss": 0.4105,
+      "step": 3551
+    },
+    {
+      "epoch": 0.8453620515261498,
+      "grad_norm": 0.42844621258294174,
+      "learning_rate": 6.474521407881116e-06,
+      "loss": 0.3035,
+      "step": 3552
+    },
+    {
+      "epoch": 0.8456000475992146,
+      "grad_norm": 0.37487158122933406,
+      "learning_rate": 6.472679430653771e-06,
+      "loss": 0.3028,
+      "step": 3553
+    },
+    {
+      "epoch": 0.8458380436722794,
+      "grad_norm": 0.3697192895630358,
+      "learning_rate": 6.470837234549831e-06,
+      "loss": 0.3821,
+      "step": 3554
+    },
+    {
+      "epoch": 0.8460760397453442,
+      "grad_norm": 0.39897676938087634,
+      "learning_rate": 6.468994819843093e-06,
+      "loss": 0.3787,
+      "step": 3555
+    },
+    {
+      "epoch": 0.846314035818409,
+      "grad_norm": 0.370440881250878,
+      "learning_rate": 6.467152186807381e-06,
+      "loss": 0.3008,
+      "step": 3556
+    },
+    {
+      "epoch": 0.8465520318914738,
+      "grad_norm": 0.36810682890311036,
+      "learning_rate": 6.4653093357165605e-06,
+      "loss": 0.3324,
+      "step": 3557
+    },
+    {
+      "epoch": 0.8467900279645386,
+      "grad_norm": 0.44038968365113795,
+      "learning_rate": 6.463466266844523e-06,
+      "loss": 0.4133,
+      "step": 3558
+    },
+    {
+      "epoch": 0.8470280240376034,
+      "grad_norm": 0.42146397355664883,
+      "learning_rate": 6.461622980465192e-06,
+      "loss": 0.3262,
+      "step": 3559
+    },
+    {
+      "epoch": 0.8472660201106682,
+      "grad_norm": 0.3892143606840479,
+      "learning_rate": 6.459779476852528e-06,
+      "loss": 0.2951,
+      "step": 3560
+    },
+    {
+      "epoch": 0.8475040161837329,
+      "grad_norm": 0.3606861577032436,
+      "learning_rate": 6.45793575628052e-06,
+      "loss": 0.3468,
+      "step": 3561
+    },
+    {
+      "epoch": 0.8477420122567978,
+      "grad_norm": 0.39838162217271655,
+      "learning_rate": 6.456091819023192e-06,
+      "loss": 0.3662,
+      "step": 3562
+    },
+    {
+      "epoch": 0.8479800083298625,
+      "grad_norm": 0.40545284210053256,
+      "learning_rate": 6.454247665354596e-06,
+      "loss": 0.3092,
+      "step": 3563
+    },
+    {
+      "epoch": 0.8482180044029274,
+      "grad_norm": 0.3745103557445221,
+      "learning_rate": 6.452403295548822e-06,
+      "loss": 0.3259,
+      "step": 3564
+    },
+    {
+      "epoch": 0.8484560004759921,
+      "grad_norm": 0.36452519630195096,
+      "learning_rate": 6.450558709879988e-06,
+      "loss": 0.3935,
+      "step": 3565
+    },
+    {
+      "epoch": 0.848693996549057,
+      "grad_norm": 0.3753670451371452,
+      "learning_rate": 6.448713908622244e-06,
+      "loss": 0.364,
+      "step": 3566
+    },
+    {
+      "epoch": 0.8489319926221217,
+      "grad_norm": 0.4415121020870584,
+      "learning_rate": 6.446868892049774e-06,
+      "loss": 0.2946,
+      "step": 3567
+    },
+    {
+      "epoch": 0.8491699886951866,
+      "grad_norm": 0.3586879855069214,
+      "learning_rate": 6.445023660436792e-06,
+      "loss": 0.3556,
+      "step": 3568
+    },
+    {
+      "epoch": 0.8494079847682513,
+      "grad_norm": 0.388777137976686,
+      "learning_rate": 6.443178214057546e-06,
+      "loss": 0.4276,
+      "step": 3569
+    },
+    {
+      "epoch": 0.8496459808413162,
+      "grad_norm": 0.37232395659576667,
+      "learning_rate": 6.441332553186317e-06,
+      "loss": 0.2855,
+      "step": 3570
+    },
+    {
+      "epoch": 0.8498839769143809,
+      "grad_norm": 0.36549499530200263,
+      "learning_rate": 6.439486678097412e-06,
+      "loss": 0.2883,
+      "step": 3571
+    },
+    {
+      "epoch": 0.8501219729874457,
+      "grad_norm": 0.4015119927722684,
+      "learning_rate": 6.437640589065175e-06,
+      "loss": 0.4052,
+      "step": 3572
+    },
+    {
+      "epoch": 0.8503599690605105,
+      "grad_norm": 0.39600433293408427,
+      "learning_rate": 6.435794286363981e-06,
+      "loss": 0.3315,
+      "step": 3573
+    },
+    {
+      "epoch": 0.8505979651335753,
+      "grad_norm": 0.34745554675127294,
+      "learning_rate": 6.4339477702682365e-06,
+      "loss": 0.3379,
+      "step": 3574
+    },
+    {
+      "epoch": 0.8508359612066401,
+      "grad_norm": 0.376990594604472,
+      "learning_rate": 6.4321010410523785e-06,
+      "loss": 0.3708,
+      "step": 3575
+    },
+    {
+      "epoch": 0.8510739572797049,
+      "grad_norm": 0.37580912173380343,
+      "learning_rate": 6.430254098990879e-06,
+      "loss": 0.3867,
+      "step": 3576
+    },
+    {
+      "epoch": 0.8513119533527697,
+      "grad_norm": 0.36492592391634565,
+      "learning_rate": 6.428406944358236e-06,
+      "loss": 0.3329,
+      "step": 3577
+    },
+    {
+      "epoch": 0.8515499494258345,
+      "grad_norm": 0.4561243150373411,
+      "learning_rate": 6.426559577428986e-06,
+      "loss": 0.3443,
+      "step": 3578
+    },
+    {
+      "epoch": 0.8517879454988992,
+      "grad_norm": 0.37361088859679636,
+      "learning_rate": 6.42471199847769e-06,
+      "loss": 0.3821,
+      "step": 3579
+    },
+    {
+      "epoch": 0.8520259415719641,
+      "grad_norm": 0.3858002858596615,
+      "learning_rate": 6.422864207778946e-06,
+      "loss": 0.3575,
+      "step": 3580
+    },
+    {
+      "epoch": 0.8522639376450288,
+      "grad_norm": 0.591361749000545,
+      "learning_rate": 6.42101620560738e-06,
+      "loss": 0.2932,
+      "step": 3581
+    },
+    {
+      "epoch": 0.8525019337180937,
+      "grad_norm": 0.3862018455711387,
+      "learning_rate": 6.4191679922376514e-06,
+      "loss": 0.332,
+      "step": 3582
+    },
+    {
+      "epoch": 0.8527399297911584,
+      "grad_norm": 0.3477194194120392,
+      "learning_rate": 6.417319567944451e-06,
+      "loss": 0.4039,
+      "step": 3583
+    },
+    {
+      "epoch": 0.8529779258642233,
+      "grad_norm": 0.3867671638518174,
+      "learning_rate": 6.4154709330025014e-06,
+      "loss": 0.3283,
+      "step": 3584
+    },
+    {
+      "epoch": 0.853215921937288,
+      "grad_norm": 0.3789296076013951,
+      "learning_rate": 6.413622087686553e-06,
+      "loss": 0.3197,
+      "step": 3585
+    },
+    {
+      "epoch": 0.8534539180103529,
+      "grad_norm": 0.35204003022469893,
+      "learning_rate": 6.411773032271391e-06,
+      "loss": 0.3575,
+      "step": 3586
+    },
+    {
+      "epoch": 0.8536919140834176,
+      "grad_norm": 0.4135524819567001,
+      "learning_rate": 6.4099237670318295e-06,
+      "loss": 0.3739,
+      "step": 3587
+    },
+    {
+      "epoch": 0.8539299101564825,
+      "grad_norm": 0.36478421612897816,
+      "learning_rate": 6.408074292242719e-06,
+      "loss": 0.314,
+      "step": 3588
+    },
+    {
+      "epoch": 0.8541679062295472,
+      "grad_norm": 0.3991900857385473,
+      "learning_rate": 6.4062246081789316e-06,
+      "loss": 0.3417,
+      "step": 3589
+    },
+    {
+      "epoch": 0.854405902302612,
+      "grad_norm": 0.35893254692737425,
+      "learning_rate": 6.40437471511538e-06,
+      "loss": 0.3911,
+      "step": 3590
+    },
+    {
+      "epoch": 0.8546438983756768,
+      "grad_norm": 0.35453540240239667,
+      "learning_rate": 6.402524613327005e-06,
+      "loss": 0.3469,
+      "step": 3591
+    },
+    {
+      "epoch": 0.8548818944487416,
+      "grad_norm": 0.43997392067335855,
+      "learning_rate": 6.400674303088774e-06,
+      "loss": 0.2904,
+      "step": 3592
+    },
+    {
+      "epoch": 0.8551198905218064,
+      "grad_norm": 0.3932703368817032,
+      "learning_rate": 6.398823784675692e-06,
+      "loss": 0.3453,
+      "step": 3593
+    },
+    {
+      "epoch": 0.8553578865948712,
+      "grad_norm": 0.38331111856702316,
+      "learning_rate": 6.3969730583627895e-06,
+      "loss": 0.4321,
+      "step": 3594
+    },
+    {
+      "epoch": 0.855595882667936,
+      "grad_norm": 0.36082426472750595,
+      "learning_rate": 6.39512212442513e-06,
+      "loss": 0.303,
+      "step": 3595
+    },
+    {
+      "epoch": 0.8558338787410008,
+      "grad_norm": 0.36640398807283237,
+      "learning_rate": 6.39327098313781e-06,
+      "loss": 0.3424,
+      "step": 3596
+    },
+    {
+      "epoch": 0.8560718748140655,
+      "grad_norm": 0.39345559455136353,
+      "learning_rate": 6.391419634775955e-06,
+      "loss": 0.3943,
+      "step": 3597
+    },
+    {
+      "epoch": 0.8563098708871304,
+      "grad_norm": 0.38176260350590113,
+      "learning_rate": 6.3895680796147195e-06,
+      "loss": 0.3916,
+      "step": 3598
+    },
+    {
+      "epoch": 0.8565478669601951,
+      "grad_norm": 0.3701812101754324,
+      "learning_rate": 6.387716317929291e-06,
+      "loss": 0.3091,
+      "step": 3599
+    },
+    {
+      "epoch": 0.85678586303326,
+      "grad_norm": 0.37515785937868595,
+      "learning_rate": 6.385864349994887e-06,
+      "loss": 0.3352,
+      "step": 3600
+    },
+    {
+      "epoch": 0.8570238591063247,
+      "grad_norm": 0.3811040883533929,
+      "learning_rate": 6.384012176086756e-06,
+      "loss": 0.4348,
+      "step": 3601
+    },
+    {
+      "epoch": 0.8572618551793896,
+      "grad_norm": 0.38277086910509656,
+      "learning_rate": 6.382159796480176e-06,
+      "loss": 0.3105,
+      "step": 3602
+    },
+    {
+      "epoch": 0.8574998512524543,
+      "grad_norm": 0.4049398160651801,
+      "learning_rate": 6.380307211450459e-06,
+      "loss": 0.2908,
+      "step": 3603
+    },
+    {
+      "epoch": 0.8577378473255192,
+      "grad_norm": 0.37259835586706896,
+      "learning_rate": 6.3784544212729425e-06,
+      "loss": 0.3591,
+      "step": 3604
+    },
+    {
+      "epoch": 0.8579758433985839,
+      "grad_norm": 0.39620623445763115,
+      "learning_rate": 6.376601426222998e-06,
+      "loss": 0.417,
+      "step": 3605
+    },
+    {
+      "epoch": 0.8582138394716488,
+      "grad_norm": 0.38565150714963886,
+      "learning_rate": 6.374748226576026e-06,
+      "loss": 0.3028,
+      "step": 3606
+    },
+    {
+      "epoch": 0.8584518355447135,
+      "grad_norm": 0.38396283701575223,
+      "learning_rate": 6.372894822607459e-06,
+      "loss": 0.3637,
+      "step": 3607
+    },
+    {
+      "epoch": 0.8586898316177783,
+      "grad_norm": 0.36500484893052054,
+      "learning_rate": 6.371041214592756e-06,
+      "loss": 0.3775,
+      "step": 3608
+    },
+    {
+      "epoch": 0.8589278276908431,
+      "grad_norm": 0.3661904393716665,
+      "learning_rate": 6.369187402807409e-06,
+      "loss": 0.3382,
+      "step": 3609
+    },
+    {
+      "epoch": 0.8591658237639079,
+      "grad_norm": 0.3635402605710434,
+      "learning_rate": 6.3673333875269435e-06,
+      "loss": 0.3263,
+      "step": 3610
+    },
+    {
+      "epoch": 0.8594038198369727,
+      "grad_norm": 0.34166628776145125,
+      "learning_rate": 6.3654791690269115e-06,
+      "loss": 0.3689,
+      "step": 3611
+    },
+    {
+      "epoch": 0.8596418159100375,
+      "grad_norm": 0.40702525148527946,
+      "learning_rate": 6.363624747582895e-06,
+      "loss": 0.3635,
+      "step": 3612
+    },
+    {
+      "epoch": 0.8598798119831023,
+      "grad_norm": 0.3788739157966428,
+      "learning_rate": 6.361770123470506e-06,
+      "loss": 0.3199,
+      "step": 3613
+    },
+    {
+      "epoch": 0.8601178080561671,
+      "grad_norm": 0.4010995049473161,
+      "learning_rate": 6.359915296965386e-06,
+      "loss": 0.3378,
+      "step": 3614
+    },
+    {
+      "epoch": 0.8603558041292318,
+      "grad_norm": 0.3976190901244627,
+      "learning_rate": 6.3580602683432114e-06,
+      "loss": 0.3947,
+      "step": 3615
+    },
+    {
+      "epoch": 0.8605938002022967,
+      "grad_norm": 0.41040643943188415,
+      "learning_rate": 6.356205037879683e-06,
+      "loss": 0.353,
+      "step": 3616
+    },
+    {
+      "epoch": 0.8608317962753614,
+      "grad_norm": 0.40129402029771805,
+      "learning_rate": 6.354349605850537e-06,
+      "loss": 0.3115,
+      "step": 3617
+    },
+    {
+      "epoch": 0.8610697923484263,
+      "grad_norm": 0.37800238180174833,
+      "learning_rate": 6.352493972531535e-06,
+      "loss": 0.3461,
+      "step": 3618
+    },
+    {
+      "epoch": 0.861307788421491,
+      "grad_norm": 0.35601066827900923,
+      "learning_rate": 6.350638138198468e-06,
+      "loss": 0.3927,
+      "step": 3619
+    },
+    {
+      "epoch": 0.8615457844945559,
+      "grad_norm": 0.4294833642760016,
+      "learning_rate": 6.348782103127161e-06,
+      "loss": 0.3186,
+      "step": 3620
+    },
+    {
+      "epoch": 0.8617837805676206,
+      "grad_norm": 0.39885340739339903,
+      "learning_rate": 6.346925867593468e-06,
+      "loss": 0.3481,
+      "step": 3621
+    },
+    {
+      "epoch": 0.8620217766406855,
+      "grad_norm": 0.381073935397439,
+      "learning_rate": 6.345069431873267e-06,
+      "loss": 0.3632,
+      "step": 3622
+    },
+    {
+      "epoch": 0.8622597727137502,
+      "grad_norm": 0.3716616071286771,
+      "learning_rate": 6.3432127962424724e-06,
+      "loss": 0.3735,
+      "step": 3623
+    },
+    {
+      "epoch": 0.8624977687868151,
+      "grad_norm": 0.3936052335005196,
+      "learning_rate": 6.341355960977029e-06,
+      "loss": 0.3002,
+      "step": 3624
+    },
+    {
+      "epoch": 0.8627357648598798,
+      "grad_norm": 0.35707110608872256,
+      "learning_rate": 6.3394989263529075e-06,
+      "loss": 0.3286,
+      "step": 3625
+    },
+    {
+      "epoch": 0.8629737609329446,
+      "grad_norm": 0.37387621795612574,
+      "learning_rate": 6.337641692646106e-06,
+      "loss": 0.4234,
+      "step": 3626
+    },
+    {
+      "epoch": 0.8632117570060094,
+      "grad_norm": 0.36608540038166837,
+      "learning_rate": 6.335784260132656e-06,
+      "loss": 0.3326,
+      "step": 3627
+    },
+    {
+      "epoch": 0.8634497530790742,
+      "grad_norm": 0.40136651207935303,
+      "learning_rate": 6.33392662908862e-06,
+      "loss": 0.3229,
+      "step": 3628
+    },
+    {
+      "epoch": 0.863687749152139,
+      "grad_norm": 0.38445657475293393,
+      "learning_rate": 6.332068799790088e-06,
+      "loss": 0.3608,
+      "step": 3629
+    },
+    {
+      "epoch": 0.8639257452252038,
+      "grad_norm": 0.4134034698432213,
+      "learning_rate": 6.330210772513179e-06,
+      "loss": 0.3888,
+      "step": 3630
+    },
+    {
+      "epoch": 0.8641637412982686,
+      "grad_norm": 0.3403794858258773,
+      "learning_rate": 6.32835254753404e-06,
+      "loss": 0.3299,
+      "step": 3631
+    },
+    {
+      "epoch": 0.8644017373713334,
+      "grad_norm": 0.400740147025239,
+      "learning_rate": 6.3264941251288524e-06,
+      "loss": 0.3519,
+      "step": 3632
+    },
+    {
+      "epoch": 0.8646397334443982,
+      "grad_norm": 0.3696283238042445,
+      "learning_rate": 6.324635505573821e-06,
+      "loss": 0.4068,
+      "step": 3633
+    },
+    {
+      "epoch": 0.864877729517463,
+      "grad_norm": 0.3648815050871534,
+      "learning_rate": 6.3227766891451834e-06,
+      "loss": 0.3229,
+      "step": 3634
+    },
+    {
+      "epoch": 0.8651157255905277,
+      "grad_norm": 0.3684339728207634,
+      "learning_rate": 6.3209176761192056e-06,
+      "loss": 0.2937,
+      "step": 3635
+    },
+    {
+      "epoch": 0.8653537216635926,
+      "grad_norm": 0.3412830815800425,
+      "learning_rate": 6.319058466772183e-06,
+      "loss": 0.3547,
+      "step": 3636
+    },
+    {
+      "epoch": 0.8655917177366573,
+      "grad_norm": 0.4151380109178161,
+      "learning_rate": 6.317199061380442e-06,
+      "loss": 0.4081,
+      "step": 3637
+    },
+    {
+      "epoch": 0.8658297138097222,
+      "grad_norm": 0.3807487486762807,
+      "learning_rate": 6.3153394602203335e-06,
+      "loss": 0.2962,
+      "step": 3638
+    },
+    {
+      "epoch": 0.8660677098827869,
+      "grad_norm": 0.3818685065331829,
+      "learning_rate": 6.313479663568241e-06,
+      "loss": 0.3313,
+      "step": 3639
+    },
+    {
+      "epoch": 0.8663057059558518,
+      "grad_norm": 0.39854588070720026,
+      "learning_rate": 6.311619671700577e-06,
+      "loss": 0.3949,
+      "step": 3640
+    },
+    {
+      "epoch": 0.8665437020289165,
+      "grad_norm": 0.39040880832128244,
+      "learning_rate": 6.309759484893781e-06,
+      "loss": 0.3463,
+      "step": 3641
+    },
+    {
+      "epoch": 0.8667816981019814,
+      "grad_norm": 0.3808134094910771,
+      "learning_rate": 6.3078991034243246e-06,
+      "loss": 0.3069,
+      "step": 3642
+    },
+    {
+      "epoch": 0.8670196941750461,
+      "grad_norm": 0.42611867594264174,
+      "learning_rate": 6.306038527568703e-06,
+      "loss": 0.3499,
+      "step": 3643
+    },
+    {
+      "epoch": 0.867257690248111,
+      "grad_norm": 0.38268539446297695,
+      "learning_rate": 6.304177757603449e-06,
+      "loss": 0.381,
+      "step": 3644
+    },
+    {
+      "epoch": 0.8674956863211757,
+      "grad_norm": 0.39139862991744423,
+      "learning_rate": 6.302316793805117e-06,
+      "loss": 0.311,
+      "step": 3645
+    },
+    {
+      "epoch": 0.8677336823942405,
+      "grad_norm": 0.34134245528800944,
+      "learning_rate": 6.300455636450291e-06,
+      "loss": 0.3326,
+      "step": 3646
+    },
+    {
+      "epoch": 0.8679716784673053,
+      "grad_norm": 0.3603534975781144,
+      "learning_rate": 6.298594285815585e-06,
+      "loss": 0.3793,
+      "step": 3647
+    },
+    {
+      "epoch": 0.8682096745403701,
+      "grad_norm": 0.3614007549196889,
+      "learning_rate": 6.296732742177644e-06,
+      "loss": 0.3392,
+      "step": 3648
+    },
+    {
+      "epoch": 0.8684476706134349,
+      "grad_norm": 0.4195748954343944,
+      "learning_rate": 6.294871005813137e-06,
+      "loss": 0.3134,
+      "step": 3649
+    },
+    {
+      "epoch": 0.8686856666864997,
+      "grad_norm": 0.39119705289423146,
+      "learning_rate": 6.293009076998763e-06,
+      "loss": 0.3361,
+      "step": 3650
+    },
+    {
+      "epoch": 0.8689236627595645,
+      "grad_norm": 0.4029024677261677,
+      "learning_rate": 6.291146956011255e-06,
+      "loss": 0.3868,
+      "step": 3651
+    },
+    {
+      "epoch": 0.8691616588326293,
+      "grad_norm": 0.3811799244974477,
+      "learning_rate": 6.289284643127367e-06,
+      "loss": 0.3218,
+      "step": 3652
+    },
+    {
+      "epoch": 0.869399654905694,
+      "grad_norm": 0.4261716940666211,
+      "learning_rate": 6.287422138623886e-06,
+      "loss": 0.3203,
+      "step": 3653
+    },
+    {
+      "epoch": 0.8696376509787589,
+      "grad_norm": 0.39445236227113656,
+      "learning_rate": 6.285559442777624e-06,
+      "loss": 0.4,
+      "step": 3654
+    },
+    {
+      "epoch": 0.8698756470518236,
+      "grad_norm": 0.4319574272153776,
+      "learning_rate": 6.283696555865429e-06,
+      "loss": 0.3435,
+      "step": 3655
+    },
+    {
+      "epoch": 0.8701136431248885,
+      "grad_norm": 0.4546433169970997,
+      "learning_rate": 6.281833478164168e-06,
+      "loss": 0.3051,
+      "step": 3656
+    },
+    {
+      "epoch": 0.8703516391979532,
+      "grad_norm": 0.38154852645595494,
+      "learning_rate": 6.279970209950738e-06,
+      "loss": 0.3498,
+      "step": 3657
+    },
+    {
+      "epoch": 0.8705896352710181,
+      "grad_norm": 0.3548764531740995,
+      "learning_rate": 6.278106751502073e-06,
+      "loss": 0.4182,
+      "step": 3658
+    },
+    {
+      "epoch": 0.8708276313440828,
+      "grad_norm": 0.3759251982640629,
+      "learning_rate": 6.2762431030951255e-06,
+      "loss": 0.3106,
+      "step": 3659
+    },
+    {
+      "epoch": 0.8710656274171477,
+      "grad_norm": 0.3613232709245374,
+      "learning_rate": 6.274379265006879e-06,
+      "loss": 0.2905,
+      "step": 3660
+    },
+    {
+      "epoch": 0.8713036234902124,
+      "grad_norm": 0.3751051295735669,
+      "learning_rate": 6.272515237514349e-06,
+      "loss": 0.358,
+      "step": 3661
+    },
+    {
+      "epoch": 0.8715416195632772,
+      "grad_norm": 0.36121575959400304,
+      "learning_rate": 6.270651020894572e-06,
+      "loss": 0.4264,
+      "step": 3662
+    },
+    {
+      "epoch": 0.871779615636342,
+      "grad_norm": 0.3780337870363989,
+      "learning_rate": 6.2687866154246204e-06,
+      "loss": 0.3072,
+      "step": 3663
+    },
+    {
+      "epoch": 0.8720176117094068,
+      "grad_norm": 0.4035208753610946,
+      "learning_rate": 6.266922021381588e-06,
+      "loss": 0.3603,
+      "step": 3664
+    },
+    {
+      "epoch": 0.8722556077824716,
+      "grad_norm": 0.3977305240744663,
+      "learning_rate": 6.265057239042602e-06,
+      "loss": 0.4096,
+      "step": 3665
+    },
+    {
+      "epoch": 0.8724936038555364,
+      "grad_norm": 0.3826214538741579,
+      "learning_rate": 6.263192268684814e-06,
+      "loss": 0.3427,
+      "step": 3666
+    },
+    {
+      "epoch": 0.8727315999286012,
+      "grad_norm": 0.40584625486221326,
+      "learning_rate": 6.2613271105854065e-06,
+      "loss": 0.3293,
+      "step": 3667
+    },
+    {
+      "epoch": 0.872969596001666,
+      "grad_norm": 0.6329608207460291,
+      "learning_rate": 6.259461765021584e-06,
+      "loss": 0.3824,
+      "step": 3668
+    },
+    {
+      "epoch": 0.8732075920747308,
+      "grad_norm": 0.3898204136913921,
+      "learning_rate": 6.257596232270587e-06,
+      "loss": 0.411,
+      "step": 3669
+    },
+    {
+      "epoch": 0.8734455881477956,
+      "grad_norm": 0.36641905581111767,
+      "learning_rate": 6.255730512609679e-06,
+      "loss": 0.3169,
+      "step": 3670
+    },
+    {
+      "epoch": 0.8736835842208603,
+      "grad_norm": 0.4829470116833935,
+      "learning_rate": 6.25386460631615e-06,
+      "loss": 0.3299,
+      "step": 3671
+    },
+    {
+      "epoch": 0.8739215802939252,
+      "grad_norm": 0.3562617699548361,
+      "learning_rate": 6.2519985136673235e-06,
+      "loss": 0.3818,
+      "step": 3672
+    },
+    {
+      "epoch": 0.8741595763669899,
+      "grad_norm": 0.3887628157005426,
+      "learning_rate": 6.250132234940543e-06,
+      "loss": 0.3696,
+      "step": 3673
+    },
+    {
+      "epoch": 0.8743975724400548,
+      "grad_norm": 0.39340605103932813,
+      "learning_rate": 6.248265770413187e-06,
+      "loss": 0.3048,
+      "step": 3674
+    },
+    {
+      "epoch": 0.8746355685131195,
+      "grad_norm": 0.3573200120544181,
+      "learning_rate": 6.2463991203626565e-06,
+      "loss": 0.3333,
+      "step": 3675
+    },
+    {
+      "epoch": 0.8748735645861844,
+      "grad_norm": 0.3773239240601227,
+      "learning_rate": 6.244532285066382e-06,
+      "loss": 0.4064,
+      "step": 3676
+    },
+    {
+      "epoch": 0.8751115606592491,
+      "grad_norm": 0.41082700756267004,
+      "learning_rate": 6.2426652648018215e-06,
+      "loss": 0.3286,
+      "step": 3677
+    },
+    {
+      "epoch": 0.875349556732314,
+      "grad_norm": 0.38054049215364216,
+      "learning_rate": 6.2407980598464615e-06,
+      "loss": 0.2894,
+      "step": 3678
+    },
+    {
+      "epoch": 0.8755875528053787,
+      "grad_norm": 0.3678209031947387,
+      "learning_rate": 6.238930670477813e-06,
+      "loss": 0.3526,
+      "step": 3679
+    },
+    {
+      "epoch": 0.8758255488784435,
+      "grad_norm": 0.40127333894221406,
+      "learning_rate": 6.237063096973418e-06,
+      "loss": 0.3743,
+      "step": 3680
+    },
+    {
+      "epoch": 0.8760635449515083,
+      "grad_norm": 0.3582362616157782,
+      "learning_rate": 6.235195339610842e-06,
+      "loss": 0.3023,
+      "step": 3681
+    },
+    {
+      "epoch": 0.8763015410245731,
+      "grad_norm": 0.44829796994435595,
+      "learning_rate": 6.233327398667682e-06,
+      "loss": 0.3699,
+      "step": 3682
+    },
+    {
+      "epoch": 0.8765395370976379,
+      "grad_norm": 0.3966518210291202,
+      "learning_rate": 6.2314592744215605e-06,
+      "loss": 0.4058,
+      "step": 3683
+    },
+    {
+      "epoch": 0.8767775331707027,
+      "grad_norm": 0.390658057800168,
+      "learning_rate": 6.229590967150124e-06,
+      "loss": 0.3136,
+      "step": 3684
+    },
+    {
+      "epoch": 0.8770155292437675,
+      "grad_norm": 0.4308954670330036,
+      "learning_rate": 6.227722477131053e-06,
+      "loss": 0.2851,
+      "step": 3685
+    },
+    {
+      "epoch": 0.8772535253168323,
+      "grad_norm": 0.442296326588899,
+      "learning_rate": 6.225853804642048e-06,
+      "loss": 0.3569,
+      "step": 3686
+    },
+    {
+      "epoch": 0.8774915213898971,
+      "grad_norm": 0.36301211552233426,
+      "learning_rate": 6.223984949960843e-06,
+      "loss": 0.3704,
+      "step": 3687
+    },
+    {
+      "epoch": 0.8777295174629619,
+      "grad_norm": 0.39720988569190824,
+      "learning_rate": 6.2221159133651946e-06,
+      "loss": 0.301,
+      "step": 3688
+    },
+    {
+      "epoch": 0.8779675135360266,
+      "grad_norm": 0.3772685192299192,
+      "learning_rate": 6.220246695132887e-06,
+      "loss": 0.3498,
+      "step": 3689
+    },
+    {
+      "epoch": 0.8782055096090915,
+      "grad_norm": 0.35844628342397195,
+      "learning_rate": 6.218377295541733e-06,
+      "loss": 0.3919,
+      "step": 3690
+    },
+    {
+      "epoch": 0.8784435056821562,
+      "grad_norm": 0.39072910342836975,
+      "learning_rate": 6.21650771486957e-06,
+      "loss": 0.3412,
+      "step": 3691
+    },
+    {
+      "epoch": 0.8786815017552211,
+      "grad_norm": 0.3934761558248729,
+      "learning_rate": 6.214637953394268e-06,
+      "loss": 0.3195,
+      "step": 3692
+    },
+    {
+      "epoch": 0.8789194978282858,
+      "grad_norm": 0.41802350381203424,
+      "learning_rate": 6.212768011393717e-06,
+      "loss": 0.3613,
+      "step": 3693
+    },
+    {
+      "epoch": 0.8791574939013507,
+      "grad_norm": 0.3673222836006008,
+      "learning_rate": 6.2108978891458374e-06,
+      "loss": 0.4147,
+      "step": 3694
+    },
+    {
+      "epoch": 0.8793954899744154,
+      "grad_norm": 0.3792040026583229,
+      "learning_rate": 6.2090275869285735e-06,
+      "loss": 0.3214,
+      "step": 3695
+    },
+    {
+      "epoch": 0.8796334860474803,
+      "grad_norm": 0.39223693432096834,
+      "learning_rate": 6.207157105019902e-06,
+      "loss": 0.3345,
+      "step": 3696
+    },
+    {
+      "epoch": 0.879871482120545,
+      "grad_norm": 0.36686866689029535,
+      "learning_rate": 6.205286443697821e-06,
+      "loss": 0.4232,
+      "step": 3697
+    },
+    {
+      "epoch": 0.8801094781936099,
+      "grad_norm": 0.3987162836648213,
+      "learning_rate": 6.2034156032403555e-06,
+      "loss": 0.3754,
+      "step": 3698
+    },
+    {
+      "epoch": 0.8803474742666746,
+      "grad_norm": 0.4458906881177551,
+      "learning_rate": 6.201544583925562e-06,
+      "loss": 0.3313,
+      "step": 3699
+    },
+    {
+      "epoch": 0.8805854703397394,
+      "grad_norm": 0.3774384264045408,
+      "learning_rate": 6.199673386031518e-06,
+      "loss": 0.3442,
+      "step": 3700
+    },
+    {
+      "epoch": 0.8808234664128042,
+      "grad_norm": 0.34013768584119114,
+      "learning_rate": 6.197802009836331e-06,
+      "loss": 0.4061,
+      "step": 3701
+    },
+    {
+      "epoch": 0.881061462485869,
+      "grad_norm": 0.37237424732062263,
+      "learning_rate": 6.195930455618132e-06,
+      "loss": 0.333,
+      "step": 3702
+    },
+    {
+      "epoch": 0.8812994585589338,
+      "grad_norm": 0.3960116351456216,
+      "learning_rate": 6.194058723655083e-06,
+      "loss": 0.3069,
+      "step": 3703
+    },
+    {
+      "epoch": 0.8815374546319986,
+      "grad_norm": 0.39926411331947254,
+      "learning_rate": 6.192186814225367e-06,
+      "loss": 0.3738,
+      "step": 3704
+    },
+    {
+      "epoch": 0.8817754507050634,
+      "grad_norm": 0.44748846445351975,
+      "learning_rate": 6.190314727607196e-06,
+      "loss": 0.3749,
+      "step": 3705
+    },
+    {
+      "epoch": 0.8820134467781282,
+      "grad_norm": 0.41088375205239147,
+      "learning_rate": 6.188442464078811e-06,
+      "loss": 0.29,
+      "step": 3706
+    },
+    {
+      "epoch": 0.882251442851193,
+      "grad_norm": 0.36272393711398615,
+      "learning_rate": 6.1865700239184755e-06,
+      "loss": 0.337,
+      "step": 3707
+    },
+    {
+      "epoch": 0.8824894389242578,
+      "grad_norm": 0.38722534388780844,
+      "learning_rate": 6.184697407404478e-06,
+      "loss": 0.4223,
+      "step": 3708
+    },
+    {
+      "epoch": 0.8827274349973225,
+      "grad_norm": 0.4042014513980286,
+      "learning_rate": 6.18282461481514e-06,
+      "loss": 0.3241,
+      "step": 3709
+    },
+    {
+      "epoch": 0.8829654310703874,
+      "grad_norm": 0.4002874055607258,
+      "learning_rate": 6.180951646428801e-06,
+      "loss": 0.3142,
+      "step": 3710
+    },
+    {
+      "epoch": 0.8832034271434521,
+      "grad_norm": 0.39939392997960865,
+      "learning_rate": 6.179078502523834e-06,
+      "loss": 0.3774,
+      "step": 3711
+    },
+    {
+      "epoch": 0.883441423216517,
+      "grad_norm": 0.36300470644879695,
+      "learning_rate": 6.177205183378629e-06,
+      "loss": 0.4179,
+      "step": 3712
+    },
+    {
+      "epoch": 0.8836794192895817,
+      "grad_norm": 0.36795015375713797,
+      "learning_rate": 6.1753316892716156e-06,
+      "loss": 0.3032,
+      "step": 3713
+    },
+    {
+      "epoch": 0.8839174153626466,
+      "grad_norm": 0.4120733964613225,
+      "learning_rate": 6.173458020481234e-06,
+      "loss": 0.3473,
+      "step": 3714
+    },
+    {
+      "epoch": 0.8841554114357113,
+      "grad_norm": 0.38507546779235274,
+      "learning_rate": 6.171584177285962e-06,
+      "loss": 0.4031,
+      "step": 3715
+    },
+    {
+      "epoch": 0.8843934075087762,
+      "grad_norm": 0.37436623493188265,
+      "learning_rate": 6.1697101599642976e-06,
+      "loss": 0.3287,
+      "step": 3716
+    },
+    {
+      "epoch": 0.8846314035818409,
+      "grad_norm": 0.37601841360225746,
+      "learning_rate": 6.167835968794766e-06,
+      "loss": 0.2813,
+      "step": 3717
+    },
+    {
+      "epoch": 0.8848693996549057,
+      "grad_norm": 0.36926230908935936,
+      "learning_rate": 6.165961604055917e-06,
+      "loss": 0.365,
+      "step": 3718
+    },
+    {
+      "epoch": 0.8851073957279705,
+      "grad_norm": 0.36487906680665755,
+      "learning_rate": 6.1640870660263295e-06,
+      "loss": 0.3984,
+      "step": 3719
+    },
+    {
+      "epoch": 0.8853453918010353,
+      "grad_norm": 0.37472610973420145,
+      "learning_rate": 6.162212354984607e-06,
+      "loss": 0.3098,
+      "step": 3720
+    },
+    {
+      "epoch": 0.8855833878741001,
+      "grad_norm": 0.37234322290198935,
+      "learning_rate": 6.160337471209377e-06,
+      "loss": 0.3439,
+      "step": 3721
+    },
+    {
+      "epoch": 0.8858213839471649,
+      "grad_norm": 0.3877763209468878,
+      "learning_rate": 6.158462414979292e-06,
+      "loss": 0.3869,
+      "step": 3722
+    },
+    {
+      "epoch": 0.8860593800202297,
+      "grad_norm": 0.4034026778107165,
+      "learning_rate": 6.156587186573033e-06,
+      "loss": 0.3543,
+      "step": 3723
+    },
+    {
+      "epoch": 0.8862973760932945,
+      "grad_norm": 0.3700540316755882,
+      "learning_rate": 6.154711786269307e-06,
+      "loss": 0.3153,
+      "step": 3724
+    },
+    {
+      "epoch": 0.8865353721663592,
+      "grad_norm": 0.39243565604036756,
+      "learning_rate": 6.152836214346843e-06,
+      "loss": 0.3664,
+      "step": 3725
+    },
+    {
+      "epoch": 0.8867733682394241,
+      "grad_norm": 0.36077090161516945,
+      "learning_rate": 6.150960471084397e-06,
+      "loss": 0.4108,
+      "step": 3726
+    },
+    {
+      "epoch": 0.8870113643124888,
+      "grad_norm": 0.37912503567699524,
+      "learning_rate": 6.149084556760753e-06,
+      "loss": 0.3269,
+      "step": 3727
+    },
+    {
+      "epoch": 0.8872493603855537,
+      "grad_norm": 0.4517235850955721,
+      "learning_rate": 6.147208471654715e-06,
+      "loss": 0.3127,
+      "step": 3728
+    },
+    {
+      "epoch": 0.8874873564586184,
+      "grad_norm": 0.387038812957885,
+      "learning_rate": 6.145332216045119e-06,
+      "loss": 0.3647,
+      "step": 3729
+    },
+    {
+      "epoch": 0.8877253525316833,
+      "grad_norm": 0.3779239208107375,
+      "learning_rate": 6.143455790210822e-06,
+      "loss": 0.4041,
+      "step": 3730
+    },
+    {
+      "epoch": 0.887963348604748,
+      "grad_norm": 0.3732475038469887,
+      "learning_rate": 6.1415791944307056e-06,
+      "loss": 0.3049,
+      "step": 3731
+    },
+    {
+      "epoch": 0.8882013446778129,
+      "grad_norm": 0.4145109701498706,
+      "learning_rate": 6.13970242898368e-06,
+      "loss": 0.3649,
+      "step": 3732
+    },
+    {
+      "epoch": 0.8884393407508776,
+      "grad_norm": 0.3649394561399619,
+      "learning_rate": 6.137825494148678e-06,
+      "loss": 0.4286,
+      "step": 3733
+    },
+    {
+      "epoch": 0.8886773368239425,
+      "grad_norm": 0.35279202905563867,
+      "learning_rate": 6.1359483902046605e-06,
+      "loss": 0.3256,
+      "step": 3734
+    },
+    {
+      "epoch": 0.8889153328970072,
+      "grad_norm": 0.3814399015745296,
+      "learning_rate": 6.134071117430609e-06,
+      "loss": 0.3111,
+      "step": 3735
+    },
+    {
+      "epoch": 0.889153328970072,
+      "grad_norm": 0.387544011842372,
+      "learning_rate": 6.132193676105533e-06,
+      "loss": 0.3563,
+      "step": 3736
+    },
+    {
+      "epoch": 0.8893913250431368,
+      "grad_norm": 0.357184072343786,
+      "learning_rate": 6.1303160665084705e-06,
+      "loss": 0.3986,
+      "step": 3737
+    },
+    {
+      "epoch": 0.8896293211162016,
+      "grad_norm": 0.378840936122068,
+      "learning_rate": 6.1284382889184756e-06,
+      "loss": 0.3211,
+      "step": 3738
+    },
+    {
+      "epoch": 0.8898673171892664,
+      "grad_norm": 0.36510963419962517,
+      "learning_rate": 6.126560343614636e-06,
+      "loss": 0.327,
+      "step": 3739
+    },
+    {
+      "epoch": 0.8901053132623312,
+      "grad_norm": 0.359174171499247,
+      "learning_rate": 6.1246822308760575e-06,
+      "loss": 0.4083,
+      "step": 3740
+    },
+    {
+      "epoch": 0.890343309335396,
+      "grad_norm": 0.45021032420547086,
+      "learning_rate": 6.122803950981878e-06,
+      "loss": 0.3493,
+      "step": 3741
+    },
+    {
+      "epoch": 0.8905813054084608,
+      "grad_norm": 0.35860397415514117,
+      "learning_rate": 6.1209255042112546e-06,
+      "loss": 0.2942,
+      "step": 3742
+    },
+    {
+      "epoch": 0.8908193014815255,
+      "grad_norm": 0.36924714501202766,
+      "learning_rate": 6.119046890843371e-06,
+      "loss": 0.3465,
+      "step": 3743
+    },
+    {
+      "epoch": 0.8910572975545904,
+      "grad_norm": 0.3562121440352722,
+      "learning_rate": 6.117168111157435e-06,
+      "loss": 0.4262,
+      "step": 3744
+    },
+    {
+      "epoch": 0.8912952936276551,
+      "grad_norm": 0.35320499238862774,
+      "learning_rate": 6.115289165432681e-06,
+      "loss": 0.3027,
+      "step": 3745
+    },
+    {
+      "epoch": 0.89153328970072,
+      "grad_norm": 0.3759235321559563,
+      "learning_rate": 6.113410053948364e-06,
+      "loss": 0.3309,
+      "step": 3746
+    },
+    {
+      "epoch": 0.8917712857737847,
+      "grad_norm": 0.36596971834289177,
+      "learning_rate": 6.111530776983771e-06,
+      "loss": 0.3858,
+      "step": 3747
+    },
+    {
+      "epoch": 0.8920092818468496,
+      "grad_norm": 0.3707386283118514,
+      "learning_rate": 6.109651334818204e-06,
+      "loss": 0.3554,
+      "step": 3748
+    },
+    {
+      "epoch": 0.8922472779199143,
+      "grad_norm": 0.34061647689147134,
+      "learning_rate": 6.1077717277309986e-06,
+      "loss": 0.2951,
+      "step": 3749
+    },
+    {
+      "epoch": 0.8924852739929792,
+      "grad_norm": 0.36573408045045774,
+      "learning_rate": 6.1058919560015106e-06,
+      "loss": 0.3549,
+      "step": 3750
+    },
+    {
+      "epoch": 0.8927232700660439,
+      "grad_norm": 0.3364732547422088,
+      "learning_rate": 6.104012019909119e-06,
+      "loss": 0.4163,
+      "step": 3751
+    },
+    {
+      "epoch": 0.8929612661391088,
+      "grad_norm": 0.37430490741861316,
+      "learning_rate": 6.102131919733229e-06,
+      "loss": 0.3171,
+      "step": 3752
+    },
+    {
+      "epoch": 0.8931992622121735,
+      "grad_norm": 0.3694758086583808,
+      "learning_rate": 6.1002516557532684e-06,
+      "loss": 0.3298,
+      "step": 3753
+    },
+    {
+      "epoch": 0.8934372582852383,
+      "grad_norm": 0.34593061138576553,
+      "learning_rate": 6.098371228248695e-06,
+      "loss": 0.373,
+      "step": 3754
+    },
+    {
+      "epoch": 0.8936752543583031,
+      "grad_norm": 0.37157546314792966,
+      "learning_rate": 6.096490637498985e-06,
+      "loss": 0.3566,
+      "step": 3755
+    },
+    {
+      "epoch": 0.8939132504313679,
+      "grad_norm": 0.4080002638011446,
+      "learning_rate": 6.09460988378364e-06,
+      "loss": 0.2772,
+      "step": 3756
+    },
+    {
+      "epoch": 0.8941512465044327,
+      "grad_norm": 0.39409190673689626,
+      "learning_rate": 6.092728967382186e-06,
+      "loss": 0.3462,
+      "step": 3757
+    },
+    {
+      "epoch": 0.8943892425774975,
+      "grad_norm": 0.3707276172179733,
+      "learning_rate": 6.090847888574176e-06,
+      "loss": 0.3948,
+      "step": 3758
+    },
+    {
+      "epoch": 0.8946272386505623,
+      "grad_norm": 0.3609993458216574,
+      "learning_rate": 6.088966647639183e-06,
+      "loss": 0.3002,
+      "step": 3759
+    },
+    {
+      "epoch": 0.8948652347236271,
+      "grad_norm": 0.3955638153510142,
+      "learning_rate": 6.087085244856805e-06,
+      "loss": 0.3046,
+      "step": 3760
+    },
+    {
+      "epoch": 0.8951032307966919,
+      "grad_norm": 0.3904254667786321,
+      "learning_rate": 6.0852036805066684e-06,
+      "loss": 0.3435,
+      "step": 3761
+    },
+    {
+      "epoch": 0.8953412268697567,
+      "grad_norm": 0.3708619740535345,
+      "learning_rate": 6.083321954868416e-06,
+      "loss": 0.3902,
+      "step": 3762
+    },
+    {
+      "epoch": 0.8955792229428214,
+      "grad_norm": 0.3966041731350864,
+      "learning_rate": 6.0814400682217236e-06,
+      "loss": 0.2959,
+      "step": 3763
+    },
+    {
+      "epoch": 0.8958172190158863,
+      "grad_norm": 0.4221574195250755,
+      "learning_rate": 6.0795580208462824e-06,
+      "loss": 0.3404,
+      "step": 3764
+    },
+    {
+      "epoch": 0.896055215088951,
+      "grad_norm": 0.3593405180380063,
+      "learning_rate": 6.077675813021812e-06,
+      "loss": 0.3972,
+      "step": 3765
+    },
+    {
+      "epoch": 0.8962932111620159,
+      "grad_norm": 0.39581975922524365,
+      "learning_rate": 6.075793445028056e-06,
+      "loss": 0.344,
+      "step": 3766
+    },
+    {
+      "epoch": 0.8965312072350806,
+      "grad_norm": 0.3955579182890299,
+      "learning_rate": 6.073910917144778e-06,
+      "loss": 0.3158,
+      "step": 3767
+    },
+    {
+      "epoch": 0.8967692033081455,
+      "grad_norm": 0.3797132012723383,
+      "learning_rate": 6.072028229651773e-06,
+      "loss": 0.3482,
+      "step": 3768
+    },
+    {
+      "epoch": 0.8970071993812102,
+      "grad_norm": 0.410201975103764,
+      "learning_rate": 6.0701453828288524e-06,
+      "loss": 0.4076,
+      "step": 3769
+    },
+    {
+      "epoch": 0.8972451954542751,
+      "grad_norm": 0.4137362624937579,
+      "learning_rate": 6.068262376955854e-06,
+      "loss": 0.2942,
+      "step": 3770
+    },
+    {
+      "epoch": 0.8974831915273398,
+      "grad_norm": 0.3409390585560632,
+      "learning_rate": 6.066379212312638e-06,
+      "loss": 0.3169,
+      "step": 3771
+    },
+    {
+      "epoch": 0.8977211876004046,
+      "grad_norm": 0.3649824690116523,
+      "learning_rate": 6.06449588917909e-06,
+      "loss": 0.418,
+      "step": 3772
+    },
+    {
+      "epoch": 0.8979591836734694,
+      "grad_norm": 0.4238883491721038,
+      "learning_rate": 6.062612407835117e-06,
+      "loss": 0.3662,
+      "step": 3773
+    },
+    {
+      "epoch": 0.8981971797465342,
+      "grad_norm": 0.3737577013120221,
+      "learning_rate": 6.060728768560654e-06,
+      "loss": 0.3084,
+      "step": 3774
+    },
+    {
+      "epoch": 0.898435175819599,
+      "grad_norm": 0.37762209485191345,
+      "learning_rate": 6.058844971635654e-06,
+      "loss": 0.3556,
+      "step": 3775
+    },
+    {
+      "epoch": 0.8986731718926638,
+      "grad_norm": 0.34826957290855015,
+      "learning_rate": 6.056961017340097e-06,
+      "loss": 0.3975,
+      "step": 3776
+    },
+    {
+      "epoch": 0.8989111679657286,
+      "grad_norm": 0.4161304452441926,
+      "learning_rate": 6.055076905953986e-06,
+      "loss": 0.2959,
+      "step": 3777
+    },
+    {
+      "epoch": 0.8991491640387934,
+      "grad_norm": 0.35944427744861646,
+      "learning_rate": 6.053192637757346e-06,
+      "loss": 0.2953,
+      "step": 3778
+    },
+    {
+      "epoch": 0.8993871601118582,
+      "grad_norm": 0.3534735542841618,
+      "learning_rate": 6.051308213030224e-06,
+      "loss": 0.3666,
+      "step": 3779
+    },
+    {
+      "epoch": 0.899625156184923,
+      "grad_norm": 0.4729648450816866,
+      "learning_rate": 6.049423632052693e-06,
+      "loss": 0.3592,
+      "step": 3780
+    },
+    {
+      "epoch": 0.8998631522579877,
+      "grad_norm": 0.40762763698787485,
+      "learning_rate": 6.0475388951048486e-06,
+      "loss": 0.298,
+      "step": 3781
+    },
+    {
+      "epoch": 0.9001011483310526,
+      "grad_norm": 0.3810584850880683,
+      "learning_rate": 6.045654002466809e-06,
+      "loss": 0.3393,
+      "step": 3782
+    },
+    {
+      "epoch": 0.9003391444041173,
+      "grad_norm": 0.4559084065779489,
+      "learning_rate": 6.043768954418719e-06,
+      "loss": 0.4002,
+      "step": 3783
+    },
+    {
+      "epoch": 0.9005771404771822,
+      "grad_norm": 0.43908602771497135,
+      "learning_rate": 6.041883751240739e-06,
+      "loss": 0.3226,
+      "step": 3784
+    },
+    {
+      "epoch": 0.9008151365502469,
+      "grad_norm": 0.3930802632673325,
+      "learning_rate": 6.03999839321306e-06,
+      "loss": 0.3064,
+      "step": 3785
+    },
+    {
+      "epoch": 0.9010531326233118,
+      "grad_norm": 0.33579273743850313,
+      "learning_rate": 6.03811288061589e-06,
+      "loss": 0.368,
+      "step": 3786
+    },
+    {
+      "epoch": 0.9012911286963765,
+      "grad_norm": 0.39169821582510483,
+      "learning_rate": 6.036227213729464e-06,
+      "loss": 0.3959,
+      "step": 3787
+    },
+    {
+      "epoch": 0.9015291247694414,
+      "grad_norm": 0.43853941776440175,
+      "learning_rate": 6.03434139283404e-06,
+      "loss": 0.3188,
+      "step": 3788
+    },
+    {
+      "epoch": 0.9017671208425061,
+      "grad_norm": 0.3689294208498079,
+      "learning_rate": 6.032455418209895e-06,
+      "loss": 0.351,
+      "step": 3789
+    },
+    {
+      "epoch": 0.902005116915571,
+      "grad_norm": 0.340196421020995,
+      "learning_rate": 6.030569290137335e-06,
+      "loss": 0.3988,
+      "step": 3790
+    },
+    {
+      "epoch": 0.9022431129886357,
+      "grad_norm": 0.39137456638518625,
+      "learning_rate": 6.028683008896683e-06,
+      "loss": 0.3547,
+      "step": 3791
+    },
+    {
+      "epoch": 0.9024811090617005,
+      "grad_norm": 0.43319110374923225,
+      "learning_rate": 6.026796574768288e-06,
+      "loss": 0.3019,
+      "step": 3792
+    },
+    {
+      "epoch": 0.9027191051347653,
+      "grad_norm": 0.3486367504488833,
+      "learning_rate": 6.024909988032519e-06,
+      "loss": 0.3584,
+      "step": 3793
+    },
+    {
+      "epoch": 0.9029571012078301,
+      "grad_norm": 0.36094039333364036,
+      "learning_rate": 6.0230232489697725e-06,
+      "loss": 0.412,
+      "step": 3794
+    },
+    {
+      "epoch": 0.9031950972808949,
+      "grad_norm": 0.3615716508779515,
+      "learning_rate": 6.021136357860461e-06,
+      "loss": 0.3103,
+      "step": 3795
+    },
+    {
+      "epoch": 0.9034330933539596,
+      "grad_norm": 0.3767985308127309,
+      "learning_rate": 6.0192493149850255e-06,
+      "loss": 0.3292,
+      "step": 3796
+    },
+    {
+      "epoch": 0.9036710894270245,
+      "grad_norm": 0.36959666250758044,
+      "learning_rate": 6.017362120623928e-06,
+      "loss": 0.3781,
+      "step": 3797
+    },
+    {
+      "epoch": 0.9039090855000892,
+      "grad_norm": 0.3571566018380154,
+      "learning_rate": 6.015474775057649e-06,
+      "loss": 0.3583,
+      "step": 3798
+    },
+    {
+      "epoch": 0.904147081573154,
+      "grad_norm": 0.38935644719203016,
+      "learning_rate": 6.013587278566698e-06,
+      "loss": 0.3145,
+      "step": 3799
+    },
+    {
+      "epoch": 0.9043850776462188,
+      "grad_norm": 0.3701781411050491,
+      "learning_rate": 6.011699631431603e-06,
+      "loss": 0.3197,
+      "step": 3800
+    },
+    {
+      "epoch": 0.9046230737192836,
+      "grad_norm": 0.3906408890697199,
+      "learning_rate": 6.0098118339329124e-06,
+      "loss": 0.4124,
+      "step": 3801
+    },
+    {
+      "epoch": 0.9048610697923484,
+      "grad_norm": 0.369451923410181,
+      "learning_rate": 6.007923886351202e-06,
+      "loss": 0.3269,
+      "step": 3802
+    },
+    {
+      "epoch": 0.9050990658654132,
+      "grad_norm": 0.37174979302655076,
+      "learning_rate": 6.00603578896707e-06,
+      "loss": 0.3283,
+      "step": 3803
+    },
+    {
+      "epoch": 0.905337061938478,
+      "grad_norm": 0.37058723879913413,
+      "learning_rate": 6.004147542061129e-06,
+      "loss": 0.3723,
+      "step": 3804
+    },
+    {
+      "epoch": 0.9055750580115428,
+      "grad_norm": 0.37650352346249066,
+      "learning_rate": 6.0022591459140235e-06,
+      "loss": 0.3471,
+      "step": 3805
+    },
+    {
+      "epoch": 0.9058130540846075,
+      "grad_norm": 0.37453899744942026,
+      "learning_rate": 6.000370600806415e-06,
+      "loss": 0.3212,
+      "step": 3806
+    },
+    {
+      "epoch": 0.9060510501576724,
+      "grad_norm": 0.3869447902100158,
+      "learning_rate": 5.9984819070189845e-06,
+      "loss": 0.3526,
+      "step": 3807
+    },
+    {
+      "epoch": 0.9062890462307371,
+      "grad_norm": 0.3527049994912703,
+      "learning_rate": 5.9965930648324425e-06,
+      "loss": 0.3878,
+      "step": 3808
+    },
+    {
+      "epoch": 0.906527042303802,
+      "grad_norm": 0.40048440151462644,
+      "learning_rate": 5.9947040745275174e-06,
+      "loss": 0.3268,
+      "step": 3809
+    },
+    {
+      "epoch": 0.9067650383768667,
+      "grad_norm": 0.36101747024521835,
+      "learning_rate": 5.992814936384958e-06,
+      "loss": 0.3299,
+      "step": 3810
+    },
+    {
+      "epoch": 0.9070030344499316,
+      "grad_norm": 0.3884060451475327,
+      "learning_rate": 5.990925650685539e-06,
+      "loss": 0.3604,
+      "step": 3811
+    },
+    {
+      "epoch": 0.9072410305229963,
+      "grad_norm": 0.38945955785689845,
+      "learning_rate": 5.9890362177100535e-06,
+      "loss": 0.3968,
+      "step": 3812
+    },
+    {
+      "epoch": 0.9074790265960612,
+      "grad_norm": 0.38256970785987937,
+      "learning_rate": 5.987146637739319e-06,
+      "loss": 0.285,
+      "step": 3813
+    },
+    {
+      "epoch": 0.9077170226691259,
+      "grad_norm": 0.3725122594244136,
+      "learning_rate": 5.985256911054171e-06,
+      "loss": 0.3099,
+      "step": 3814
+    },
+    {
+      "epoch": 0.9079550187421908,
+      "grad_norm": 0.3832707895396545,
+      "learning_rate": 5.983367037935473e-06,
+      "loss": 0.3996,
+      "step": 3815
+    },
+    {
+      "epoch": 0.9081930148152555,
+      "grad_norm": 0.3833096459028676,
+      "learning_rate": 5.9814770186641065e-06,
+      "loss": 0.3466,
+      "step": 3816
+    },
+    {
+      "epoch": 0.9084310108883203,
+      "grad_norm": 0.36397044311226656,
+      "learning_rate": 5.979586853520974e-06,
+      "loss": 0.2958,
+      "step": 3817
+    },
+    {
+      "epoch": 0.9086690069613851,
+      "grad_norm": 0.36891177007878145,
+      "learning_rate": 5.977696542787003e-06,
+      "loss": 0.3562,
+      "step": 3818
+    },
+    {
+      "epoch": 0.9089070030344499,
+      "grad_norm": 0.3840511476005476,
+      "learning_rate": 5.975806086743137e-06,
+      "loss": 0.3974,
+      "step": 3819
+    },
+    {
+      "epoch": 0.9091449991075147,
+      "grad_norm": 0.36836219306960205,
+      "learning_rate": 5.973915485670348e-06,
+      "loss": 0.2928,
+      "step": 3820
+    },
+    {
+      "epoch": 0.9093829951805795,
+      "grad_norm": 0.3684371158231424,
+      "learning_rate": 5.972024739849622e-06,
+      "loss": 0.3035,
+      "step": 3821
+    },
+    {
+      "epoch": 0.9096209912536443,
+      "grad_norm": 0.3623595279659431,
+      "learning_rate": 5.970133849561973e-06,
+      "loss": 0.3729,
+      "step": 3822
+    },
+    {
+      "epoch": 0.9098589873267091,
+      "grad_norm": 0.3813529115463126,
+      "learning_rate": 5.968242815088435e-06,
+      "loss": 0.3771,
+      "step": 3823
+    },
+    {
+      "epoch": 0.9100969833997739,
+      "grad_norm": 0.4428663575301828,
+      "learning_rate": 5.9663516367100614e-06,
+      "loss": 0.3069,
+      "step": 3824
+    },
+    {
+      "epoch": 0.9103349794728387,
+      "grad_norm": 0.38649551932626275,
+      "learning_rate": 5.964460314707928e-06,
+      "loss": 0.3451,
+      "step": 3825
+    },
+    {
+      "epoch": 0.9105729755459034,
+      "grad_norm": 0.35224183137486065,
+      "learning_rate": 5.962568849363133e-06,
+      "loss": 0.4268,
+      "step": 3826
+    },
+    {
+      "epoch": 0.9108109716189683,
+      "grad_norm": 0.3468956951171978,
+      "learning_rate": 5.960677240956792e-06,
+      "loss": 0.3078,
+      "step": 3827
+    },
+    {
+      "epoch": 0.911048967692033,
+      "grad_norm": 0.3542112048407511,
+      "learning_rate": 5.958785489770049e-06,
+      "loss": 0.2749,
+      "step": 3828
+    },
+    {
+      "epoch": 0.9112869637650979,
+      "grad_norm": 0.3674351382002242,
+      "learning_rate": 5.956893596084061e-06,
+      "loss": 0.3578,
+      "step": 3829
+    },
+    {
+      "epoch": 0.9115249598381626,
+      "grad_norm": 0.35983162039990363,
+      "learning_rate": 5.955001560180015e-06,
+      "loss": 0.392,
+      "step": 3830
+    },
+    {
+      "epoch": 0.9117629559112275,
+      "grad_norm": 0.36869673312884654,
+      "learning_rate": 5.953109382339111e-06,
+      "loss": 0.296,
+      "step": 3831
+    },
+    {
+      "epoch": 0.9120009519842922,
+      "grad_norm": 0.3937419732099826,
+      "learning_rate": 5.951217062842573e-06,
+      "loss": 0.3287,
+      "step": 3832
+    },
+    {
+      "epoch": 0.9122389480573571,
+      "grad_norm": 0.39019517526976494,
+      "learning_rate": 5.949324601971648e-06,
+      "loss": 0.4424,
+      "step": 3833
+    },
+    {
+      "epoch": 0.9124769441304218,
+      "grad_norm": 0.3944112522424989,
+      "learning_rate": 5.947432000007601e-06,
+      "loss": 0.3025,
+      "step": 3834
+    },
+    {
+      "epoch": 0.9127149402034866,
+      "grad_norm": 0.3906256947087767,
+      "learning_rate": 5.9455392572317225e-06,
+      "loss": 0.3067,
+      "step": 3835
+    },
+    {
+      "epoch": 0.9129529362765514,
+      "grad_norm": 0.36209911312794113,
+      "learning_rate": 5.9436463739253154e-06,
+      "loss": 0.3718,
+      "step": 3836
+    },
+    {
+      "epoch": 0.9131909323496162,
+      "grad_norm": 0.37733100352844173,
+      "learning_rate": 5.9417533503697155e-06,
+      "loss": 0.398,
+      "step": 3837
+    },
+    {
+      "epoch": 0.913428928422681,
+      "grad_norm": 0.3835158676308841,
+      "learning_rate": 5.939860186846269e-06,
+      "loss": 0.3215,
+      "step": 3838
+    },
+    {
+      "epoch": 0.9136669244957458,
+      "grad_norm": 0.367565730076443,
+      "learning_rate": 5.937966883636348e-06,
+      "loss": 0.3263,
+      "step": 3839
+    },
+    {
+      "epoch": 0.9139049205688106,
+      "grad_norm": 0.35666300012315477,
+      "learning_rate": 5.936073441021344e-06,
+      "loss": 0.3827,
+      "step": 3840
+    },
+    {
+      "epoch": 0.9141429166418754,
+      "grad_norm": 0.3638886300120447,
+      "learning_rate": 5.934179859282668e-06,
+      "loss": 0.3401,
+      "step": 3841
+    },
+    {
+      "epoch": 0.9143809127149402,
+      "grad_norm": 0.3412227559234039,
+      "learning_rate": 5.932286138701756e-06,
+      "loss": 0.3269,
+      "step": 3842
+    },
+    {
+      "epoch": 0.914618908788005,
+      "grad_norm": 0.36540816931263137,
+      "learning_rate": 5.930392279560059e-06,
+      "loss": 0.3647,
+      "step": 3843
+    },
+    {
+      "epoch": 0.9148569048610697,
+      "grad_norm": 0.346958869901011,
+      "learning_rate": 5.928498282139053e-06,
+      "loss": 0.3772,
+      "step": 3844
+    },
+    {
+      "epoch": 0.9150949009341346,
+      "grad_norm": 0.3794111378758652,
+      "learning_rate": 5.926604146720232e-06,
+      "loss": 0.3163,
+      "step": 3845
+    },
+    {
+      "epoch": 0.9153328970071993,
+      "grad_norm": 0.39132485441773784,
+      "learning_rate": 5.9247098735851125e-06,
+      "loss": 0.3245,
+      "step": 3846
+    },
+    {
+      "epoch": 0.9155708930802642,
+      "grad_norm": 0.36279235336046384,
+      "learning_rate": 5.922815463015229e-06,
+      "loss": 0.3925,
+      "step": 3847
+    },
+    {
+      "epoch": 0.9158088891533289,
+      "grad_norm": 0.380778919905592,
+      "learning_rate": 5.920920915292138e-06,
+      "loss": 0.3313,
+      "step": 3848
+    },
+    {
+      "epoch": 0.9160468852263938,
+      "grad_norm": 0.3651654131810896,
+      "learning_rate": 5.919026230697418e-06,
+      "loss": 0.2968,
+      "step": 3849
+    },
+    {
+      "epoch": 0.9162848812994585,
+      "grad_norm": 0.3670633414101187,
+      "learning_rate": 5.917131409512663e-06,
+      "loss": 0.3495,
+      "step": 3850
+    },
+    {
+      "epoch": 0.9165228773725234,
+      "grad_norm": 0.34896717104695807,
+      "learning_rate": 5.915236452019491e-06,
+      "loss": 0.3772,
+      "step": 3851
+    },
+    {
+      "epoch": 0.9167608734455881,
+      "grad_norm": 0.3780403642407919,
+      "learning_rate": 5.913341358499543e-06,
+      "loss": 0.3302,
+      "step": 3852
+    },
+    {
+      "epoch": 0.916998869518653,
+      "grad_norm": 0.36534779843005166,
+      "learning_rate": 5.911446129234473e-06,
+      "loss": 0.3154,
+      "step": 3853
+    },
+    {
+      "epoch": 0.9172368655917177,
+      "grad_norm": 0.36166515125650645,
+      "learning_rate": 5.909550764505959e-06,
+      "loss": 0.3428,
+      "step": 3854
+    },
+    {
+      "epoch": 0.9174748616647825,
+      "grad_norm": 0.36662895628646563,
+      "learning_rate": 5.907655264595701e-06,
+      "loss": 0.3726,
+      "step": 3855
+    },
+    {
+      "epoch": 0.9177128577378473,
+      "grad_norm": 0.41478425489418713,
+      "learning_rate": 5.905759629785417e-06,
+      "loss": 0.306,
+      "step": 3856
+    },
+    {
+      "epoch": 0.9179508538109121,
+      "grad_norm": 0.442141948179443,
+      "learning_rate": 5.903863860356843e-06,
+      "loss": 0.3664,
+      "step": 3857
+    },
+    {
+      "epoch": 0.9181888498839769,
+      "grad_norm": 0.3835994090422323,
+      "learning_rate": 5.9019679565917396e-06,
+      "loss": 0.4351,
+      "step": 3858
+    },
+    {
+      "epoch": 0.9184268459570417,
+      "grad_norm": 0.3916877069146898,
+      "learning_rate": 5.900071918771885e-06,
+      "loss": 0.3322,
+      "step": 3859
+    },
+    {
+      "epoch": 0.9186648420301065,
+      "grad_norm": 0.4452770801712549,
+      "learning_rate": 5.898175747179077e-06,
+      "loss": 0.3111,
+      "step": 3860
+    },
+    {
+      "epoch": 0.9189028381031713,
+      "grad_norm": 0.3810069457057913,
+      "learning_rate": 5.896279442095132e-06,
+      "loss": 0.3757,
+      "step": 3861
+    },
+    {
+      "epoch": 0.919140834176236,
+      "grad_norm": 0.42405450068532397,
+      "learning_rate": 5.894383003801889e-06,
+      "loss": 0.3957,
+      "step": 3862
+    },
+    {
+      "epoch": 0.9193788302493009,
+      "grad_norm": 0.3860106901737225,
+      "learning_rate": 5.892486432581206e-06,
+      "loss": 0.3128,
+      "step": 3863
+    },
+    {
+      "epoch": 0.9196168263223656,
+      "grad_norm": 0.35769077891896367,
+      "learning_rate": 5.890589728714959e-06,
+      "loss": 0.3283,
+      "step": 3864
+    },
+    {
+      "epoch": 0.9198548223954305,
+      "grad_norm": 0.3580135745616656,
+      "learning_rate": 5.8886928924850484e-06,
+      "loss": 0.4148,
+      "step": 3865
+    },
+    {
+      "epoch": 0.9200928184684952,
+      "grad_norm": 0.390966292937549,
+      "learning_rate": 5.886795924173388e-06,
+      "loss": 0.3515,
+      "step": 3866
+    },
+    {
+      "epoch": 0.9203308145415601,
+      "grad_norm": 0.3582229975022328,
+      "learning_rate": 5.884898824061914e-06,
+      "loss": 0.2942,
+      "step": 3867
+    },
+    {
+      "epoch": 0.9205688106146248,
+      "grad_norm": 0.383565757640599,
+      "learning_rate": 5.883001592432583e-06,
+      "loss": 0.3574,
+      "step": 3868
+    },
+    {
+      "epoch": 0.9208068066876897,
+      "grad_norm": 0.41449449325824644,
+      "learning_rate": 5.881104229567373e-06,
+      "loss": 0.3743,
+      "step": 3869
+    },
+    {
+      "epoch": 0.9210448027607544,
+      "grad_norm": 0.4022815267702927,
+      "learning_rate": 5.879206735748275e-06,
+      "loss": 0.3004,
+      "step": 3870
+    },
+    {
+      "epoch": 0.9212827988338192,
+      "grad_norm": 0.37800390773235576,
+      "learning_rate": 5.877309111257306e-06,
+      "loss": 0.3007,
+      "step": 3871
+    },
+    {
+      "epoch": 0.921520794906884,
+      "grad_norm": 0.3826886690510641,
+      "learning_rate": 5.8754113563765e-06,
+      "loss": 0.3896,
+      "step": 3872
+    },
+    {
+      "epoch": 0.9217587909799488,
+      "grad_norm": 0.39813220759183665,
+      "learning_rate": 5.873513471387911e-06,
+      "loss": 0.3316,
+      "step": 3873
+    },
+    {
+      "epoch": 0.9219967870530136,
+      "grad_norm": 0.3865193341498503,
+      "learning_rate": 5.871615456573608e-06,
+      "loss": 0.302,
+      "step": 3874
+    },
+    {
+      "epoch": 0.9222347831260784,
+      "grad_norm": 0.37725905957142175,
+      "learning_rate": 5.8697173122156885e-06,
+      "loss": 0.3364,
+      "step": 3875
+    },
+    {
+      "epoch": 0.9224727791991432,
+      "grad_norm": 0.3896841573344773,
+      "learning_rate": 5.8678190385962585e-06,
+      "loss": 0.4141,
+      "step": 3876
+    },
+    {
+      "epoch": 0.922710775272208,
+      "grad_norm": 0.3717555754024187,
+      "learning_rate": 5.8659206359974495e-06,
+      "loss": 0.3271,
+      "step": 3877
+    },
+    {
+      "epoch": 0.9229487713452728,
+      "grad_norm": 0.38884985169390957,
+      "learning_rate": 5.864022104701413e-06,
+      "loss": 0.3206,
+      "step": 3878
+    },
+    {
+      "epoch": 0.9231867674183376,
+      "grad_norm": 0.40651050344993656,
+      "learning_rate": 5.862123444990319e-06,
+      "loss": 0.3614,
+      "step": 3879
+    },
+    {
+      "epoch": 0.9234247634914023,
+      "grad_norm": 0.3699122959015574,
+      "learning_rate": 5.8602246571463506e-06,
+      "loss": 0.3953,
+      "step": 3880
+    },
+    {
+      "epoch": 0.9236627595644672,
+      "grad_norm": 0.35373619100502907,
+      "learning_rate": 5.858325741451718e-06,
+      "loss": 0.3106,
+      "step": 3881
+    },
+    {
+      "epoch": 0.9239007556375319,
+      "grad_norm": 0.3822556148266497,
+      "learning_rate": 5.8564266981886465e-06,
+      "loss": 0.3551,
+      "step": 3882
+    },
+    {
+      "epoch": 0.9241387517105968,
+      "grad_norm": 0.3367352347831046,
+      "learning_rate": 5.854527527639381e-06,
+      "loss": 0.4214,
+      "step": 3883
+    },
+    {
+      "epoch": 0.9243767477836615,
+      "grad_norm": 0.41091234068434845,
+      "learning_rate": 5.852628230086184e-06,
+      "loss": 0.3105,
+      "step": 3884
+    },
+    {
+      "epoch": 0.9246147438567264,
+      "grad_norm": 0.39373294006122483,
+      "learning_rate": 5.850728805811339e-06,
+      "loss": 0.3234,
+      "step": 3885
+    },
+    {
+      "epoch": 0.9248527399297911,
+      "grad_norm": 0.3427571664260684,
+      "learning_rate": 5.8488292550971485e-06,
+      "loss": 0.3525,
+      "step": 3886
+    },
+    {
+      "epoch": 0.925090736002856,
+      "grad_norm": 0.35439854536041354,
+      "learning_rate": 5.84692957822593e-06,
+      "loss": 0.3836,
+      "step": 3887
+    },
+    {
+      "epoch": 0.9253287320759207,
+      "grad_norm": 0.396795073329766,
+      "learning_rate": 5.845029775480026e-06,
+      "loss": 0.295,
+      "step": 3888
+    },
+    {
+      "epoch": 0.9255667281489856,
+      "grad_norm": 0.3777068010438854,
+      "learning_rate": 5.843129847141792e-06,
+      "loss": 0.3076,
+      "step": 3889
+    },
+    {
+      "epoch": 0.9258047242220503,
+      "grad_norm": 0.35763041328215195,
+      "learning_rate": 5.841229793493604e-06,
+      "loss": 0.4072,
+      "step": 3890
+    },
+    {
+      "epoch": 0.9260427202951151,
+      "grad_norm": 0.38017084454143346,
+      "learning_rate": 5.839329614817858e-06,
+      "loss": 0.3474,
+      "step": 3891
+    },
+    {
+      "epoch": 0.9262807163681799,
+      "grad_norm": 0.35675610940815455,
+      "learning_rate": 5.837429311396967e-06,
+      "loss": 0.2979,
+      "step": 3892
+    },
+    {
+      "epoch": 0.9265187124412447,
+      "grad_norm": 0.3793380743409526,
+      "learning_rate": 5.835528883513364e-06,
+      "loss": 0.3703,
+      "step": 3893
+    },
+    {
+      "epoch": 0.9267567085143095,
+      "grad_norm": 0.36281165596625997,
+      "learning_rate": 5.833628331449498e-06,
+      "loss": 0.4068,
+      "step": 3894
+    },
+    {
+      "epoch": 0.9269947045873743,
+      "grad_norm": 0.3691042976161106,
+      "learning_rate": 5.831727655487839e-06,
+      "loss": 0.3091,
+      "step": 3895
+    },
+    {
+      "epoch": 0.9272327006604391,
+      "grad_norm": 0.344360877845291,
+      "learning_rate": 5.829826855910875e-06,
+      "loss": 0.3287,
+      "step": 3896
+    },
+    {
+      "epoch": 0.9274706967335039,
+      "grad_norm": 0.3740607315809258,
+      "learning_rate": 5.827925933001111e-06,
+      "loss": 0.3885,
+      "step": 3897
+    },
+    {
+      "epoch": 0.9277086928065686,
+      "grad_norm": 0.3685370821262386,
+      "learning_rate": 5.826024887041071e-06,
+      "loss": 0.3755,
+      "step": 3898
+    },
+    {
+      "epoch": 0.9279466888796335,
+      "grad_norm": 0.3760368000609693,
+      "learning_rate": 5.8241237183132986e-06,
+      "loss": 0.289,
+      "step": 3899
+    },
+    {
+      "epoch": 0.9281846849526982,
+      "grad_norm": 0.3705899806280766,
+      "learning_rate": 5.822222427100354e-06,
+      "loss": 0.3575,
+      "step": 3900
+    },
+    {
+      "epoch": 0.9284226810257631,
+      "grad_norm": 0.39603394670641257,
+      "learning_rate": 5.820321013684815e-06,
+      "loss": 0.4147,
+      "step": 3901
+    },
+    {
+      "epoch": 0.9286606770988278,
+      "grad_norm": 0.36521058651075333,
+      "learning_rate": 5.818419478349281e-06,
+      "loss": 0.3325,
+      "step": 3902
+    },
+    {
+      "epoch": 0.9288986731718927,
+      "grad_norm": 0.3439794302882938,
+      "learning_rate": 5.816517821376365e-06,
+      "loss": 0.3189,
+      "step": 3903
+    },
+    {
+      "epoch": 0.9291366692449574,
+      "grad_norm": 0.37141750525341993,
+      "learning_rate": 5.814616043048702e-06,
+      "loss": 0.3497,
+      "step": 3904
+    },
+    {
+      "epoch": 0.9293746653180223,
+      "grad_norm": 0.38016142601364294,
+      "learning_rate": 5.8127141436489395e-06,
+      "loss": 0.3838,
+      "step": 3905
+    },
+    {
+      "epoch": 0.929612661391087,
+      "grad_norm": 0.3606163418991101,
+      "learning_rate": 5.810812123459753e-06,
+      "loss": 0.2855,
+      "step": 3906
+    },
+    {
+      "epoch": 0.9298506574641519,
+      "grad_norm": 0.36356185335551205,
+      "learning_rate": 5.808909982763825e-06,
+      "loss": 0.3472,
+      "step": 3907
+    },
+    {
+      "epoch": 0.9300886535372166,
+      "grad_norm": 0.36536907609964997,
+      "learning_rate": 5.807007721843862e-06,
+      "loss": 0.421,
+      "step": 3908
+    },
+    {
+      "epoch": 0.9303266496102814,
+      "grad_norm": 0.4097160409903546,
+      "learning_rate": 5.805105340982586e-06,
+      "loss": 0.3065,
+      "step": 3909
+    },
+    {
+      "epoch": 0.9305646456833462,
+      "grad_norm": 0.32639160343180673,
+      "learning_rate": 5.803202840462741e-06,
+      "loss": 0.2853,
+      "step": 3910
+    },
+    {
+      "epoch": 0.930802641756411,
+      "grad_norm": 0.34431820740085145,
+      "learning_rate": 5.801300220567083e-06,
+      "loss": 0.3654,
+      "step": 3911
+    },
+    {
+      "epoch": 0.9310406378294758,
+      "grad_norm": 0.3393942154458769,
+      "learning_rate": 5.799397481578388e-06,
+      "loss": 0.3814,
+      "step": 3912
+    },
+    {
+      "epoch": 0.9312786339025406,
+      "grad_norm": 0.3650148623295529,
+      "learning_rate": 5.797494623779453e-06,
+      "loss": 0.2883,
+      "step": 3913
+    },
+    {
+      "epoch": 0.9315166299756054,
+      "grad_norm": 0.40074023711292456,
+      "learning_rate": 5.795591647453086e-06,
+      "loss": 0.3447,
+      "step": 3914
+    },
+    {
+      "epoch": 0.9317546260486702,
+      "grad_norm": 0.38051255748001506,
+      "learning_rate": 5.79368855288212e-06,
+      "loss": 0.3804,
+      "step": 3915
+    },
+    {
+      "epoch": 0.931992622121735,
+      "grad_norm": 0.37149987585322053,
+      "learning_rate": 5.7917853403493985e-06,
+      "loss": 0.3397,
+      "step": 3916
+    },
+    {
+      "epoch": 0.9322306181947998,
+      "grad_norm": 0.37572329221656625,
+      "learning_rate": 5.7898820101377885e-06,
+      "loss": 0.2941,
+      "step": 3917
+    },
+    {
+      "epoch": 0.9324686142678645,
+      "grad_norm": 0.3971060021730622,
+      "learning_rate": 5.787978562530172e-06,
+      "loss": 0.3382,
+      "step": 3918
+    },
+    {
+      "epoch": 0.9327066103409294,
+      "grad_norm": 0.39918849730220163,
+      "learning_rate": 5.786074997809445e-06,
+      "loss": 0.4049,
+      "step": 3919
+    },
+    {
+      "epoch": 0.9329446064139941,
+      "grad_norm": 0.386940462855031,
+      "learning_rate": 5.784171316258528e-06,
+      "loss": 0.3133,
+      "step": 3920
+    },
+    {
+      "epoch": 0.933182602487059,
+      "grad_norm": 0.39373148829016735,
+      "learning_rate": 5.782267518160354e-06,
+      "loss": 0.3161,
+      "step": 3921
+    },
+    {
+      "epoch": 0.9334205985601237,
+      "grad_norm": 0.4027047525256417,
+      "learning_rate": 5.780363603797874e-06,
+      "loss": 0.3806,
+      "step": 3922
+    },
+    {
+      "epoch": 0.9336585946331886,
+      "grad_norm": 0.436189477052603,
+      "learning_rate": 5.778459573454058e-06,
+      "loss": 0.3396,
+      "step": 3923
+    },
+    {
+      "epoch": 0.9338965907062533,
+      "grad_norm": 0.37159344865326444,
+      "learning_rate": 5.776555427411891e-06,
+      "loss": 0.2909,
+      "step": 3924
+    },
+    {
+      "epoch": 0.9341345867793182,
+      "grad_norm": 0.35999558789407055,
+      "learning_rate": 5.774651165954377e-06,
+      "loss": 0.3357,
+      "step": 3925
+    },
+    {
+      "epoch": 0.9343725828523829,
+      "grad_norm": 0.3450624176837847,
+      "learning_rate": 5.772746789364534e-06,
+      "loss": 0.4188,
+      "step": 3926
+    },
+    {
+      "epoch": 0.9346105789254477,
+      "grad_norm": 0.375979183281526,
+      "learning_rate": 5.770842297925402e-06,
+      "loss": 0.3443,
+      "step": 3927
+    },
+    {
+      "epoch": 0.9348485749985125,
+      "grad_norm": 0.3646206735183569,
+      "learning_rate": 5.768937691920036e-06,
+      "loss": 0.2904,
+      "step": 3928
+    },
+    {
+      "epoch": 0.9350865710715773,
+      "grad_norm": 0.37338768923849774,
+      "learning_rate": 5.767032971631506e-06,
+      "loss": 0.3965,
+      "step": 3929
+    },
+    {
+      "epoch": 0.9353245671446421,
+      "grad_norm": 0.36862713467496905,
+      "learning_rate": 5.7651281373429e-06,
+      "loss": 0.3441,
+      "step": 3930
+    },
+    {
+      "epoch": 0.9355625632177069,
+      "grad_norm": 0.36754831493989937,
+      "learning_rate": 5.763223189337324e-06,
+      "loss": 0.325,
+      "step": 3931
+    },
+    {
+      "epoch": 0.9358005592907717,
+      "grad_norm": 0.37946091566468837,
+      "learning_rate": 5.761318127897903e-06,
+      "loss": 0.3237,
+      "step": 3932
+    },
+    {
+      "epoch": 0.9360385553638365,
+      "grad_norm": 0.381077983146763,
+      "learning_rate": 5.759412953307771e-06,
+      "loss": 0.3836,
+      "step": 3933
+    },
+    {
+      "epoch": 0.9362765514369012,
+      "grad_norm": 0.3797233859010859,
+      "learning_rate": 5.75750766585009e-06,
+      "loss": 0.3368,
+      "step": 3934
+    },
+    {
+      "epoch": 0.9365145475099661,
+      "grad_norm": 0.4055187658650143,
+      "learning_rate": 5.7556022658080276e-06,
+      "loss": 0.3064,
+      "step": 3935
+    },
+    {
+      "epoch": 0.9367525435830308,
+      "grad_norm": 0.3428864215245841,
+      "learning_rate": 5.753696753464778e-06,
+      "loss": 0.3785,
+      "step": 3936
+    },
+    {
+      "epoch": 0.9369905396560957,
+      "grad_norm": 0.38066009287080843,
+      "learning_rate": 5.751791129103545e-06,
+      "loss": 0.4031,
+      "step": 3937
+    },
+    {
+      "epoch": 0.9372285357291604,
+      "grad_norm": 0.38910361084548145,
+      "learning_rate": 5.749885393007552e-06,
+      "loss": 0.3196,
+      "step": 3938
+    },
+    {
+      "epoch": 0.9374665318022253,
+      "grad_norm": 0.4447865941468255,
+      "learning_rate": 5.747979545460036e-06,
+      "loss": 0.3112,
+      "step": 3939
+    },
+    {
+      "epoch": 0.93770452787529,
+      "grad_norm": 0.36280386489269756,
+      "learning_rate": 5.746073586744258e-06,
+      "loss": 0.3824,
+      "step": 3940
+    },
+    {
+      "epoch": 0.9379425239483549,
+      "grad_norm": 0.3715191063601824,
+      "learning_rate": 5.744167517143486e-06,
+      "loss": 0.351,
+      "step": 3941
+    },
+    {
+      "epoch": 0.9381805200214196,
+      "grad_norm": 0.3916498754924406,
+      "learning_rate": 5.742261336941013e-06,
+      "loss": 0.3221,
+      "step": 3942
+    },
+    {
+      "epoch": 0.9384185160944845,
+      "grad_norm": 0.38502974009058716,
+      "learning_rate": 5.740355046420142e-06,
+      "loss": 0.4041,
+      "step": 3943
+    },
+    {
+      "epoch": 0.9386565121675492,
+      "grad_norm": 0.5425497961680991,
+      "learning_rate": 5.738448645864195e-06,
+      "loss": 0.415,
+      "step": 3944
+    },
+    {
+      "epoch": 0.938894508240614,
+      "grad_norm": 0.3914605628286909,
+      "learning_rate": 5.736542135556512e-06,
+      "loss": 0.3102,
+      "step": 3945
+    },
+    {
+      "epoch": 0.9391325043136788,
+      "grad_norm": 0.3930678191979905,
+      "learning_rate": 5.7346355157804455e-06,
+      "loss": 0.3227,
+      "step": 3946
+    },
+    {
+      "epoch": 0.9393705003867436,
+      "grad_norm": 0.3660355806904165,
+      "learning_rate": 5.732728786819368e-06,
+      "loss": 0.3873,
+      "step": 3947
+    },
+    {
+      "epoch": 0.9396084964598084,
+      "grad_norm": 0.3579948522758407,
+      "learning_rate": 5.730821948956665e-06,
+      "loss": 0.366,
+      "step": 3948
+    },
+    {
+      "epoch": 0.9398464925328732,
+      "grad_norm": 0.35994213666298,
+      "learning_rate": 5.7289150024757415e-06,
+      "loss": 0.2896,
+      "step": 3949
+    },
+    {
+      "epoch": 0.940084488605938,
+      "grad_norm": 0.3660284695948094,
+      "learning_rate": 5.727007947660016e-06,
+      "loss": 0.3578,
+      "step": 3950
+    },
+    {
+      "epoch": 0.9403224846790028,
+      "grad_norm": 0.39417668974001635,
+      "learning_rate": 5.725100784792924e-06,
+      "loss": 0.3948,
+      "step": 3951
+    },
+    {
+      "epoch": 0.9405604807520676,
+      "grad_norm": 0.382207382206984,
+      "learning_rate": 5.723193514157918e-06,
+      "loss": 0.3124,
+      "step": 3952
+    },
+    {
+      "epoch": 0.9407984768251324,
+      "grad_norm": 0.4024628893156382,
+      "learning_rate": 5.721286136038463e-06,
+      "loss": 0.2865,
+      "step": 3953
+    },
+    {
+      "epoch": 0.9410364728981971,
+      "grad_norm": 0.386924814297485,
+      "learning_rate": 5.719378650718046e-06,
+      "loss": 0.364,
+      "step": 3954
+    },
+    {
+      "epoch": 0.941274468971262,
+      "grad_norm": 0.376900110000449,
+      "learning_rate": 5.717471058480165e-06,
+      "loss": 0.3686,
+      "step": 3955
+    },
+    {
+      "epoch": 0.9415124650443267,
+      "grad_norm": 0.3832996377113013,
+      "learning_rate": 5.7155633596083345e-06,
+      "loss": 0.2966,
+      "step": 3956
+    },
+    {
+      "epoch": 0.9417504611173916,
+      "grad_norm": 0.3689062510414365,
+      "learning_rate": 5.713655554386088e-06,
+      "loss": 0.3313,
+      "step": 3957
+    },
+    {
+      "epoch": 0.9419884571904563,
+      "grad_norm": 0.36517279149138915,
+      "learning_rate": 5.71174764309697e-06,
+      "loss": 0.3935,
+      "step": 3958
+    },
+    {
+      "epoch": 0.9422264532635212,
+      "grad_norm": 0.37317314065760315,
+      "learning_rate": 5.709839626024545e-06,
+      "loss": 0.2822,
+      "step": 3959
+    },
+    {
+      "epoch": 0.9424644493365859,
+      "grad_norm": 0.39141777366177366,
+      "learning_rate": 5.70793150345239e-06,
+      "loss": 0.3011,
+      "step": 3960
+    },
+    {
+      "epoch": 0.9427024454096508,
+      "grad_norm": 0.3705484724661079,
+      "learning_rate": 5.706023275664101e-06,
+      "loss": 0.3703,
+      "step": 3961
+    },
+    {
+      "epoch": 0.9429404414827155,
+      "grad_norm": 0.37822124229893755,
+      "learning_rate": 5.704114942943286e-06,
+      "loss": 0.3911,
+      "step": 3962
+    },
+    {
+      "epoch": 0.9431784375557803,
+      "grad_norm": 0.4070014262482564,
+      "learning_rate": 5.702206505573572e-06,
+      "loss": 0.3071,
+      "step": 3963
+    },
+    {
+      "epoch": 0.9434164336288451,
+      "grad_norm": 0.36329996164851147,
+      "learning_rate": 5.7002979638386005e-06,
+      "loss": 0.3356,
+      "step": 3964
+    },
+    {
+      "epoch": 0.9436544297019099,
+      "grad_norm": 0.36649962919002427,
+      "learning_rate": 5.698389318022026e-06,
+      "loss": 0.4008,
+      "step": 3965
+    },
+    {
+      "epoch": 0.9438924257749747,
+      "grad_norm": 0.3638109792735608,
+      "learning_rate": 5.696480568407523e-06,
+      "loss": 0.3156,
+      "step": 3966
+    },
+    {
+      "epoch": 0.9441304218480395,
+      "grad_norm": 0.3605646832595526,
+      "learning_rate": 5.694571715278775e-06,
+      "loss": 0.2936,
+      "step": 3967
+    },
+    {
+      "epoch": 0.9443684179211043,
+      "grad_norm": 0.3453303049010467,
+      "learning_rate": 5.692662758919489e-06,
+      "loss": 0.3573,
+      "step": 3968
+    },
+    {
+      "epoch": 0.9446064139941691,
+      "grad_norm": 0.3772126922641312,
+      "learning_rate": 5.690753699613382e-06,
+      "loss": 0.4241,
+      "step": 3969
+    },
+    {
+      "epoch": 0.9448444100672339,
+      "grad_norm": 0.445093770628227,
+      "learning_rate": 5.688844537644186e-06,
+      "loss": 0.3324,
+      "step": 3970
+    },
+    {
+      "epoch": 0.9450824061402987,
+      "grad_norm": 0.3699388236702028,
+      "learning_rate": 5.686935273295649e-06,
+      "loss": 0.3115,
+      "step": 3971
+    },
+    {
+      "epoch": 0.9453204022133634,
+      "grad_norm": 0.3897628984455295,
+      "learning_rate": 5.685025906851539e-06,
+      "loss": 0.3806,
+      "step": 3972
+    },
+    {
+      "epoch": 0.9455583982864283,
+      "grad_norm": 0.4016257052922472,
+      "learning_rate": 5.6831164385956314e-06,
+      "loss": 0.3558,
+      "step": 3973
+    },
+    {
+      "epoch": 0.945796394359493,
+      "grad_norm": 0.3631492629536495,
+      "learning_rate": 5.681206868811721e-06,
+      "loss": 0.299,
+      "step": 3974
+    },
+    {
+      "epoch": 0.9460343904325579,
+      "grad_norm": 0.3617842355634617,
+      "learning_rate": 5.679297197783617e-06,
+      "loss": 0.3351,
+      "step": 3975
+    },
+    {
+      "epoch": 0.9462723865056226,
+      "grad_norm": 0.37153756822155454,
+      "learning_rate": 5.677387425795146e-06,
+      "loss": 0.389,
+      "step": 3976
+    },
+    {
+      "epoch": 0.9465103825786875,
+      "grad_norm": 0.37351547703057986,
+      "learning_rate": 5.675477553130145e-06,
+      "loss": 0.3245,
+      "step": 3977
+    },
+    {
+      "epoch": 0.9467483786517522,
+      "grad_norm": 0.38350414771635083,
+      "learning_rate": 5.6735675800724695e-06,
+      "loss": 0.3055,
+      "step": 3978
+    },
+    {
+      "epoch": 0.9469863747248171,
+      "grad_norm": 0.42323362399946074,
+      "learning_rate": 5.671657506905989e-06,
+      "loss": 0.3664,
+      "step": 3979
+    },
+    {
+      "epoch": 0.9472243707978818,
+      "grad_norm": 0.3736129312742057,
+      "learning_rate": 5.669747333914586e-06,
+      "loss": 0.3772,
+      "step": 3980
+    },
+    {
+      "epoch": 0.9474623668709466,
+      "grad_norm": 0.3940418326123638,
+      "learning_rate": 5.66783706138216e-06,
+      "loss": 0.3301,
+      "step": 3981
+    },
+    {
+      "epoch": 0.9477003629440114,
+      "grad_norm": 0.3647257614559865,
+      "learning_rate": 5.665926689592626e-06,
+      "loss": 0.3509,
+      "step": 3982
+    },
+    {
+      "epoch": 0.9479383590170762,
+      "grad_norm": 0.35718234674396576,
+      "learning_rate": 5.664016218829911e-06,
+      "loss": 0.4159,
+      "step": 3983
+    },
+    {
+      "epoch": 0.948176355090141,
+      "grad_norm": 0.35559817666437965,
+      "learning_rate": 5.6621056493779605e-06,
+      "loss": 0.2914,
+      "step": 3984
+    },
+    {
+      "epoch": 0.9484143511632058,
+      "grad_norm": 0.4155897468492624,
+      "learning_rate": 5.660194981520729e-06,
+      "loss": 0.3059,
+      "step": 3985
+    },
+    {
+      "epoch": 0.9486523472362706,
+      "grad_norm": 0.39058764215986536,
+      "learning_rate": 5.658284215542191e-06,
+      "loss": 0.3828,
+      "step": 3986
+    },
+    {
+      "epoch": 0.9488903433093354,
+      "grad_norm": 0.36296631175689215,
+      "learning_rate": 5.656373351726334e-06,
+      "loss": 0.3917,
+      "step": 3987
+    },
+    {
+      "epoch": 0.9491283393824002,
+      "grad_norm": 0.3654376052680258,
+      "learning_rate": 5.654462390357159e-06,
+      "loss": 0.3176,
+      "step": 3988
+    },
+    {
+      "epoch": 0.949366335455465,
+      "grad_norm": 0.37351319837559777,
+      "learning_rate": 5.652551331718681e-06,
+      "loss": 0.3476,
+      "step": 3989
+    },
+    {
+      "epoch": 0.9496043315285297,
+      "grad_norm": 0.3654444319811079,
+      "learning_rate": 5.6506401760949335e-06,
+      "loss": 0.4022,
+      "step": 3990
+    },
+    {
+      "epoch": 0.9498423276015946,
+      "grad_norm": 0.37676497486164945,
+      "learning_rate": 5.6487289237699595e-06,
+      "loss": 0.3443,
+      "step": 3991
+    },
+    {
+      "epoch": 0.9500803236746593,
+      "grad_norm": 0.3816522699094411,
+      "learning_rate": 5.646817575027819e-06,
+      "loss": 0.2994,
+      "step": 3992
+    },
+    {
+      "epoch": 0.9503183197477242,
+      "grad_norm": 0.36418061297866533,
+      "learning_rate": 5.6449061301525845e-06,
+      "loss": 0.3747,
+      "step": 3993
+    },
+    {
+      "epoch": 0.9505563158207889,
+      "grad_norm": 0.3457838501298794,
+      "learning_rate": 5.642994589428344e-06,
+      "loss": 0.3956,
+      "step": 3994
+    },
+    {
+      "epoch": 0.9507943118938538,
+      "grad_norm": 0.3814926599582243,
+      "learning_rate": 5.641082953139201e-06,
+      "loss": 0.3113,
+      "step": 3995
+    },
+    {
+      "epoch": 0.9510323079669185,
+      "grad_norm": 0.38468757516147417,
+      "learning_rate": 5.639171221569273e-06,
+      "loss": 0.3233,
+      "step": 3996
+    },
+    {
+      "epoch": 0.9512703040399834,
+      "grad_norm": 0.3676820335135461,
+      "learning_rate": 5.637259395002688e-06,
+      "loss": 0.409,
+      "step": 3997
+    },
+    {
+      "epoch": 0.9515083001130481,
+      "grad_norm": 0.366933898548463,
+      "learning_rate": 5.635347473723592e-06,
+      "loss": 0.3456,
+      "step": 3998
+    },
+    {
+      "epoch": 0.951746296186113,
+      "grad_norm": 0.3470289154843928,
+      "learning_rate": 5.633435458016144e-06,
+      "loss": 0.3005,
+      "step": 3999
+    },
+    {
+      "epoch": 0.9519842922591777,
+      "grad_norm": 0.3685253402972838,
+      "learning_rate": 5.631523348164517e-06,
+      "loss": 0.3329,
+      "step": 4000
+    },
+    {
+      "epoch": 0.9522222883322425,
+      "grad_norm": 0.37326221272675375,
+      "learning_rate": 5.629611144452896e-06,
+      "loss": 0.3898,
+      "step": 4001
+    },
+    {
+      "epoch": 0.9524602844053073,
+      "grad_norm": 0.369692070324989,
+      "learning_rate": 5.627698847165484e-06,
+      "loss": 0.3069,
+      "step": 4002
+    },
+    {
+      "epoch": 0.9526982804783721,
+      "grad_norm": 0.4315390716578853,
+      "learning_rate": 5.625786456586493e-06,
+      "loss": 0.33,
+      "step": 4003
+    },
+    {
+      "epoch": 0.9529362765514369,
+      "grad_norm": 0.4310155102266023,
+      "learning_rate": 5.623873973000156e-06,
+      "loss": 0.4028,
+      "step": 4004
+    },
+    {
+      "epoch": 0.9531742726245017,
+      "grad_norm": 0.3751833494613908,
+      "learning_rate": 5.621961396690712e-06,
+      "loss": 0.3683,
+      "step": 4005
+    },
+    {
+      "epoch": 0.9534122686975665,
+      "grad_norm": 0.42057861372299066,
+      "learning_rate": 5.620048727942416e-06,
+      "loss": 0.2942,
+      "step": 4006
+    },
+    {
+      "epoch": 0.9536502647706313,
+      "grad_norm": 0.3625924700652696,
+      "learning_rate": 5.618135967039542e-06,
+      "loss": 0.3401,
+      "step": 4007
+    },
+    {
+      "epoch": 0.953888260843696,
+      "grad_norm": 0.3542760242464296,
+      "learning_rate": 5.616223114266369e-06,
+      "loss": 0.417,
+      "step": 4008
+    },
+    {
+      "epoch": 0.9541262569167609,
+      "grad_norm": 0.40855459494512736,
+      "learning_rate": 5.614310169907199e-06,
+      "loss": 0.3232,
+      "step": 4009
+    },
+    {
+      "epoch": 0.9543642529898256,
+      "grad_norm": 0.37105385201006647,
+      "learning_rate": 5.61239713424634e-06,
+      "loss": 0.2834,
+      "step": 4010
+    },
+    {
+      "epoch": 0.9546022490628905,
+      "grad_norm": 0.37855213750129024,
+      "learning_rate": 5.610484007568117e-06,
+      "loss": 0.3644,
+      "step": 4011
+    },
+    {
+      "epoch": 0.9548402451359552,
+      "grad_norm": 0.36744333500958176,
+      "learning_rate": 5.608570790156867e-06,
+      "loss": 0.3968,
+      "step": 4012
+    },
+    {
+      "epoch": 0.9550782412090201,
+      "grad_norm": 0.40177284275871733,
+      "learning_rate": 5.606657482296943e-06,
+      "loss": 0.3354,
+      "step": 4013
+    },
+    {
+      "epoch": 0.9553162372820848,
+      "grad_norm": 0.40131947230064974,
+      "learning_rate": 5.6047440842727075e-06,
+      "loss": 0.3262,
+      "step": 4014
+    },
+    {
+      "epoch": 0.9555542333551497,
+      "grad_norm": 0.35010111030696944,
+      "learning_rate": 5.602830596368543e-06,
+      "loss": 0.3737,
+      "step": 4015
+    },
+    {
+      "epoch": 0.9557922294282144,
+      "grad_norm": 0.3939903727222176,
+      "learning_rate": 5.600917018868835e-06,
+      "loss": 0.3344,
+      "step": 4016
+    },
+    {
+      "epoch": 0.9560302255012793,
+      "grad_norm": 0.4081024631555434,
+      "learning_rate": 5.599003352057994e-06,
+      "loss": 0.3028,
+      "step": 4017
+    },
+    {
+      "epoch": 0.956268221574344,
+      "grad_norm": 0.38240903753305877,
+      "learning_rate": 5.597089596220437e-06,
+      "loss": 0.3847,
+      "step": 4018
+    },
+    {
+      "epoch": 0.9565062176474088,
+      "grad_norm": 0.39578792437353716,
+      "learning_rate": 5.595175751640595e-06,
+      "loss": 0.3914,
+      "step": 4019
+    },
+    {
+      "epoch": 0.9567442137204736,
+      "grad_norm": 0.43574336288913884,
+      "learning_rate": 5.593261818602912e-06,
+      "loss": 0.33,
+      "step": 4020
+    },
+    {
+      "epoch": 0.9569822097935384,
+      "grad_norm": 0.3944916284608389,
+      "learning_rate": 5.5913477973918465e-06,
+      "loss": 0.3162,
+      "step": 4021
+    },
+    {
+      "epoch": 0.9572202058666032,
+      "grad_norm": 0.3667260519177921,
+      "learning_rate": 5.589433688291867e-06,
+      "loss": 0.3784,
+      "step": 4022
+    },
+    {
+      "epoch": 0.957458201939668,
+      "grad_norm": 0.38390699290292674,
+      "learning_rate": 5.5875194915874605e-06,
+      "loss": 0.3512,
+      "step": 4023
+    },
+    {
+      "epoch": 0.9576961980127328,
+      "grad_norm": 0.37193449103190873,
+      "learning_rate": 5.585605207563124e-06,
+      "loss": 0.2912,
+      "step": 4024
+    },
+    {
+      "epoch": 0.9579341940857976,
+      "grad_norm": 0.3699623027804539,
+      "learning_rate": 5.583690836503366e-06,
+      "loss": 0.3425,
+      "step": 4025
+    },
+    {
+      "epoch": 0.9581721901588623,
+      "grad_norm": 0.37064282931321874,
+      "learning_rate": 5.58177637869271e-06,
+      "loss": 0.3992,
+      "step": 4026
+    },
+    {
+      "epoch": 0.9584101862319272,
+      "grad_norm": 0.37635147922447526,
+      "learning_rate": 5.5798618344156916e-06,
+      "loss": 0.3117,
+      "step": 4027
+    },
+    {
+      "epoch": 0.9586481823049919,
+      "grad_norm": 0.3987637365857944,
+      "learning_rate": 5.577947203956858e-06,
+      "loss": 0.3143,
+      "step": 4028
+    },
+    {
+      "epoch": 0.9588861783780568,
+      "grad_norm": 0.4035395300731761,
+      "learning_rate": 5.576032487600773e-06,
+      "loss": 0.378,
+      "step": 4029
+    },
+    {
+      "epoch": 0.9591241744511215,
+      "grad_norm": 0.4101548946799405,
+      "learning_rate": 5.5741176856320105e-06,
+      "loss": 0.3796,
+      "step": 4030
+    },
+    {
+      "epoch": 0.9593621705241864,
+      "grad_norm": 0.3845855259199741,
+      "learning_rate": 5.5722027983351565e-06,
+      "loss": 0.3068,
+      "step": 4031
+    },
+    {
+      "epoch": 0.9596001665972511,
+      "grad_norm": 0.3609449815477264,
+      "learning_rate": 5.570287825994812e-06,
+      "loss": 0.3256,
+      "step": 4032
+    },
+    {
+      "epoch": 0.959838162670316,
+      "grad_norm": 0.3452672734583461,
+      "learning_rate": 5.568372768895588e-06,
+      "loss": 0.4226,
+      "step": 4033
+    },
+    {
+      "epoch": 0.9600761587433807,
+      "grad_norm": 0.3712187021978535,
+      "learning_rate": 5.566457627322109e-06,
+      "loss": 0.3293,
+      "step": 4034
+    },
+    {
+      "epoch": 0.9603141548164456,
+      "grad_norm": 0.35303931097547897,
+      "learning_rate": 5.5645424015590144e-06,
+      "loss": 0.28,
+      "step": 4035
+    },
+    {
+      "epoch": 0.9605521508895103,
+      "grad_norm": 0.3633441697287166,
+      "learning_rate": 5.562627091890951e-06,
+      "loss": 0.3538,
+      "step": 4036
+    },
+    {
+      "epoch": 0.9607901469625751,
+      "grad_norm": 0.41159227713226276,
+      "learning_rate": 5.560711698602584e-06,
+      "loss": 0.383,
+      "step": 4037
+    },
+    {
+      "epoch": 0.9610281430356399,
+      "grad_norm": 0.35928803089226324,
+      "learning_rate": 5.558796221978587e-06,
+      "loss": 0.2985,
+      "step": 4038
+    },
+    {
+      "epoch": 0.9612661391087047,
+      "grad_norm": 0.44144609493808085,
+      "learning_rate": 5.556880662303648e-06,
+      "loss": 0.3703,
+      "step": 4039
+    },
+    {
+      "epoch": 0.9615041351817695,
+      "grad_norm": 0.3803784240581862,
+      "learning_rate": 5.554965019862466e-06,
+      "loss": 0.3911,
+      "step": 4040
+    },
+    {
+      "epoch": 0.9617421312548343,
+      "grad_norm": 0.36481000247944867,
+      "learning_rate": 5.553049294939752e-06,
+      "loss": 0.3311,
+      "step": 4041
+    },
+    {
+      "epoch": 0.9619801273278991,
+      "grad_norm": 0.3593387405042659,
+      "learning_rate": 5.551133487820231e-06,
+      "loss": 0.2952,
+      "step": 4042
+    },
+    {
+      "epoch": 0.9622181234009639,
+      "grad_norm": 0.3625692759696533,
+      "learning_rate": 5.54921759878864e-06,
+      "loss": 0.3734,
+      "step": 4043
+    },
+    {
+      "epoch": 0.9624561194740286,
+      "grad_norm": 0.4202028000178488,
+      "learning_rate": 5.547301628129726e-06,
+      "loss": 0.3948,
+      "step": 4044
+    },
+    {
+      "epoch": 0.9626941155470935,
+      "grad_norm": 0.3783279956511722,
+      "learning_rate": 5.545385576128252e-06,
+      "loss": 0.3263,
+      "step": 4045
+    },
+    {
+      "epoch": 0.9629321116201582,
+      "grad_norm": 0.3926107017682296,
+      "learning_rate": 5.54346944306899e-06,
+      "loss": 0.3155,
+      "step": 4046
+    },
+    {
+      "epoch": 0.9631701076932231,
+      "grad_norm": 0.3926267406486093,
+      "learning_rate": 5.541553229236721e-06,
+      "loss": 0.367,
+      "step": 4047
+    },
+    {
+      "epoch": 0.9634081037662878,
+      "grad_norm": 0.38364235306865657,
+      "learning_rate": 5.539636934916247e-06,
+      "loss": 0.3193,
+      "step": 4048
+    },
+    {
+      "epoch": 0.9636460998393527,
+      "grad_norm": 0.38507849318568027,
+      "learning_rate": 5.537720560392373e-06,
+      "loss": 0.3056,
+      "step": 4049
+    },
+    {
+      "epoch": 0.9638840959124174,
+      "grad_norm": 0.37301941582141107,
+      "learning_rate": 5.535804105949922e-06,
+      "loss": 0.3405,
+      "step": 4050
+    },
+    {
+      "epoch": 0.9641220919854823,
+      "grad_norm": 0.35296914488364023,
+      "learning_rate": 5.533887571873725e-06,
+      "loss": 0.4018,
+      "step": 4051
+    },
+    {
+      "epoch": 0.964360088058547,
+      "grad_norm": 0.3588491551483761,
+      "learning_rate": 5.531970958448628e-06,
+      "loss": 0.3004,
+      "step": 4052
+    },
+    {
+      "epoch": 0.9645980841316119,
+      "grad_norm": 0.39575643827764756,
+      "learning_rate": 5.530054265959486e-06,
+      "loss": 0.2866,
+      "step": 4053
+    },
+    {
+      "epoch": 0.9648360802046766,
+      "grad_norm": 0.3466594386599281,
+      "learning_rate": 5.528137494691167e-06,
+      "loss": 0.3955,
+      "step": 4054
+    },
+    {
+      "epoch": 0.9650740762777414,
+      "grad_norm": 0.373285637369822,
+      "learning_rate": 5.52622064492855e-06,
+      "loss": 0.3484,
+      "step": 4055
+    },
+    {
+      "epoch": 0.9653120723508062,
+      "grad_norm": 0.42113681485363846,
+      "learning_rate": 5.524303716956528e-06,
+      "loss": 0.3046,
+      "step": 4056
+    },
+    {
+      "epoch": 0.965550068423871,
+      "grad_norm": 0.39214970328364873,
+      "learning_rate": 5.522386711060002e-06,
+      "loss": 0.3518,
+      "step": 4057
+    },
+    {
+      "epoch": 0.9657880644969358,
+      "grad_norm": 0.3636119541231596,
+      "learning_rate": 5.520469627523889e-06,
+      "loss": 0.3726,
+      "step": 4058
+    },
+    {
+      "epoch": 0.9660260605700006,
+      "grad_norm": 0.38531522777192306,
+      "learning_rate": 5.518552466633112e-06,
+      "loss": 0.3341,
+      "step": 4059
+    },
+    {
+      "epoch": 0.9662640566430654,
+      "grad_norm": 0.3689829393306755,
+      "learning_rate": 5.516635228672612e-06,
+      "loss": 0.3131,
+      "step": 4060
+    },
+    {
+      "epoch": 0.9665020527161302,
+      "grad_norm": 0.37131422956593985,
+      "learning_rate": 5.514717913927336e-06,
+      "loss": 0.3755,
+      "step": 4061
+    },
+    {
+      "epoch": 0.966740048789195,
+      "grad_norm": 0.36698419750198635,
+      "learning_rate": 5.5128005226822435e-06,
+      "loss": 0.4112,
+      "step": 4062
+    },
+    {
+      "epoch": 0.9669780448622598,
+      "grad_norm": 0.37767384948653765,
+      "learning_rate": 5.510883055222307e-06,
+      "loss": 0.3098,
+      "step": 4063
+    },
+    {
+      "epoch": 0.9672160409353245,
+      "grad_norm": 0.3831387096070846,
+      "learning_rate": 5.508965511832509e-06,
+      "loss": 0.3377,
+      "step": 4064
+    },
+    {
+      "epoch": 0.9674540370083894,
+      "grad_norm": 0.3413991536604678,
+      "learning_rate": 5.507047892797846e-06,
+      "loss": 0.3926,
+      "step": 4065
+    },
+    {
+      "epoch": 0.9676920330814541,
+      "grad_norm": 0.4405648042369594,
+      "learning_rate": 5.505130198403324e-06,
+      "loss": 0.3568,
+      "step": 4066
+    },
+    {
+      "epoch": 0.967930029154519,
+      "grad_norm": 0.36599437275948865,
+      "learning_rate": 5.503212428933956e-06,
+      "loss": 0.3054,
+      "step": 4067
+    },
+    {
+      "epoch": 0.9681680252275837,
+      "grad_norm": 0.4337574365319751,
+      "learning_rate": 5.501294584674771e-06,
+      "loss": 0.3363,
+      "step": 4068
+    },
+    {
+      "epoch": 0.9684060213006486,
+      "grad_norm": 0.3602978619539388,
+      "learning_rate": 5.499376665910812e-06,
+      "loss": 0.4053,
+      "step": 4069
+    },
+    {
+      "epoch": 0.9686440173737133,
+      "grad_norm": 0.3642524483127621,
+      "learning_rate": 5.497458672927124e-06,
+      "loss": 0.3262,
+      "step": 4070
+    },
+    {
+      "epoch": 0.9688820134467782,
+      "grad_norm": 0.3990749674732015,
+      "learning_rate": 5.495540606008772e-06,
+      "loss": 0.3322,
+      "step": 4071
+    },
+    {
+      "epoch": 0.9691200095198429,
+      "grad_norm": 0.3452572444028366,
+      "learning_rate": 5.493622465440828e-06,
+      "loss": 0.401,
+      "step": 4072
+    },
+    {
+      "epoch": 0.9693580055929077,
+      "grad_norm": 0.3947407561093372,
+      "learning_rate": 5.491704251508373e-06,
+      "loss": 0.3727,
+      "step": 4073
+    },
+    {
+      "epoch": 0.9695960016659725,
+      "grad_norm": 0.35872991795969333,
+      "learning_rate": 5.489785964496503e-06,
+      "loss": 0.2752,
+      "step": 4074
+    },
+    {
+      "epoch": 0.9698339977390373,
+      "grad_norm": 0.3810135107843619,
+      "learning_rate": 5.48786760469032e-06,
+      "loss": 0.3631,
+      "step": 4075
+    },
+    {
+      "epoch": 0.9700719938121021,
+      "grad_norm": 0.36736070326138315,
+      "learning_rate": 5.485949172374944e-06,
+      "loss": 0.4297,
+      "step": 4076
+    },
+    {
+      "epoch": 0.9703099898851669,
+      "grad_norm": 0.38917535599652914,
+      "learning_rate": 5.484030667835496e-06,
+      "loss": 0.3338,
+      "step": 4077
+    },
+    {
+      "epoch": 0.9705479859582317,
+      "grad_norm": 0.4162578135435146,
+      "learning_rate": 5.482112091357119e-06,
+      "loss": 0.3201,
+      "step": 4078
+    },
+    {
+      "epoch": 0.9707859820312965,
+      "grad_norm": 0.4446346807732351,
+      "learning_rate": 5.480193443224957e-06,
+      "loss": 0.3781,
+      "step": 4079
+    },
+    {
+      "epoch": 0.9710239781043613,
+      "grad_norm": 0.363773543353832,
+      "learning_rate": 5.478274723724172e-06,
+      "loss": 0.363,
+      "step": 4080
+    },
+    {
+      "epoch": 0.9712619741774261,
+      "grad_norm": 0.44939319983262227,
+      "learning_rate": 5.47635593313993e-06,
+      "loss": 0.3207,
+      "step": 4081
+    },
+    {
+      "epoch": 0.9714999702504908,
+      "grad_norm": 0.38983047666671894,
+      "learning_rate": 5.47443707175741e-06,
+      "loss": 0.3369,
+      "step": 4082
+    },
+    {
+      "epoch": 0.9717379663235557,
+      "grad_norm": 0.3609085790607123,
+      "learning_rate": 5.472518139861806e-06,
+      "loss": 0.4184,
+      "step": 4083
+    },
+    {
+      "epoch": 0.9719759623966204,
+      "grad_norm": 0.4326297185681936,
+      "learning_rate": 5.470599137738315e-06,
+      "loss": 0.3157,
+      "step": 4084
+    },
+    {
+      "epoch": 0.9722139584696853,
+      "grad_norm": 0.3863516117664841,
+      "learning_rate": 5.468680065672152e-06,
+      "loss": 0.3215,
+      "step": 4085
+    },
+    {
+      "epoch": 0.97245195454275,
+      "grad_norm": 0.37735302540928745,
+      "learning_rate": 5.466760923948536e-06,
+      "loss": 0.4003,
+      "step": 4086
+    },
+    {
+      "epoch": 0.9726899506158149,
+      "grad_norm": 0.3777977585862572,
+      "learning_rate": 5.464841712852701e-06,
+      "loss": 0.4035,
+      "step": 4087
+    },
+    {
+      "epoch": 0.9729279466888796,
+      "grad_norm": 0.3884798640154453,
+      "learning_rate": 5.462922432669886e-06,
+      "loss": 0.2875,
+      "step": 4088
+    },
+    {
+      "epoch": 0.9731659427619445,
+      "grad_norm": 0.3968066470866644,
+      "learning_rate": 5.461003083685346e-06,
+      "loss": 0.3509,
+      "step": 4089
+    },
+    {
+      "epoch": 0.9734039388350092,
+      "grad_norm": 0.37546658100029856,
+      "learning_rate": 5.459083666184344e-06,
+      "loss": 0.4117,
+      "step": 4090
+    },
+    {
+      "epoch": 0.973641934908074,
+      "grad_norm": 0.4033600486378232,
+      "learning_rate": 5.4571641804521505e-06,
+      "loss": 0.3023,
+      "step": 4091
+    },
+    {
+      "epoch": 0.9738799309811388,
+      "grad_norm": 0.366298156488569,
+      "learning_rate": 5.4552446267740515e-06,
+      "loss": 0.2914,
+      "step": 4092
+    },
+    {
+      "epoch": 0.9741179270542036,
+      "grad_norm": 0.3820600291897777,
+      "learning_rate": 5.453325005435337e-06,
+      "loss": 0.3516,
+      "step": 4093
+    },
+    {
+      "epoch": 0.9743559231272684,
+      "grad_norm": 0.3654555161312593,
+      "learning_rate": 5.451405316721313e-06,
+      "loss": 0.3877,
+      "step": 4094
+    },
+    {
+      "epoch": 0.9745939192003332,
+      "grad_norm": 0.40260654542095964,
+      "learning_rate": 5.449485560917291e-06,
+      "loss": 0.2995,
+      "step": 4095
+    },
+    {
+      "epoch": 0.974831915273398,
+      "grad_norm": 0.3968641036674078,
+      "learning_rate": 5.4475657383085955e-06,
+      "loss": 0.306,
+      "step": 4096
+    },
+    {
+      "epoch": 0.9750699113464628,
+      "grad_norm": 0.3903922997930381,
+      "learning_rate": 5.44564584918056e-06,
+      "loss": 0.3827,
+      "step": 4097
+    },
+    {
+      "epoch": 0.9753079074195276,
+      "grad_norm": 0.439867110372997,
+      "learning_rate": 5.443725893818524e-06,
+      "loss": 0.3308,
+      "step": 4098
+    },
+    {
+      "epoch": 0.9755459034925924,
+      "grad_norm": 0.4024543253971147,
+      "learning_rate": 5.441805872507846e-06,
+      "loss": 0.2899,
+      "step": 4099
+    },
+    {
+      "epoch": 0.9757838995656571,
+      "grad_norm": 0.3732267890164506,
+      "learning_rate": 5.439885785533884e-06,
+      "loss": 0.3557,
+      "step": 4100
+    },
+    {
+      "epoch": 0.976021895638722,
+      "grad_norm": 0.38784018489625477,
+      "learning_rate": 5.437965633182012e-06,
+      "loss": 0.3825,
+      "step": 4101
+    },
+    {
+      "epoch": 0.9762598917117867,
+      "grad_norm": 0.35874607034227307,
+      "learning_rate": 5.436045415737613e-06,
+      "loss": 0.2903,
+      "step": 4102
+    },
+    {
+      "epoch": 0.9764978877848516,
+      "grad_norm": 0.35448408385767743,
+      "learning_rate": 5.434125133486078e-06,
+      "loss": 0.3117,
+      "step": 4103
+    },
+    {
+      "epoch": 0.9767358838579163,
+      "grad_norm": 0.36167261584165866,
+      "learning_rate": 5.432204786712807e-06,
+      "loss": 0.3822,
+      "step": 4104
+    },
+    {
+      "epoch": 0.9769738799309812,
+      "grad_norm": 0.36236014204816985,
+      "learning_rate": 5.430284375703213e-06,
+      "loss": 0.3937,
+      "step": 4105
+    },
+    {
+      "epoch": 0.9772118760040459,
+      "grad_norm": 0.3615011221235271,
+      "learning_rate": 5.428363900742717e-06,
+      "loss": 0.2882,
+      "step": 4106
+    },
+    {
+      "epoch": 0.9774498720771108,
+      "grad_norm": 0.36358536744553127,
+      "learning_rate": 5.426443362116746e-06,
+      "loss": 0.3445,
+      "step": 4107
+    },
+    {
+      "epoch": 0.9776878681501755,
+      "grad_norm": 0.3499471450758627,
+      "learning_rate": 5.424522760110744e-06,
+      "loss": 0.3819,
+      "step": 4108
+    },
+    {
+      "epoch": 0.9779258642232403,
+      "grad_norm": 0.3783758065626236,
+      "learning_rate": 5.422602095010157e-06,
+      "loss": 0.3003,
+      "step": 4109
+    },
+    {
+      "epoch": 0.9781638602963051,
+      "grad_norm": 0.3485194689281019,
+      "learning_rate": 5.420681367100443e-06,
+      "loss": 0.3274,
+      "step": 4110
+    },
+    {
+      "epoch": 0.9784018563693699,
+      "grad_norm": 0.3560157044900479,
+      "learning_rate": 5.418760576667071e-06,
+      "loss": 0.3514,
+      "step": 4111
+    },
+    {
+      "epoch": 0.9786398524424347,
+      "grad_norm": 0.36557640443599765,
+      "learning_rate": 5.416839723995518e-06,
+      "loss": 0.3701,
+      "step": 4112
+    },
+    {
+      "epoch": 0.9788778485154995,
+      "grad_norm": 0.4127125165049302,
+      "learning_rate": 5.41491880937127e-06,
+      "loss": 0.3,
+      "step": 4113
+    },
+    {
+      "epoch": 0.9791158445885643,
+      "grad_norm": 0.3946777818047238,
+      "learning_rate": 5.4129978330798224e-06,
+      "loss": 0.3319,
+      "step": 4114
+    },
+    {
+      "epoch": 0.9793538406616291,
+      "grad_norm": 0.35875313580725443,
+      "learning_rate": 5.41107679540668e-06,
+      "loss": 0.3716,
+      "step": 4115
+    },
+    {
+      "epoch": 0.9795918367346939,
+      "grad_norm": 0.3583640218186905,
+      "learning_rate": 5.409155696637357e-06,
+      "loss": 0.3203,
+      "step": 4116
+    },
+    {
+      "epoch": 0.9798298328077587,
+      "grad_norm": 0.43393683642612224,
+      "learning_rate": 5.4072345370573745e-06,
+      "loss": 0.3299,
+      "step": 4117
+    },
+    {
+      "epoch": 0.9800678288808234,
+      "grad_norm": 0.3689652477197836,
+      "learning_rate": 5.405313316952265e-06,
+      "loss": 0.3645,
+      "step": 4118
+    },
+    {
+      "epoch": 0.9803058249538883,
+      "grad_norm": 0.38248656505142614,
+      "learning_rate": 5.403392036607568e-06,
+      "loss": 0.4105,
+      "step": 4119
+    },
+    {
+      "epoch": 0.980543821026953,
+      "grad_norm": 0.34515772392085886,
+      "learning_rate": 5.401470696308838e-06,
+      "loss": 0.3192,
+      "step": 4120
+    },
+    {
+      "epoch": 0.9807818171000179,
+      "grad_norm": 0.349261482848996,
+      "learning_rate": 5.399549296341629e-06,
+      "loss": 0.3257,
+      "step": 4121
+    },
+    {
+      "epoch": 0.9810198131730826,
+      "grad_norm": 0.3802296455540952,
+      "learning_rate": 5.39762783699151e-06,
+      "loss": 0.3968,
+      "step": 4122
+    },
+    {
+      "epoch": 0.9812578092461475,
+      "grad_norm": 0.3615234129840896,
+      "learning_rate": 5.395706318544059e-06,
+      "loss": 0.3298,
+      "step": 4123
+    },
+    {
+      "epoch": 0.9814958053192122,
+      "grad_norm": 0.4070437397173655,
+      "learning_rate": 5.393784741284858e-06,
+      "loss": 0.3196,
+      "step": 4124
+    },
+    {
+      "epoch": 0.9817338013922771,
+      "grad_norm": 0.3920234501678305,
+      "learning_rate": 5.391863105499505e-06,
+      "loss": 0.3159,
+      "step": 4125
+    },
+    {
+      "epoch": 0.9819717974653418,
+      "grad_norm": 0.3990009179740692,
+      "learning_rate": 5.389941411473598e-06,
+      "loss": 0.4174,
+      "step": 4126
+    },
+    {
+      "epoch": 0.9822097935384067,
+      "grad_norm": 0.41580376912089045,
+      "learning_rate": 5.3880196594927514e-06,
+      "loss": 0.2869,
+      "step": 4127
+    },
+    {
+      "epoch": 0.9824477896114714,
+      "grad_norm": 0.41224321601260766,
+      "learning_rate": 5.3860978498425845e-06,
+      "loss": 0.2956,
+      "step": 4128
+    },
+    {
+      "epoch": 0.9826857856845362,
+      "grad_norm": 0.37553862338123545,
+      "learning_rate": 5.384175982808726e-06,
+      "loss": 0.4063,
+      "step": 4129
+    },
+    {
+      "epoch": 0.982923781757601,
+      "grad_norm": 0.3883006441711229,
+      "learning_rate": 5.382254058676812e-06,
+      "loss": 0.3706,
+      "step": 4130
+    },
+    {
+      "epoch": 0.9831617778306658,
+      "grad_norm": 0.35991795831667905,
+      "learning_rate": 5.380332077732489e-06,
+      "loss": 0.2823,
+      "step": 4131
+    },
+    {
+      "epoch": 0.9833997739037306,
+      "grad_norm": 0.34116602859971285,
+      "learning_rate": 5.378410040261408e-06,
+      "loss": 0.3513,
+      "step": 4132
+    },
+    {
+      "epoch": 0.9836377699767954,
+      "grad_norm": 0.4029763010820559,
+      "learning_rate": 5.376487946549235e-06,
+      "loss": 0.3987,
+      "step": 4133
+    },
+    {
+      "epoch": 0.9838757660498602,
+      "grad_norm": 0.3868100898163792,
+      "learning_rate": 5.374565796881639e-06,
+      "loss": 0.2773,
+      "step": 4134
+    },
+    {
+      "epoch": 0.984113762122925,
+      "grad_norm": 0.3428976994977603,
+      "learning_rate": 5.3726435915442986e-06,
+      "loss": 0.2696,
+      "step": 4135
+    },
+    {
+      "epoch": 0.9843517581959897,
+      "grad_norm": 0.37266226267641545,
+      "learning_rate": 5.370721330822904e-06,
+      "loss": 0.3714,
+      "step": 4136
+    },
+    {
+      "epoch": 0.9845897542690546,
+      "grad_norm": 0.34598524004373593,
+      "learning_rate": 5.368799015003146e-06,
+      "loss": 0.384,
+      "step": 4137
+    },
+    {
+      "epoch": 0.9848277503421193,
+      "grad_norm": 0.4192229780499278,
+      "learning_rate": 5.366876644370733e-06,
+      "loss": 0.273,
+      "step": 4138
+    },
+    {
+      "epoch": 0.9850657464151842,
+      "grad_norm": 0.3486972025471823,
+      "learning_rate": 5.364954219211372e-06,
+      "loss": 0.337,
+      "step": 4139
+    },
+    {
+      "epoch": 0.9853037424882489,
+      "grad_norm": 0.36821440890479656,
+      "learning_rate": 5.363031739810787e-06,
+      "loss": 0.4031,
+      "step": 4140
+    },
+    {
+      "epoch": 0.9855417385613138,
+      "grad_norm": 0.39566913279734434,
+      "learning_rate": 5.361109206454704e-06,
+      "loss": 0.3342,
+      "step": 4141
+    },
+    {
+      "epoch": 0.9857797346343785,
+      "grad_norm": 0.4004101088241766,
+      "learning_rate": 5.359186619428861e-06,
+      "loss": 0.2989,
+      "step": 4142
+    },
+    {
+      "epoch": 0.9860177307074434,
+      "grad_norm": 0.41706306261687004,
+      "learning_rate": 5.357263979018999e-06,
+      "loss": 0.3715,
+      "step": 4143
+    },
+    {
+      "epoch": 0.9862557267805081,
+      "grad_norm": 0.38390634367368964,
+      "learning_rate": 5.355341285510872e-06,
+      "loss": 0.4016,
+      "step": 4144
+    },
+    {
+      "epoch": 0.986493722853573,
+      "grad_norm": 0.4119688642491026,
+      "learning_rate": 5.35341853919024e-06,
+      "loss": 0.3256,
+      "step": 4145
+    },
+    {
+      "epoch": 0.9867317189266377,
+      "grad_norm": 0.3966809599743582,
+      "learning_rate": 5.351495740342868e-06,
+      "loss": 0.3151,
+      "step": 4146
+    },
+    {
+      "epoch": 0.9869697149997025,
+      "grad_norm": 0.37458039305346175,
+      "learning_rate": 5.349572889254535e-06,
+      "loss": 0.3874,
+      "step": 4147
+    },
+    {
+      "epoch": 0.9872077110727673,
+      "grad_norm": 0.35624311527568975,
+      "learning_rate": 5.347649986211022e-06,
+      "loss": 0.3685,
+      "step": 4148
+    },
+    {
+      "epoch": 0.9874457071458321,
+      "grad_norm": 0.41927466293047233,
+      "learning_rate": 5.34572703149812e-06,
+      "loss": 0.3162,
+      "step": 4149
+    },
+    {
+      "epoch": 0.9876837032188969,
+      "grad_norm": 0.3880202059036109,
+      "learning_rate": 5.343804025401628e-06,
+      "loss": 0.3319,
+      "step": 4150
+    },
+    {
+      "epoch": 0.9879216992919617,
+      "grad_norm": 0.36850162612727255,
+      "learning_rate": 5.3418809682073546e-06,
+      "loss": 0.4004,
+      "step": 4151
+    },
+    {
+      "epoch": 0.9881596953650265,
+      "grad_norm": 0.3884277693954743,
+      "learning_rate": 5.339957860201111e-06,
+      "loss": 0.2997,
+      "step": 4152
+    },
+    {
+      "epoch": 0.9883976914380913,
+      "grad_norm": 0.3670588779035891,
+      "learning_rate": 5.338034701668717e-06,
+      "loss": 0.3034,
+      "step": 4153
+    },
+    {
+      "epoch": 0.988635687511156,
+      "grad_norm": 0.35848269208319117,
+      "learning_rate": 5.336111492896005e-06,
+      "loss": 0.3707,
+      "step": 4154
+    },
+    {
+      "epoch": 0.9888736835842209,
+      "grad_norm": 0.39501627370624287,
+      "learning_rate": 5.334188234168811e-06,
+      "loss": 0.3591,
+      "step": 4155
+    },
+    {
+      "epoch": 0.9891116796572856,
+      "grad_norm": 0.3924138317274099,
+      "learning_rate": 5.332264925772979e-06,
+      "loss": 0.3118,
+      "step": 4156
+    },
+    {
+      "epoch": 0.9893496757303505,
+      "grad_norm": 0.40183975512658543,
+      "learning_rate": 5.330341567994357e-06,
+      "loss": 0.3298,
+      "step": 4157
+    },
+    {
+      "epoch": 0.9895876718034152,
+      "grad_norm": 0.3911218716929348,
+      "learning_rate": 5.328418161118807e-06,
+      "loss": 0.3957,
+      "step": 4158
+    },
+    {
+      "epoch": 0.9898256678764801,
+      "grad_norm": 0.39644737968164145,
+      "learning_rate": 5.326494705432194e-06,
+      "loss": 0.3453,
+      "step": 4159
+    },
+    {
+      "epoch": 0.9900636639495448,
+      "grad_norm": 0.37820532138547525,
+      "learning_rate": 5.324571201220388e-06,
+      "loss": 0.303,
+      "step": 4160
+    },
+    {
+      "epoch": 0.9903016600226097,
+      "grad_norm": 0.3680837268102721,
+      "learning_rate": 5.322647648769275e-06,
+      "loss": 0.3685,
+      "step": 4161
+    },
+    {
+      "epoch": 0.9905396560956744,
+      "grad_norm": 0.398138838445303,
+      "learning_rate": 5.320724048364736e-06,
+      "loss": 0.3738,
+      "step": 4162
+    },
+    {
+      "epoch": 0.9907776521687393,
+      "grad_norm": 0.3627540114538959,
+      "learning_rate": 5.3188004002926715e-06,
+      "loss": 0.3166,
+      "step": 4163
+    },
+    {
+      "epoch": 0.991015648241804,
+      "grad_norm": 0.3694127802681357,
+      "learning_rate": 5.316876704838981e-06,
+      "loss": 0.3229,
+      "step": 4164
+    },
+    {
+      "epoch": 0.9912536443148688,
+      "grad_norm": 0.37299601489965684,
+      "learning_rate": 5.314952962289574e-06,
+      "loss": 0.3883,
+      "step": 4165
+    },
+    {
+      "epoch": 0.9914916403879336,
+      "grad_norm": 0.3572974513846591,
+      "learning_rate": 5.3130291729303625e-06,
+      "loss": 0.3368,
+      "step": 4166
+    },
+    {
+      "epoch": 0.9917296364609984,
+      "grad_norm": 0.3730121718719649,
+      "learning_rate": 5.311105337047273e-06,
+      "loss": 0.2987,
+      "step": 4167
+    },
+    {
+      "epoch": 0.9919676325340632,
+      "grad_norm": 0.34901771416498145,
+      "learning_rate": 5.3091814549262345e-06,
+      "loss": 0.3522,
+      "step": 4168
+    },
+    {
+      "epoch": 0.992205628607128,
+      "grad_norm": 0.3810404289861147,
+      "learning_rate": 5.3072575268531835e-06,
+      "loss": 0.404,
+      "step": 4169
+    },
+    {
+      "epoch": 0.9924436246801928,
+      "grad_norm": 0.35455196776033093,
+      "learning_rate": 5.305333553114061e-06,
+      "loss": 0.2767,
+      "step": 4170
+    },
+    {
+      "epoch": 0.9926816207532576,
+      "grad_norm": 0.3861290164430468,
+      "learning_rate": 5.303409533994821e-06,
+      "loss": 0.319,
+      "step": 4171
+    },
+    {
+      "epoch": 0.9929196168263223,
+      "grad_norm": 0.3494911589336109,
+      "learning_rate": 5.301485469781418e-06,
+      "loss": 0.3845,
+      "step": 4172
+    },
+    {
+      "epoch": 0.9931576128993872,
+      "grad_norm": 0.3522517568562557,
+      "learning_rate": 5.299561360759815e-06,
+      "loss": 0.3433,
+      "step": 4173
+    },
+    {
+      "epoch": 0.9933956089724519,
+      "grad_norm": 0.39675837098549627,
+      "learning_rate": 5.297637207215982e-06,
+      "loss": 0.3027,
+      "step": 4174
+    },
+    {
+      "epoch": 0.9936336050455168,
+      "grad_norm": 0.37785284486500015,
+      "learning_rate": 5.295713009435898e-06,
+      "loss": 0.3505,
+      "step": 4175
+    },
+    {
+      "epoch": 0.9938716011185815,
+      "grad_norm": 0.3903018829163007,
+      "learning_rate": 5.293788767705544e-06,
+      "loss": 0.4349,
+      "step": 4176
+    },
+    {
+      "epoch": 0.9941095971916464,
+      "grad_norm": 0.36388379671761967,
+      "learning_rate": 5.291864482310913e-06,
+      "loss": 0.3286,
+      "step": 4177
+    },
+    {
+      "epoch": 0.9943475932647111,
+      "grad_norm": 0.359114110954639,
+      "learning_rate": 5.289940153537999e-06,
+      "loss": 0.3227,
+      "step": 4178
+    },
+    {
+      "epoch": 0.994585589337776,
+      "grad_norm": 0.3408663433796706,
+      "learning_rate": 5.2880157816728055e-06,
+      "loss": 0.4037,
+      "step": 4179
+    },
+    {
+      "epoch": 0.9948235854108407,
+      "grad_norm": 0.3827699859436007,
+      "learning_rate": 5.2860913670013415e-06,
+      "loss": 0.363,
+      "step": 4180
+    },
+    {
+      "epoch": 0.9950615814839056,
+      "grad_norm": 0.34748568520852696,
+      "learning_rate": 5.2841669098096215e-06,
+      "loss": 0.2973,
+      "step": 4181
+    },
+    {
+      "epoch": 0.9952995775569703,
+      "grad_norm": 0.3573147481932218,
+      "learning_rate": 5.2822424103836715e-06,
+      "loss": 0.3185,
+      "step": 4182
+    },
+    {
+      "epoch": 0.9955375736300351,
+      "grad_norm": 0.35989079023901194,
+      "learning_rate": 5.280317869009514e-06,
+      "loss": 0.3781,
+      "step": 4183
+    },
+    {
+      "epoch": 0.9957755697030999,
+      "grad_norm": 0.43015819010371176,
+      "learning_rate": 5.278393285973189e-06,
+      "loss": 0.3283,
+      "step": 4184
+    },
+    {
+      "epoch": 0.9960135657761647,
+      "grad_norm": 0.36962744882859006,
+      "learning_rate": 5.276468661560733e-06,
+      "loss": 0.2961,
+      "step": 4185
+    },
+    {
+      "epoch": 0.9962515618492295,
+      "grad_norm": 0.351978630130564,
+      "learning_rate": 5.274543996058195e-06,
+      "loss": 0.3664,
+      "step": 4186
+    },
+    {
+      "epoch": 0.9964895579222943,
+      "grad_norm": 0.3936832601718924,
+      "learning_rate": 5.272619289751627e-06,
+      "loss": 0.4083,
+      "step": 4187
+    },
+    {
+      "epoch": 0.9967275539953591,
+      "grad_norm": 0.4103960278310374,
+      "learning_rate": 5.270694542927089e-06,
+      "loss": 0.3127,
+      "step": 4188
+    },
+    {
+      "epoch": 0.9969655500684239,
+      "grad_norm": 0.3889947229714806,
+      "learning_rate": 5.268769755870643e-06,
+      "loss": 0.3141,
+      "step": 4189
+    },
+    {
+      "epoch": 0.9972035461414887,
+      "grad_norm": 0.39726469757369814,
+      "learning_rate": 5.266844928868364e-06,
+      "loss": 0.3777,
+      "step": 4190
+    },
+    {
+      "epoch": 0.9974415422145535,
+      "grad_norm": 0.3797443444152874,
+      "learning_rate": 5.264920062206328e-06,
+      "loss": 0.3664,
+      "step": 4191
+    },
+    {
+      "epoch": 0.9976795382876182,
+      "grad_norm": 0.3731570350476907,
+      "learning_rate": 5.262995156170616e-06,
+      "loss": 0.3014,
+      "step": 4192
+    },
+    {
+      "epoch": 0.9979175343606831,
+      "grad_norm": 0.3635536818396404,
+      "learning_rate": 5.261070211047318e-06,
+      "loss": 0.335,
+      "step": 4193
+    },
+    {
+      "epoch": 0.9981555304337478,
+      "grad_norm": 0.37957844831809195,
+      "learning_rate": 5.2591452271225276e-06,
+      "loss": 0.385,
+      "step": 4194
+    },
+    {
+      "epoch": 0.9983935265068127,
+      "grad_norm": 0.4412946371353015,
+      "learning_rate": 5.2572202046823455e-06,
+      "loss": 0.333,
+      "step": 4195
+    },
+    {
+      "epoch": 0.9986315225798774,
+      "grad_norm": 0.3949114004805517,
+      "learning_rate": 5.255295144012877e-06,
+      "loss": 0.327,
+      "step": 4196
+    },
+    {
+      "epoch": 0.9988695186529423,
+      "grad_norm": 0.37599046677243475,
+      "learning_rate": 5.253370045400236e-06,
+      "loss": 0.4231,
+      "step": 4197
+    },
+    {
+      "epoch": 0.999107514726007,
+      "grad_norm": 0.36211875864414783,
+      "learning_rate": 5.2514449091305375e-06,
+      "loss": 0.3453,
+      "step": 4198
+    },
+    {
+      "epoch": 0.9993455107990719,
+      "grad_norm": 0.4080030072391759,
+      "learning_rate": 5.249519735489904e-06,
+      "loss": 0.2976,
+      "step": 4199
+    },
+    {
+      "epoch": 0.9995835068721366,
+      "grad_norm": 0.4151890843838898,
+      "learning_rate": 5.247594524764466e-06,
+      "loss": 0.3492,
+      "step": 4200
+    },
+    {
+      "epoch": 0.9998215029452014,
+      "grad_norm": 0.3667853249254919,
+      "learning_rate": 5.2456692772403565e-06,
+      "loss": 0.4018,
+      "step": 4201
+    },
+    {
+      "epoch": 1.0002379960730647,
+      "grad_norm": 0.34868915612115237,
+      "learning_rate": 5.243743993203715e-06,
+      "loss": 0.461,
+      "step": 4202
+    },
+    {
+      "epoch": 1.0004759921461295,
+      "grad_norm": 0.3543503974639351,
+      "learning_rate": 5.241818672940684e-06,
+      "loss": 0.3138,
+      "step": 4203
+    },
+    {
+      "epoch": 1.0007139882191944,
+      "grad_norm": 0.35172615371783905,
+      "learning_rate": 5.239893316737419e-06,
+      "loss": 0.3275,
+      "step": 4204
+    },
+    {
+      "epoch": 1.0009519842922592,
+      "grad_norm": 0.37691116884826764,
+      "learning_rate": 5.23796792488007e-06,
+      "loss": 0.3566,
+      "step": 4205
+    },
+    {
+      "epoch": 1.001189980365324,
+      "grad_norm": 0.3409709908376464,
+      "learning_rate": 5.236042497654802e-06,
+      "loss": 0.3944,
+      "step": 4206
+    },
+    {
+      "epoch": 1.0014279764383887,
+      "grad_norm": 0.36845210293609737,
+      "learning_rate": 5.2341170353477795e-06,
+      "loss": 0.2646,
+      "step": 4207
+    },
+    {
+      "epoch": 1.0016659725114536,
+      "grad_norm": 0.359375740170696,
+      "learning_rate": 5.232191538245173e-06,
+      "loss": 0.319,
+      "step": 4208
+    },
+    {
+      "epoch": 1.0019039685845184,
+      "grad_norm": 0.3920628761903601,
+      "learning_rate": 5.23026600663316e-06,
+      "loss": 0.3928,
+      "step": 4209
+    },
+    {
+      "epoch": 1.002141964657583,
+      "grad_norm": 0.41390620119344756,
+      "learning_rate": 5.2283404407979225e-06,
+      "loss": 0.3013,
+      "step": 4210
+    },
+    {
+      "epoch": 1.0023799607306478,
+      "grad_norm": 0.4220811528276169,
+      "learning_rate": 5.226414841025645e-06,
+      "loss": 0.2726,
+      "step": 4211
+    },
+    {
+      "epoch": 1.0026179568037128,
+      "grad_norm": 0.41557061395608813,
+      "learning_rate": 5.2244892076025225e-06,
+      "loss": 0.365,
+      "step": 4212
+    },
+    {
+      "epoch": 1.0028559528767775,
+      "grad_norm": 0.40671365818027116,
+      "learning_rate": 5.22256354081475e-06,
+      "loss": 0.3638,
+      "step": 4213
+    },
+    {
+      "epoch": 1.0030939489498423,
+      "grad_norm": 0.38827259080751864,
+      "learning_rate": 5.220637840948528e-06,
+      "loss": 0.3195,
+      "step": 4214
+    },
+    {
+      "epoch": 1.003331945022907,
+      "grad_norm": 0.46007325776629265,
+      "learning_rate": 5.218712108290065e-06,
+      "loss": 0.3423,
+      "step": 4215
+    },
+    {
+      "epoch": 1.003569941095972,
+      "grad_norm": 0.3927299977513199,
+      "learning_rate": 5.216786343125572e-06,
+      "loss": 0.3815,
+      "step": 4216
+    },
+    {
+      "epoch": 1.0038079371690367,
+      "grad_norm": 0.3777948807064563,
+      "learning_rate": 5.214860545741266e-06,
+      "loss": 0.3139,
+      "step": 4217
+    },
+    {
+      "epoch": 1.0040459332421015,
+      "grad_norm": 0.4096489630411397,
+      "learning_rate": 5.212934716423368e-06,
+      "loss": 0.2923,
+      "step": 4218
+    },
+    {
+      "epoch": 1.0042839293151662,
+      "grad_norm": 0.4428443016282393,
+      "learning_rate": 5.211008855458103e-06,
+      "loss": 0.3226,
+      "step": 4219
+    },
+    {
+      "epoch": 1.0045219253882312,
+      "grad_norm": 0.3548101722105468,
+      "learning_rate": 5.209082963131702e-06,
+      "loss": 0.3717,
+      "step": 4220
+    },
+    {
+      "epoch": 1.004759921461296,
+      "grad_norm": 0.3661098943341066,
+      "learning_rate": 5.2071570397303995e-06,
+      "loss": 0.3291,
+      "step": 4221
+    },
+    {
+      "epoch": 1.0049979175343606,
+      "grad_norm": 0.39846837704199717,
+      "learning_rate": 5.2052310855404356e-06,
+      "loss": 0.2979,
+      "step": 4222
+    },
+    {
+      "epoch": 1.0052359136074254,
+      "grad_norm": 0.44041097154948994,
+      "learning_rate": 5.203305100848056e-06,
+      "loss": 0.3587,
+      "step": 4223
+    },
+    {
+      "epoch": 1.0054739096804903,
+      "grad_norm": 0.3696154903241395,
+      "learning_rate": 5.20137908593951e-06,
+      "loss": 0.357,
+      "step": 4224
+    },
+    {
+      "epoch": 1.005711905753555,
+      "grad_norm": 0.3589543414638921,
+      "learning_rate": 5.1994530411010495e-06,
+      "loss": 0.3017,
+      "step": 4225
+    },
+    {
+      "epoch": 1.0059499018266198,
+      "grad_norm": 0.4534518628656481,
+      "learning_rate": 5.1975269666189325e-06,
+      "loss": 0.3259,
+      "step": 4226
+    },
+    {
+      "epoch": 1.0061878978996845,
+      "grad_norm": 0.37753721010737906,
+      "learning_rate": 5.195600862779421e-06,
+      "loss": 0.3984,
+      "step": 4227
+    },
+    {
+      "epoch": 1.0064258939727495,
+      "grad_norm": 0.35907800320760047,
+      "learning_rate": 5.193674729868781e-06,
+      "loss": 0.3162,
+      "step": 4228
+    },
+    {
+      "epoch": 1.0066638900458142,
+      "grad_norm": 0.38681209195212035,
+      "learning_rate": 5.191748568173288e-06,
+      "loss": 0.3052,
+      "step": 4229
+    },
+    {
+      "epoch": 1.006901886118879,
+      "grad_norm": 0.38975003350282217,
+      "learning_rate": 5.1898223779792125e-06,
+      "loss": 0.369,
+      "step": 4230
+    },
+    {
+      "epoch": 1.0071398821919437,
+      "grad_norm": 0.37451302101771794,
+      "learning_rate": 5.187896159572836e-06,
+      "loss": 0.3867,
+      "step": 4231
+    },
+    {
+      "epoch": 1.0073778782650087,
+      "grad_norm": 0.3746202575214494,
+      "learning_rate": 5.185969913240442e-06,
+      "loss": 0.3123,
+      "step": 4232
+    },
+    {
+      "epoch": 1.0076158743380734,
+      "grad_norm": 0.3399200166882226,
+      "learning_rate": 5.184043639268318e-06,
+      "loss": 0.3423,
+      "step": 4233
+    },
+    {
+      "epoch": 1.0078538704111382,
+      "grad_norm": 0.37484361667395105,
+      "learning_rate": 5.1821173379427566e-06,
+      "loss": 0.3874,
+      "step": 4234
+    },
+    {
+      "epoch": 1.008091866484203,
+      "grad_norm": 0.38264504872383276,
+      "learning_rate": 5.180191009550053e-06,
+      "loss": 0.3665,
+      "step": 4235
+    },
+    {
+      "epoch": 1.0083298625572679,
+      "grad_norm": 0.3905871905210143,
+      "learning_rate": 5.178264654376507e-06,
+      "loss": 0.3112,
+      "step": 4236
+    },
+    {
+      "epoch": 1.0085678586303326,
+      "grad_norm": 0.39456565680861116,
+      "learning_rate": 5.176338272708424e-06,
+      "loss": 0.3265,
+      "step": 4237
+    },
+    {
+      "epoch": 1.0088058547033973,
+      "grad_norm": 0.36221965002648904,
+      "learning_rate": 5.174411864832111e-06,
+      "loss": 0.3884,
+      "step": 4238
+    },
+    {
+      "epoch": 1.009043850776462,
+      "grad_norm": 0.40646971651183245,
+      "learning_rate": 5.172485431033882e-06,
+      "loss": 0.2814,
+      "step": 4239
+    },
+    {
+      "epoch": 1.009281846849527,
+      "grad_norm": 0.40606667706481386,
+      "learning_rate": 5.17055897160005e-06,
+      "loss": 0.295,
+      "step": 4240
+    },
+    {
+      "epoch": 1.0095198429225918,
+      "grad_norm": 0.4187704484376356,
+      "learning_rate": 5.168632486816934e-06,
+      "loss": 0.3786,
+      "step": 4241
+    },
+    {
+      "epoch": 1.0097578389956565,
+      "grad_norm": 0.38137336025761603,
+      "learning_rate": 5.1667059769708595e-06,
+      "loss": 0.3093,
+      "step": 4242
+    },
+    {
+      "epoch": 1.0099958350687213,
+      "grad_norm": 0.4262476491330625,
+      "learning_rate": 5.1647794423481516e-06,
+      "loss": 0.2841,
+      "step": 4243
+    },
+    {
+      "epoch": 1.0102338311417862,
+      "grad_norm": 0.42194011631004263,
+      "learning_rate": 5.1628528832351436e-06,
+      "loss": 0.3438,
+      "step": 4244
+    },
+    {
+      "epoch": 1.010471827214851,
+      "grad_norm": 0.37173229226099785,
+      "learning_rate": 5.160926299918167e-06,
+      "loss": 0.3863,
+      "step": 4245
+    },
+    {
+      "epoch": 1.0107098232879157,
+      "grad_norm": 0.3611360955387892,
+      "learning_rate": 5.158999692683563e-06,
+      "loss": 0.2911,
+      "step": 4246
+    },
+    {
+      "epoch": 1.0109478193609804,
+      "grad_norm": 0.37976056901327965,
+      "learning_rate": 5.15707306181767e-06,
+      "loss": 0.3135,
+      "step": 4247
+    },
+    {
+      "epoch": 1.0111858154340454,
+      "grad_norm": 0.40048201279174483,
+      "learning_rate": 5.155146407606835e-06,
+      "loss": 0.3757,
+      "step": 4248
+    },
+    {
+      "epoch": 1.0114238115071101,
+      "grad_norm": 0.36075054638382786,
+      "learning_rate": 5.153219730337406e-06,
+      "loss": 0.3147,
+      "step": 4249
+    },
+    {
+      "epoch": 1.0116618075801749,
+      "grad_norm": 0.3679232730274379,
+      "learning_rate": 5.151293030295732e-06,
+      "loss": 0.2865,
+      "step": 4250
+    },
+    {
+      "epoch": 1.0118998036532396,
+      "grad_norm": 0.3581758351944555,
+      "learning_rate": 5.149366307768173e-06,
+      "loss": 0.3218,
+      "step": 4251
+    },
+    {
+      "epoch": 1.0121377997263046,
+      "grad_norm": 0.3823224717214005,
+      "learning_rate": 5.147439563041086e-06,
+      "loss": 0.3987,
+      "step": 4252
+    },
+    {
+      "epoch": 1.0123757957993693,
+      "grad_norm": 0.33371251114397465,
+      "learning_rate": 5.145512796400831e-06,
+      "loss": 0.2999,
+      "step": 4253
+    },
+    {
+      "epoch": 1.012613791872434,
+      "grad_norm": 0.3823340185719183,
+      "learning_rate": 5.143586008133776e-06,
+      "loss": 0.2941,
+      "step": 4254
+    },
+    {
+      "epoch": 1.0128517879454988,
+      "grad_norm": 0.382504369490985,
+      "learning_rate": 5.141659198526287e-06,
+      "loss": 0.3752,
+      "step": 4255
+    },
+    {
+      "epoch": 1.0130897840185638,
+      "grad_norm": 0.3814481433528493,
+      "learning_rate": 5.139732367864736e-06,
+      "loss": 0.3442,
+      "step": 4256
+    },
+    {
+      "epoch": 1.0133277800916285,
+      "grad_norm": 0.3964801873649591,
+      "learning_rate": 5.137805516435499e-06,
+      "loss": 0.2744,
+      "step": 4257
+    },
+    {
+      "epoch": 1.0135657761646932,
+      "grad_norm": 0.35135187470588386,
+      "learning_rate": 5.135878644524953e-06,
+      "loss": 0.3428,
+      "step": 4258
+    },
+    {
+      "epoch": 1.013803772237758,
+      "grad_norm": 0.43647825401352464,
+      "learning_rate": 5.13395175241948e-06,
+      "loss": 0.4011,
+      "step": 4259
+    },
+    {
+      "epoch": 1.014041768310823,
+      "grad_norm": 0.48740933719008334,
+      "learning_rate": 5.132024840405462e-06,
+      "loss": 0.3116,
+      "step": 4260
+    },
+    {
+      "epoch": 1.0142797643838877,
+      "grad_norm": 0.3754855598282161,
+      "learning_rate": 5.130097908769287e-06,
+      "loss": 0.2721,
+      "step": 4261
+    },
+    {
+      "epoch": 1.0145177604569524,
+      "grad_norm": 0.37968335966178834,
+      "learning_rate": 5.128170957797345e-06,
+      "loss": 0.3164,
+      "step": 4262
+    },
+    {
+      "epoch": 1.0147557565300172,
+      "grad_norm": 0.38142166044175096,
+      "learning_rate": 5.126243987776026e-06,
+      "loss": 0.3873,
+      "step": 4263
+    },
+    {
+      "epoch": 1.0149937526030821,
+      "grad_norm": 0.374445975748934,
+      "learning_rate": 5.124316998991728e-06,
+      "loss": 0.2786,
+      "step": 4264
+    },
+    {
+      "epoch": 1.0152317486761469,
+      "grad_norm": 0.42403267311698933,
+      "learning_rate": 5.122389991730848e-06,
+      "loss": 0.3162,
+      "step": 4265
+    },
+    {
+      "epoch": 1.0154697447492116,
+      "grad_norm": 0.3964607797578093,
+      "learning_rate": 5.120462966279789e-06,
+      "loss": 0.3996,
+      "step": 4266
+    },
+    {
+      "epoch": 1.0157077408222763,
+      "grad_norm": 0.363101926710214,
+      "learning_rate": 5.118535922924952e-06,
+      "loss": 0.3341,
+      "step": 4267
+    },
+    {
+      "epoch": 1.0159457368953413,
+      "grad_norm": 0.3751042986845883,
+      "learning_rate": 5.1166088619527445e-06,
+      "loss": 0.2738,
+      "step": 4268
+    },
+    {
+      "epoch": 1.016183732968406,
+      "grad_norm": 0.36892262345896815,
+      "learning_rate": 5.114681783649575e-06,
+      "loss": 0.3194,
+      "step": 4269
+    },
+    {
+      "epoch": 1.0164217290414708,
+      "grad_norm": 0.34748547029304633,
+      "learning_rate": 5.112754688301855e-06,
+      "loss": 0.3562,
+      "step": 4270
+    },
+    {
+      "epoch": 1.0166597251145355,
+      "grad_norm": 0.38763196396366484,
+      "learning_rate": 5.110827576196e-06,
+      "loss": 0.307,
+      "step": 4271
+    },
+    {
+      "epoch": 1.0168977211876005,
+      "grad_norm": 0.37714618148409906,
+      "learning_rate": 5.1089004476184255e-06,
+      "loss": 0.3069,
+      "step": 4272
+    },
+    {
+      "epoch": 1.0171357172606652,
+      "grad_norm": 0.3771535074762055,
+      "learning_rate": 5.10697330285555e-06,
+      "loss": 0.3587,
+      "step": 4273
+    },
+    {
+      "epoch": 1.01737371333373,
+      "grad_norm": 0.3691570859798567,
+      "learning_rate": 5.105046142193796e-06,
+      "loss": 0.3495,
+      "step": 4274
+    },
+    {
+      "epoch": 1.0176117094067947,
+      "grad_norm": 0.4171393197587053,
+      "learning_rate": 5.103118965919586e-06,
+      "loss": 0.2785,
+      "step": 4275
+    },
+    {
+      "epoch": 1.0178497054798596,
+      "grad_norm": 0.3718279478108766,
+      "learning_rate": 5.101191774319346e-06,
+      "loss": 0.3278,
+      "step": 4276
+    },
+    {
+      "epoch": 1.0180877015529244,
+      "grad_norm": 0.3762497178178657,
+      "learning_rate": 5.099264567679505e-06,
+      "loss": 0.3814,
+      "step": 4277
+    },
+    {
+      "epoch": 1.0183256976259891,
+      "grad_norm": 0.3925916540388656,
+      "learning_rate": 5.097337346286494e-06,
+      "loss": 0.3096,
+      "step": 4278
+    },
+    {
+      "epoch": 1.0185636936990539,
+      "grad_norm": 0.4026406915827254,
+      "learning_rate": 5.095410110426746e-06,
+      "loss": 0.2751,
+      "step": 4279
+    },
+    {
+      "epoch": 1.0188016897721188,
+      "grad_norm": 0.3737121034084029,
+      "learning_rate": 5.093482860386695e-06,
+      "loss": 0.3442,
+      "step": 4280
+    },
+    {
+      "epoch": 1.0190396858451836,
+      "grad_norm": 0.3662525796650136,
+      "learning_rate": 5.091555596452777e-06,
+      "loss": 0.3694,
+      "step": 4281
+    },
+    {
+      "epoch": 1.0192776819182483,
+      "grad_norm": 0.37406892875546127,
+      "learning_rate": 5.089628318911434e-06,
+      "loss": 0.2785,
+      "step": 4282
+    },
+    {
+      "epoch": 1.019515677991313,
+      "grad_norm": 0.362735438961647,
+      "learning_rate": 5.0877010280491045e-06,
+      "loss": 0.3289,
+      "step": 4283
+    },
+    {
+      "epoch": 1.019753674064378,
+      "grad_norm": 0.3737293250336674,
+      "learning_rate": 5.085773724152232e-06,
+      "loss": 0.3725,
+      "step": 4284
+    },
+    {
+      "epoch": 1.0199916701374427,
+      "grad_norm": 0.37431985674627993,
+      "learning_rate": 5.083846407507263e-06,
+      "loss": 0.3119,
+      "step": 4285
+    },
+    {
+      "epoch": 1.0202296662105075,
+      "grad_norm": 0.39281524554349173,
+      "learning_rate": 5.0819190784006444e-06,
+      "loss": 0.3318,
+      "step": 4286
+    },
+    {
+      "epoch": 1.0204676622835722,
+      "grad_norm": 0.4178241632215565,
+      "learning_rate": 5.079991737118823e-06,
+      "loss": 0.3592,
+      "step": 4287
+    },
+    {
+      "epoch": 1.0207056583566372,
+      "grad_norm": 0.37007927511322464,
+      "learning_rate": 5.0780643839482515e-06,
+      "loss": 0.3696,
+      "step": 4288
+    },
+    {
+      "epoch": 1.020943654429702,
+      "grad_norm": 0.393635301072028,
+      "learning_rate": 5.076137019175381e-06,
+      "loss": 0.2891,
+      "step": 4289
+    },
+    {
+      "epoch": 1.0211816505027667,
+      "grad_norm": 0.37703876121952784,
+      "learning_rate": 5.074209643086666e-06,
+      "loss": 0.33,
+      "step": 4290
+    },
+    {
+      "epoch": 1.0214196465758314,
+      "grad_norm": 0.42049817515097926,
+      "learning_rate": 5.072282255968561e-06,
+      "loss": 0.3764,
+      "step": 4291
+    },
+    {
+      "epoch": 1.0216576426488964,
+      "grad_norm": 0.38019207344494804,
+      "learning_rate": 5.070354858107526e-06,
+      "loss": 0.31,
+      "step": 4292
+    },
+    {
+      "epoch": 1.021895638721961,
+      "grad_norm": 0.4040481113943476,
+      "learning_rate": 5.068427449790019e-06,
+      "loss": 0.2868,
+      "step": 4293
+    },
+    {
+      "epoch": 1.0221336347950258,
+      "grad_norm": 0.370686703144065,
+      "learning_rate": 5.0665000313024995e-06,
+      "loss": 0.3369,
+      "step": 4294
+    },
+    {
+      "epoch": 1.0223716308680906,
+      "grad_norm": 0.39346612528862,
+      "learning_rate": 5.064572602931428e-06,
+      "loss": 0.3823,
+      "step": 4295
+    },
+    {
+      "epoch": 1.0226096269411555,
+      "grad_norm": 0.3717514479143731,
+      "learning_rate": 5.0626451649632725e-06,
+      "loss": 0.3087,
+      "step": 4296
+    },
+    {
+      "epoch": 1.0228476230142203,
+      "grad_norm": 0.4092985098023293,
+      "learning_rate": 5.060717717684496e-06,
+      "loss": 0.3115,
+      "step": 4297
+    },
+    {
+      "epoch": 1.023085619087285,
+      "grad_norm": 0.3934618504996616,
+      "learning_rate": 5.058790261381563e-06,
+      "loss": 0.3801,
+      "step": 4298
+    },
+    {
+      "epoch": 1.0233236151603498,
+      "grad_norm": 0.3492191091572103,
+      "learning_rate": 5.056862796340944e-06,
+      "loss": 0.3489,
+      "step": 4299
+    },
+    {
+      "epoch": 1.0235616112334147,
+      "grad_norm": 0.4051063697173622,
+      "learning_rate": 5.054935322849107e-06,
+      "loss": 0.2997,
+      "step": 4300
+    },
+    {
+      "epoch": 1.0237996073064795,
+      "grad_norm": 0.3469654129695318,
+      "learning_rate": 5.053007841192522e-06,
+      "loss": 0.3036,
+      "step": 4301
+    },
+    {
+      "epoch": 1.0240376033795442,
+      "grad_norm": 0.38270686628936906,
+      "learning_rate": 5.05108035165766e-06,
+      "loss": 0.3964,
+      "step": 4302
+    },
+    {
+      "epoch": 1.024275599452609,
+      "grad_norm": 0.36001024577676244,
+      "learning_rate": 5.049152854530994e-06,
+      "loss": 0.2792,
+      "step": 4303
+    },
+    {
+      "epoch": 1.024513595525674,
+      "grad_norm": 0.37961661632016525,
+      "learning_rate": 5.047225350098999e-06,
+      "loss": 0.3038,
+      "step": 4304
+    },
+    {
+      "epoch": 1.0247515915987386,
+      "grad_norm": 0.3494836771560282,
+      "learning_rate": 5.045297838648145e-06,
+      "loss": 0.3919,
+      "step": 4305
+    },
+    {
+      "epoch": 1.0249895876718034,
+      "grad_norm": 0.36305044333379766,
+      "learning_rate": 5.043370320464915e-06,
+      "loss": 0.3636,
+      "step": 4306
+    },
+    {
+      "epoch": 1.025227583744868,
+      "grad_norm": 0.4258357687705437,
+      "learning_rate": 5.041442795835783e-06,
+      "loss": 0.2803,
+      "step": 4307
+    },
+    {
+      "epoch": 1.025465579817933,
+      "grad_norm": 0.4160868010268067,
+      "learning_rate": 5.039515265047224e-06,
+      "loss": 0.3107,
+      "step": 4308
+    },
+    {
+      "epoch": 1.0257035758909978,
+      "grad_norm": 0.4342418241767286,
+      "learning_rate": 5.037587728385719e-06,
+      "loss": 0.4278,
+      "step": 4309
+    },
+    {
+      "epoch": 1.0259415719640625,
+      "grad_norm": 0.37103949650763524,
+      "learning_rate": 5.035660186137749e-06,
+      "loss": 0.3015,
+      "step": 4310
+    },
+    {
+      "epoch": 1.0261795680371273,
+      "grad_norm": 0.38074037486983087,
+      "learning_rate": 5.033732638589793e-06,
+      "loss": 0.2781,
+      "step": 4311
+    },
+    {
+      "epoch": 1.0264175641101922,
+      "grad_norm": 0.4189038925958272,
+      "learning_rate": 5.0318050860283306e-06,
+      "loss": 0.3664,
+      "step": 4312
+    },
+    {
+      "epoch": 1.026655560183257,
+      "grad_norm": 0.3951890393473658,
+      "learning_rate": 5.029877528739848e-06,
+      "loss": 0.3446,
+      "step": 4313
+    },
+    {
+      "epoch": 1.0268935562563217,
+      "grad_norm": 0.3983002055289879,
+      "learning_rate": 5.0279499670108245e-06,
+      "loss": 0.3065,
+      "step": 4314
+    },
+    {
+      "epoch": 1.0271315523293865,
+      "grad_norm": 0.39958104133807376,
+      "learning_rate": 5.0260224011277445e-06,
+      "loss": 0.3262,
+      "step": 4315
+    },
+    {
+      "epoch": 1.0273695484024514,
+      "grad_norm": 0.4123932451298326,
+      "learning_rate": 5.0240948313770934e-06,
+      "loss": 0.3763,
+      "step": 4316
+    },
+    {
+      "epoch": 1.0276075444755162,
+      "grad_norm": 0.39306243304232114,
+      "learning_rate": 5.022167258045353e-06,
+      "loss": 0.3214,
+      "step": 4317
+    },
+    {
+      "epoch": 1.027845540548581,
+      "grad_norm": 0.3761676443591786,
+      "learning_rate": 5.0202396814190095e-06,
+      "loss": 0.2848,
+      "step": 4318
+    },
+    {
+      "epoch": 1.0280835366216456,
+      "grad_norm": 0.38223410956180287,
+      "learning_rate": 5.018312101784548e-06,
+      "loss": 0.3257,
+      "step": 4319
+    },
+    {
+      "epoch": 1.0283215326947106,
+      "grad_norm": 0.3597507159599325,
+      "learning_rate": 5.016384519428456e-06,
+      "loss": 0.3752,
+      "step": 4320
+    },
+    {
+      "epoch": 1.0285595287677753,
+      "grad_norm": 0.328025805606429,
+      "learning_rate": 5.0144569346372185e-06,
+      "loss": 0.2937,
+      "step": 4321
+    },
+    {
+      "epoch": 1.02879752484084,
+      "grad_norm": 0.4660913256009466,
+      "learning_rate": 5.012529347697322e-06,
+      "loss": 0.3239,
+      "step": 4322
+    },
+    {
+      "epoch": 1.0290355209139048,
+      "grad_norm": 0.383898516473306,
+      "learning_rate": 5.010601758895257e-06,
+      "loss": 0.3778,
+      "step": 4323
+    },
+    {
+      "epoch": 1.0292735169869698,
+      "grad_norm": 0.36767695340931256,
+      "learning_rate": 5.008674168517507e-06,
+      "loss": 0.3279,
+      "step": 4324
+    },
+    {
+      "epoch": 1.0295115130600345,
+      "grad_norm": 0.37694685767170544,
+      "learning_rate": 5.006746576850562e-06,
+      "loss": 0.3108,
+      "step": 4325
+    },
+    {
+      "epoch": 1.0297495091330993,
+      "grad_norm": 0.37325893976856545,
+      "learning_rate": 5.004818984180907e-06,
+      "loss": 0.3387,
+      "step": 4326
+    },
+    {
+      "epoch": 1.029987505206164,
+      "grad_norm": 0.38537315515694565,
+      "learning_rate": 5.002891390795033e-06,
+      "loss": 0.4258,
+      "step": 4327
+    },
+    {
+      "epoch": 1.030225501279229,
+      "grad_norm": 0.3589773496139594,
+      "learning_rate": 5.0009637969794255e-06,
+      "loss": 0.2959,
+      "step": 4328
+    },
+    {
+      "epoch": 1.0304634973522937,
+      "grad_norm": 0.37090708026971275,
+      "learning_rate": 4.9990362030205745e-06,
+      "loss": 0.2818,
+      "step": 4329
+    },
+    {
+      "epoch": 1.0307014934253584,
+      "grad_norm": 0.3798707951977427,
+      "learning_rate": 4.997108609204968e-06,
+      "loss": 0.3882,
+      "step": 4330
+    },
+    {
+      "epoch": 1.0309394894984232,
+      "grad_norm": 0.3786611077695848,
+      "learning_rate": 4.995181015819094e-06,
+      "loss": 0.3903,
+      "step": 4331
+    },
+    {
+      "epoch": 1.0311774855714881,
+      "grad_norm": 0.37672618523404194,
+      "learning_rate": 4.99325342314944e-06,
+      "loss": 0.3122,
+      "step": 4332
+    },
+    {
+      "epoch": 1.0314154816445529,
+      "grad_norm": 0.391170036332283,
+      "learning_rate": 4.991325831482494e-06,
+      "loss": 0.3309,
+      "step": 4333
+    },
+    {
+      "epoch": 1.0316534777176176,
+      "grad_norm": 0.4742272257387251,
+      "learning_rate": 4.989398241104745e-06,
+      "loss": 0.3744,
+      "step": 4334
+    },
+    {
+      "epoch": 1.0318914737906824,
+      "grad_norm": 0.381528588265086,
+      "learning_rate": 4.987470652302679e-06,
+      "loss": 0.3236,
+      "step": 4335
+    },
+    {
+      "epoch": 1.0321294698637473,
+      "grad_norm": 0.3909739709474945,
+      "learning_rate": 4.985543065362782e-06,
+      "loss": 0.3005,
+      "step": 4336
+    },
+    {
+      "epoch": 1.032367465936812,
+      "grad_norm": 0.4030798587204337,
+      "learning_rate": 4.983615480571546e-06,
+      "loss": 0.3292,
+      "step": 4337
+    },
+    {
+      "epoch": 1.0326054620098768,
+      "grad_norm": 0.39016102552178,
+      "learning_rate": 4.981687898215454e-06,
+      "loss": 0.3811,
+      "step": 4338
+    },
+    {
+      "epoch": 1.0328434580829415,
+      "grad_norm": 0.35718795349461896,
+      "learning_rate": 4.979760318580993e-06,
+      "loss": 0.297,
+      "step": 4339
+    },
+    {
+      "epoch": 1.0330814541560065,
+      "grad_norm": 1.5662065706280457,
+      "learning_rate": 4.97783274195465e-06,
+      "loss": 0.3122,
+      "step": 4340
+    },
+    {
+      "epoch": 1.0333194502290712,
+      "grad_norm": 0.4031583976642777,
+      "learning_rate": 4.97590516862291e-06,
+      "loss": 0.3742,
+      "step": 4341
+    },
+    {
+      "epoch": 1.033557446302136,
+      "grad_norm": 0.3666936736763186,
+      "learning_rate": 4.973977598872257e-06,
+      "loss": 0.3361,
+      "step": 4342
+    },
+    {
+      "epoch": 1.0337954423752007,
+      "grad_norm": 0.3870156161445879,
+      "learning_rate": 4.9720500329891755e-06,
+      "loss": 0.2907,
+      "step": 4343
+    },
+    {
+      "epoch": 1.0340334384482657,
+      "grad_norm": 0.3943973816109862,
+      "learning_rate": 4.9701224712601526e-06,
+      "loss": 0.3316,
+      "step": 4344
+    },
+    {
+      "epoch": 1.0342714345213304,
+      "grad_norm": 0.37045738296936315,
+      "learning_rate": 4.9681949139716686e-06,
+      "loss": 0.3745,
+      "step": 4345
+    },
+    {
+      "epoch": 1.0345094305943952,
+      "grad_norm": 0.3526903041685496,
+      "learning_rate": 4.966267361410209e-06,
+      "loss": 0.2671,
+      "step": 4346
+    },
+    {
+      "epoch": 1.03474742666746,
+      "grad_norm": 0.4097067620359685,
+      "learning_rate": 4.964339813862252e-06,
+      "loss": 0.3019,
+      "step": 4347
+    },
+    {
+      "epoch": 1.0349854227405249,
+      "grad_norm": 0.3907934312556052,
+      "learning_rate": 4.962412271614282e-06,
+      "loss": 0.3721,
+      "step": 4348
+    },
+    {
+      "epoch": 1.0352234188135896,
+      "grad_norm": 0.39946066906031263,
+      "learning_rate": 4.9604847349527775e-06,
+      "loss": 0.3449,
+      "step": 4349
+    },
+    {
+      "epoch": 1.0354614148866543,
+      "grad_norm": 0.35694279142285595,
+      "learning_rate": 4.958557204164219e-06,
+      "loss": 0.2945,
+      "step": 4350
+    },
+    {
+      "epoch": 1.035699410959719,
+      "grad_norm": 0.36264389505281575,
+      "learning_rate": 4.956629679535086e-06,
+      "loss": 0.3114,
+      "step": 4351
+    },
+    {
+      "epoch": 1.035937407032784,
+      "grad_norm": 0.3837297046950812,
+      "learning_rate": 4.954702161351856e-06,
+      "loss": 0.398,
+      "step": 4352
+    },
+    {
+      "epoch": 1.0361754031058488,
+      "grad_norm": 0.3723581905698087,
+      "learning_rate": 4.952774649901004e-06,
+      "loss": 0.3005,
+      "step": 4353
+    },
+    {
+      "epoch": 1.0364133991789135,
+      "grad_norm": 0.3668387139380813,
+      "learning_rate": 4.950847145469008e-06,
+      "loss": 0.3008,
+      "step": 4354
+    },
+    {
+      "epoch": 1.0366513952519782,
+      "grad_norm": 0.36162042038397063,
+      "learning_rate": 4.948919648342342e-06,
+      "loss": 0.3702,
+      "step": 4355
+    },
+    {
+      "epoch": 1.0368893913250432,
+      "grad_norm": 0.3603460188850127,
+      "learning_rate": 4.946992158807481e-06,
+      "loss": 0.3833,
+      "step": 4356
+    },
+    {
+      "epoch": 1.037127387398108,
+      "grad_norm": 0.3744254359291755,
+      "learning_rate": 4.945064677150893e-06,
+      "loss": 0.281,
+      "step": 4357
+    },
+    {
+      "epoch": 1.0373653834711727,
+      "grad_norm": 0.3842915853972375,
+      "learning_rate": 4.943137203659056e-06,
+      "loss": 0.346,
+      "step": 4358
+    },
+    {
+      "epoch": 1.0376033795442374,
+      "grad_norm": 0.3822698279560983,
+      "learning_rate": 4.941209738618437e-06,
+      "loss": 0.4112,
+      "step": 4359
+    },
+    {
+      "epoch": 1.0378413756173024,
+      "grad_norm": 0.40057285679058796,
+      "learning_rate": 4.939282282315505e-06,
+      "loss": 0.2888,
+      "step": 4360
+    },
+    {
+      "epoch": 1.0380793716903671,
+      "grad_norm": 0.382452358107358,
+      "learning_rate": 4.937354835036728e-06,
+      "loss": 0.2905,
+      "step": 4361
+    },
+    {
+      "epoch": 1.0383173677634319,
+      "grad_norm": 0.3850074270426723,
+      "learning_rate": 4.935427397068573e-06,
+      "loss": 0.3644,
+      "step": 4362
+    },
+    {
+      "epoch": 1.0385553638364966,
+      "grad_norm": 0.34238196519140435,
+      "learning_rate": 4.933499968697503e-06,
+      "loss": 0.3405,
+      "step": 4363
+    },
+    {
+      "epoch": 1.0387933599095616,
+      "grad_norm": 0.36795318780076897,
+      "learning_rate": 4.931572550209983e-06,
+      "loss": 0.2819,
+      "step": 4364
+    },
+    {
+      "epoch": 1.0390313559826263,
+      "grad_norm": 0.3974431339386504,
+      "learning_rate": 4.929645141892475e-06,
+      "loss": 0.3485,
+      "step": 4365
+    },
+    {
+      "epoch": 1.039269352055691,
+      "grad_norm": 0.3861629647354368,
+      "learning_rate": 4.92771774403144e-06,
+      "loss": 0.4002,
+      "step": 4366
+    },
+    {
+      "epoch": 1.0395073481287558,
+      "grad_norm": 0.3874473883648921,
+      "learning_rate": 4.925790356913337e-06,
+      "loss": 0.3324,
+      "step": 4367
+    },
+    {
+      "epoch": 1.0397453442018207,
+      "grad_norm": 0.39344198259301555,
+      "learning_rate": 4.923862980824622e-06,
+      "loss": 0.2779,
+      "step": 4368
+    },
+    {
+      "epoch": 1.0399833402748855,
+      "grad_norm": 0.4195324906692714,
+      "learning_rate": 4.921935616051751e-06,
+      "loss": 0.3325,
+      "step": 4369
+    },
+    {
+      "epoch": 1.0402213363479502,
+      "grad_norm": 0.3908083748851083,
+      "learning_rate": 4.920008262881177e-06,
+      "loss": 0.3987,
+      "step": 4370
+    },
+    {
+      "epoch": 1.040459332421015,
+      "grad_norm": 0.3951373300669574,
+      "learning_rate": 4.918080921599356e-06,
+      "loss": 0.2903,
+      "step": 4371
+    },
+    {
+      "epoch": 1.04069732849408,
+      "grad_norm": 0.40344550596748047,
+      "learning_rate": 4.9161535924927375e-06,
+      "loss": 0.3174,
+      "step": 4372
+    },
+    {
+      "epoch": 1.0409353245671447,
+      "grad_norm": 0.3856931317543428,
+      "learning_rate": 4.914226275847768e-06,
+      "loss": 0.3455,
+      "step": 4373
+    },
+    {
+      "epoch": 1.0411733206402094,
+      "grad_norm": 0.36331854787145207,
+      "learning_rate": 4.912298971950897e-06,
+      "loss": 0.3507,
+      "step": 4374
+    },
+    {
+      "epoch": 1.0414113167132741,
+      "grad_norm": 0.370895311159966,
+      "learning_rate": 4.910371681088568e-06,
+      "loss": 0.3087,
+      "step": 4375
+    },
+    {
+      "epoch": 1.041649312786339,
+      "grad_norm": 0.3771524826161335,
+      "learning_rate": 4.908444403547224e-06,
+      "loss": 0.3444,
+      "step": 4376
+    },
+    {
+      "epoch": 1.0418873088594038,
+      "grad_norm": 0.41144922180124954,
+      "learning_rate": 4.906517139613307e-06,
+      "loss": 0.4052,
+      "step": 4377
+    },
+    {
+      "epoch": 1.0421253049324686,
+      "grad_norm": 0.3646312683356659,
+      "learning_rate": 4.9045898895732555e-06,
+      "loss": 0.296,
+      "step": 4378
+    },
+    {
+      "epoch": 1.0423633010055333,
+      "grad_norm": 0.38904303184845496,
+      "learning_rate": 4.902662653713507e-06,
+      "loss": 0.2914,
+      "step": 4379
+    },
+    {
+      "epoch": 1.0426012970785983,
+      "grad_norm": 0.38168935291772754,
+      "learning_rate": 4.900735432320496e-06,
+      "loss": 0.3709,
+      "step": 4380
+    },
+    {
+      "epoch": 1.042839293151663,
+      "grad_norm": 0.3730634093577279,
+      "learning_rate": 4.898808225680656e-06,
+      "loss": 0.358,
+      "step": 4381
+    },
+    {
+      "epoch": 1.0430772892247278,
+      "grad_norm": 0.38559125950569123,
+      "learning_rate": 4.8968810340804166e-06,
+      "loss": 0.2736,
+      "step": 4382
+    },
+    {
+      "epoch": 1.0433152852977925,
+      "grad_norm": 0.38852084763466593,
+      "learning_rate": 4.894953857806207e-06,
+      "loss": 0.3232,
+      "step": 4383
+    },
+    {
+      "epoch": 1.0435532813708575,
+      "grad_norm": 0.3922704604899118,
+      "learning_rate": 4.893026697144451e-06,
+      "loss": 0.3953,
+      "step": 4384
+    },
+    {
+      "epoch": 1.0437912774439222,
+      "grad_norm": 0.3464747600169571,
+      "learning_rate": 4.891099552381575e-06,
+      "loss": 0.3247,
+      "step": 4385
+    },
+    {
+      "epoch": 1.044029273516987,
+      "grad_norm": 0.3814240246283631,
+      "learning_rate": 4.8891724238040004e-06,
+      "loss": 0.2907,
+      "step": 4386
+    },
+    {
+      "epoch": 1.0442672695900517,
+      "grad_norm": 0.3771293957546087,
+      "learning_rate": 4.887245311698146e-06,
+      "loss": 0.355,
+      "step": 4387
+    },
+    {
+      "epoch": 1.0445052656631166,
+      "grad_norm": 0.3868624148021604,
+      "learning_rate": 4.8853182163504265e-06,
+      "loss": 0.3455,
+      "step": 4388
+    },
+    {
+      "epoch": 1.0447432617361814,
+      "grad_norm": 0.4023587442678937,
+      "learning_rate": 4.883391138047258e-06,
+      "loss": 0.3055,
+      "step": 4389
+    },
+    {
+      "epoch": 1.0449812578092461,
+      "grad_norm": 0.3743342358321569,
+      "learning_rate": 4.8814640770750495e-06,
+      "loss": 0.3441,
+      "step": 4390
+    },
+    {
+      "epoch": 1.0452192538823109,
+      "grad_norm": 0.3908606318397311,
+      "learning_rate": 4.879537033720212e-06,
+      "loss": 0.3837,
+      "step": 4391
+    },
+    {
+      "epoch": 1.0454572499553758,
+      "grad_norm": 0.3927864891120312,
+      "learning_rate": 4.877610008269153e-06,
+      "loss": 0.3233,
+      "step": 4392
+    },
+    {
+      "epoch": 1.0456952460284406,
+      "grad_norm": 0.37891845327252005,
+      "learning_rate": 4.875683001008274e-06,
+      "loss": 0.3127,
+      "step": 4393
+    },
+    {
+      "epoch": 1.0459332421015053,
+      "grad_norm": 0.40161470551225587,
+      "learning_rate": 4.873756012223977e-06,
+      "loss": 0.3332,
+      "step": 4394
+    },
+    {
+      "epoch": 1.04617123817457,
+      "grad_norm": 0.436580051432871,
+      "learning_rate": 4.871829042202658e-06,
+      "loss": 0.3743,
+      "step": 4395
+    },
+    {
+      "epoch": 1.046409234247635,
+      "grad_norm": 0.4008401671583846,
+      "learning_rate": 4.8699020912307155e-06,
+      "loss": 0.303,
+      "step": 4396
+    },
+    {
+      "epoch": 1.0466472303206997,
+      "grad_norm": 0.39091168851981944,
+      "learning_rate": 4.86797515959454e-06,
+      "loss": 0.3068,
+      "step": 4397
+    },
+    {
+      "epoch": 1.0468852263937645,
+      "grad_norm": 0.4171630119960414,
+      "learning_rate": 4.866048247580521e-06,
+      "loss": 0.3736,
+      "step": 4398
+    },
+    {
+      "epoch": 1.0471232224668292,
+      "grad_norm": 0.3830177263756179,
+      "learning_rate": 4.864121355475047e-06,
+      "loss": 0.3746,
+      "step": 4399
+    },
+    {
+      "epoch": 1.0473612185398942,
+      "grad_norm": 0.37078719399047205,
+      "learning_rate": 4.862194483564501e-06,
+      "loss": 0.2977,
+      "step": 4400
+    },
+    {
+      "epoch": 1.047599214612959,
+      "grad_norm": 0.36764015487272417,
+      "learning_rate": 4.8602676321352646e-06,
+      "loss": 0.3353,
+      "step": 4401
+    },
+    {
+      "epoch": 1.0478372106860236,
+      "grad_norm": 0.3791879193271118,
+      "learning_rate": 4.858340801473715e-06,
+      "loss": 0.3836,
+      "step": 4402
+    },
+    {
+      "epoch": 1.0480752067590884,
+      "grad_norm": 0.34908273436476017,
+      "learning_rate": 4.856413991866225e-06,
+      "loss": 0.2991,
+      "step": 4403
+    },
+    {
+      "epoch": 1.0483132028321533,
+      "grad_norm": 0.3726740932651389,
+      "learning_rate": 4.85448720359917e-06,
+      "loss": 0.3106,
+      "step": 4404
+    },
+    {
+      "epoch": 1.048551198905218,
+      "grad_norm": 0.4036355116456374,
+      "learning_rate": 4.852560436958916e-06,
+      "loss": 0.3737,
+      "step": 4405
+    },
+    {
+      "epoch": 1.0487891949782828,
+      "grad_norm": 0.36697835588631594,
+      "learning_rate": 4.850633692231828e-06,
+      "loss": 0.3501,
+      "step": 4406
+    },
+    {
+      "epoch": 1.0490271910513476,
+      "grad_norm": 0.36549398149960405,
+      "learning_rate": 4.848706969704269e-06,
+      "loss": 0.285,
+      "step": 4407
+    },
+    {
+      "epoch": 1.0492651871244125,
+      "grad_norm": 0.38262831908848516,
+      "learning_rate": 4.846780269662597e-06,
+      "loss": 0.3424,
+      "step": 4408
+    },
+    {
+      "epoch": 1.0495031831974773,
+      "grad_norm": 0.427528345212208,
+      "learning_rate": 4.8448535923931675e-06,
+      "loss": 0.3711,
+      "step": 4409
+    },
+    {
+      "epoch": 1.049741179270542,
+      "grad_norm": 0.35527661752034534,
+      "learning_rate": 4.842926938182332e-06,
+      "loss": 0.3213,
+      "step": 4410
+    },
+    {
+      "epoch": 1.0499791753436067,
+      "grad_norm": 0.40048394169356094,
+      "learning_rate": 4.84100030731644e-06,
+      "loss": 0.2941,
+      "step": 4411
+    },
+    {
+      "epoch": 1.0502171714166717,
+      "grad_norm": 0.3892299648187431,
+      "learning_rate": 4.8390737000818326e-06,
+      "loss": 0.3235,
+      "step": 4412
+    },
+    {
+      "epoch": 1.0504551674897364,
+      "grad_norm": 0.3772914101337546,
+      "learning_rate": 4.837147116764857e-06,
+      "loss": 0.346,
+      "step": 4413
+    },
+    {
+      "epoch": 1.0506931635628012,
+      "grad_norm": 0.37840783896597235,
+      "learning_rate": 4.835220557651849e-06,
+      "loss": 0.2795,
+      "step": 4414
+    },
+    {
+      "epoch": 1.050931159635866,
+      "grad_norm": 0.38825533133807016,
+      "learning_rate": 4.833294023029142e-06,
+      "loss": 0.3024,
+      "step": 4415
+    },
+    {
+      "epoch": 1.0511691557089309,
+      "grad_norm": 0.3865655826669927,
+      "learning_rate": 4.831367513183068e-06,
+      "loss": 0.3913,
+      "step": 4416
+    },
+    {
+      "epoch": 1.0514071517819956,
+      "grad_norm": 0.35292818007412385,
+      "learning_rate": 4.829441028399952e-06,
+      "loss": 0.322,
+      "step": 4417
+    },
+    {
+      "epoch": 1.0516451478550604,
+      "grad_norm": 0.3668531742705322,
+      "learning_rate": 4.827514568966119e-06,
+      "loss": 0.298,
+      "step": 4418
+    },
+    {
+      "epoch": 1.051883143928125,
+      "grad_norm": 0.3749278980780594,
+      "learning_rate": 4.825588135167889e-06,
+      "loss": 0.3288,
+      "step": 4419
+    },
+    {
+      "epoch": 1.05212114000119,
+      "grad_norm": 0.3622324145932679,
+      "learning_rate": 4.823661727291577e-06,
+      "loss": 0.389,
+      "step": 4420
+    },
+    {
+      "epoch": 1.0523591360742548,
+      "grad_norm": 0.396752910815969,
+      "learning_rate": 4.821735345623494e-06,
+      "loss": 0.2953,
+      "step": 4421
+    },
+    {
+      "epoch": 1.0525971321473195,
+      "grad_norm": 0.3758408909876523,
+      "learning_rate": 4.819808990449949e-06,
+      "loss": 0.303,
+      "step": 4422
+    },
+    {
+      "epoch": 1.0528351282203843,
+      "grad_norm": 0.4081539657673205,
+      "learning_rate": 4.817882662057246e-06,
+      "loss": 0.3709,
+      "step": 4423
+    },
+    {
+      "epoch": 1.0530731242934492,
+      "grad_norm": 0.349815937390151,
+      "learning_rate": 4.815956360731684e-06,
+      "loss": 0.349,
+      "step": 4424
+    },
+    {
+      "epoch": 1.053311120366514,
+      "grad_norm": 0.34903612236748616,
+      "learning_rate": 4.814030086759561e-06,
+      "loss": 0.2885,
+      "step": 4425
+    },
+    {
+      "epoch": 1.0535491164395787,
+      "grad_norm": 0.39164825325025976,
+      "learning_rate": 4.812103840427165e-06,
+      "loss": 0.3512,
+      "step": 4426
+    },
+    {
+      "epoch": 1.0537871125126435,
+      "grad_norm": 0.36669475356586617,
+      "learning_rate": 4.810177622020788e-06,
+      "loss": 0.3903,
+      "step": 4427
+    },
+    {
+      "epoch": 1.0540251085857084,
+      "grad_norm": 0.3504650544959202,
+      "learning_rate": 4.808251431826713e-06,
+      "loss": 0.2901,
+      "step": 4428
+    },
+    {
+      "epoch": 1.0542631046587732,
+      "grad_norm": 0.34674534781413774,
+      "learning_rate": 4.806325270131219e-06,
+      "loss": 0.2899,
+      "step": 4429
+    },
+    {
+      "epoch": 1.054501100731838,
+      "grad_norm": 0.3777786304373956,
+      "learning_rate": 4.80439913722058e-06,
+      "loss": 0.3304,
+      "step": 4430
+    },
+    {
+      "epoch": 1.0547390968049026,
+      "grad_norm": 0.37181874792865827,
+      "learning_rate": 4.802473033381069e-06,
+      "loss": 0.4006,
+      "step": 4431
+    },
+    {
+      "epoch": 1.0549770928779676,
+      "grad_norm": 0.38052787907987795,
+      "learning_rate": 4.800546958898952e-06,
+      "loss": 0.3153,
+      "step": 4432
+    },
+    {
+      "epoch": 1.0552150889510323,
+      "grad_norm": 0.40481761082909645,
+      "learning_rate": 4.798620914060492e-06,
+      "loss": 0.3188,
+      "step": 4433
+    },
+    {
+      "epoch": 1.055453085024097,
+      "grad_norm": 0.3588854522073126,
+      "learning_rate": 4.7966948991519446e-06,
+      "loss": 0.3802,
+      "step": 4434
+    },
+    {
+      "epoch": 1.0556910810971618,
+      "grad_norm": 0.3773596458270944,
+      "learning_rate": 4.794768914459565e-06,
+      "loss": 0.3467,
+      "step": 4435
+    },
+    {
+      "epoch": 1.0559290771702268,
+      "grad_norm": 0.4156895917585443,
+      "learning_rate": 4.792842960269603e-06,
+      "loss": 0.3065,
+      "step": 4436
+    },
+    {
+      "epoch": 1.0561670732432915,
+      "grad_norm": 0.3609593155405436,
+      "learning_rate": 4.790917036868301e-06,
+      "loss": 0.3456,
+      "step": 4437
+    },
+    {
+      "epoch": 1.0564050693163562,
+      "grad_norm": 0.37963675817335796,
+      "learning_rate": 4.7889911445419e-06,
+      "loss": 0.3859,
+      "step": 4438
+    },
+    {
+      "epoch": 1.056643065389421,
+      "grad_norm": 0.38339295139402263,
+      "learning_rate": 4.787065283576633e-06,
+      "loss": 0.3096,
+      "step": 4439
+    },
+    {
+      "epoch": 1.056881061462486,
+      "grad_norm": 0.4148001262695314,
+      "learning_rate": 4.785139454258734e-06,
+      "loss": 0.3063,
+      "step": 4440
+    },
+    {
+      "epoch": 1.0571190575355507,
+      "grad_norm": 0.3849351006849719,
+      "learning_rate": 4.783213656874428e-06,
+      "loss": 0.354,
+      "step": 4441
+    },
+    {
+      "epoch": 1.0573570536086154,
+      "grad_norm": 0.3904733075529352,
+      "learning_rate": 4.781287891709936e-06,
+      "loss": 0.3307,
+      "step": 4442
+    },
+    {
+      "epoch": 1.0575950496816802,
+      "grad_norm": 0.39073356763740286,
+      "learning_rate": 4.779362159051474e-06,
+      "loss": 0.3202,
+      "step": 4443
+    },
+    {
+      "epoch": 1.0578330457547451,
+      "grad_norm": 0.382841095559944,
+      "learning_rate": 4.777436459185252e-06,
+      "loss": 0.3363,
+      "step": 4444
+    },
+    {
+      "epoch": 1.0580710418278099,
+      "grad_norm": 0.37809442113241104,
+      "learning_rate": 4.775510792397479e-06,
+      "loss": 0.365,
+      "step": 4445
+    },
+    {
+      "epoch": 1.0583090379008746,
+      "grad_norm": 0.36463101257216596,
+      "learning_rate": 4.773585158974356e-06,
+      "loss": 0.3039,
+      "step": 4446
+    },
+    {
+      "epoch": 1.0585470339739393,
+      "grad_norm": 0.4116743791608411,
+      "learning_rate": 4.77165955920208e-06,
+      "loss": 0.3284,
+      "step": 4447
+    },
+    {
+      "epoch": 1.0587850300470043,
+      "grad_norm": 0.42410933724473493,
+      "learning_rate": 4.769733993366842e-06,
+      "loss": 0.3708,
+      "step": 4448
+    },
+    {
+      "epoch": 1.059023026120069,
+      "grad_norm": 0.37205012492946415,
+      "learning_rate": 4.76780846175483e-06,
+      "loss": 0.3242,
+      "step": 4449
+    },
+    {
+      "epoch": 1.0592610221931338,
+      "grad_norm": 0.3472944005021577,
+      "learning_rate": 4.765882964652223e-06,
+      "loss": 0.2834,
+      "step": 4450
+    },
+    {
+      "epoch": 1.0594990182661985,
+      "grad_norm": 0.420700058102123,
+      "learning_rate": 4.7639575023452e-06,
+      "loss": 0.3451,
+      "step": 4451
+    },
+    {
+      "epoch": 1.0597370143392635,
+      "grad_norm": 0.3850537500639467,
+      "learning_rate": 4.762032075119932e-06,
+      "loss": 0.4028,
+      "step": 4452
+    },
+    {
+      "epoch": 1.0599750104123282,
+      "grad_norm": 0.35606255517837127,
+      "learning_rate": 4.760106683262582e-06,
+      "loss": 0.3206,
+      "step": 4453
+    },
+    {
+      "epoch": 1.060213006485393,
+      "grad_norm": 0.4207878717161147,
+      "learning_rate": 4.758181327059316e-06,
+      "loss": 0.2905,
+      "step": 4454
+    },
+    {
+      "epoch": 1.0604510025584577,
+      "grad_norm": 0.39178183992943744,
+      "learning_rate": 4.756256006796287e-06,
+      "loss": 0.3826,
+      "step": 4455
+    },
+    {
+      "epoch": 1.0606889986315227,
+      "grad_norm": 0.3829442776223385,
+      "learning_rate": 4.754330722759645e-06,
+      "loss": 0.3628,
+      "step": 4456
+    },
+    {
+      "epoch": 1.0609269947045874,
+      "grad_norm": 0.3954498442186975,
+      "learning_rate": 4.7524054752355345e-06,
+      "loss": 0.3066,
+      "step": 4457
+    },
+    {
+      "epoch": 1.0611649907776521,
+      "grad_norm": 0.393762593771785,
+      "learning_rate": 4.750480264510097e-06,
+      "loss": 0.3186,
+      "step": 4458
+    },
+    {
+      "epoch": 1.0614029868507169,
+      "grad_norm": 0.36465796218337193,
+      "learning_rate": 4.748555090869464e-06,
+      "loss": 0.3876,
+      "step": 4459
+    },
+    {
+      "epoch": 1.0616409829237818,
+      "grad_norm": 0.3664333642625829,
+      "learning_rate": 4.746629954599766e-06,
+      "loss": 0.3216,
+      "step": 4460
+    },
+    {
+      "epoch": 1.0618789789968466,
+      "grad_norm": 0.37283914856358796,
+      "learning_rate": 4.744704855987125e-06,
+      "loss": 0.2674,
+      "step": 4461
+    },
+    {
+      "epoch": 1.0621169750699113,
+      "grad_norm": 0.3929643060999283,
+      "learning_rate": 4.742779795317657e-06,
+      "loss": 0.3625,
+      "step": 4462
+    },
+    {
+      "epoch": 1.062354971142976,
+      "grad_norm": 0.3639867047560858,
+      "learning_rate": 4.740854772877475e-06,
+      "loss": 0.3875,
+      "step": 4463
+    },
+    {
+      "epoch": 1.062592967216041,
+      "grad_norm": 0.34468157426611506,
+      "learning_rate": 4.738929788952685e-06,
+      "loss": 0.3047,
+      "step": 4464
+    },
+    {
+      "epoch": 1.0628309632891058,
+      "grad_norm": 0.407693623350532,
+      "learning_rate": 4.737004843829387e-06,
+      "loss": 0.3164,
+      "step": 4465
+    },
+    {
+      "epoch": 1.0630689593621705,
+      "grad_norm": 0.3953104625961449,
+      "learning_rate": 4.735079937793675e-06,
+      "loss": 0.3848,
+      "step": 4466
+    },
+    {
+      "epoch": 1.0633069554352352,
+      "grad_norm": 0.35620972393509376,
+      "learning_rate": 4.733155071131636e-06,
+      "loss": 0.3204,
+      "step": 4467
+    },
+    {
+      "epoch": 1.0635449515083002,
+      "grad_norm": 0.37941109344701723,
+      "learning_rate": 4.731230244129357e-06,
+      "loss": 0.2926,
+      "step": 4468
+    },
+    {
+      "epoch": 1.063782947581365,
+      "grad_norm": 0.36519045880745366,
+      "learning_rate": 4.729305457072913e-06,
+      "loss": 0.3606,
+      "step": 4469
+    },
+    {
+      "epoch": 1.0640209436544297,
+      "grad_norm": 0.3575723913390525,
+      "learning_rate": 4.727380710248375e-06,
+      "loss": 0.3744,
+      "step": 4470
+    },
+    {
+      "epoch": 1.0642589397274944,
+      "grad_norm": 0.3567610268225815,
+      "learning_rate": 4.725456003941805e-06,
+      "loss": 0.2866,
+      "step": 4471
+    },
+    {
+      "epoch": 1.0644969358005594,
+      "grad_norm": 0.37128027520359136,
+      "learning_rate": 4.723531338439268e-06,
+      "loss": 0.3214,
+      "step": 4472
+    },
+    {
+      "epoch": 1.0647349318736241,
+      "grad_norm": 0.3820839576708276,
+      "learning_rate": 4.721606714026812e-06,
+      "loss": 0.3902,
+      "step": 4473
+    },
+    {
+      "epoch": 1.0649729279466889,
+      "grad_norm": 0.4089191938756884,
+      "learning_rate": 4.7196821309904865e-06,
+      "loss": 0.334,
+      "step": 4474
+    },
+    {
+      "epoch": 1.0652109240197536,
+      "grad_norm": 0.3728452645121599,
+      "learning_rate": 4.717757589616331e-06,
+      "loss": 0.2551,
+      "step": 4475
+    },
+    {
+      "epoch": 1.0654489200928186,
+      "grad_norm": 0.37511700011909915,
+      "learning_rate": 4.715833090190379e-06,
+      "loss": 0.3443,
+      "step": 4476
+    },
+    {
+      "epoch": 1.0656869161658833,
+      "grad_norm": 0.37701885649521155,
+      "learning_rate": 4.713908632998661e-06,
+      "loss": 0.3962,
+      "step": 4477
+    },
+    {
+      "epoch": 1.065924912238948,
+      "grad_norm": 0.3616386275749788,
+      "learning_rate": 4.711984218327197e-06,
+      "loss": 0.3166,
+      "step": 4478
+    },
+    {
+      "epoch": 1.0661629083120128,
+      "grad_norm": 0.38456233240401916,
+      "learning_rate": 4.710059846462003e-06,
+      "loss": 0.3029,
+      "step": 4479
+    },
+    {
+      "epoch": 1.0664009043850777,
+      "grad_norm": 0.3789566875155259,
+      "learning_rate": 4.708135517689088e-06,
+      "loss": 0.3717,
+      "step": 4480
+    },
+    {
+      "epoch": 1.0666389004581425,
+      "grad_norm": 0.3768525328415589,
+      "learning_rate": 4.706211232294456e-06,
+      "loss": 0.3438,
+      "step": 4481
+    },
+    {
+      "epoch": 1.0668768965312072,
+      "grad_norm": 0.3783650867548586,
+      "learning_rate": 4.704286990564103e-06,
+      "loss": 0.2877,
+      "step": 4482
+    },
+    {
+      "epoch": 1.067114892604272,
+      "grad_norm": 0.41564337367706855,
+      "learning_rate": 4.702362792784019e-06,
+      "loss": 0.3331,
+      "step": 4483
+    },
+    {
+      "epoch": 1.067352888677337,
+      "grad_norm": 0.43075275803922985,
+      "learning_rate": 4.700438639240186e-06,
+      "loss": 0.3629,
+      "step": 4484
+    },
+    {
+      "epoch": 1.0675908847504016,
+      "grad_norm": 0.36040007984808253,
+      "learning_rate": 4.698514530218584e-06,
+      "loss": 0.3061,
+      "step": 4485
+    },
+    {
+      "epoch": 1.0678288808234664,
+      "grad_norm": 0.404962256464901,
+      "learning_rate": 4.6965904660051804e-06,
+      "loss": 0.3088,
+      "step": 4486
+    },
+    {
+      "epoch": 1.0680668768965311,
+      "grad_norm": 0.3697507889870145,
+      "learning_rate": 4.6946664468859395e-06,
+      "loss": 0.3457,
+      "step": 4487
+    },
+    {
+      "epoch": 1.068304872969596,
+      "grad_norm": 0.4036468754442713,
+      "learning_rate": 4.692742473146818e-06,
+      "loss": 0.3603,
+      "step": 4488
+    },
+    {
+      "epoch": 1.0685428690426608,
+      "grad_norm": 0.4042242479571805,
+      "learning_rate": 4.690818545073767e-06,
+      "loss": 0.2794,
+      "step": 4489
+    },
+    {
+      "epoch": 1.0687808651157256,
+      "grad_norm": 0.4099157906703565,
+      "learning_rate": 4.688894662952729e-06,
+      "loss": 0.3127,
+      "step": 4490
+    },
+    {
+      "epoch": 1.0690188611887903,
+      "grad_norm": 0.3468763641985224,
+      "learning_rate": 4.686970827069639e-06,
+      "loss": 0.3843,
+      "step": 4491
+    },
+    {
+      "epoch": 1.0692568572618553,
+      "grad_norm": 0.37635304439254064,
+      "learning_rate": 4.68504703771043e-06,
+      "loss": 0.3453,
+      "step": 4492
+    },
+    {
+      "epoch": 1.06949485333492,
+      "grad_norm": 0.39528560108604954,
+      "learning_rate": 4.683123295161021e-06,
+      "loss": 0.2812,
+      "step": 4493
+    },
+    {
+      "epoch": 1.0697328494079847,
+      "grad_norm": 0.4283777664794825,
+      "learning_rate": 4.6811995997073285e-06,
+      "loss": 0.3331,
+      "step": 4494
+    },
+    {
+      "epoch": 1.0699708454810495,
+      "grad_norm": 0.3710567512234336,
+      "learning_rate": 4.679275951635264e-06,
+      "loss": 0.3875,
+      "step": 4495
+    },
+    {
+      "epoch": 1.0702088415541144,
+      "grad_norm": 0.36243117057229535,
+      "learning_rate": 4.6773523512307275e-06,
+      "loss": 0.2728,
+      "step": 4496
+    },
+    {
+      "epoch": 1.0704468376271792,
+      "grad_norm": 0.37892611536648607,
+      "learning_rate": 4.675428798779613e-06,
+      "loss": 0.3035,
+      "step": 4497
+    },
+    {
+      "epoch": 1.070684833700244,
+      "grad_norm": 0.36860351230116295,
+      "learning_rate": 4.673505294567809e-06,
+      "loss": 0.3669,
+      "step": 4498
+    },
+    {
+      "epoch": 1.0709228297733087,
+      "grad_norm": 0.36740997263006975,
+      "learning_rate": 4.6715818388811945e-06,
+      "loss": 0.3238,
+      "step": 4499
+    },
+    {
+      "epoch": 1.0711608258463736,
+      "grad_norm": 0.3994604434926032,
+      "learning_rate": 4.669658432005644e-06,
+      "loss": 0.296,
+      "step": 4500
+    },
+    {
+      "epoch": 1.0713988219194384,
+      "grad_norm": 0.3698668250700613,
+      "learning_rate": 4.667735074227024e-06,
+      "loss": 0.3494,
+      "step": 4501
+    },
+    {
+      "epoch": 1.071636817992503,
+      "grad_norm": 0.440292753533519,
+      "learning_rate": 4.66581176583119e-06,
+      "loss": 0.4327,
+      "step": 4502
+    },
+    {
+      "epoch": 1.0718748140655678,
+      "grad_norm": 0.3719781722435344,
+      "learning_rate": 4.663888507103996e-06,
+      "loss": 0.3222,
+      "step": 4503
+    },
+    {
+      "epoch": 1.0721128101386328,
+      "grad_norm": 0.4082136221681287,
+      "learning_rate": 4.6619652983312844e-06,
+      "loss": 0.2975,
+      "step": 4504
+    },
+    {
+      "epoch": 1.0723508062116975,
+      "grad_norm": 0.360658966423447,
+      "learning_rate": 4.660042139798892e-06,
+      "loss": 0.342,
+      "step": 4505
+    },
+    {
+      "epoch": 1.0725888022847623,
+      "grad_norm": 0.3593445634450902,
+      "learning_rate": 4.658119031792648e-06,
+      "loss": 0.3406,
+      "step": 4506
+    },
+    {
+      "epoch": 1.072826798357827,
+      "grad_norm": 0.40080585396608215,
+      "learning_rate": 4.6561959745983724e-06,
+      "loss": 0.2878,
+      "step": 4507
+    },
+    {
+      "epoch": 1.073064794430892,
+      "grad_norm": 0.38452336744074866,
+      "learning_rate": 4.65427296850188e-06,
+      "loss": 0.3205,
+      "step": 4508
+    },
+    {
+      "epoch": 1.0733027905039567,
+      "grad_norm": 0.40724996130914426,
+      "learning_rate": 4.652350013788979e-06,
+      "loss": 0.4083,
+      "step": 4509
+    },
+    {
+      "epoch": 1.0735407865770215,
+      "grad_norm": 0.4014847828369795,
+      "learning_rate": 4.650427110745467e-06,
+      "loss": 0.3173,
+      "step": 4510
+    },
+    {
+      "epoch": 1.0737787826500862,
+      "grad_norm": 0.38549635848058716,
+      "learning_rate": 4.648504259657132e-06,
+      "loss": 0.2915,
+      "step": 4511
+    },
+    {
+      "epoch": 1.0740167787231512,
+      "grad_norm": 0.3852457361423471,
+      "learning_rate": 4.646581460809762e-06,
+      "loss": 0.3348,
+      "step": 4512
+    },
+    {
+      "epoch": 1.074254774796216,
+      "grad_norm": 0.3712671658962982,
+      "learning_rate": 4.644658714489129e-06,
+      "loss": 0.3808,
+      "step": 4513
+    },
+    {
+      "epoch": 1.0744927708692806,
+      "grad_norm": 0.43356148305108094,
+      "learning_rate": 4.642736020981002e-06,
+      "loss": 0.2974,
+      "step": 4514
+    },
+    {
+      "epoch": 1.0747307669423454,
+      "grad_norm": 0.3800025715731251,
+      "learning_rate": 4.64081338057114e-06,
+      "loss": 0.3053,
+      "step": 4515
+    },
+    {
+      "epoch": 1.0749687630154103,
+      "grad_norm": 0.4199497942695732,
+      "learning_rate": 4.638890793545297e-06,
+      "loss": 0.3937,
+      "step": 4516
+    },
+    {
+      "epoch": 1.075206759088475,
+      "grad_norm": 0.35073357189494836,
+      "learning_rate": 4.636968260189214e-06,
+      "loss": 0.3195,
+      "step": 4517
+    },
+    {
+      "epoch": 1.0754447551615398,
+      "grad_norm": 0.4014979008525594,
+      "learning_rate": 4.635045780788629e-06,
+      "loss": 0.2757,
+      "step": 4518
+    },
+    {
+      "epoch": 1.0756827512346046,
+      "grad_norm": 0.4018578448257378,
+      "learning_rate": 4.63312335562927e-06,
+      "loss": 0.3202,
+      "step": 4519
+    },
+    {
+      "epoch": 1.0759207473076695,
+      "grad_norm": 0.3853979552410293,
+      "learning_rate": 4.6312009849968544e-06,
+      "loss": 0.3796,
+      "step": 4520
+    },
+    {
+      "epoch": 1.0761587433807343,
+      "grad_norm": 0.3725529344423733,
+      "learning_rate": 4.629278669177098e-06,
+      "loss": 0.3162,
+      "step": 4521
+    },
+    {
+      "epoch": 1.076396739453799,
+      "grad_norm": 0.3465817848520335,
+      "learning_rate": 4.627356408455701e-06,
+      "loss": 0.2863,
+      "step": 4522
+    },
+    {
+      "epoch": 1.0766347355268637,
+      "grad_norm": 0.3869139089298558,
+      "learning_rate": 4.625434203118362e-06,
+      "loss": 0.3826,
+      "step": 4523
+    },
+    {
+      "epoch": 1.0768727315999287,
+      "grad_norm": 0.3809180722646469,
+      "learning_rate": 4.623512053450767e-06,
+      "loss": 0.3408,
+      "step": 4524
+    },
+    {
+      "epoch": 1.0771107276729934,
+      "grad_norm": 0.3703176815754259,
+      "learning_rate": 4.621589959738593e-06,
+      "loss": 0.3108,
+      "step": 4525
+    },
+    {
+      "epoch": 1.0773487237460582,
+      "grad_norm": 0.37759518980753926,
+      "learning_rate": 4.619667922267514e-06,
+      "loss": 0.3366,
+      "step": 4526
+    },
+    {
+      "epoch": 1.077586719819123,
+      "grad_norm": 0.3898155012281468,
+      "learning_rate": 4.617745941323189e-06,
+      "loss": 0.4133,
+      "step": 4527
+    },
+    {
+      "epoch": 1.0778247158921879,
+      "grad_norm": 0.36255189396131254,
+      "learning_rate": 4.615824017191275e-06,
+      "loss": 0.3031,
+      "step": 4528
+    },
+    {
+      "epoch": 1.0780627119652526,
+      "grad_norm": 0.3691335793651894,
+      "learning_rate": 4.613902150157416e-06,
+      "loss": 0.2816,
+      "step": 4529
+    },
+    {
+      "epoch": 1.0783007080383173,
+      "grad_norm": 0.39196606245884325,
+      "learning_rate": 4.61198034050725e-06,
+      "loss": 0.3632,
+      "step": 4530
+    },
+    {
+      "epoch": 1.078538704111382,
+      "grad_norm": 0.39160631748508107,
+      "learning_rate": 4.610058588526404e-06,
+      "loss": 0.3471,
+      "step": 4531
+    },
+    {
+      "epoch": 1.078776700184447,
+      "grad_norm": 0.3624974985968683,
+      "learning_rate": 4.6081368945004976e-06,
+      "loss": 0.301,
+      "step": 4532
+    },
+    {
+      "epoch": 1.0790146962575118,
+      "grad_norm": 0.3921259171027228,
+      "learning_rate": 4.606215258715144e-06,
+      "loss": 0.3611,
+      "step": 4533
+    },
+    {
+      "epoch": 1.0792526923305765,
+      "grad_norm": 0.38270676417966176,
+      "learning_rate": 4.604293681455942e-06,
+      "loss": 0.3717,
+      "step": 4534
+    },
+    {
+      "epoch": 1.0794906884036413,
+      "grad_norm": 0.4878601599757468,
+      "learning_rate": 4.602372163008491e-06,
+      "loss": 0.3385,
+      "step": 4535
+    },
+    {
+      "epoch": 1.0797286844767062,
+      "grad_norm": 0.4258797436561885,
+      "learning_rate": 4.6004507036583714e-06,
+      "loss": 0.3118,
+      "step": 4536
+    },
+    {
+      "epoch": 1.079966680549771,
+      "grad_norm": 0.36842771403409225,
+      "learning_rate": 4.598529303691163e-06,
+      "loss": 0.354,
+      "step": 4537
+    },
+    {
+      "epoch": 1.0802046766228357,
+      "grad_norm": 0.38395032278946645,
+      "learning_rate": 4.596607963392431e-06,
+      "loss": 0.3878,
+      "step": 4538
+    },
+    {
+      "epoch": 1.0804426726959004,
+      "grad_norm": 0.37424827121079335,
+      "learning_rate": 4.594686683047736e-06,
+      "loss": 0.3127,
+      "step": 4539
+    },
+    {
+      "epoch": 1.0806806687689654,
+      "grad_norm": 0.36463462903952687,
+      "learning_rate": 4.592765462942627e-06,
+      "loss": 0.3135,
+      "step": 4540
+    },
+    {
+      "epoch": 1.0809186648420301,
+      "grad_norm": 0.4010499621374228,
+      "learning_rate": 4.590844303362645e-06,
+      "loss": 0.4033,
+      "step": 4541
+    },
+    {
+      "epoch": 1.0811566609150949,
+      "grad_norm": 0.35721252951546795,
+      "learning_rate": 4.5889232045933204e-06,
+      "loss": 0.3467,
+      "step": 4542
+    },
+    {
+      "epoch": 1.0813946569881596,
+      "grad_norm": 0.3883975133944084,
+      "learning_rate": 4.587002166920178e-06,
+      "loss": 0.2614,
+      "step": 4543
+    },
+    {
+      "epoch": 1.0816326530612246,
+      "grad_norm": 0.3644825130668685,
+      "learning_rate": 4.5850811906287315e-06,
+      "loss": 0.3605,
+      "step": 4544
+    },
+    {
+      "epoch": 1.0818706491342893,
+      "grad_norm": 0.3839532185379082,
+      "learning_rate": 4.583160276004483e-06,
+      "loss": 0.3859,
+      "step": 4545
+    },
+    {
+      "epoch": 1.082108645207354,
+      "grad_norm": 0.37149271895145186,
+      "learning_rate": 4.58123942333293e-06,
+      "loss": 0.2943,
+      "step": 4546
+    },
+    {
+      "epoch": 1.0823466412804188,
+      "grad_norm": 0.386864449700323,
+      "learning_rate": 4.5793186328995585e-06,
+      "loss": 0.3084,
+      "step": 4547
+    },
+    {
+      "epoch": 1.0825846373534838,
+      "grad_norm": 0.36399312445382515,
+      "learning_rate": 4.577397904989846e-06,
+      "loss": 0.3798,
+      "step": 4548
+    },
+    {
+      "epoch": 1.0828226334265485,
+      "grad_norm": 0.4192409273041034,
+      "learning_rate": 4.575477239889258e-06,
+      "loss": 0.359,
+      "step": 4549
+    },
+    {
+      "epoch": 1.0830606294996132,
+      "grad_norm": 0.4118535694398554,
+      "learning_rate": 4.5735566378832545e-06,
+      "loss": 0.3003,
+      "step": 4550
+    },
+    {
+      "epoch": 1.083298625572678,
+      "grad_norm": 0.39405326883837694,
+      "learning_rate": 4.571636099257285e-06,
+      "loss": 0.3287,
+      "step": 4551
+    },
+    {
+      "epoch": 1.083536621645743,
+      "grad_norm": 0.4282411602854888,
+      "learning_rate": 4.569715624296788e-06,
+      "loss": 0.3933,
+      "step": 4552
+    },
+    {
+      "epoch": 1.0837746177188077,
+      "grad_norm": 0.3646511346959913,
+      "learning_rate": 4.567795213287194e-06,
+      "loss": 0.3013,
+      "step": 4553
+    },
+    {
+      "epoch": 1.0840126137918724,
+      "grad_norm": 0.39611556493858835,
+      "learning_rate": 4.565874866513924e-06,
+      "loss": 0.3023,
+      "step": 4554
+    },
+    {
+      "epoch": 1.0842506098649372,
+      "grad_norm": 0.41873346499425335,
+      "learning_rate": 4.563954584262388e-06,
+      "loss": 0.3308,
+      "step": 4555
+    },
+    {
+      "epoch": 1.0844886059380021,
+      "grad_norm": 0.37895952135841016,
+      "learning_rate": 4.562034366817989e-06,
+      "loss": 0.3503,
+      "step": 4556
+    },
+    {
+      "epoch": 1.0847266020110669,
+      "grad_norm": 0.3755212007026155,
+      "learning_rate": 4.560114214466118e-06,
+      "loss": 0.2777,
+      "step": 4557
+    },
+    {
+      "epoch": 1.0849645980841316,
+      "grad_norm": 0.3686507710984981,
+      "learning_rate": 4.558194127492156e-06,
+      "loss": 0.3146,
+      "step": 4558
+    },
+    {
+      "epoch": 1.0852025941571963,
+      "grad_norm": 0.37532310578022504,
+      "learning_rate": 4.556274106181477e-06,
+      "loss": 0.4073,
+      "step": 4559
+    },
+    {
+      "epoch": 1.0854405902302613,
+      "grad_norm": 0.38070453470704196,
+      "learning_rate": 4.554354150819442e-06,
+      "loss": 0.3069,
+      "step": 4560
+    },
+    {
+      "epoch": 1.085678586303326,
+      "grad_norm": 0.38448824696456024,
+      "learning_rate": 4.552434261691405e-06,
+      "loss": 0.2876,
+      "step": 4561
+    },
+    {
+      "epoch": 1.0859165823763908,
+      "grad_norm": 0.4030821070519119,
+      "learning_rate": 4.55051443908271e-06,
+      "loss": 0.3773,
+      "step": 4562
+    },
+    {
+      "epoch": 1.0861545784494555,
+      "grad_norm": 0.35298371324439093,
+      "learning_rate": 4.5485946832786885e-06,
+      "loss": 0.3785,
+      "step": 4563
+    },
+    {
+      "epoch": 1.0863925745225205,
+      "grad_norm": 0.36012679234114514,
+      "learning_rate": 4.546674994564664e-06,
+      "loss": 0.3021,
+      "step": 4564
+    },
+    {
+      "epoch": 1.0866305705955852,
+      "grad_norm": 0.41175830720140366,
+      "learning_rate": 4.544755373225949e-06,
+      "loss": 0.3291,
+      "step": 4565
+    },
+    {
+      "epoch": 1.08686856666865,
+      "grad_norm": 0.3729710973827501,
+      "learning_rate": 4.54283581954785e-06,
+      "loss": 0.3917,
+      "step": 4566
+    },
+    {
+      "epoch": 1.0871065627417147,
+      "grad_norm": 0.38900125919653233,
+      "learning_rate": 4.540916333815658e-06,
+      "loss": 0.3425,
+      "step": 4567
+    },
+    {
+      "epoch": 1.0873445588147796,
+      "grad_norm": 0.4185005530339485,
+      "learning_rate": 4.5389969163146544e-06,
+      "loss": 0.3012,
+      "step": 4568
+    },
+    {
+      "epoch": 1.0875825548878444,
+      "grad_norm": 0.4088166803824769,
+      "learning_rate": 4.537077567330115e-06,
+      "loss": 0.3202,
+      "step": 4569
+    },
+    {
+      "epoch": 1.0878205509609091,
+      "grad_norm": 0.39177237999984393,
+      "learning_rate": 4.535158287147301e-06,
+      "loss": 0.3713,
+      "step": 4570
+    },
+    {
+      "epoch": 1.0880585470339739,
+      "grad_norm": 0.3853574514336472,
+      "learning_rate": 4.533239076051465e-06,
+      "loss": 0.3007,
+      "step": 4571
+    },
+    {
+      "epoch": 1.0882965431070388,
+      "grad_norm": 0.43082485652705493,
+      "learning_rate": 4.531319934327849e-06,
+      "loss": 0.3271,
+      "step": 4572
+    },
+    {
+      "epoch": 1.0885345391801036,
+      "grad_norm": 0.40152293574128467,
+      "learning_rate": 4.529400862261686e-06,
+      "loss": 0.3935,
+      "step": 4573
+    },
+    {
+      "epoch": 1.0887725352531683,
+      "grad_norm": 0.3820747241972296,
+      "learning_rate": 4.527481860138196e-06,
+      "loss": 0.3331,
+      "step": 4574
+    },
+    {
+      "epoch": 1.089010531326233,
+      "grad_norm": 0.3766358656424945,
+      "learning_rate": 4.525562928242592e-06,
+      "loss": 0.3045,
+      "step": 4575
+    },
+    {
+      "epoch": 1.089248527399298,
+      "grad_norm": 0.4237754758083936,
+      "learning_rate": 4.523644066860074e-06,
+      "loss": 0.3531,
+      "step": 4576
+    },
+    {
+      "epoch": 1.0894865234723627,
+      "grad_norm": 0.4304790203953773,
+      "learning_rate": 4.52172527627583e-06,
+      "loss": 0.3691,
+      "step": 4577
+    },
+    {
+      "epoch": 1.0897245195454275,
+      "grad_norm": 0.35559337217271153,
+      "learning_rate": 4.519806556775043e-06,
+      "loss": 0.3105,
+      "step": 4578
+    },
+    {
+      "epoch": 1.0899625156184922,
+      "grad_norm": 0.36470318443377925,
+      "learning_rate": 4.517887908642882e-06,
+      "loss": 0.2944,
+      "step": 4579
+    },
+    {
+      "epoch": 1.0902005116915572,
+      "grad_norm": 0.3917075725340582,
+      "learning_rate": 4.515969332164504e-06,
+      "loss": 0.3561,
+      "step": 4580
+    },
+    {
+      "epoch": 1.090438507764622,
+      "grad_norm": 0.3706508136005053,
+      "learning_rate": 4.514050827625058e-06,
+      "loss": 0.3862,
+      "step": 4581
+    },
+    {
+      "epoch": 1.0906765038376867,
+      "grad_norm": 0.38950829726986785,
+      "learning_rate": 4.512132395309681e-06,
+      "loss": 0.2885,
+      "step": 4582
+    },
+    {
+      "epoch": 1.0909144999107514,
+      "grad_norm": 0.3771455973737931,
+      "learning_rate": 4.510214035503499e-06,
+      "loss": 0.3464,
+      "step": 4583
+    },
+    {
+      "epoch": 1.0911524959838164,
+      "grad_norm": 0.36988427743457747,
+      "learning_rate": 4.508295748491628e-06,
+      "loss": 0.4167,
+      "step": 4584
+    },
+    {
+      "epoch": 1.091390492056881,
+      "grad_norm": 0.35658085377837206,
+      "learning_rate": 4.506377534559174e-06,
+      "loss": 0.3147,
+      "step": 4585
+    },
+    {
+      "epoch": 1.0916284881299458,
+      "grad_norm": 0.3761893075654618,
+      "learning_rate": 4.504459393991229e-06,
+      "loss": 0.2916,
+      "step": 4586
+    },
+    {
+      "epoch": 1.0918664842030106,
+      "grad_norm": 0.3778558101407558,
+      "learning_rate": 4.502541327072877e-06,
+      "loss": 0.3372,
+      "step": 4587
+    },
+    {
+      "epoch": 1.0921044802760755,
+      "grad_norm": 0.3819227496695729,
+      "learning_rate": 4.50062333408919e-06,
+      "loss": 0.4006,
+      "step": 4588
+    },
+    {
+      "epoch": 1.0923424763491403,
+      "grad_norm": 0.4723091940072908,
+      "learning_rate": 4.49870541532523e-06,
+      "loss": 0.3153,
+      "step": 4589
+    },
+    {
+      "epoch": 1.092580472422205,
+      "grad_norm": 0.45016072076415764,
+      "learning_rate": 4.496787571066047e-06,
+      "loss": 0.3088,
+      "step": 4590
+    },
+    {
+      "epoch": 1.0928184684952698,
+      "grad_norm": 0.3866045379739883,
+      "learning_rate": 4.494869801596679e-06,
+      "loss": 0.3687,
+      "step": 4591
+    },
+    {
+      "epoch": 1.0930564645683347,
+      "grad_norm": 0.3739372092733529,
+      "learning_rate": 4.492952107202154e-06,
+      "loss": 0.2934,
+      "step": 4592
+    },
+    {
+      "epoch": 1.0932944606413995,
+      "grad_norm": 0.35295097365438255,
+      "learning_rate": 4.49103448816749e-06,
+      "loss": 0.2993,
+      "step": 4593
+    },
+    {
+      "epoch": 1.0935324567144642,
+      "grad_norm": 0.3938263757181118,
+      "learning_rate": 4.489116944777694e-06,
+      "loss": 0.3359,
+      "step": 4594
+    },
+    {
+      "epoch": 1.093770452787529,
+      "grad_norm": 0.3831990383015829,
+      "learning_rate": 4.487199477317758e-06,
+      "loss": 0.3769,
+      "step": 4595
+    },
+    {
+      "epoch": 1.094008448860594,
+      "grad_norm": 0.35100904325779053,
+      "learning_rate": 4.485282086072666e-06,
+      "loss": 0.2762,
+      "step": 4596
+    },
+    {
+      "epoch": 1.0942464449336586,
+      "grad_norm": 0.38231100716355565,
+      "learning_rate": 4.48336477132739e-06,
+      "loss": 0.2986,
+      "step": 4597
+    },
+    {
+      "epoch": 1.0944844410067234,
+      "grad_norm": 0.36444180770894197,
+      "learning_rate": 4.4814475333668884e-06,
+      "loss": 0.3311,
+      "step": 4598
+    },
+    {
+      "epoch": 1.0947224370797881,
+      "grad_norm": 0.36930290531645005,
+      "learning_rate": 4.479530372476113e-06,
+      "loss": 0.3345,
+      "step": 4599
+    },
+    {
+      "epoch": 1.094960433152853,
+      "grad_norm": 0.38130542257095973,
+      "learning_rate": 4.477613288939999e-06,
+      "loss": 0.2754,
+      "step": 4600
+    },
+    {
+      "epoch": 1.0951984292259178,
+      "grad_norm": 0.371481003330978,
+      "learning_rate": 4.4756962830434735e-06,
+      "loss": 0.3172,
+      "step": 4601
+    },
+    {
+      "epoch": 1.0954364252989826,
+      "grad_norm": 0.3810788506730066,
+      "learning_rate": 4.4737793550714515e-06,
+      "loss": 0.3931,
+      "step": 4602
+    },
+    {
+      "epoch": 1.0956744213720473,
+      "grad_norm": 0.35314794218327844,
+      "learning_rate": 4.471862505308835e-06,
+      "loss": 0.2966,
+      "step": 4603
+    },
+    {
+      "epoch": 1.0959124174451123,
+      "grad_norm": 0.3836019074456394,
+      "learning_rate": 4.469945734040516e-06,
+      "loss": 0.331,
+      "step": 4604
+    },
+    {
+      "epoch": 1.096150413518177,
+      "grad_norm": 0.42836723354827033,
+      "learning_rate": 4.468029041551372e-06,
+      "loss": 0.3328,
+      "step": 4605
+    },
+    {
+      "epoch": 1.0963884095912417,
+      "grad_norm": 0.3480001600905687,
+      "learning_rate": 4.466112428126275e-06,
+      "loss": 0.3601,
+      "step": 4606
+    },
+    {
+      "epoch": 1.0966264056643065,
+      "grad_norm": 0.3885251422953546,
+      "learning_rate": 4.464195894050079e-06,
+      "loss": 0.2733,
+      "step": 4607
+    },
+    {
+      "epoch": 1.0968644017373714,
+      "grad_norm": 0.3940111256961415,
+      "learning_rate": 4.462279439607628e-06,
+      "loss": 0.3078,
+      "step": 4608
+    },
+    {
+      "epoch": 1.0971023978104362,
+      "grad_norm": 0.41464737325896495,
+      "learning_rate": 4.4603630650837545e-06,
+      "loss": 0.4018,
+      "step": 4609
+    },
+    {
+      "epoch": 1.097340393883501,
+      "grad_norm": 0.3657995338802541,
+      "learning_rate": 4.4584467707632804e-06,
+      "loss": 0.2893,
+      "step": 4610
+    },
+    {
+      "epoch": 1.0975783899565656,
+      "grad_norm": 0.3670886985872823,
+      "learning_rate": 4.456530556931013e-06,
+      "loss": 0.2717,
+      "step": 4611
+    },
+    {
+      "epoch": 1.0978163860296306,
+      "grad_norm": 0.3815190866043315,
+      "learning_rate": 4.454614423871749e-06,
+      "loss": 0.3265,
+      "step": 4612
+    },
+    {
+      "epoch": 1.0980543821026953,
+      "grad_norm": 0.3479008384560524,
+      "learning_rate": 4.4526983718702745e-06,
+      "loss": 0.3696,
+      "step": 4613
+    },
+    {
+      "epoch": 1.09829237817576,
+      "grad_norm": 0.3510433837303061,
+      "learning_rate": 4.450782401211362e-06,
+      "loss": 0.2636,
+      "step": 4614
+    },
+    {
+      "epoch": 1.0985303742488248,
+      "grad_norm": 0.3804726727656412,
+      "learning_rate": 4.4488665121797696e-06,
+      "loss": 0.3307,
+      "step": 4615
+    },
+    {
+      "epoch": 1.0987683703218898,
+      "grad_norm": 0.3614441700666874,
+      "learning_rate": 4.446950705060249e-06,
+      "loss": 0.3949,
+      "step": 4616
+    },
+    {
+      "epoch": 1.0990063663949545,
+      "grad_norm": 0.37288138794486936,
+      "learning_rate": 4.445034980137536e-06,
+      "loss": 0.2985,
+      "step": 4617
+    },
+    {
+      "epoch": 1.0992443624680193,
+      "grad_norm": 0.3609717776051928,
+      "learning_rate": 4.4431193376963534e-06,
+      "loss": 0.2821,
+      "step": 4618
+    },
+    {
+      "epoch": 1.099482358541084,
+      "grad_norm": 0.377067343268631,
+      "learning_rate": 4.441203778021412e-06,
+      "loss": 0.3213,
+      "step": 4619
+    },
+    {
+      "epoch": 1.099720354614149,
+      "grad_norm": 0.385028432731855,
+      "learning_rate": 4.439288301397416e-06,
+      "loss": 0.3769,
+      "step": 4620
+    },
+    {
+      "epoch": 1.0999583506872137,
+      "grad_norm": 0.3857984245884134,
+      "learning_rate": 4.437372908109049e-06,
+      "loss": 0.284,
+      "step": 4621
+    },
+    {
+      "epoch": 1.1001963467602784,
+      "grad_norm": 0.3804900796934547,
+      "learning_rate": 4.435457598440987e-06,
+      "loss": 0.3023,
+      "step": 4622
+    },
+    {
+      "epoch": 1.1004343428333432,
+      "grad_norm": 0.4095757121830929,
+      "learning_rate": 4.4335423726778914e-06,
+      "loss": 0.3569,
+      "step": 4623
+    },
+    {
+      "epoch": 1.1006723389064081,
+      "grad_norm": 0.3626590444928867,
+      "learning_rate": 4.431627231104413e-06,
+      "loss": 0.3528,
+      "step": 4624
+    },
+    {
+      "epoch": 1.1009103349794729,
+      "grad_norm": 0.3950093241561531,
+      "learning_rate": 4.429712174005189e-06,
+      "loss": 0.299,
+      "step": 4625
+    },
+    {
+      "epoch": 1.1011483310525376,
+      "grad_norm": 0.35813564107373463,
+      "learning_rate": 4.427797201664844e-06,
+      "loss": 0.3488,
+      "step": 4626
+    },
+    {
+      "epoch": 1.1013863271256024,
+      "grad_norm": 0.39623477898910664,
+      "learning_rate": 4.425882314367991e-06,
+      "loss": 0.3686,
+      "step": 4627
+    },
+    {
+      "epoch": 1.1016243231986673,
+      "grad_norm": 0.3536379619561069,
+      "learning_rate": 4.423967512399228e-06,
+      "loss": 0.3094,
+      "step": 4628
+    },
+    {
+      "epoch": 1.101862319271732,
+      "grad_norm": 0.361987975704904,
+      "learning_rate": 4.4220527960431435e-06,
+      "loss": 0.2832,
+      "step": 4629
+    },
+    {
+      "epoch": 1.1021003153447968,
+      "grad_norm": 0.40248707316608023,
+      "learning_rate": 4.420138165584311e-06,
+      "loss": 0.3405,
+      "step": 4630
+    },
+    {
+      "epoch": 1.1023383114178615,
+      "grad_norm": 0.37274174398754245,
+      "learning_rate": 4.418223621307293e-06,
+      "loss": 0.3721,
+      "step": 4631
+    },
+    {
+      "epoch": 1.1025763074909265,
+      "grad_norm": 0.3929251247694433,
+      "learning_rate": 4.416309163496635e-06,
+      "loss": 0.2833,
+      "step": 4632
+    },
+    {
+      "epoch": 1.1028143035639912,
+      "grad_norm": 0.41200450576439623,
+      "learning_rate": 4.414394792436877e-06,
+      "loss": 0.3211,
+      "step": 4633
+    },
+    {
+      "epoch": 1.103052299637056,
+      "grad_norm": 0.39715739923858967,
+      "learning_rate": 4.41248050841254e-06,
+      "loss": 0.3868,
+      "step": 4634
+    },
+    {
+      "epoch": 1.1032902957101207,
+      "grad_norm": 0.36901909132216004,
+      "learning_rate": 4.410566311708134e-06,
+      "loss": 0.316,
+      "step": 4635
+    },
+    {
+      "epoch": 1.1035282917831857,
+      "grad_norm": 0.3902124213524561,
+      "learning_rate": 4.408652202608156e-06,
+      "loss": 0.2951,
+      "step": 4636
+    },
+    {
+      "epoch": 1.1037662878562504,
+      "grad_norm": 0.37677094100062875,
+      "learning_rate": 4.40673818139709e-06,
+      "loss": 0.3694,
+      "step": 4637
+    },
+    {
+      "epoch": 1.1040042839293152,
+      "grad_norm": 0.4185542623961448,
+      "learning_rate": 4.404824248359407e-06,
+      "loss": 0.3527,
+      "step": 4638
+    },
+    {
+      "epoch": 1.10424228000238,
+      "grad_norm": 0.3668392732421244,
+      "learning_rate": 4.402910403779564e-06,
+      "loss": 0.273,
+      "step": 4639
+    },
+    {
+      "epoch": 1.1044802760754449,
+      "grad_norm": 0.370730106826423,
+      "learning_rate": 4.400996647942007e-06,
+      "loss": 0.3284,
+      "step": 4640
+    },
+    {
+      "epoch": 1.1047182721485096,
+      "grad_norm": 0.3592174118196979,
+      "learning_rate": 4.399082981131166e-06,
+      "loss": 0.393,
+      "step": 4641
+    },
+    {
+      "epoch": 1.1049562682215743,
+      "grad_norm": 0.36645767089673803,
+      "learning_rate": 4.39716940363146e-06,
+      "loss": 0.3077,
+      "step": 4642
+    },
+    {
+      "epoch": 1.105194264294639,
+      "grad_norm": 0.35544644161118766,
+      "learning_rate": 4.395255915727294e-06,
+      "loss": 0.278,
+      "step": 4643
+    },
+    {
+      "epoch": 1.105432260367704,
+      "grad_norm": 0.3834171151203786,
+      "learning_rate": 4.39334251770306e-06,
+      "loss": 0.3348,
+      "step": 4644
+    },
+    {
+      "epoch": 1.1056702564407688,
+      "grad_norm": 0.3825948031558651,
+      "learning_rate": 4.391429209843135e-06,
+      "loss": 0.387,
+      "step": 4645
+    },
+    {
+      "epoch": 1.1059082525138335,
+      "grad_norm": 0.35003853431694987,
+      "learning_rate": 4.389515992431884e-06,
+      "loss": 0.3046,
+      "step": 4646
+    },
+    {
+      "epoch": 1.1061462485868983,
+      "grad_norm": 0.3825762439469127,
+      "learning_rate": 4.387602865753661e-06,
+      "loss": 0.327,
+      "step": 4647
+    },
+    {
+      "epoch": 1.1063842446599632,
+      "grad_norm": 0.3975773348300809,
+      "learning_rate": 4.385689830092802e-06,
+      "loss": 0.3667,
+      "step": 4648
+    },
+    {
+      "epoch": 1.106622240733028,
+      "grad_norm": 0.4272073165238033,
+      "learning_rate": 4.383776885733631e-06,
+      "loss": 0.337,
+      "step": 4649
+    },
+    {
+      "epoch": 1.1068602368060927,
+      "grad_norm": 0.38783082489646,
+      "learning_rate": 4.3818640329604594e-06,
+      "loss": 0.3064,
+      "step": 4650
+    },
+    {
+      "epoch": 1.1070982328791574,
+      "grad_norm": 0.37650129964931706,
+      "learning_rate": 4.3799512720575845e-06,
+      "loss": 0.3432,
+      "step": 4651
+    },
+    {
+      "epoch": 1.1073362289522224,
+      "grad_norm": 0.3741163747344154,
+      "learning_rate": 4.3780386033092905e-06,
+      "loss": 0.3701,
+      "step": 4652
+    },
+    {
+      "epoch": 1.1075742250252871,
+      "grad_norm": 0.37940812325594914,
+      "learning_rate": 4.376126026999846e-06,
+      "loss": 0.3167,
+      "step": 4653
+    },
+    {
+      "epoch": 1.1078122210983519,
+      "grad_norm": 0.3866628177506814,
+      "learning_rate": 4.374213543413508e-06,
+      "loss": 0.3018,
+      "step": 4654
+    },
+    {
+      "epoch": 1.1080502171714166,
+      "grad_norm": 0.4190491021927373,
+      "learning_rate": 4.372301152834518e-06,
+      "loss": 0.3635,
+      "step": 4655
+    },
+    {
+      "epoch": 1.1082882132444816,
+      "grad_norm": 0.36038485452576885,
+      "learning_rate": 4.370388855547106e-06,
+      "loss": 0.3567,
+      "step": 4656
+    },
+    {
+      "epoch": 1.1085262093175463,
+      "grad_norm": 0.36962141459506104,
+      "learning_rate": 4.368476651835485e-06,
+      "loss": 0.2713,
+      "step": 4657
+    },
+    {
+      "epoch": 1.108764205390611,
+      "grad_norm": 0.3760349048148074,
+      "learning_rate": 4.366564541983858e-06,
+      "loss": 0.315,
+      "step": 4658
+    },
+    {
+      "epoch": 1.1090022014636758,
+      "grad_norm": 0.39751754536485306,
+      "learning_rate": 4.36465252627641e-06,
+      "loss": 0.3915,
+      "step": 4659
+    },
+    {
+      "epoch": 1.1092401975367407,
+      "grad_norm": 0.380234301198992,
+      "learning_rate": 4.362740604997312e-06,
+      "loss": 0.3096,
+      "step": 4660
+    },
+    {
+      "epoch": 1.1094781936098055,
+      "grad_norm": 0.3650822635588656,
+      "learning_rate": 4.360828778430728e-06,
+      "loss": 0.2777,
+      "step": 4661
+    },
+    {
+      "epoch": 1.1097161896828702,
+      "grad_norm": 0.38870114316759063,
+      "learning_rate": 4.358917046860799e-06,
+      "loss": 0.3885,
+      "step": 4662
+    },
+    {
+      "epoch": 1.109954185755935,
+      "grad_norm": 0.3821913095469407,
+      "learning_rate": 4.357005410571657e-06,
+      "loss": 0.3813,
+      "step": 4663
+    },
+    {
+      "epoch": 1.110192181829,
+      "grad_norm": 0.38268196285227385,
+      "learning_rate": 4.355093869847418e-06,
+      "loss": 0.2743,
+      "step": 4664
+    },
+    {
+      "epoch": 1.1104301779020647,
+      "grad_norm": 0.3950888429234955,
+      "learning_rate": 4.353182424972184e-06,
+      "loss": 0.3504,
+      "step": 4665
+    },
+    {
+      "epoch": 1.1106681739751294,
+      "grad_norm": 0.38723636956923085,
+      "learning_rate": 4.351271076230042e-06,
+      "loss": 0.4136,
+      "step": 4666
+    },
+    {
+      "epoch": 1.1109061700481941,
+      "grad_norm": 0.3745288707524673,
+      "learning_rate": 4.349359823905068e-06,
+      "loss": 0.3286,
+      "step": 4667
+    },
+    {
+      "epoch": 1.111144166121259,
+      "grad_norm": 0.35266375814406953,
+      "learning_rate": 4.34744866828132e-06,
+      "loss": 0.2839,
+      "step": 4668
+    },
+    {
+      "epoch": 1.1113821621943238,
+      "grad_norm": 0.4116398445629615,
+      "learning_rate": 4.345537609642843e-06,
+      "loss": 0.2851,
+      "step": 4669
+    },
+    {
+      "epoch": 1.1116201582673886,
+      "grad_norm": 0.4227512343669503,
+      "learning_rate": 4.343626648273667e-06,
+      "loss": 0.3862,
+      "step": 4670
+    },
+    {
+      "epoch": 1.1118581543404533,
+      "grad_norm": 0.40325487336775656,
+      "learning_rate": 4.34171578445781e-06,
+      "loss": 0.3086,
+      "step": 4671
+    },
+    {
+      "epoch": 1.1120961504135183,
+      "grad_norm": 0.41223011157726425,
+      "learning_rate": 4.339805018479273e-06,
+      "loss": 0.2965,
+      "step": 4672
+    },
+    {
+      "epoch": 1.112334146486583,
+      "grad_norm": 0.3738155892736554,
+      "learning_rate": 4.337894350622043e-06,
+      "loss": 0.371,
+      "step": 4673
+    },
+    {
+      "epoch": 1.1125721425596478,
+      "grad_norm": 0.3812405808634638,
+      "learning_rate": 4.335983781170089e-06,
+      "loss": 0.3477,
+      "step": 4674
+    },
+    {
+      "epoch": 1.1128101386327125,
+      "grad_norm": 0.4077123636219675,
+      "learning_rate": 4.334073310407375e-06,
+      "loss": 0.283,
+      "step": 4675
+    },
+    {
+      "epoch": 1.1130481347057772,
+      "grad_norm": 0.36774441909673067,
+      "learning_rate": 4.332162938617841e-06,
+      "loss": 0.3361,
+      "step": 4676
+    },
+    {
+      "epoch": 1.1132861307788422,
+      "grad_norm": 0.38011910226148504,
+      "learning_rate": 4.3302526660854155e-06,
+      "loss": 0.4107,
+      "step": 4677
+    },
+    {
+      "epoch": 1.113524126851907,
+      "grad_norm": 0.37483811120325883,
+      "learning_rate": 4.3283424930940135e-06,
+      "loss": 0.3076,
+      "step": 4678
+    },
+    {
+      "epoch": 1.1137621229249717,
+      "grad_norm": 0.4000942474412326,
+      "learning_rate": 4.326432419927532e-06,
+      "loss": 0.3053,
+      "step": 4679
+    },
+    {
+      "epoch": 1.1140001189980366,
+      "grad_norm": 0.4085016676353177,
+      "learning_rate": 4.324522446869856e-06,
+      "loss": 0.3721,
+      "step": 4680
+    },
+    {
+      "epoch": 1.1142381150711014,
+      "grad_norm": 0.38122936015480213,
+      "learning_rate": 4.322612574204856e-06,
+      "loss": 0.3561,
+      "step": 4681
+    },
+    {
+      "epoch": 1.1144761111441661,
+      "grad_norm": 0.46561579745978454,
+      "learning_rate": 4.320702802216384e-06,
+      "loss": 0.2734,
+      "step": 4682
+    },
+    {
+      "epoch": 1.1147141072172309,
+      "grad_norm": 0.38133591698111996,
+      "learning_rate": 4.318793131188281e-06,
+      "loss": 0.3366,
+      "step": 4683
+    },
+    {
+      "epoch": 1.1149521032902956,
+      "grad_norm": 0.39153072129909505,
+      "learning_rate": 4.316883561404371e-06,
+      "loss": 0.3929,
+      "step": 4684
+    },
+    {
+      "epoch": 1.1151900993633606,
+      "grad_norm": 0.39464520028456496,
+      "learning_rate": 4.314974093148464e-06,
+      "loss": 0.2922,
+      "step": 4685
+    },
+    {
+      "epoch": 1.1154280954364253,
+      "grad_norm": 0.3582950558941121,
+      "learning_rate": 4.313064726704352e-06,
+      "loss": 0.2963,
+      "step": 4686
+    },
+    {
+      "epoch": 1.11566609150949,
+      "grad_norm": 0.39622662281594123,
+      "learning_rate": 4.311155462355817e-06,
+      "loss": 0.3557,
+      "step": 4687
+    },
+    {
+      "epoch": 1.115904087582555,
+      "grad_norm": 0.38590265523350004,
+      "learning_rate": 4.309246300386619e-06,
+      "loss": 0.3763,
+      "step": 4688
+    },
+    {
+      "epoch": 1.1161420836556197,
+      "grad_norm": 0.3951607366057468,
+      "learning_rate": 4.3073372410805115e-06,
+      "loss": 0.2872,
+      "step": 4689
+    },
+    {
+      "epoch": 1.1163800797286845,
+      "grad_norm": 0.3600542205066233,
+      "learning_rate": 4.305428284721225e-06,
+      "loss": 0.3088,
+      "step": 4690
+    },
+    {
+      "epoch": 1.1166180758017492,
+      "grad_norm": 0.37643095384884356,
+      "learning_rate": 4.303519431592479e-06,
+      "loss": 0.3968,
+      "step": 4691
+    },
+    {
+      "epoch": 1.116856071874814,
+      "grad_norm": 0.36518409589510575,
+      "learning_rate": 4.301610681977975e-06,
+      "loss": 0.3333,
+      "step": 4692
+    },
+    {
+      "epoch": 1.117094067947879,
+      "grad_norm": 0.37667222528849537,
+      "learning_rate": 4.2997020361614e-06,
+      "loss": 0.2865,
+      "step": 4693
+    },
+    {
+      "epoch": 1.1173320640209436,
+      "grad_norm": 0.4077874035405591,
+      "learning_rate": 4.297793494426429e-06,
+      "loss": 0.3299,
+      "step": 4694
+    },
+    {
+      "epoch": 1.1175700600940084,
+      "grad_norm": 0.36323609120530465,
+      "learning_rate": 4.295885057056716e-06,
+      "loss": 0.4048,
+      "step": 4695
+    },
+    {
+      "epoch": 1.1178080561670733,
+      "grad_norm": 0.36319522250442954,
+      "learning_rate": 4.293976724335901e-06,
+      "loss": 0.2963,
+      "step": 4696
+    },
+    {
+      "epoch": 1.118046052240138,
+      "grad_norm": 0.4272106433249216,
+      "learning_rate": 4.292068496547612e-06,
+      "loss": 0.3082,
+      "step": 4697
+    },
+    {
+      "epoch": 1.1182840483132028,
+      "grad_norm": 0.3620957198196407,
+      "learning_rate": 4.290160373975457e-06,
+      "loss": 0.338,
+      "step": 4698
+    },
+    {
+      "epoch": 1.1185220443862676,
+      "grad_norm": 0.36206405693132543,
+      "learning_rate": 4.2882523569030325e-06,
+      "loss": 0.3244,
+      "step": 4699
+    },
+    {
+      "epoch": 1.1187600404593323,
+      "grad_norm": 0.40476613713113396,
+      "learning_rate": 4.286344445613914e-06,
+      "loss": 0.2865,
+      "step": 4700
+    },
+    {
+      "epoch": 1.1189980365323973,
+      "grad_norm": 0.4027176453027199,
+      "learning_rate": 4.2844366403916654e-06,
+      "loss": 0.3403,
+      "step": 4701
+    },
+    {
+      "epoch": 1.119236032605462,
+      "grad_norm": 0.37078100855116464,
+      "learning_rate": 4.282528941519836e-06,
+      "loss": 0.389,
+      "step": 4702
+    },
+    {
+      "epoch": 1.1194740286785267,
+      "grad_norm": 0.35909975767598074,
+      "learning_rate": 4.280621349281954e-06,
+      "loss": 0.2934,
+      "step": 4703
+    },
+    {
+      "epoch": 1.1197120247515917,
+      "grad_norm": 0.39195993134735413,
+      "learning_rate": 4.278713863961538e-06,
+      "loss": 0.2757,
+      "step": 4704
+    },
+    {
+      "epoch": 1.1199500208246564,
+      "grad_norm": 0.4917626986677953,
+      "learning_rate": 4.276806485842084e-06,
+      "loss": 0.3648,
+      "step": 4705
+    },
+    {
+      "epoch": 1.1201880168977212,
+      "grad_norm": 0.3819285748588736,
+      "learning_rate": 4.274899215207077e-06,
+      "loss": 0.3353,
+      "step": 4706
+    },
+    {
+      "epoch": 1.120426012970786,
+      "grad_norm": 0.3671927065772577,
+      "learning_rate": 4.272992052339986e-06,
+      "loss": 0.2612,
+      "step": 4707
+    },
+    {
+      "epoch": 1.1206640090438507,
+      "grad_norm": 0.44096283257819413,
+      "learning_rate": 4.271084997524261e-06,
+      "loss": 0.3343,
+      "step": 4708
+    },
+    {
+      "epoch": 1.1209020051169156,
+      "grad_norm": 0.41801562980887697,
+      "learning_rate": 4.269178051043336e-06,
+      "loss": 0.3761,
+      "step": 4709
+    },
+    {
+      "epoch": 1.1211400011899804,
+      "grad_norm": 0.3605543613421053,
+      "learning_rate": 4.2672712131806334e-06,
+      "loss": 0.2884,
+      "step": 4710
+    },
+    {
+      "epoch": 1.121377997263045,
+      "grad_norm": 0.4186671411554116,
+      "learning_rate": 4.265364484219556e-06,
+      "loss": 0.286,
+      "step": 4711
+    },
+    {
+      "epoch": 1.12161599333611,
+      "grad_norm": 0.3763168545269733,
+      "learning_rate": 4.263457864443491e-06,
+      "loss": 0.3482,
+      "step": 4712
+    },
+    {
+      "epoch": 1.1218539894091748,
+      "grad_norm": 0.3640333763753411,
+      "learning_rate": 4.261551354135807e-06,
+      "loss": 0.3576,
+      "step": 4713
+    },
+    {
+      "epoch": 1.1220919854822395,
+      "grad_norm": 0.3807965516234715,
+      "learning_rate": 4.259644953579861e-06,
+      "loss": 0.2808,
+      "step": 4714
+    },
+    {
+      "epoch": 1.1223299815553043,
+      "grad_norm": 0.38737302259739137,
+      "learning_rate": 4.2577386630589875e-06,
+      "loss": 0.3466,
+      "step": 4715
+    },
+    {
+      "epoch": 1.122567977628369,
+      "grad_norm": 0.3709834921935741,
+      "learning_rate": 4.255832482856514e-06,
+      "loss": 0.3879,
+      "step": 4716
+    },
+    {
+      "epoch": 1.122805973701434,
+      "grad_norm": 0.397332853820356,
+      "learning_rate": 4.253926413255743e-06,
+      "loss": 0.3053,
+      "step": 4717
+    },
+    {
+      "epoch": 1.1230439697744987,
+      "grad_norm": 0.37589401628573843,
+      "learning_rate": 4.252020454539965e-06,
+      "loss": 0.3125,
+      "step": 4718
+    },
+    {
+      "epoch": 1.1232819658475635,
+      "grad_norm": 0.39174326299097356,
+      "learning_rate": 4.250114606992451e-06,
+      "loss": 0.3167,
+      "step": 4719
+    },
+    {
+      "epoch": 1.1235199619206284,
+      "grad_norm": 0.3617505028367288,
+      "learning_rate": 4.248208870896456e-06,
+      "loss": 0.3868,
+      "step": 4720
+    },
+    {
+      "epoch": 1.1237579579936932,
+      "grad_norm": 0.401359881455322,
+      "learning_rate": 4.246303246535224e-06,
+      "loss": 0.2771,
+      "step": 4721
+    },
+    {
+      "epoch": 1.123995954066758,
+      "grad_norm": 0.3950415349640952,
+      "learning_rate": 4.244397734191973e-06,
+      "loss": 0.3337,
+      "step": 4722
+    },
+    {
+      "epoch": 1.1242339501398226,
+      "grad_norm": 0.40327719323124134,
+      "learning_rate": 4.242492334149911e-06,
+      "loss": 0.4014,
+      "step": 4723
+    },
+    {
+      "epoch": 1.1244719462128874,
+      "grad_norm": 0.37522594034975987,
+      "learning_rate": 4.2405870466922295e-06,
+      "loss": 0.3311,
+      "step": 4724
+    },
+    {
+      "epoch": 1.1247099422859523,
+      "grad_norm": 0.3900606525913724,
+      "learning_rate": 4.2386818721021e-06,
+      "loss": 0.2906,
+      "step": 4725
+    },
+    {
+      "epoch": 1.124947938359017,
+      "grad_norm": 0.36868405071537597,
+      "learning_rate": 4.236776810662677e-06,
+      "loss": 0.3458,
+      "step": 4726
+    },
+    {
+      "epoch": 1.1251859344320818,
+      "grad_norm": 0.3751466988447734,
+      "learning_rate": 4.2348718626571024e-06,
+      "loss": 0.3748,
+      "step": 4727
+    },
+    {
+      "epoch": 1.1254239305051468,
+      "grad_norm": 0.3635206406113209,
+      "learning_rate": 4.232967028368498e-06,
+      "loss": 0.27,
+      "step": 4728
+    },
+    {
+      "epoch": 1.1256619265782115,
+      "grad_norm": 0.3790908884920694,
+      "learning_rate": 4.231062308079965e-06,
+      "loss": 0.2931,
+      "step": 4729
+    },
+    {
+      "epoch": 1.1258999226512763,
+      "grad_norm": 0.3699061956945172,
+      "learning_rate": 4.229157702074598e-06,
+      "loss": 0.3573,
+      "step": 4730
+    },
+    {
+      "epoch": 1.126137918724341,
+      "grad_norm": 0.3541326251134355,
+      "learning_rate": 4.227253210635467e-06,
+      "loss": 0.3273,
+      "step": 4731
+    },
+    {
+      "epoch": 1.1263759147974057,
+      "grad_norm": 0.3706068751730151,
+      "learning_rate": 4.225348834045625e-06,
+      "loss": 0.3089,
+      "step": 4732
+    },
+    {
+      "epoch": 1.1266139108704707,
+      "grad_norm": 0.36912447443811713,
+      "learning_rate": 4.223444572588111e-06,
+      "loss": 0.3112,
+      "step": 4733
+    },
+    {
+      "epoch": 1.1268519069435354,
+      "grad_norm": 0.38272894471100344,
+      "learning_rate": 4.221540426545943e-06,
+      "loss": 0.3732,
+      "step": 4734
+    },
+    {
+      "epoch": 1.1270899030166002,
+      "grad_norm": 0.3659819182595056,
+      "learning_rate": 4.219636396202127e-06,
+      "loss": 0.3125,
+      "step": 4735
+    },
+    {
+      "epoch": 1.1273278990896651,
+      "grad_norm": 0.4029280672903165,
+      "learning_rate": 4.217732481839647e-06,
+      "loss": 0.2993,
+      "step": 4736
+    },
+    {
+      "epoch": 1.1275658951627299,
+      "grad_norm": 0.3471586840183988,
+      "learning_rate": 4.215828683741473e-06,
+      "loss": 0.3264,
+      "step": 4737
+    },
+    {
+      "epoch": 1.1278038912357946,
+      "grad_norm": 0.40496516884275047,
+      "learning_rate": 4.2139250021905564e-06,
+      "loss": 0.3911,
+      "step": 4738
+    },
+    {
+      "epoch": 1.1280418873088593,
+      "grad_norm": 0.3604417470905285,
+      "learning_rate": 4.212021437469831e-06,
+      "loss": 0.2733,
+      "step": 4739
+    },
+    {
+      "epoch": 1.128279883381924,
+      "grad_norm": 0.4045782091345007,
+      "learning_rate": 4.210117989862213e-06,
+      "loss": 0.3166,
+      "step": 4740
+    },
+    {
+      "epoch": 1.128517879454989,
+      "grad_norm": 0.42245205724517865,
+      "learning_rate": 4.208214659650603e-06,
+      "loss": 0.3806,
+      "step": 4741
+    },
+    {
+      "epoch": 1.1287558755280538,
+      "grad_norm": 0.37772806607684767,
+      "learning_rate": 4.206311447117883e-06,
+      "loss": 0.3229,
+      "step": 4742
+    },
+    {
+      "epoch": 1.1289938716011185,
+      "grad_norm": 0.45042461811956686,
+      "learning_rate": 4.204408352546914e-06,
+      "loss": 0.2872,
+      "step": 4743
+    },
+    {
+      "epoch": 1.1292318676741835,
+      "grad_norm": 0.41146418750242547,
+      "learning_rate": 4.202505376220548e-06,
+      "loss": 0.3595,
+      "step": 4744
+    },
+    {
+      "epoch": 1.1294698637472482,
+      "grad_norm": 0.37141506761563525,
+      "learning_rate": 4.2006025184216124e-06,
+      "loss": 0.387,
+      "step": 4745
+    },
+    {
+      "epoch": 1.129707859820313,
+      "grad_norm": 0.4033866279526767,
+      "learning_rate": 4.198699779432918e-06,
+      "loss": 0.2798,
+      "step": 4746
+    },
+    {
+      "epoch": 1.1299458558933777,
+      "grad_norm": 0.39808863942798545,
+      "learning_rate": 4.19679715953726e-06,
+      "loss": 0.3075,
+      "step": 4747
+    },
+    {
+      "epoch": 1.1301838519664424,
+      "grad_norm": 0.35618340589269987,
+      "learning_rate": 4.194894659017415e-06,
+      "loss": 0.3694,
+      "step": 4748
+    },
+    {
+      "epoch": 1.1304218480395074,
+      "grad_norm": 0.3632549519200726,
+      "learning_rate": 4.192992278156141e-06,
+      "loss": 0.3462,
+      "step": 4749
+    },
+    {
+      "epoch": 1.1306598441125721,
+      "grad_norm": 0.4018427851047768,
+      "learning_rate": 4.191090017236177e-06,
+      "loss": 0.2608,
+      "step": 4750
+    },
+    {
+      "epoch": 1.1308978401856369,
+      "grad_norm": 0.3889863500503954,
+      "learning_rate": 4.189187876540249e-06,
+      "loss": 0.3352,
+      "step": 4751
+    },
+    {
+      "epoch": 1.1311358362587018,
+      "grad_norm": 0.3495459311201199,
+      "learning_rate": 4.187285856351061e-06,
+      "loss": 0.3973,
+      "step": 4752
+    },
+    {
+      "epoch": 1.1313738323317666,
+      "grad_norm": 0.3678770041190173,
+      "learning_rate": 4.1853839569513015e-06,
+      "loss": 0.295,
+      "step": 4753
+    },
+    {
+      "epoch": 1.1316118284048313,
+      "grad_norm": 0.4160690563278992,
+      "learning_rate": 4.1834821786236375e-06,
+      "loss": 0.3,
+      "step": 4754
+    },
+    {
+      "epoch": 1.131849824477896,
+      "grad_norm": 0.3814427610026273,
+      "learning_rate": 4.181580521650722e-06,
+      "loss": 0.3727,
+      "step": 4755
+    },
+    {
+      "epoch": 1.1320878205509608,
+      "grad_norm": 0.35113445654594827,
+      "learning_rate": 4.179678986315185e-06,
+      "loss": 0.317,
+      "step": 4756
+    },
+    {
+      "epoch": 1.1323258166240258,
+      "grad_norm": 0.38027676840048824,
+      "learning_rate": 4.177777572899647e-06,
+      "loss": 0.2754,
+      "step": 4757
+    },
+    {
+      "epoch": 1.1325638126970905,
+      "grad_norm": 0.3539448580155885,
+      "learning_rate": 4.175876281686702e-06,
+      "loss": 0.3154,
+      "step": 4758
+    },
+    {
+      "epoch": 1.1328018087701552,
+      "grad_norm": 0.3741958790250975,
+      "learning_rate": 4.17397511295893e-06,
+      "loss": 0.4037,
+      "step": 4759
+    },
+    {
+      "epoch": 1.1330398048432202,
+      "grad_norm": 0.35759297486340763,
+      "learning_rate": 4.17207406699889e-06,
+      "loss": 0.2913,
+      "step": 4760
+    },
+    {
+      "epoch": 1.133277800916285,
+      "grad_norm": 0.41006094302343415,
+      "learning_rate": 4.170173144089127e-06,
+      "loss": 0.3066,
+      "step": 4761
+    },
+    {
+      "epoch": 1.1335157969893497,
+      "grad_norm": 0.3764612848174041,
+      "learning_rate": 4.168272344512163e-06,
+      "loss": 0.3449,
+      "step": 4762
+    },
+    {
+      "epoch": 1.1337537930624144,
+      "grad_norm": 0.3676606468861143,
+      "learning_rate": 4.1663716685505026e-06,
+      "loss": 0.353,
+      "step": 4763
+    },
+    {
+      "epoch": 1.1339917891354792,
+      "grad_norm": 0.43991654949679093,
+      "learning_rate": 4.164471116486638e-06,
+      "loss": 0.3147,
+      "step": 4764
+    },
+    {
+      "epoch": 1.1342297852085441,
+      "grad_norm": 0.389125286946846,
+      "learning_rate": 4.162570688603035e-06,
+      "loss": 0.2964,
+      "step": 4765
+    },
+    {
+      "epoch": 1.1344677812816089,
+      "grad_norm": 0.3953023155670322,
+      "learning_rate": 4.160670385182144e-06,
+      "loss": 0.3646,
+      "step": 4766
+    },
+    {
+      "epoch": 1.1347057773546736,
+      "grad_norm": 0.3803124764959841,
+      "learning_rate": 4.158770206506398e-06,
+      "loss": 0.3455,
+      "step": 4767
+    },
+    {
+      "epoch": 1.1349437734277386,
+      "grad_norm": 0.3932564435551,
+      "learning_rate": 4.1568701528582105e-06,
+      "loss": 0.3009,
+      "step": 4768
+    },
+    {
+      "epoch": 1.1351817695008033,
+      "grad_norm": 0.37337164388298766,
+      "learning_rate": 4.154970224519976e-06,
+      "loss": 0.3482,
+      "step": 4769
+    },
+    {
+      "epoch": 1.135419765573868,
+      "grad_norm": 0.3899160813781137,
+      "learning_rate": 4.15307042177407e-06,
+      "loss": 0.4065,
+      "step": 4770
+    },
+    {
+      "epoch": 1.1356577616469328,
+      "grad_norm": 0.35579178905990316,
+      "learning_rate": 4.151170744902852e-06,
+      "loss": 0.3079,
+      "step": 4771
+    },
+    {
+      "epoch": 1.1358957577199975,
+      "grad_norm": 0.37945121820654776,
+      "learning_rate": 4.149271194188662e-06,
+      "loss": 0.3105,
+      "step": 4772
+    },
+    {
+      "epoch": 1.1361337537930625,
+      "grad_norm": 0.4165547341391943,
+      "learning_rate": 4.147371769913817e-06,
+      "loss": 0.3696,
+      "step": 4773
+    },
+    {
+      "epoch": 1.1363717498661272,
+      "grad_norm": 0.376295683219943,
+      "learning_rate": 4.145472472360621e-06,
+      "loss": 0.3747,
+      "step": 4774
+    },
+    {
+      "epoch": 1.136609745939192,
+      "grad_norm": 0.3886386600985121,
+      "learning_rate": 4.143573301811355e-06,
+      "loss": 0.2783,
+      "step": 4775
+    },
+    {
+      "epoch": 1.136847742012257,
+      "grad_norm": 0.40835168558501084,
+      "learning_rate": 4.141674258548284e-06,
+      "loss": 0.3384,
+      "step": 4776
+    },
+    {
+      "epoch": 1.1370857380853217,
+      "grad_norm": 0.371459747825813,
+      "learning_rate": 4.13977534285365e-06,
+      "loss": 0.383,
+      "step": 4777
+    },
+    {
+      "epoch": 1.1373237341583864,
+      "grad_norm": 0.3944390627645514,
+      "learning_rate": 4.137876555009684e-06,
+      "loss": 0.3055,
+      "step": 4778
+    },
+    {
+      "epoch": 1.1375617302314511,
+      "grad_norm": 0.41245043405705123,
+      "learning_rate": 4.135977895298588e-06,
+      "loss": 0.3053,
+      "step": 4779
+    },
+    {
+      "epoch": 1.1377997263045159,
+      "grad_norm": 0.3846823535751807,
+      "learning_rate": 4.134079364002552e-06,
+      "loss": 0.3395,
+      "step": 4780
+    },
+    {
+      "epoch": 1.1380377223775808,
+      "grad_norm": 0.39696399724800313,
+      "learning_rate": 4.132180961403744e-06,
+      "loss": 0.3519,
+      "step": 4781
+    },
+    {
+      "epoch": 1.1382757184506456,
+      "grad_norm": 0.39325849294481,
+      "learning_rate": 4.130282687784315e-06,
+      "loss": 0.2742,
+      "step": 4782
+    },
+    {
+      "epoch": 1.1385137145237103,
+      "grad_norm": 0.40331167287236297,
+      "learning_rate": 4.1283845434263935e-06,
+      "loss": 0.3036,
+      "step": 4783
+    },
+    {
+      "epoch": 1.1387517105967753,
+      "grad_norm": 0.4029530512494824,
+      "learning_rate": 4.12648652861209e-06,
+      "loss": 0.3796,
+      "step": 4784
+    },
+    {
+      "epoch": 1.13898970666984,
+      "grad_norm": 0.3698197939582225,
+      "learning_rate": 4.1245886436235e-06,
+      "loss": 0.3043,
+      "step": 4785
+    },
+    {
+      "epoch": 1.1392277027429047,
+      "grad_norm": 0.37697533259800486,
+      "learning_rate": 4.122690888742694e-06,
+      "loss": 0.2891,
+      "step": 4786
+    },
+    {
+      "epoch": 1.1394656988159695,
+      "grad_norm": 0.39557375149649615,
+      "learning_rate": 4.120793264251726e-06,
+      "loss": 0.3232,
+      "step": 4787
+    },
+    {
+      "epoch": 1.1397036948890342,
+      "grad_norm": 0.364829176455308,
+      "learning_rate": 4.118895770432629e-06,
+      "loss": 0.3722,
+      "step": 4788
+    },
+    {
+      "epoch": 1.1399416909620992,
+      "grad_norm": 0.3778167261258978,
+      "learning_rate": 4.1169984075674184e-06,
+      "loss": 0.2911,
+      "step": 4789
+    },
+    {
+      "epoch": 1.140179687035164,
+      "grad_norm": 0.44220258438829385,
+      "learning_rate": 4.115101175938088e-06,
+      "loss": 0.3305,
+      "step": 4790
+    },
+    {
+      "epoch": 1.1404176831082287,
+      "grad_norm": 0.36291623619174723,
+      "learning_rate": 4.113204075826614e-06,
+      "loss": 0.3847,
+      "step": 4791
+    },
+    {
+      "epoch": 1.1406556791812936,
+      "grad_norm": 0.36770447829250785,
+      "learning_rate": 4.111307107514953e-06,
+      "loss": 0.3274,
+      "step": 4792
+    },
+    {
+      "epoch": 1.1408936752543584,
+      "grad_norm": 0.35815092054120873,
+      "learning_rate": 4.109410271285042e-06,
+      "loss": 0.2641,
+      "step": 4793
+    },
+    {
+      "epoch": 1.141131671327423,
+      "grad_norm": 0.36035533257546154,
+      "learning_rate": 4.107513567418796e-06,
+      "loss": 0.3437,
+      "step": 4794
+    },
+    {
+      "epoch": 1.1413696674004878,
+      "grad_norm": 0.3963850334927369,
+      "learning_rate": 4.105616996198113e-06,
+      "loss": 0.3981,
+      "step": 4795
+    },
+    {
+      "epoch": 1.1416076634735526,
+      "grad_norm": 0.3367518459228409,
+      "learning_rate": 4.10372055790487e-06,
+      "loss": 0.3095,
+      "step": 4796
+    },
+    {
+      "epoch": 1.1418456595466175,
+      "grad_norm": 0.38016156127534406,
+      "learning_rate": 4.101824252820926e-06,
+      "loss": 0.3133,
+      "step": 4797
+    },
+    {
+      "epoch": 1.1420836556196823,
+      "grad_norm": 0.3912648891515814,
+      "learning_rate": 4.099928081228115e-06,
+      "loss": 0.3705,
+      "step": 4798
+    },
+    {
+      "epoch": 1.142321651692747,
+      "grad_norm": 0.37387297859975754,
+      "learning_rate": 4.09803204340826e-06,
+      "loss": 0.3113,
+      "step": 4799
+    },
+    {
+      "epoch": 1.142559647765812,
+      "grad_norm": 0.3845831972941408,
+      "learning_rate": 4.096136139643158e-06,
+      "loss": 0.3138,
+      "step": 4800
+    },
+    {
+      "epoch": 1.1427976438388767,
+      "grad_norm": 0.3742150865042121,
+      "learning_rate": 4.094240370214585e-06,
+      "loss": 0.3117,
+      "step": 4801
+    },
+    {
+      "epoch": 1.1430356399119415,
+      "grad_norm": 0.39531922526195684,
+      "learning_rate": 4.0923447354043e-06,
+      "loss": 0.3866,
+      "step": 4802
+    },
+    {
+      "epoch": 1.1432736359850062,
+      "grad_norm": 0.36395182823896827,
+      "learning_rate": 4.090449235494043e-06,
+      "loss": 0.2797,
+      "step": 4803
+    },
+    {
+      "epoch": 1.143511632058071,
+      "grad_norm": 0.36855555988204036,
+      "learning_rate": 4.088553870765529e-06,
+      "loss": 0.2781,
+      "step": 4804
+    },
+    {
+      "epoch": 1.143749628131136,
+      "grad_norm": 0.37855067257091296,
+      "learning_rate": 4.086658641500458e-06,
+      "loss": 0.3791,
+      "step": 4805
+    },
+    {
+      "epoch": 1.1439876242042006,
+      "grad_norm": 0.3872640810576243,
+      "learning_rate": 4.0847635479805095e-06,
+      "loss": 0.3539,
+      "step": 4806
+    },
+    {
+      "epoch": 1.1442256202772654,
+      "grad_norm": 0.47053726164577175,
+      "learning_rate": 4.082868590487339e-06,
+      "loss": 0.2798,
+      "step": 4807
+    },
+    {
+      "epoch": 1.1444636163503303,
+      "grad_norm": 0.36992308839073973,
+      "learning_rate": 4.080973769302584e-06,
+      "loss": 0.3663,
+      "step": 4808
+    },
+    {
+      "epoch": 1.144701612423395,
+      "grad_norm": 0.3820408245781083,
+      "learning_rate": 4.079079084707864e-06,
+      "loss": 0.3765,
+      "step": 4809
+    },
+    {
+      "epoch": 1.1449396084964598,
+      "grad_norm": 0.36508657238373526,
+      "learning_rate": 4.077184536984773e-06,
+      "loss": 0.2741,
+      "step": 4810
+    },
+    {
+      "epoch": 1.1451776045695246,
+      "grad_norm": 0.3625603860149019,
+      "learning_rate": 4.07529012641489e-06,
+      "loss": 0.2954,
+      "step": 4811
+    },
+    {
+      "epoch": 1.1454156006425893,
+      "grad_norm": 0.3529289410508482,
+      "learning_rate": 4.073395853279768e-06,
+      "loss": 0.3098,
+      "step": 4812
+    },
+    {
+      "epoch": 1.1456535967156543,
+      "grad_norm": 0.3698173348748085,
+      "learning_rate": 4.071501717860947e-06,
+      "loss": 0.387,
+      "step": 4813
+    },
+    {
+      "epoch": 1.145891592788719,
+      "grad_norm": 0.36193766430504687,
+      "learning_rate": 4.069607720439942e-06,
+      "loss": 0.269,
+      "step": 4814
+    },
+    {
+      "epoch": 1.1461295888617837,
+      "grad_norm": 0.39085304104980867,
+      "learning_rate": 4.067713861298246e-06,
+      "loss": 0.3279,
+      "step": 4815
+    },
+    {
+      "epoch": 1.1463675849348487,
+      "grad_norm": 0.3703336609246225,
+      "learning_rate": 4.0658201407173335e-06,
+      "loss": 0.3719,
+      "step": 4816
+    },
+    {
+      "epoch": 1.1466055810079134,
+      "grad_norm": 0.3642397937157319,
+      "learning_rate": 4.063926558978657e-06,
+      "loss": 0.3411,
+      "step": 4817
+    },
+    {
+      "epoch": 1.1468435770809782,
+      "grad_norm": 0.3570075003209498,
+      "learning_rate": 4.062033116363653e-06,
+      "loss": 0.2743,
+      "step": 4818
+    },
+    {
+      "epoch": 1.147081573154043,
+      "grad_norm": 0.3926987214457056,
+      "learning_rate": 4.060139813153732e-06,
+      "loss": 0.338,
+      "step": 4819
+    },
+    {
+      "epoch": 1.1473195692271076,
+      "grad_norm": 0.36780611870947877,
+      "learning_rate": 4.058246649630286e-06,
+      "loss": 0.3727,
+      "step": 4820
+    },
+    {
+      "epoch": 1.1475575653001726,
+      "grad_norm": 0.3501727672950386,
+      "learning_rate": 4.056353626074685e-06,
+      "loss": 0.2804,
+      "step": 4821
+    },
+    {
+      "epoch": 1.1477955613732373,
+      "grad_norm": 0.4059745265643401,
+      "learning_rate": 4.05446074276828e-06,
+      "loss": 0.2883,
+      "step": 4822
+    },
+    {
+      "epoch": 1.148033557446302,
+      "grad_norm": 0.4002543534646528,
+      "learning_rate": 4.0525679999924e-06,
+      "loss": 0.3783,
+      "step": 4823
+    },
+    {
+      "epoch": 1.148271553519367,
+      "grad_norm": 0.37811094840803156,
+      "learning_rate": 4.050675398028354e-06,
+      "loss": 0.3345,
+      "step": 4824
+    },
+    {
+      "epoch": 1.1485095495924318,
+      "grad_norm": 0.39558594949102993,
+      "learning_rate": 4.048782937157427e-06,
+      "loss": 0.2961,
+      "step": 4825
+    },
+    {
+      "epoch": 1.1487475456654965,
+      "grad_norm": 0.39775175118133455,
+      "learning_rate": 4.04689061766089e-06,
+      "loss": 0.3206,
+      "step": 4826
+    },
+    {
+      "epoch": 1.1489855417385613,
+      "grad_norm": 0.37445723489979743,
+      "learning_rate": 4.044998439819986e-06,
+      "loss": 0.3986,
+      "step": 4827
+    },
+    {
+      "epoch": 1.149223537811626,
+      "grad_norm": 0.3739855904232471,
+      "learning_rate": 4.043106403915938e-06,
+      "loss": 0.2836,
+      "step": 4828
+    },
+    {
+      "epoch": 1.149461533884691,
+      "grad_norm": 0.4283283093929408,
+      "learning_rate": 4.041214510229952e-06,
+      "loss": 0.3011,
+      "step": 4829
+    },
+    {
+      "epoch": 1.1496995299577557,
+      "grad_norm": 0.40090894213533174,
+      "learning_rate": 4.0393227590432085e-06,
+      "loss": 0.3815,
+      "step": 4830
+    },
+    {
+      "epoch": 1.1499375260308204,
+      "grad_norm": 0.3617108318222815,
+      "learning_rate": 4.037431150636868e-06,
+      "loss": 0.3683,
+      "step": 4831
+    },
+    {
+      "epoch": 1.1501755221038854,
+      "grad_norm": 0.405107436573071,
+      "learning_rate": 4.0355396852920735e-06,
+      "loss": 0.2949,
+      "step": 4832
+    },
+    {
+      "epoch": 1.1504135181769501,
+      "grad_norm": 0.4091009038760894,
+      "learning_rate": 4.03364836328994e-06,
+      "loss": 0.2922,
+      "step": 4833
+    },
+    {
+      "epoch": 1.1506515142500149,
+      "grad_norm": 0.3723010060588749,
+      "learning_rate": 4.0317571849115665e-06,
+      "loss": 0.3649,
+      "step": 4834
+    },
+    {
+      "epoch": 1.1508895103230796,
+      "grad_norm": 0.3486410506502387,
+      "learning_rate": 4.029866150438029e-06,
+      "loss": 0.2946,
+      "step": 4835
+    },
+    {
+      "epoch": 1.1511275063961444,
+      "grad_norm": 0.40140072119081,
+      "learning_rate": 4.02797526015038e-06,
+      "loss": 0.2545,
+      "step": 4836
+    },
+    {
+      "epoch": 1.1513655024692093,
+      "grad_norm": 0.36966831779652004,
+      "learning_rate": 4.026084514329656e-06,
+      "loss": 0.3484,
+      "step": 4837
+    },
+    {
+      "epoch": 1.151603498542274,
+      "grad_norm": 0.3744156869841182,
+      "learning_rate": 4.024193913256865e-06,
+      "loss": 0.3852,
+      "step": 4838
+    },
+    {
+      "epoch": 1.1518414946153388,
+      "grad_norm": 0.3862547285091858,
+      "learning_rate": 4.022303457212998e-06,
+      "loss": 0.2861,
+      "step": 4839
+    },
+    {
+      "epoch": 1.1520794906884038,
+      "grad_norm": 0.3873257265237817,
+      "learning_rate": 4.020413146479026e-06,
+      "loss": 0.3121,
+      "step": 4840
+    },
+    {
+      "epoch": 1.1523174867614685,
+      "grad_norm": 0.4522160266820443,
+      "learning_rate": 4.018522981335894e-06,
+      "loss": 0.3734,
+      "step": 4841
+    },
+    {
+      "epoch": 1.1525554828345332,
+      "grad_norm": 0.3643381212709832,
+      "learning_rate": 4.0166329620645275e-06,
+      "loss": 0.3121,
+      "step": 4842
+    },
+    {
+      "epoch": 1.152793478907598,
+      "grad_norm": 0.40744044121578465,
+      "learning_rate": 4.01474308894583e-06,
+      "loss": 0.302,
+      "step": 4843
+    },
+    {
+      "epoch": 1.1530314749806627,
+      "grad_norm": 0.3799665007235963,
+      "learning_rate": 4.012853362260683e-06,
+      "loss": 0.325,
+      "step": 4844
+    },
+    {
+      "epoch": 1.1532694710537277,
+      "grad_norm": 0.40864748898009884,
+      "learning_rate": 4.010963782289948e-06,
+      "loss": 0.3931,
+      "step": 4845
+    },
+    {
+      "epoch": 1.1535074671267924,
+      "grad_norm": 0.39417564672662864,
+      "learning_rate": 4.009074349314462e-06,
+      "loss": 0.3143,
+      "step": 4846
+    },
+    {
+      "epoch": 1.1537454631998572,
+      "grad_norm": 0.42660463794138176,
+      "learning_rate": 4.007185063615043e-06,
+      "loss": 0.2997,
+      "step": 4847
+    },
+    {
+      "epoch": 1.1539834592729221,
+      "grad_norm": 0.45151329355721237,
+      "learning_rate": 4.005295925472484e-06,
+      "loss": 0.3525,
+      "step": 4848
+    },
+    {
+      "epoch": 1.1542214553459869,
+      "grad_norm": 0.3534689223030634,
+      "learning_rate": 4.003406935167558e-06,
+      "loss": 0.3348,
+      "step": 4849
+    },
+    {
+      "epoch": 1.1544594514190516,
+      "grad_norm": 0.3741669360663959,
+      "learning_rate": 4.001518092981017e-06,
+      "loss": 0.2907,
+      "step": 4850
+    },
+    {
+      "epoch": 1.1546974474921163,
+      "grad_norm": 0.3795133464854993,
+      "learning_rate": 3.999629399193589e-06,
+      "loss": 0.3346,
+      "step": 4851
+    },
+    {
+      "epoch": 1.154935443565181,
+      "grad_norm": 0.4063978193238941,
+      "learning_rate": 3.997740854085979e-06,
+      "loss": 0.369,
+      "step": 4852
+    },
+    {
+      "epoch": 1.155173439638246,
+      "grad_norm": 0.35078236527156986,
+      "learning_rate": 3.995852457938871e-06,
+      "loss": 0.3191,
+      "step": 4853
+    },
+    {
+      "epoch": 1.1554114357113108,
+      "grad_norm": 0.38716927785738015,
+      "learning_rate": 3.993964211032931e-06,
+      "loss": 0.2937,
+      "step": 4854
+    },
+    {
+      "epoch": 1.1556494317843755,
+      "grad_norm": 0.41552238495054705,
+      "learning_rate": 3.992076113648797e-06,
+      "loss": 0.3745,
+      "step": 4855
+    },
+    {
+      "epoch": 1.1558874278574405,
+      "grad_norm": 0.3909495828147318,
+      "learning_rate": 3.990188166067088e-06,
+      "loss": 0.3445,
+      "step": 4856
+    },
+    {
+      "epoch": 1.1561254239305052,
+      "grad_norm": 0.3773512559043716,
+      "learning_rate": 3.9883003685684e-06,
+      "loss": 0.2753,
+      "step": 4857
+    },
+    {
+      "epoch": 1.15636342000357,
+      "grad_norm": 0.3815969618990709,
+      "learning_rate": 3.9864127214333035e-06,
+      "loss": 0.3294,
+      "step": 4858
+    },
+    {
+      "epoch": 1.1566014160766347,
+      "grad_norm": 0.3682808970267619,
+      "learning_rate": 3.984525224942352e-06,
+      "loss": 0.3877,
+      "step": 4859
+    },
+    {
+      "epoch": 1.1568394121496994,
+      "grad_norm": 0.38593825650163915,
+      "learning_rate": 3.982637879376075e-06,
+      "loss": 0.3136,
+      "step": 4860
+    },
+    {
+      "epoch": 1.1570774082227644,
+      "grad_norm": 0.4087706279356301,
+      "learning_rate": 3.980750685014975e-06,
+      "loss": 0.2707,
+      "step": 4861
+    },
+    {
+      "epoch": 1.1573154042958291,
+      "grad_norm": 0.36407650716725964,
+      "learning_rate": 3.978863642139541e-06,
+      "loss": 0.3644,
+      "step": 4862
+    },
+    {
+      "epoch": 1.1575534003688939,
+      "grad_norm": 0.391041663164892,
+      "learning_rate": 3.97697675103023e-06,
+      "loss": 0.3702,
+      "step": 4863
+    },
+    {
+      "epoch": 1.1577913964419588,
+      "grad_norm": 0.36440676261441685,
+      "learning_rate": 3.975090011967483e-06,
+      "loss": 0.2939,
+      "step": 4864
+    },
+    {
+      "epoch": 1.1580293925150236,
+      "grad_norm": 0.3944390664922029,
+      "learning_rate": 3.973203425231715e-06,
+      "loss": 0.3345,
+      "step": 4865
+    },
+    {
+      "epoch": 1.1582673885880883,
+      "grad_norm": 0.37043202122893515,
+      "learning_rate": 3.971316991103319e-06,
+      "loss": 0.3966,
+      "step": 4866
+    },
+    {
+      "epoch": 1.158505384661153,
+      "grad_norm": 0.3596933682030246,
+      "learning_rate": 3.969430709862665e-06,
+      "loss": 0.3262,
+      "step": 4867
+    },
+    {
+      "epoch": 1.1587433807342178,
+      "grad_norm": 0.415916141016302,
+      "learning_rate": 3.967544581790105e-06,
+      "loss": 0.2955,
+      "step": 4868
+    },
+    {
+      "epoch": 1.1589813768072827,
+      "grad_norm": 0.364949249649275,
+      "learning_rate": 3.965658607165961e-06,
+      "loss": 0.3108,
+      "step": 4869
+    },
+    {
+      "epoch": 1.1592193728803475,
+      "grad_norm": 0.4038748930409246,
+      "learning_rate": 3.9637727862705375e-06,
+      "loss": 0.3811,
+      "step": 4870
+    },
+    {
+      "epoch": 1.1594573689534122,
+      "grad_norm": 0.37034415569974766,
+      "learning_rate": 3.961887119384111e-06,
+      "loss": 0.2626,
+      "step": 4871
+    },
+    {
+      "epoch": 1.1596953650264772,
+      "grad_norm": 0.37614882290717155,
+      "learning_rate": 3.960001606786942e-06,
+      "loss": 0.2926,
+      "step": 4872
+    },
+    {
+      "epoch": 1.159933361099542,
+      "grad_norm": 0.6816306788345337,
+      "learning_rate": 3.958116248759262e-06,
+      "loss": 0.3556,
+      "step": 4873
+    },
+    {
+      "epoch": 1.1601713571726067,
+      "grad_norm": 0.3601925064970214,
+      "learning_rate": 3.9562310455812825e-06,
+      "loss": 0.3386,
+      "step": 4874
+    },
+    {
+      "epoch": 1.1604093532456714,
+      "grad_norm": 0.37272600080408436,
+      "learning_rate": 3.9543459975331914e-06,
+      "loss": 0.2821,
+      "step": 4875
+    },
+    {
+      "epoch": 1.1606473493187361,
+      "grad_norm": 0.37365911547388664,
+      "learning_rate": 3.952461104895153e-06,
+      "loss": 0.304,
+      "step": 4876
+    },
+    {
+      "epoch": 1.160885345391801,
+      "grad_norm": 0.4220848745812085,
+      "learning_rate": 3.95057636794731e-06,
+      "loss": 0.3819,
+      "step": 4877
+    },
+    {
+      "epoch": 1.1611233414648658,
+      "grad_norm": 0.36664046373827575,
+      "learning_rate": 3.9486917869697795e-06,
+      "loss": 0.3174,
+      "step": 4878
+    },
+    {
+      "epoch": 1.1613613375379306,
+      "grad_norm": 0.36434472833349824,
+      "learning_rate": 3.9468073622426574e-06,
+      "loss": 0.3156,
+      "step": 4879
+    },
+    {
+      "epoch": 1.1615993336109953,
+      "grad_norm": 0.39980314450070703,
+      "learning_rate": 3.944923094046016e-06,
+      "loss": 0.3815,
+      "step": 4880
+    },
+    {
+      "epoch": 1.1618373296840603,
+      "grad_norm": 0.3825456910870827,
+      "learning_rate": 3.9430389826599026e-06,
+      "loss": 0.3578,
+      "step": 4881
+    },
+    {
+      "epoch": 1.162075325757125,
+      "grad_norm": 0.3901463596231784,
+      "learning_rate": 3.9411550283643465e-06,
+      "loss": 0.2817,
+      "step": 4882
+    },
+    {
+      "epoch": 1.1623133218301898,
+      "grad_norm": 0.4095148078840206,
+      "learning_rate": 3.939271231439348e-06,
+      "loss": 0.3111,
+      "step": 4883
+    },
+    {
+      "epoch": 1.1625513179032545,
+      "grad_norm": 0.40126458626689326,
+      "learning_rate": 3.937387592164884e-06,
+      "loss": 0.4164,
+      "step": 4884
+    },
+    {
+      "epoch": 1.1627893139763195,
+      "grad_norm": 0.40524146316425336,
+      "learning_rate": 3.935504110820912e-06,
+      "loss": 0.3031,
+      "step": 4885
+    },
+    {
+      "epoch": 1.1630273100493842,
+      "grad_norm": 0.41484043466115905,
+      "learning_rate": 3.933620787687365e-06,
+      "loss": 0.2683,
+      "step": 4886
+    },
+    {
+      "epoch": 1.163265306122449,
+      "grad_norm": 0.42894419912735265,
+      "learning_rate": 3.931737623044149e-06,
+      "loss": 0.3446,
+      "step": 4887
+    },
+    {
+      "epoch": 1.1635033021955137,
+      "grad_norm": 0.37569764295018193,
+      "learning_rate": 3.929854617171149e-06,
+      "loss": 0.3875,
+      "step": 4888
+    },
+    {
+      "epoch": 1.1637412982685786,
+      "grad_norm": 0.38169828943839607,
+      "learning_rate": 3.927971770348228e-06,
+      "loss": 0.2981,
+      "step": 4889
+    },
+    {
+      "epoch": 1.1639792943416434,
+      "grad_norm": 0.4326907379906986,
+      "learning_rate": 3.9260890828552225e-06,
+      "loss": 0.321,
+      "step": 4890
+    },
+    {
+      "epoch": 1.1642172904147081,
+      "grad_norm": 0.37216325656330895,
+      "learning_rate": 3.924206554971947e-06,
+      "loss": 0.3721,
+      "step": 4891
+    },
+    {
+      "epoch": 1.1644552864877729,
+      "grad_norm": 0.38522417456420815,
+      "learning_rate": 3.92232418697819e-06,
+      "loss": 0.3308,
+      "step": 4892
+    },
+    {
+      "epoch": 1.1646932825608378,
+      "grad_norm": 0.40316171495665337,
+      "learning_rate": 3.920441979153721e-06,
+      "loss": 0.2768,
+      "step": 4893
+    },
+    {
+      "epoch": 1.1649312786339026,
+      "grad_norm": 0.37987776460120803,
+      "learning_rate": 3.918559931778277e-06,
+      "loss": 0.3307,
+      "step": 4894
+    },
+    {
+      "epoch": 1.1651692747069673,
+      "grad_norm": 0.3813096868034741,
+      "learning_rate": 3.916678045131584e-06,
+      "loss": 0.3603,
+      "step": 4895
+    },
+    {
+      "epoch": 1.165407270780032,
+      "grad_norm": 0.351364277393361,
+      "learning_rate": 3.914796319493333e-06,
+      "loss": 0.2873,
+      "step": 4896
+    },
+    {
+      "epoch": 1.165645266853097,
+      "grad_norm": 0.3926597142637336,
+      "learning_rate": 3.912914755143196e-06,
+      "loss": 0.3285,
+      "step": 4897
+    },
+    {
+      "epoch": 1.1658832629261617,
+      "grad_norm": 0.3905874686541914,
+      "learning_rate": 3.911033352360818e-06,
+      "loss": 0.3833,
+      "step": 4898
+    },
+    {
+      "epoch": 1.1661212589992265,
+      "grad_norm": 0.35936466590080296,
+      "learning_rate": 3.909152111425825e-06,
+      "loss": 0.3381,
+      "step": 4899
+    },
+    {
+      "epoch": 1.1663592550722912,
+      "grad_norm": 0.3725108986050583,
+      "learning_rate": 3.907271032617815e-06,
+      "loss": 0.2671,
+      "step": 4900
+    },
+    {
+      "epoch": 1.1665972511453562,
+      "grad_norm": 0.3835107673363921,
+      "learning_rate": 3.905390116216362e-06,
+      "loss": 0.3108,
+      "step": 4901
+    },
+    {
+      "epoch": 1.166835247218421,
+      "grad_norm": 0.3562168456730705,
+      "learning_rate": 3.9035093625010164e-06,
+      "loss": 0.379,
+      "step": 4902
+    },
+    {
+      "epoch": 1.1670732432914857,
+      "grad_norm": 0.34281015797938996,
+      "learning_rate": 3.901628771751306e-06,
+      "loss": 0.3064,
+      "step": 4903
+    },
+    {
+      "epoch": 1.1673112393645504,
+      "grad_norm": 0.40233559709709515,
+      "learning_rate": 3.899748344246732e-06,
+      "loss": 0.3005,
+      "step": 4904
+    },
+    {
+      "epoch": 1.1675492354376154,
+      "grad_norm": 0.422059793045895,
+      "learning_rate": 3.897868080266774e-06,
+      "loss": 0.3804,
+      "step": 4905
+    },
+    {
+      "epoch": 1.16778723151068,
+      "grad_norm": 0.4089575743526671,
+      "learning_rate": 3.895987980090884e-06,
+      "loss": 0.332,
+      "step": 4906
+    },
+    {
+      "epoch": 1.1680252275837448,
+      "grad_norm": 0.3718624537861076,
+      "learning_rate": 3.894108043998492e-06,
+      "loss": 0.2945,
+      "step": 4907
+    },
+    {
+      "epoch": 1.1682632236568096,
+      "grad_norm": 0.4028263680996291,
+      "learning_rate": 3.8922282722690006e-06,
+      "loss": 0.3229,
+      "step": 4908
+    },
+    {
+      "epoch": 1.1685012197298745,
+      "grad_norm": 0.45190473479232407,
+      "learning_rate": 3.890348665181796e-06,
+      "loss": 0.3799,
+      "step": 4909
+    },
+    {
+      "epoch": 1.1687392158029393,
+      "grad_norm": 0.37951311105400887,
+      "learning_rate": 3.888469223016231e-06,
+      "loss": 0.3311,
+      "step": 4910
+    },
+    {
+      "epoch": 1.168977211876004,
+      "grad_norm": 0.398817375650266,
+      "learning_rate": 3.886589946051637e-06,
+      "loss": 0.3005,
+      "step": 4911
+    },
+    {
+      "epoch": 1.1692152079490687,
+      "grad_norm": 0.3788321557640797,
+      "learning_rate": 3.884710834567321e-06,
+      "loss": 0.3668,
+      "step": 4912
+    },
+    {
+      "epoch": 1.1694532040221337,
+      "grad_norm": 0.3679850869739202,
+      "learning_rate": 3.882831888842566e-06,
+      "loss": 0.4159,
+      "step": 4913
+    },
+    {
+      "epoch": 1.1696912000951984,
+      "grad_norm": 0.38977532532858084,
+      "learning_rate": 3.880953109156631e-06,
+      "loss": 0.2969,
+      "step": 4914
+    },
+    {
+      "epoch": 1.1699291961682632,
+      "grad_norm": 0.3673051411020341,
+      "learning_rate": 3.879074495788746e-06,
+      "loss": 0.3382,
+      "step": 4915
+    },
+    {
+      "epoch": 1.170167192241328,
+      "grad_norm": 0.40474641214394,
+      "learning_rate": 3.8771960490181226e-06,
+      "loss": 0.3647,
+      "step": 4916
+    },
+    {
+      "epoch": 1.1704051883143929,
+      "grad_norm": 0.3677584145288695,
+      "learning_rate": 3.875317769123943e-06,
+      "loss": 0.3355,
+      "step": 4917
+    },
+    {
+      "epoch": 1.1706431843874576,
+      "grad_norm": 0.38928549279284,
+      "learning_rate": 3.873439656385367e-06,
+      "loss": 0.2852,
+      "step": 4918
+    },
+    {
+      "epoch": 1.1708811804605224,
+      "grad_norm": 0.40919373160433137,
+      "learning_rate": 3.871561711081526e-06,
+      "loss": 0.3662,
+      "step": 4919
+    },
+    {
+      "epoch": 1.171119176533587,
+      "grad_norm": 0.3911851721735495,
+      "learning_rate": 3.869683933491533e-06,
+      "loss": 0.3833,
+      "step": 4920
+    },
+    {
+      "epoch": 1.171357172606652,
+      "grad_norm": 0.382574197170867,
+      "learning_rate": 3.8678063238944674e-06,
+      "loss": 0.31,
+      "step": 4921
+    },
+    {
+      "epoch": 1.1715951686797168,
+      "grad_norm": 0.37949281565255283,
+      "learning_rate": 3.865928882569392e-06,
+      "loss": 0.3181,
+      "step": 4922
+    },
+    {
+      "epoch": 1.1718331647527815,
+      "grad_norm": 0.38403965708871085,
+      "learning_rate": 3.86405160979534e-06,
+      "loss": 0.3739,
+      "step": 4923
+    },
+    {
+      "epoch": 1.1720711608258463,
+      "grad_norm": 0.38512569570032895,
+      "learning_rate": 3.8621745058513225e-06,
+      "loss": 0.3229,
+      "step": 4924
+    },
+    {
+      "epoch": 1.1723091568989112,
+      "grad_norm": 0.3957163771288543,
+      "learning_rate": 3.8602975710163205e-06,
+      "loss": 0.2844,
+      "step": 4925
+    },
+    {
+      "epoch": 1.172547152971976,
+      "grad_norm": 0.39701056868163465,
+      "learning_rate": 3.858420805569295e-06,
+      "loss": 0.338,
+      "step": 4926
+    },
+    {
+      "epoch": 1.1727851490450407,
+      "grad_norm": 0.41141027404117647,
+      "learning_rate": 3.856544209789179e-06,
+      "loss": 0.4213,
+      "step": 4927
+    },
+    {
+      "epoch": 1.1730231451181055,
+      "grad_norm": 0.356743827281772,
+      "learning_rate": 3.854667783954882e-06,
+      "loss": 0.3054,
+      "step": 4928
+    },
+    {
+      "epoch": 1.1732611411911704,
+      "grad_norm": 0.36020835139169016,
+      "learning_rate": 3.852791528345286e-06,
+      "loss": 0.2547,
+      "step": 4929
+    },
+    {
+      "epoch": 1.1734991372642352,
+      "grad_norm": 0.39353940700048456,
+      "learning_rate": 3.85091544323925e-06,
+      "loss": 0.3495,
+      "step": 4930
+    },
+    {
+      "epoch": 1.1737371333373,
+      "grad_norm": 0.3739662096404696,
+      "learning_rate": 3.849039528915605e-06,
+      "loss": 0.3416,
+      "step": 4931
+    },
+    {
+      "epoch": 1.1739751294103646,
+      "grad_norm": 0.4021690846442395,
+      "learning_rate": 3.847163785653159e-06,
+      "loss": 0.2926,
+      "step": 4932
+    },
+    {
+      "epoch": 1.1742131254834296,
+      "grad_norm": 0.35310815717245986,
+      "learning_rate": 3.845288213730695e-06,
+      "loss": 0.3188,
+      "step": 4933
+    },
+    {
+      "epoch": 1.1744511215564943,
+      "grad_norm": 0.37742127338548703,
+      "learning_rate": 3.843412813426967e-06,
+      "loss": 0.3857,
+      "step": 4934
+    },
+    {
+      "epoch": 1.174689117629559,
+      "grad_norm": 0.36700195619184406,
+      "learning_rate": 3.84153758502071e-06,
+      "loss": 0.3065,
+      "step": 4935
+    },
+    {
+      "epoch": 1.1749271137026238,
+      "grad_norm": 0.37862390020379844,
+      "learning_rate": 3.839662528790625e-06,
+      "loss": 0.2822,
+      "step": 4936
+    },
+    {
+      "epoch": 1.1751651097756888,
+      "grad_norm": 0.37919008323525716,
+      "learning_rate": 3.837787645015395e-06,
+      "loss": 0.3459,
+      "step": 4937
+    },
+    {
+      "epoch": 1.1754031058487535,
+      "grad_norm": 0.3805425537038319,
+      "learning_rate": 3.835912933973671e-06,
+      "loss": 0.4043,
+      "step": 4938
+    },
+    {
+      "epoch": 1.1756411019218183,
+      "grad_norm": 0.3812253553869978,
+      "learning_rate": 3.834038395944084e-06,
+      "loss": 0.2991,
+      "step": 4939
+    },
+    {
+      "epoch": 1.175879097994883,
+      "grad_norm": 0.3885217222382259,
+      "learning_rate": 3.832164031205237e-06,
+      "loss": 0.3469,
+      "step": 4940
+    },
+    {
+      "epoch": 1.176117094067948,
+      "grad_norm": 0.3496306031416618,
+      "learning_rate": 3.830289840035705e-06,
+      "loss": 0.39,
+      "step": 4941
+    },
+    {
+      "epoch": 1.1763550901410127,
+      "grad_norm": 0.3725490258794375,
+      "learning_rate": 3.82841582271404e-06,
+      "loss": 0.3009,
+      "step": 4942
+    },
+    {
+      "epoch": 1.1765930862140774,
+      "grad_norm": 0.3961940913795775,
+      "learning_rate": 3.8265419795187675e-06,
+      "loss": 0.3007,
+      "step": 4943
+    },
+    {
+      "epoch": 1.1768310822871422,
+      "grad_norm": 0.40324342500628063,
+      "learning_rate": 3.824668310728387e-06,
+      "loss": 0.3132,
+      "step": 4944
+    },
+    {
+      "epoch": 1.1770690783602071,
+      "grad_norm": 0.36079064416493883,
+      "learning_rate": 3.822794816621371e-06,
+      "loss": 0.4038,
+      "step": 4945
+    },
+    {
+      "epoch": 1.1773070744332719,
+      "grad_norm": 0.40154958109483985,
+      "learning_rate": 3.8209214974761685e-06,
+      "loss": 0.2703,
+      "step": 4946
+    },
+    {
+      "epoch": 1.1775450705063366,
+      "grad_norm": 0.40470924949834836,
+      "learning_rate": 3.819048353571201e-06,
+      "loss": 0.3144,
+      "step": 4947
+    },
+    {
+      "epoch": 1.1777830665794013,
+      "grad_norm": 0.4251133239570864,
+      "learning_rate": 3.817175385184861e-06,
+      "loss": 0.3686,
+      "step": 4948
+    },
+    {
+      "epoch": 1.1780210626524663,
+      "grad_norm": 0.3824740738540186,
+      "learning_rate": 3.815302592595522e-06,
+      "loss": 0.3313,
+      "step": 4949
+    },
+    {
+      "epoch": 1.178259058725531,
+      "grad_norm": 0.43168535718039347,
+      "learning_rate": 3.813429976081526e-06,
+      "loss": 0.2672,
+      "step": 4950
+    },
+    {
+      "epoch": 1.1784970547985958,
+      "grad_norm": 0.3844314375612966,
+      "learning_rate": 3.8115575359211905e-06,
+      "loss": 0.3406,
+      "step": 4951
+    },
+    {
+      "epoch": 1.1787350508716605,
+      "grad_norm": 0.38263513845378433,
+      "learning_rate": 3.809685272392804e-06,
+      "loss": 0.3849,
+      "step": 4952
+    },
+    {
+      "epoch": 1.1789730469447255,
+      "grad_norm": 0.3613570750787706,
+      "learning_rate": 3.8078131857746346e-06,
+      "loss": 0.2964,
+      "step": 4953
+    },
+    {
+      "epoch": 1.1792110430177902,
+      "grad_norm": 0.38485653076788723,
+      "learning_rate": 3.8059412763449187e-06,
+      "loss": 0.3033,
+      "step": 4954
+    },
+    {
+      "epoch": 1.179449039090855,
+      "grad_norm": 0.3490639146215179,
+      "learning_rate": 3.804069544381869e-06,
+      "loss": 0.3481,
+      "step": 4955
+    },
+    {
+      "epoch": 1.1796870351639197,
+      "grad_norm": 0.3856291558218839,
+      "learning_rate": 3.802197990163671e-06,
+      "loss": 0.3471,
+      "step": 4956
+    },
+    {
+      "epoch": 1.1799250312369847,
+      "grad_norm": 0.39550282927113356,
+      "learning_rate": 3.8003266139684832e-06,
+      "loss": 0.2598,
+      "step": 4957
+    },
+    {
+      "epoch": 1.1801630273100494,
+      "grad_norm": 0.3862155603983133,
+      "learning_rate": 3.798455416074439e-06,
+      "loss": 0.3378,
+      "step": 4958
+    },
+    {
+      "epoch": 1.1804010233831141,
+      "grad_norm": 0.3839437629883702,
+      "learning_rate": 3.7965843967596453e-06,
+      "loss": 0.3982,
+      "step": 4959
+    },
+    {
+      "epoch": 1.1806390194561789,
+      "grad_norm": 0.4078427772915639,
+      "learning_rate": 3.7947135563021814e-06,
+      "loss": 0.2971,
+      "step": 4960
+    },
+    {
+      "epoch": 1.1808770155292438,
+      "grad_norm": 0.35307107716990815,
+      "learning_rate": 3.7928428949800996e-06,
+      "loss": 0.2988,
+      "step": 4961
+    },
+    {
+      "epoch": 1.1811150116023086,
+      "grad_norm": 0.3992799282030228,
+      "learning_rate": 3.7909724130714277e-06,
+      "loss": 0.3522,
+      "step": 4962
+    },
+    {
+      "epoch": 1.1813530076753733,
+      "grad_norm": 0.381054998500261,
+      "learning_rate": 3.7891021108541642e-06,
+      "loss": 0.3713,
+      "step": 4963
+    },
+    {
+      "epoch": 1.181591003748438,
+      "grad_norm": 0.40058461166635667,
+      "learning_rate": 3.787231988606284e-06,
+      "loss": 0.3104,
+      "step": 4964
+    },
+    {
+      "epoch": 1.181828999821503,
+      "grad_norm": 0.36686697436707344,
+      "learning_rate": 3.785362046605732e-06,
+      "loss": 0.3072,
+      "step": 4965
+    },
+    {
+      "epoch": 1.1820669958945678,
+      "grad_norm": 0.4149878680339965,
+      "learning_rate": 3.7834922851304297e-06,
+      "loss": 0.4061,
+      "step": 4966
+    },
+    {
+      "epoch": 1.1823049919676325,
+      "grad_norm": 0.3695370875718639,
+      "learning_rate": 3.7816227044582687e-06,
+      "loss": 0.3289,
+      "step": 4967
+    },
+    {
+      "epoch": 1.1825429880406972,
+      "grad_norm": 0.3596942240313325,
+      "learning_rate": 3.7797533048671146e-06,
+      "loss": 0.2954,
+      "step": 4968
+    },
+    {
+      "epoch": 1.1827809841137622,
+      "grad_norm": 0.39500031375483247,
+      "learning_rate": 3.7778840866348075e-06,
+      "loss": 0.3261,
+      "step": 4969
+    },
+    {
+      "epoch": 1.183018980186827,
+      "grad_norm": 0.41374777726632095,
+      "learning_rate": 3.7760150500391584e-06,
+      "loss": 0.3944,
+      "step": 4970
+    },
+    {
+      "epoch": 1.1832569762598917,
+      "grad_norm": 0.3537235485993441,
+      "learning_rate": 3.7741461953579527e-06,
+      "loss": 0.2971,
+      "step": 4971
+    },
+    {
+      "epoch": 1.1834949723329564,
+      "grad_norm": 0.3748434274707176,
+      "learning_rate": 3.772277522868949e-06,
+      "loss": 0.3244,
+      "step": 4972
+    },
+    {
+      "epoch": 1.1837329684060214,
+      "grad_norm": 0.38166375878188813,
+      "learning_rate": 3.770409032849878e-06,
+      "loss": 0.3598,
+      "step": 4973
+    },
+    {
+      "epoch": 1.1839709644790861,
+      "grad_norm": 0.3922577890411974,
+      "learning_rate": 3.7685407255784424e-06,
+      "loss": 0.3348,
+      "step": 4974
+    },
+    {
+      "epoch": 1.1842089605521509,
+      "grad_norm": 0.3845300676159576,
+      "learning_rate": 3.766672601332319e-06,
+      "loss": 0.2835,
+      "step": 4975
+    },
+    {
+      "epoch": 1.1844469566252156,
+      "grad_norm": 0.3864290171184251,
+      "learning_rate": 3.76480466038916e-06,
+      "loss": 0.3286,
+      "step": 4976
+    },
+    {
+      "epoch": 1.1846849526982806,
+      "grad_norm": 0.40164860660518437,
+      "learning_rate": 3.7629369030265834e-06,
+      "loss": 0.4308,
+      "step": 4977
+    },
+    {
+      "epoch": 1.1849229487713453,
+      "grad_norm": 0.38101030100174693,
+      "learning_rate": 3.7610693295221885e-06,
+      "loss": 0.278,
+      "step": 4978
+    },
+    {
+      "epoch": 1.18516094484441,
+      "grad_norm": 0.3827769312151021,
+      "learning_rate": 3.7592019401535397e-06,
+      "loss": 0.2915,
+      "step": 4979
+    },
+    {
+      "epoch": 1.1853989409174748,
+      "grad_norm": 0.39348943008486564,
+      "learning_rate": 3.7573347351981785e-06,
+      "loss": 0.3532,
+      "step": 4980
+    },
+    {
+      "epoch": 1.1856369369905397,
+      "grad_norm": 0.35366811317076075,
+      "learning_rate": 3.7554677149336186e-06,
+      "loss": 0.355,
+      "step": 4981
+    },
+    {
+      "epoch": 1.1858749330636045,
+      "grad_norm": 0.3730839491699928,
+      "learning_rate": 3.7536008796373447e-06,
+      "loss": 0.2738,
+      "step": 4982
+    },
+    {
+      "epoch": 1.1861129291366692,
+      "grad_norm": 0.36612366371157234,
+      "learning_rate": 3.7517342295868142e-06,
+      "loss": 0.3053,
+      "step": 4983
+    },
+    {
+      "epoch": 1.186350925209734,
+      "grad_norm": 0.3823419705705663,
+      "learning_rate": 3.7498677650594585e-06,
+      "loss": 0.3915,
+      "step": 4984
+    },
+    {
+      "epoch": 1.186588921282799,
+      "grad_norm": 0.36531157147493043,
+      "learning_rate": 3.7480014863326786e-06,
+      "loss": 0.2942,
+      "step": 4985
+    },
+    {
+      "epoch": 1.1868269173558637,
+      "grad_norm": 0.401505598688106,
+      "learning_rate": 3.746135393683851e-06,
+      "loss": 0.2788,
+      "step": 4986
+    },
+    {
+      "epoch": 1.1870649134289284,
+      "grad_norm": 0.38093798207572305,
+      "learning_rate": 3.7442694873903236e-06,
+      "loss": 0.3348,
+      "step": 4987
+    },
+    {
+      "epoch": 1.1873029095019931,
+      "grad_norm": 0.38132115350268836,
+      "learning_rate": 3.742403767729414e-06,
+      "loss": 0.3639,
+      "step": 4988
+    },
+    {
+      "epoch": 1.187540905575058,
+      "grad_norm": 0.39346856748575865,
+      "learning_rate": 3.740538234978417e-06,
+      "loss": 0.2893,
+      "step": 4989
+    },
+    {
+      "epoch": 1.1877789016481228,
+      "grad_norm": 0.3613470905129032,
+      "learning_rate": 3.7386728894145965e-06,
+      "loss": 0.3158,
+      "step": 4990
+    },
+    {
+      "epoch": 1.1880168977211876,
+      "grad_norm": 0.3715632977391001,
+      "learning_rate": 3.7368077313151866e-06,
+      "loss": 0.4092,
+      "step": 4991
+    },
+    {
+      "epoch": 1.1882548937942523,
+      "grad_norm": 0.43216241244974535,
+      "learning_rate": 3.7349427609573985e-06,
+      "loss": 0.3197,
+      "step": 4992
+    },
+    {
+      "epoch": 1.1884928898673173,
+      "grad_norm": 0.38373124295883176,
+      "learning_rate": 3.7330779786184122e-06,
+      "loss": 0.3042,
+      "step": 4993
+    },
+    {
+      "epoch": 1.188730885940382,
+      "grad_norm": 0.36301878619729605,
+      "learning_rate": 3.731213384575381e-06,
+      "loss": 0.3551,
+      "step": 4994
+    },
+    {
+      "epoch": 1.1889688820134467,
+      "grad_norm": 0.3633370905617256,
+      "learning_rate": 3.7293489791054293e-06,
+      "loss": 0.381,
+      "step": 4995
+    },
+    {
+      "epoch": 1.1892068780865115,
+      "grad_norm": 0.36132719462940693,
+      "learning_rate": 3.727484762485653e-06,
+      "loss": 0.2943,
+      "step": 4996
+    },
+    {
+      "epoch": 1.1894448741595764,
+      "grad_norm": 0.3626513454326024,
+      "learning_rate": 3.7256207349931216e-06,
+      "loss": 0.3158,
+      "step": 4997
+    },
+    {
+      "epoch": 1.1896828702326412,
+      "grad_norm": 0.39265659273339876,
+      "learning_rate": 3.7237568969048766e-06,
+      "loss": 0.3878,
+      "step": 4998
+    },
+    {
+      "epoch": 1.189920866305706,
+      "grad_norm": 0.35277130854248157,
+      "learning_rate": 3.7218932484979287e-06,
+      "loss": 0.338,
+      "step": 4999
+    },
+    {
+      "epoch": 1.1901588623787707,
+      "grad_norm": 0.39473980233793865,
+      "learning_rate": 3.7200297900492632e-06,
+      "loss": 0.2937,
+      "step": 5000
+    },
+    {
+      "epoch": 1.1903968584518356,
+      "grad_norm": 0.3618280922673368,
+      "learning_rate": 3.7181665218358354e-06,
+      "loss": 0.3374,
+      "step": 5001
+    },
+    {
+      "epoch": 1.1906348545249004,
+      "grad_norm": 0.39899118970440267,
+      "learning_rate": 3.7163034441345725e-06,
+      "loss": 0.3833,
+      "step": 5002
+    },
+    {
+      "epoch": 1.190872850597965,
+      "grad_norm": 0.36708868631705965,
+      "learning_rate": 3.7144405572223762e-06,
+      "loss": 0.3217,
+      "step": 5003
+    },
+    {
+      "epoch": 1.1911108466710298,
+      "grad_norm": 0.36066225989423434,
+      "learning_rate": 3.7125778613761164e-06,
+      "loss": 0.3079,
+      "step": 5004
+    },
+    {
+      "epoch": 1.1913488427440948,
+      "grad_norm": 0.4031849072858775,
+      "learning_rate": 3.710715356872634e-06,
+      "loss": 0.3519,
+      "step": 5005
+    },
+    {
+      "epoch": 1.1915868388171595,
+      "grad_norm": 0.39768511346134466,
+      "learning_rate": 3.708853043988746e-06,
+      "loss": 0.3612,
+      "step": 5006
+    },
+    {
+      "epoch": 1.1918248348902243,
+      "grad_norm": 0.3746114220221446,
+      "learning_rate": 3.7069909230012376e-06,
+      "loss": 0.3005,
+      "step": 5007
+    },
+    {
+      "epoch": 1.192062830963289,
+      "grad_norm": 0.3935100573355243,
+      "learning_rate": 3.705128994186865e-06,
+      "loss": 0.3071,
+      "step": 5008
+    },
+    {
+      "epoch": 1.192300827036354,
+      "grad_norm": 0.3544256856276741,
+      "learning_rate": 3.7032672578223583e-06,
+      "loss": 0.3868,
+      "step": 5009
+    },
+    {
+      "epoch": 1.1925388231094187,
+      "grad_norm": 0.37426633165695083,
+      "learning_rate": 3.701405714184416e-06,
+      "loss": 0.2953,
+      "step": 5010
+    },
+    {
+      "epoch": 1.1927768191824835,
+      "grad_norm": 0.42034731649414875,
+      "learning_rate": 3.699544363549711e-06,
+      "loss": 0.3005,
+      "step": 5011
+    },
+    {
+      "epoch": 1.1930148152555482,
+      "grad_norm": 0.4082503637885382,
+      "learning_rate": 3.6976832061948845e-06,
+      "loss": 0.3477,
+      "step": 5012
+    },
+    {
+      "epoch": 1.1932528113286132,
+      "grad_norm": 0.8416333609801906,
+      "learning_rate": 3.695822242396552e-06,
+      "loss": 0.3687,
+      "step": 5013
+    },
+    {
+      "epoch": 1.193490807401678,
+      "grad_norm": 0.37912063370018584,
+      "learning_rate": 3.693961472431298e-06,
+      "loss": 0.2979,
+      "step": 5014
+    },
+    {
+      "epoch": 1.1937288034747426,
+      "grad_norm": 0.401423257824879,
+      "learning_rate": 3.6921008965756775e-06,
+      "loss": 0.3199,
+      "step": 5015
+    },
+    {
+      "epoch": 1.1939667995478074,
+      "grad_norm": 0.3800891163492014,
+      "learning_rate": 3.690240515106221e-06,
+      "loss": 0.3765,
+      "step": 5016
+    },
+    {
+      "epoch": 1.1942047956208723,
+      "grad_norm": 0.3669803455039553,
+      "learning_rate": 3.6883803282994256e-06,
+      "loss": 0.3246,
+      "step": 5017
+    },
+    {
+      "epoch": 1.194442791693937,
+      "grad_norm": 0.4200743670629511,
+      "learning_rate": 3.6865203364317605e-06,
+      "loss": 0.2893,
+      "step": 5018
+    },
+    {
+      "epoch": 1.1946807877670018,
+      "grad_norm": 0.37720436817760933,
+      "learning_rate": 3.6846605397796677e-06,
+      "loss": 0.3496,
+      "step": 5019
+    },
+    {
+      "epoch": 1.1949187838400666,
+      "grad_norm": 0.38267177297360994,
+      "learning_rate": 3.6828009386195592e-06,
+      "loss": 0.3647,
+      "step": 5020
+    },
+    {
+      "epoch": 1.1951567799131315,
+      "grad_norm": 0.3855032217341476,
+      "learning_rate": 3.680941533227817e-06,
+      "loss": 0.2893,
+      "step": 5021
+    },
+    {
+      "epoch": 1.1953947759861963,
+      "grad_norm": 0.41661785772795806,
+      "learning_rate": 3.679082323880795e-06,
+      "loss": 0.2966,
+      "step": 5022
+    },
+    {
+      "epoch": 1.195632772059261,
+      "grad_norm": 0.38996824463256474,
+      "learning_rate": 3.6772233108548182e-06,
+      "loss": 0.3682,
+      "step": 5023
+    },
+    {
+      "epoch": 1.1958707681323257,
+      "grad_norm": 0.3579485476112444,
+      "learning_rate": 3.6753644944261806e-06,
+      "loss": 0.3461,
+      "step": 5024
+    },
+    {
+      "epoch": 1.1961087642053907,
+      "grad_norm": 0.38986130206284325,
+      "learning_rate": 3.6735058748711492e-06,
+      "loss": 0.2968,
+      "step": 5025
+    },
+    {
+      "epoch": 1.1963467602784554,
+      "grad_norm": 0.3703854439689402,
+      "learning_rate": 3.6716474524659608e-06,
+      "loss": 0.3479,
+      "step": 5026
+    },
+    {
+      "epoch": 1.1965847563515202,
+      "grad_norm": 0.4016888133573442,
+      "learning_rate": 3.669789227486823e-06,
+      "loss": 0.3763,
+      "step": 5027
+    },
+    {
+      "epoch": 1.196822752424585,
+      "grad_norm": 0.3767473882222918,
+      "learning_rate": 3.667931200209913e-06,
+      "loss": 0.2947,
+      "step": 5028
+    },
+    {
+      "epoch": 1.1970607484976499,
+      "grad_norm": 0.39627616648735325,
+      "learning_rate": 3.6660733709113805e-06,
+      "loss": 0.2915,
+      "step": 5029
+    },
+    {
+      "epoch": 1.1972987445707146,
+      "grad_norm": 0.3824120166417069,
+      "learning_rate": 3.664215739867345e-06,
+      "loss": 0.3831,
+      "step": 5030
+    },
+    {
+      "epoch": 1.1975367406437794,
+      "grad_norm": 0.3796189625607475,
+      "learning_rate": 3.662358307353897e-06,
+      "loss": 0.3594,
+      "step": 5031
+    },
+    {
+      "epoch": 1.197774736716844,
+      "grad_norm": 0.4057709147757602,
+      "learning_rate": 3.6605010736470945e-06,
+      "loss": 0.3109,
+      "step": 5032
+    },
+    {
+      "epoch": 1.198012732789909,
+      "grad_norm": 0.38925059478299473,
+      "learning_rate": 3.6586440390229705e-06,
+      "loss": 0.3187,
+      "step": 5033
+    },
+    {
+      "epoch": 1.1982507288629738,
+      "grad_norm": 0.3795304534109489,
+      "learning_rate": 3.656787203757527e-06,
+      "loss": 0.4042,
+      "step": 5034
+    },
+    {
+      "epoch": 1.1984887249360385,
+      "grad_norm": 0.33530850356764125,
+      "learning_rate": 3.654930568126734e-06,
+      "loss": 0.2998,
+      "step": 5035
+    },
+    {
+      "epoch": 1.1987267210091033,
+      "grad_norm": 0.43587793237382827,
+      "learning_rate": 3.6530741324065343e-06,
+      "loss": 0.3011,
+      "step": 5036
+    },
+    {
+      "epoch": 1.1989647170821682,
+      "grad_norm": 0.392590497330817,
+      "learning_rate": 3.65121789687284e-06,
+      "loss": 0.3432,
+      "step": 5037
+    },
+    {
+      "epoch": 1.199202713155233,
+      "grad_norm": 0.4238986802151149,
+      "learning_rate": 3.6493618618015335e-06,
+      "loss": 0.3859,
+      "step": 5038
+    },
+    {
+      "epoch": 1.1994407092282977,
+      "grad_norm": 0.3718207166623785,
+      "learning_rate": 3.647506027468467e-06,
+      "loss": 0.2891,
+      "step": 5039
+    },
+    {
+      "epoch": 1.1996787053013624,
+      "grad_norm": 0.39411079054515186,
+      "learning_rate": 3.645650394149465e-06,
+      "loss": 0.3328,
+      "step": 5040
+    },
+    {
+      "epoch": 1.1999167013744274,
+      "grad_norm": 0.36581910992126304,
+      "learning_rate": 3.6437949621203184e-06,
+      "loss": 0.3807,
+      "step": 5041
+    },
+    {
+      "epoch": 1.2001546974474921,
+      "grad_norm": 0.3661757831955536,
+      "learning_rate": 3.6419397316567902e-06,
+      "loss": 0.3113,
+      "step": 5042
+    },
+    {
+      "epoch": 1.2003926935205569,
+      "grad_norm": 0.45796517113773294,
+      "learning_rate": 3.640084703034616e-06,
+      "loss": 0.3018,
+      "step": 5043
+    },
+    {
+      "epoch": 1.2006306895936216,
+      "grad_norm": 0.35520419192547475,
+      "learning_rate": 3.6382298765294978e-06,
+      "loss": 0.3468,
+      "step": 5044
+    },
+    {
+      "epoch": 1.2008686856666866,
+      "grad_norm": 0.3770376085730245,
+      "learning_rate": 3.6363752524171083e-06,
+      "loss": 0.3914,
+      "step": 5045
+    },
+    {
+      "epoch": 1.2011066817397513,
+      "grad_norm": 0.37954607416403163,
+      "learning_rate": 3.6345208309730885e-06,
+      "loss": 0.3111,
+      "step": 5046
+    },
+    {
+      "epoch": 1.201344677812816,
+      "grad_norm": 0.3863432844778005,
+      "learning_rate": 3.632666612473056e-06,
+      "loss": 0.319,
+      "step": 5047
+    },
+    {
+      "epoch": 1.2015826738858808,
+      "grad_norm": 0.4500532032678951,
+      "learning_rate": 3.630812597192591e-06,
+      "loss": 0.3771,
+      "step": 5048
+    },
+    {
+      "epoch": 1.2018206699589458,
+      "grad_norm": 0.37110161136850156,
+      "learning_rate": 3.628958785407246e-06,
+      "loss": 0.3539,
+      "step": 5049
+    },
+    {
+      "epoch": 1.2020586660320105,
+      "grad_norm": 0.38550398130650293,
+      "learning_rate": 3.6271051773925434e-06,
+      "loss": 0.2796,
+      "step": 5050
+    },
+    {
+      "epoch": 1.2022966621050752,
+      "grad_norm": 0.41164229002791736,
+      "learning_rate": 3.6252517734239757e-06,
+      "loss": 0.3169,
+      "step": 5051
+    },
+    {
+      "epoch": 1.20253465817814,
+      "grad_norm": 0.4273076385043179,
+      "learning_rate": 3.6233985737770034e-06,
+      "loss": 0.3748,
+      "step": 5052
+    },
+    {
+      "epoch": 1.202772654251205,
+      "grad_norm": 0.38342810948189754,
+      "learning_rate": 3.6215455787270587e-06,
+      "loss": 0.3031,
+      "step": 5053
+    },
+    {
+      "epoch": 1.2030106503242697,
+      "grad_norm": 0.3815063828509451,
+      "learning_rate": 3.6196927885495426e-06,
+      "loss": 0.3084,
+      "step": 5054
+    },
+    {
+      "epoch": 1.2032486463973344,
+      "grad_norm": 0.39701942080159974,
+      "learning_rate": 3.617840203519825e-06,
+      "loss": 0.3449,
+      "step": 5055
+    },
+    {
+      "epoch": 1.2034866424703992,
+      "grad_norm": 0.3826077784262676,
+      "learning_rate": 3.6159878239132453e-06,
+      "loss": 0.3861,
+      "step": 5056
+    },
+    {
+      "epoch": 1.2037246385434641,
+      "grad_norm": 0.3761935038939176,
+      "learning_rate": 3.614135650005115e-06,
+      "loss": 0.2809,
+      "step": 5057
+    },
+    {
+      "epoch": 1.2039626346165289,
+      "grad_norm": 0.38813967758028695,
+      "learning_rate": 3.6122836820707107e-06,
+      "loss": 0.3292,
+      "step": 5058
+    },
+    {
+      "epoch": 1.2042006306895936,
+      "grad_norm": 0.4094613511832653,
+      "learning_rate": 3.6104319203852826e-06,
+      "loss": 0.3954,
+      "step": 5059
+    },
+    {
+      "epoch": 1.2044386267626583,
+      "grad_norm": 0.35131103373017636,
+      "learning_rate": 3.608580365224045e-06,
+      "loss": 0.3157,
+      "step": 5060
+    },
+    {
+      "epoch": 1.2046766228357233,
+      "grad_norm": 0.3779833656079706,
+      "learning_rate": 3.60672901686219e-06,
+      "loss": 0.3155,
+      "step": 5061
+    },
+    {
+      "epoch": 1.204914618908788,
+      "grad_norm": 0.3801342639897726,
+      "learning_rate": 3.60487787557487e-06,
+      "loss": 0.3412,
+      "step": 5062
+    },
+    {
+      "epoch": 1.2051526149818528,
+      "grad_norm": 0.3661980162834189,
+      "learning_rate": 3.603026941637212e-06,
+      "loss": 0.3519,
+      "step": 5063
+    },
+    {
+      "epoch": 1.2053906110549175,
+      "grad_norm": 0.3839709384547976,
+      "learning_rate": 3.6011762153243096e-06,
+      "loss": 0.3023,
+      "step": 5064
+    },
+    {
+      "epoch": 1.2056286071279825,
+      "grad_norm": 0.39407579614935423,
+      "learning_rate": 3.5993256969112266e-06,
+      "loss": 0.3062,
+      "step": 5065
+    },
+    {
+      "epoch": 1.2058666032010472,
+      "grad_norm": 0.3593214219415961,
+      "learning_rate": 3.5974753866729966e-06,
+      "loss": 0.4125,
+      "step": 5066
+    },
+    {
+      "epoch": 1.206104599274112,
+      "grad_norm": 0.3672574458694017,
+      "learning_rate": 3.5956252848846206e-06,
+      "loss": 0.3438,
+      "step": 5067
+    },
+    {
+      "epoch": 1.2063425953471767,
+      "grad_norm": 0.3723709621082384,
+      "learning_rate": 3.5937753918210705e-06,
+      "loss": 0.2843,
+      "step": 5068
+    },
+    {
+      "epoch": 1.2065805914202417,
+      "grad_norm": 0.3901256291535116,
+      "learning_rate": 3.5919257077572835e-06,
+      "loss": 0.3146,
+      "step": 5069
+    },
+    {
+      "epoch": 1.2068185874933064,
+      "grad_norm": 0.37899864532094263,
+      "learning_rate": 3.5900762329681717e-06,
+      "loss": 0.3923,
+      "step": 5070
+    },
+    {
+      "epoch": 1.2070565835663711,
+      "grad_norm": 0.35151540606068965,
+      "learning_rate": 3.5882269677286117e-06,
+      "loss": 0.2781,
+      "step": 5071
+    },
+    {
+      "epoch": 1.2072945796394359,
+      "grad_norm": 0.3834887945303703,
+      "learning_rate": 3.58637791231345e-06,
+      "loss": 0.3403,
+      "step": 5072
+    },
+    {
+      "epoch": 1.2075325757125008,
+      "grad_norm": 0.3971997277785191,
+      "learning_rate": 3.5845290669975015e-06,
+      "loss": 0.3797,
+      "step": 5073
+    },
+    {
+      "epoch": 1.2077705717855656,
+      "grad_norm": 0.36240585843749556,
+      "learning_rate": 3.5826804320555486e-06,
+      "loss": 0.3157,
+      "step": 5074
+    },
+    {
+      "epoch": 1.2080085678586303,
+      "grad_norm": 0.3767347339492281,
+      "learning_rate": 3.5808320077623485e-06,
+      "loss": 0.2653,
+      "step": 5075
+    },
+    {
+      "epoch": 1.208246563931695,
+      "grad_norm": 0.38456131915052816,
+      "learning_rate": 3.5789837943926208e-06,
+      "loss": 0.3516,
+      "step": 5076
+    },
+    {
+      "epoch": 1.20848456000476,
+      "grad_norm": 0.4128907315601859,
+      "learning_rate": 3.5771357922210555e-06,
+      "loss": 0.3727,
+      "step": 5077
+    },
+    {
+      "epoch": 1.2087225560778248,
+      "grad_norm": 0.3723204263394461,
+      "learning_rate": 3.5752880015223113e-06,
+      "loss": 0.2884,
+      "step": 5078
+    },
+    {
+      "epoch": 1.2089605521508895,
+      "grad_norm": 0.3637169168437252,
+      "learning_rate": 3.5734404225710157e-06,
+      "loss": 0.2807,
+      "step": 5079
+    },
+    {
+      "epoch": 1.2091985482239542,
+      "grad_norm": 0.38764628752736946,
+      "learning_rate": 3.5715930556417644e-06,
+      "loss": 0.3578,
+      "step": 5080
+    },
+    {
+      "epoch": 1.2094365442970192,
+      "grad_norm": 0.3450780601550473,
+      "learning_rate": 3.569745901009123e-06,
+      "loss": 0.3605,
+      "step": 5081
+    },
+    {
+      "epoch": 1.209674540370084,
+      "grad_norm": 0.3581139717614697,
+      "learning_rate": 3.5678989589476228e-06,
+      "loss": 0.2712,
+      "step": 5082
+    },
+    {
+      "epoch": 1.2099125364431487,
+      "grad_norm": 0.34988502902927887,
+      "learning_rate": 3.5660522297317648e-06,
+      "loss": 0.3645,
+      "step": 5083
+    },
+    {
+      "epoch": 1.2101505325162134,
+      "grad_norm": 0.3708416415228171,
+      "learning_rate": 3.5642057136360205e-06,
+      "loss": 0.3858,
+      "step": 5084
+    },
+    {
+      "epoch": 1.2103885285892784,
+      "grad_norm": 0.38942129205042847,
+      "learning_rate": 3.562359410934827e-06,
+      "loss": 0.3324,
+      "step": 5085
+    },
+    {
+      "epoch": 1.210626524662343,
+      "grad_norm": 0.3806323119990266,
+      "learning_rate": 3.560513321902591e-06,
+      "loss": 0.29,
+      "step": 5086
+    },
+    {
+      "epoch": 1.2108645207354078,
+      "grad_norm": 0.3700249289986975,
+      "learning_rate": 3.5586674468136838e-06,
+      "loss": 0.3722,
+      "step": 5087
+    },
+    {
+      "epoch": 1.2111025168084726,
+      "grad_norm": 0.3719331028186071,
+      "learning_rate": 3.5568217859424535e-06,
+      "loss": 0.3775,
+      "step": 5088
+    },
+    {
+      "epoch": 1.2113405128815375,
+      "grad_norm": 0.36400596249820083,
+      "learning_rate": 3.554976339563209e-06,
+      "loss": 0.3116,
+      "step": 5089
+    },
+    {
+      "epoch": 1.2115785089546023,
+      "grad_norm": 0.38994792469616396,
+      "learning_rate": 3.553131107950227e-06,
+      "loss": 0.3026,
+      "step": 5090
+    },
+    {
+      "epoch": 1.211816505027667,
+      "grad_norm": 0.3941320652284189,
+      "learning_rate": 3.551286091377757e-06,
+      "loss": 0.3698,
+      "step": 5091
+    },
+    {
+      "epoch": 1.2120545011007318,
+      "grad_norm": 0.4046036167301338,
+      "learning_rate": 3.549441290120013e-06,
+      "loss": 0.3097,
+      "step": 5092
+    },
+    {
+      "epoch": 1.2122924971737967,
+      "grad_norm": 0.4039353108910549,
+      "learning_rate": 3.547596704451179e-06,
+      "loss": 0.2699,
+      "step": 5093
+    },
+    {
+      "epoch": 1.2125304932468615,
+      "grad_norm": 0.4363727370434713,
+      "learning_rate": 3.545752334645405e-06,
+      "loss": 0.3303,
+      "step": 5094
+    },
+    {
+      "epoch": 1.2127684893199262,
+      "grad_norm": 0.3840952200590034,
+      "learning_rate": 3.5439081809768103e-06,
+      "loss": 0.3837,
+      "step": 5095
+    },
+    {
+      "epoch": 1.213006485392991,
+      "grad_norm": 0.37439745222911935,
+      "learning_rate": 3.5420642437194807e-06,
+      "loss": 0.3031,
+      "step": 5096
+    },
+    {
+      "epoch": 1.213244481466056,
+      "grad_norm": 0.40481232665765876,
+      "learning_rate": 3.540220523147474e-06,
+      "loss": 0.3129,
+      "step": 5097
+    },
+    {
+      "epoch": 1.2134824775391206,
+      "grad_norm": 0.3706994690208066,
+      "learning_rate": 3.53837701953481e-06,
+      "loss": 0.3574,
+      "step": 5098
+    },
+    {
+      "epoch": 1.2137204736121854,
+      "grad_norm": 0.36982293775644515,
+      "learning_rate": 3.53653373315548e-06,
+      "loss": 0.3361,
+      "step": 5099
+    },
+    {
+      "epoch": 1.2139584696852501,
+      "grad_norm": 0.44289417728869535,
+      "learning_rate": 3.534690664283441e-06,
+      "loss": 0.2839,
+      "step": 5100
+    },
+    {
+      "epoch": 1.214196465758315,
+      "grad_norm": 0.39709885081786933,
+      "learning_rate": 3.5328478131926182e-06,
+      "loss": 0.3385,
+      "step": 5101
+    },
+    {
+      "epoch": 1.2144344618313798,
+      "grad_norm": 0.386735844300816,
+      "learning_rate": 3.5310051801569077e-06,
+      "loss": 0.3861,
+      "step": 5102
+    },
+    {
+      "epoch": 1.2146724579044446,
+      "grad_norm": 0.38596849207457884,
+      "learning_rate": 3.5291627654501683e-06,
+      "loss": 0.3018,
+      "step": 5103
+    },
+    {
+      "epoch": 1.2149104539775093,
+      "grad_norm": 0.3772909341628557,
+      "learning_rate": 3.5273205693462294e-06,
+      "loss": 0.276,
+      "step": 5104
+    },
+    {
+      "epoch": 1.2151484500505743,
+      "grad_norm": 0.40038122463925,
+      "learning_rate": 3.5254785921188855e-06,
+      "loss": 0.3544,
+      "step": 5105
+    },
+    {
+      "epoch": 1.215386446123639,
+      "grad_norm": 0.394545199953025,
+      "learning_rate": 3.5236368340419015e-06,
+      "loss": 0.3795,
+      "step": 5106
+    },
+    {
+      "epoch": 1.2156244421967037,
+      "grad_norm": 0.37930731017600633,
+      "learning_rate": 3.5217952953890065e-06,
+      "loss": 0.2946,
+      "step": 5107
+    },
+    {
+      "epoch": 1.2158624382697685,
+      "grad_norm": 0.3801188952039268,
+      "learning_rate": 3.5199539764338995e-06,
+      "loss": 0.3532,
+      "step": 5108
+    },
+    {
+      "epoch": 1.2161004343428334,
+      "grad_norm": 0.3760239490498582,
+      "learning_rate": 3.518112877450247e-06,
+      "loss": 0.3925,
+      "step": 5109
+    },
+    {
+      "epoch": 1.2163384304158982,
+      "grad_norm": 0.34044640736378307,
+      "learning_rate": 3.51627199871168e-06,
+      "loss": 0.2874,
+      "step": 5110
+    },
+    {
+      "epoch": 1.216576426488963,
+      "grad_norm": 0.4167896879648559,
+      "learning_rate": 3.5144313404918e-06,
+      "loss": 0.3038,
+      "step": 5111
+    },
+    {
+      "epoch": 1.2168144225620277,
+      "grad_norm": 0.375356242770381,
+      "learning_rate": 3.512590903064175e-06,
+      "loss": 0.3309,
+      "step": 5112
+    },
+    {
+      "epoch": 1.2170524186350926,
+      "grad_norm": 0.37061202523878695,
+      "learning_rate": 3.5107506867023377e-06,
+      "loss": 0.373,
+      "step": 5113
+    },
+    {
+      "epoch": 1.2172904147081574,
+      "grad_norm": 0.36705161824794263,
+      "learning_rate": 3.508910691679791e-06,
+      "loss": 0.3056,
+      "step": 5114
+    },
+    {
+      "epoch": 1.217528410781222,
+      "grad_norm": 0.3939317720079383,
+      "learning_rate": 3.5070709182700007e-06,
+      "loss": 0.3024,
+      "step": 5115
+    },
+    {
+      "epoch": 1.2177664068542868,
+      "grad_norm": 0.4054053311852339,
+      "learning_rate": 3.5052313667464075e-06,
+      "loss": 0.3854,
+      "step": 5116
+    },
+    {
+      "epoch": 1.2180044029273518,
+      "grad_norm": 0.33993185418916133,
+      "learning_rate": 3.5033920373824125e-06,
+      "loss": 0.3238,
+      "step": 5117
+    },
+    {
+      "epoch": 1.2182423990004165,
+      "grad_norm": 0.3434027285290983,
+      "learning_rate": 3.5015529304513845e-06,
+      "loss": 0.2779,
+      "step": 5118
+    },
+    {
+      "epoch": 1.2184803950734813,
+      "grad_norm": 0.39857723088957964,
+      "learning_rate": 3.499714046226661e-06,
+      "loss": 0.3658,
+      "step": 5119
+    },
+    {
+      "epoch": 1.218718391146546,
+      "grad_norm": 0.3991383073000782,
+      "learning_rate": 3.4978753849815457e-06,
+      "loss": 0.3682,
+      "step": 5120
+    },
+    {
+      "epoch": 1.218956387219611,
+      "grad_norm": 0.3685347968394098,
+      "learning_rate": 3.4960369469893087e-06,
+      "loss": 0.2886,
+      "step": 5121
+    },
+    {
+      "epoch": 1.2191943832926757,
+      "grad_norm": 0.3833481773800204,
+      "learning_rate": 3.4941987325231873e-06,
+      "loss": 0.3273,
+      "step": 5122
+    },
+    {
+      "epoch": 1.2194323793657404,
+      "grad_norm": 0.3827498561469543,
+      "learning_rate": 3.4923607418563855e-06,
+      "loss": 0.3803,
+      "step": 5123
+    },
+    {
+      "epoch": 1.2196703754388052,
+      "grad_norm": 0.3628595561239951,
+      "learning_rate": 3.490522975262076e-06,
+      "loss": 0.3226,
+      "step": 5124
+    },
+    {
+      "epoch": 1.2199083715118701,
+      "grad_norm": 0.3427380203291989,
+      "learning_rate": 3.4886854330133947e-06,
+      "loss": 0.2499,
+      "step": 5125
+    },
+    {
+      "epoch": 1.2201463675849349,
+      "grad_norm": 0.3902097175795296,
+      "learning_rate": 3.4868481153834454e-06,
+      "loss": 0.328,
+      "step": 5126
+    },
+    {
+      "epoch": 1.2203843636579996,
+      "grad_norm": 0.3953893414153582,
+      "learning_rate": 3.485011022645301e-06,
+      "loss": 0.4049,
+      "step": 5127
+    },
+    {
+      "epoch": 1.2206223597310644,
+      "grad_norm": 0.36238449295552555,
+      "learning_rate": 3.4831741550719964e-06,
+      "loss": 0.3147,
+      "step": 5128
+    },
+    {
+      "epoch": 1.2208603558041293,
+      "grad_norm": 0.3978994446444037,
+      "learning_rate": 3.4813375129365357e-06,
+      "loss": 0.2917,
+      "step": 5129
+    },
+    {
+      "epoch": 1.221098351877194,
+      "grad_norm": 0.3669821823525373,
+      "learning_rate": 3.4795010965118926e-06,
+      "loss": 0.3574,
+      "step": 5130
+    },
+    {
+      "epoch": 1.2213363479502588,
+      "grad_norm": 0.3692786501186213,
+      "learning_rate": 3.477664906071001e-06,
+      "loss": 0.3685,
+      "step": 5131
+    },
+    {
+      "epoch": 1.2215743440233235,
+      "grad_norm": 0.3751050152516497,
+      "learning_rate": 3.4758289418867665e-06,
+      "loss": 0.2776,
+      "step": 5132
+    },
+    {
+      "epoch": 1.2218123400963885,
+      "grad_norm": 0.3598859162181419,
+      "learning_rate": 3.473993204232056e-06,
+      "loss": 0.3547,
+      "step": 5133
+    },
+    {
+      "epoch": 1.2220503361694532,
+      "grad_norm": 0.36721640794062704,
+      "learning_rate": 3.4721576933797072e-06,
+      "loss": 0.3836,
+      "step": 5134
+    },
+    {
+      "epoch": 1.222288332242518,
+      "grad_norm": 0.37753613271798336,
+      "learning_rate": 3.470322409602523e-06,
+      "loss": 0.3061,
+      "step": 5135
+    },
+    {
+      "epoch": 1.2225263283155827,
+      "grad_norm": 0.36033685671415805,
+      "learning_rate": 3.4684873531732704e-06,
+      "loss": 0.2989,
+      "step": 5136
+    },
+    {
+      "epoch": 1.2227643243886477,
+      "grad_norm": 0.4190827209156023,
+      "learning_rate": 3.4666525243646845e-06,
+      "loss": 0.3299,
+      "step": 5137
+    },
+    {
+      "epoch": 1.2230023204617124,
+      "grad_norm": 0.41563040679604657,
+      "learning_rate": 3.464817923449467e-06,
+      "loss": 0.3757,
+      "step": 5138
+    },
+    {
+      "epoch": 1.2232403165347772,
+      "grad_norm": 0.4017327361796809,
+      "learning_rate": 3.4629835507002853e-06,
+      "loss": 0.2852,
+      "step": 5139
+    },
+    {
+      "epoch": 1.223478312607842,
+      "grad_norm": 0.39000305457626766,
+      "learning_rate": 3.461149406389771e-06,
+      "loss": 0.306,
+      "step": 5140
+    },
+    {
+      "epoch": 1.2237163086809069,
+      "grad_norm": 0.3870106967773287,
+      "learning_rate": 3.4593154907905246e-06,
+      "loss": 0.3645,
+      "step": 5141
+    },
+    {
+      "epoch": 1.2239543047539716,
+      "grad_norm": 0.38791895997961046,
+      "learning_rate": 3.4574818041751113e-06,
+      "loss": 0.3239,
+      "step": 5142
+    },
+    {
+      "epoch": 1.2241923008270363,
+      "grad_norm": 0.3959723469109777,
+      "learning_rate": 3.45564834681606e-06,
+      "loss": 0.2933,
+      "step": 5143
+    },
+    {
+      "epoch": 1.224430296900101,
+      "grad_norm": 0.36905913364646237,
+      "learning_rate": 3.4538151189858717e-06,
+      "loss": 0.3453,
+      "step": 5144
+    },
+    {
+      "epoch": 1.224668292973166,
+      "grad_norm": 0.3837400246280125,
+      "learning_rate": 3.451982120957007e-06,
+      "loss": 0.3988,
+      "step": 5145
+    },
+    {
+      "epoch": 1.2249062890462308,
+      "grad_norm": 0.35298513550582294,
+      "learning_rate": 3.450149353001896e-06,
+      "loss": 0.3025,
+      "step": 5146
+    },
+    {
+      "epoch": 1.2251442851192955,
+      "grad_norm": 0.372652851908798,
+      "learning_rate": 3.4483168153929324e-06,
+      "loss": 0.2998,
+      "step": 5147
+    },
+    {
+      "epoch": 1.2253822811923603,
+      "grad_norm": 0.3571143797030416,
+      "learning_rate": 3.4464845084024767e-06,
+      "loss": 0.3579,
+      "step": 5148
+    },
+    {
+      "epoch": 1.2256202772654252,
+      "grad_norm": 0.3669338824803904,
+      "learning_rate": 3.444652432302855e-06,
+      "loss": 0.3227,
+      "step": 5149
+    },
+    {
+      "epoch": 1.22585827333849,
+      "grad_norm": 0.38715345328480644,
+      "learning_rate": 3.4428205873663584e-06,
+      "loss": 0.2854,
+      "step": 5150
+    },
+    {
+      "epoch": 1.2260962694115547,
+      "grad_norm": 0.3667959879928667,
+      "learning_rate": 3.440988973865246e-06,
+      "loss": 0.3279,
+      "step": 5151
+    },
+    {
+      "epoch": 1.2263342654846194,
+      "grad_norm": 0.39136089925706374,
+      "learning_rate": 3.4391575920717407e-06,
+      "loss": 0.4016,
+      "step": 5152
+    },
+    {
+      "epoch": 1.2265722615576844,
+      "grad_norm": 0.38499224408384597,
+      "learning_rate": 3.4373264422580305e-06,
+      "loss": 0.2915,
+      "step": 5153
+    },
+    {
+      "epoch": 1.2268102576307491,
+      "grad_norm": 0.3893740742863143,
+      "learning_rate": 3.4354955246962694e-06,
+      "loss": 0.2866,
+      "step": 5154
+    },
+    {
+      "epoch": 1.2270482537038139,
+      "grad_norm": 0.38314543939198864,
+      "learning_rate": 3.4336648396585777e-06,
+      "loss": 0.3695,
+      "step": 5155
+    },
+    {
+      "epoch": 1.2272862497768786,
+      "grad_norm": 0.3553545241693441,
+      "learning_rate": 3.4318343874170378e-06,
+      "loss": 0.3653,
+      "step": 5156
+    },
+    {
+      "epoch": 1.2275242458499434,
+      "grad_norm": 0.3721824699789484,
+      "learning_rate": 3.4300041682437046e-06,
+      "loss": 0.3055,
+      "step": 5157
+    },
+    {
+      "epoch": 1.2277622419230083,
+      "grad_norm": 0.34746845955767125,
+      "learning_rate": 3.428174182410592e-06,
+      "loss": 0.3031,
+      "step": 5158
+    },
+    {
+      "epoch": 1.228000237996073,
+      "grad_norm": 0.39404533678367354,
+      "learning_rate": 3.4263444301896805e-06,
+      "loss": 0.3667,
+      "step": 5159
+    },
+    {
+      "epoch": 1.2282382340691378,
+      "grad_norm": 0.3725519922822452,
+      "learning_rate": 3.424514911852917e-06,
+      "loss": 0.2965,
+      "step": 5160
+    },
+    {
+      "epoch": 1.2284762301422028,
+      "grad_norm": 0.38747366594584076,
+      "learning_rate": 3.4226856276722133e-06,
+      "loss": 0.2872,
+      "step": 5161
+    },
+    {
+      "epoch": 1.2287142262152675,
+      "grad_norm": 0.38316424613293093,
+      "learning_rate": 3.4208565779194467e-06,
+      "loss": 0.3519,
+      "step": 5162
+    },
+    {
+      "epoch": 1.2289522222883322,
+      "grad_norm": 0.38446797539419886,
+      "learning_rate": 3.4190277628664583e-06,
+      "loss": 0.3526,
+      "step": 5163
+    },
+    {
+      "epoch": 1.229190218361397,
+      "grad_norm": 0.3794454331861941,
+      "learning_rate": 3.417199182785055e-06,
+      "loss": 0.2841,
+      "step": 5164
+    },
+    {
+      "epoch": 1.2294282144344617,
+      "grad_norm": 0.3878874853038905,
+      "learning_rate": 3.4153708379470107e-06,
+      "loss": 0.3286,
+      "step": 5165
+    },
+    {
+      "epoch": 1.2296662105075267,
+      "grad_norm": 0.40528158013767834,
+      "learning_rate": 3.4135427286240613e-06,
+      "loss": 0.3699,
+      "step": 5166
+    },
+    {
+      "epoch": 1.2299042065805914,
+      "grad_norm": 0.3725084616775476,
+      "learning_rate": 3.41171485508791e-06,
+      "loss": 0.3351,
+      "step": 5167
+    },
+    {
+      "epoch": 1.2301422026536561,
+      "grad_norm": 0.3539613656037255,
+      "learning_rate": 3.409887217610223e-06,
+      "loss": 0.2858,
+      "step": 5168
+    },
+    {
+      "epoch": 1.230380198726721,
+      "grad_norm": 0.3923246643319525,
+      "learning_rate": 3.4080598164626333e-06,
+      "loss": 0.3168,
+      "step": 5169
+    },
+    {
+      "epoch": 1.2306181947997858,
+      "grad_norm": 0.3783741339958614,
+      "learning_rate": 3.4062326519167354e-06,
+      "loss": 0.3958,
+      "step": 5170
+    },
+    {
+      "epoch": 1.2308561908728506,
+      "grad_norm": 0.34812779553971557,
+      "learning_rate": 3.4044057242440954e-06,
+      "loss": 0.2827,
+      "step": 5171
+    },
+    {
+      "epoch": 1.2310941869459153,
+      "grad_norm": 0.3862300590706353,
+      "learning_rate": 3.402579033716238e-06,
+      "loss": 0.3068,
+      "step": 5172
+    },
+    {
+      "epoch": 1.23133218301898,
+      "grad_norm": 0.41044439207246175,
+      "learning_rate": 3.400752580604655e-06,
+      "loss": 0.3579,
+      "step": 5173
+    },
+    {
+      "epoch": 1.231570179092045,
+      "grad_norm": 0.35576711138677686,
+      "learning_rate": 3.3989263651808013e-06,
+      "loss": 0.3522,
+      "step": 5174
+    },
+    {
+      "epoch": 1.2318081751651098,
+      "grad_norm": 0.37907648138016425,
+      "learning_rate": 3.397100387716098e-06,
+      "loss": 0.2591,
+      "step": 5175
+    },
+    {
+      "epoch": 1.2320461712381745,
+      "grad_norm": 0.4047380890985508,
+      "learning_rate": 3.395274648481932e-06,
+      "loss": 0.3778,
+      "step": 5176
+    },
+    {
+      "epoch": 1.2322841673112395,
+      "grad_norm": 0.3767213089270945,
+      "learning_rate": 3.39344914774965e-06,
+      "loss": 0.3612,
+      "step": 5177
+    },
+    {
+      "epoch": 1.2325221633843042,
+      "grad_norm": 0.3618213335927789,
+      "learning_rate": 3.391623885790571e-06,
+      "loss": 0.2956,
+      "step": 5178
+    },
+    {
+      "epoch": 1.232760159457369,
+      "grad_norm": 0.37715805015205855,
+      "learning_rate": 3.3897988628759714e-06,
+      "loss": 0.3076,
+      "step": 5179
+    },
+    {
+      "epoch": 1.2329981555304337,
+      "grad_norm": 0.36354382791582407,
+      "learning_rate": 3.387974079277095e-06,
+      "loss": 0.3843,
+      "step": 5180
+    },
+    {
+      "epoch": 1.2332361516034984,
+      "grad_norm": 0.3828085727627118,
+      "learning_rate": 3.3861495352651504e-06,
+      "loss": 0.3464,
+      "step": 5181
+    },
+    {
+      "epoch": 1.2334741476765634,
+      "grad_norm": 0.40137098436265456,
+      "learning_rate": 3.3843252311113095e-06,
+      "loss": 0.2923,
+      "step": 5182
+    },
+    {
+      "epoch": 1.2337121437496281,
+      "grad_norm": 0.4180250525495085,
+      "learning_rate": 3.3825011670867086e-06,
+      "loss": 0.3647,
+      "step": 5183
+    },
+    {
+      "epoch": 1.2339501398226929,
+      "grad_norm": 0.3742865593493896,
+      "learning_rate": 3.3806773434624475e-06,
+      "loss": 0.3939,
+      "step": 5184
+    },
+    {
+      "epoch": 1.2341881358957578,
+      "grad_norm": 0.3799380469090617,
+      "learning_rate": 3.3788537605095957e-06,
+      "loss": 0.3187,
+      "step": 5185
+    },
+    {
+      "epoch": 1.2344261319688226,
+      "grad_norm": 0.3845177336003192,
+      "learning_rate": 3.37703041849918e-06,
+      "loss": 0.2822,
+      "step": 5186
+    },
+    {
+      "epoch": 1.2346641280418873,
+      "grad_norm": 0.38597150172129496,
+      "learning_rate": 3.375207317702194e-06,
+      "loss": 0.3609,
+      "step": 5187
+    },
+    {
+      "epoch": 1.234902124114952,
+      "grad_norm": 0.3806826446535519,
+      "learning_rate": 3.373384458389597e-06,
+      "loss": 0.3542,
+      "step": 5188
+    },
+    {
+      "epoch": 1.2351401201880168,
+      "grad_norm": 0.3664699692758197,
+      "learning_rate": 3.371561840832309e-06,
+      "loss": 0.2939,
+      "step": 5189
+    },
+    {
+      "epoch": 1.2353781162610817,
+      "grad_norm": 0.3855337340082918,
+      "learning_rate": 3.3697394653012185e-06,
+      "loss": 0.3241,
+      "step": 5190
+    },
+    {
+      "epoch": 1.2356161123341465,
+      "grad_norm": 0.3949003682130631,
+      "learning_rate": 3.367917332067172e-06,
+      "loss": 0.3747,
+      "step": 5191
+    },
+    {
+      "epoch": 1.2358541084072112,
+      "grad_norm": 0.3916469796647736,
+      "learning_rate": 3.3660954414009872e-06,
+      "loss": 0.2907,
+      "step": 5192
+    },
+    {
+      "epoch": 1.2360921044802762,
+      "grad_norm": 0.41103529668836525,
+      "learning_rate": 3.3642737935734403e-06,
+      "loss": 0.266,
+      "step": 5193
+    },
+    {
+      "epoch": 1.236330100553341,
+      "grad_norm": 0.4135491847771145,
+      "learning_rate": 3.3624523888552734e-06,
+      "loss": 0.3251,
+      "step": 5194
+    },
+    {
+      "epoch": 1.2365680966264057,
+      "grad_norm": 0.41366473528894604,
+      "learning_rate": 3.3606312275171928e-06,
+      "loss": 0.3663,
+      "step": 5195
+    },
+    {
+      "epoch": 1.2368060926994704,
+      "grad_norm": 0.3584651473821333,
+      "learning_rate": 3.358810309829868e-06,
+      "loss": 0.2973,
+      "step": 5196
+    },
+    {
+      "epoch": 1.2370440887725351,
+      "grad_norm": 0.44390275619305986,
+      "learning_rate": 3.356989636063932e-06,
+      "loss": 0.3176,
+      "step": 5197
+    },
+    {
+      "epoch": 1.2372820848456,
+      "grad_norm": 0.408274658661175,
+      "learning_rate": 3.3551692064899806e-06,
+      "loss": 0.388,
+      "step": 5198
+    },
+    {
+      "epoch": 1.2375200809186648,
+      "grad_norm": 0.3676982361906048,
+      "learning_rate": 3.353349021378578e-06,
+      "loss": 0.3239,
+      "step": 5199
+    },
+    {
+      "epoch": 1.2377580769917296,
+      "grad_norm": 0.38263348645456147,
+      "learning_rate": 3.3515290810002464e-06,
+      "loss": 0.3018,
+      "step": 5200
+    },
+    {
+      "epoch": 1.2379960730647945,
+      "grad_norm": 0.37234902877710807,
+      "learning_rate": 3.3497093856254757e-06,
+      "loss": 0.3607,
+      "step": 5201
+    },
+    {
+      "epoch": 1.2382340691378593,
+      "grad_norm": 0.396040248592026,
+      "learning_rate": 3.347889935524716e-06,
+      "loss": 0.4182,
+      "step": 5202
+    },
+    {
+      "epoch": 1.238472065210924,
+      "grad_norm": 0.3999889104147471,
+      "learning_rate": 3.3460707309683826e-06,
+      "loss": 0.304,
+      "step": 5203
+    },
+    {
+      "epoch": 1.2387100612839888,
+      "grad_norm": 0.40354368811724745,
+      "learning_rate": 3.3442517722268543e-06,
+      "loss": 0.2784,
+      "step": 5204
+    },
+    {
+      "epoch": 1.2389480573570535,
+      "grad_norm": 0.4088049571035965,
+      "learning_rate": 3.342433059570475e-06,
+      "loss": 0.3927,
+      "step": 5205
+    },
+    {
+      "epoch": 1.2391860534301185,
+      "grad_norm": 0.3784733136722843,
+      "learning_rate": 3.340614593269549e-06,
+      "loss": 0.3523,
+      "step": 5206
+    },
+    {
+      "epoch": 1.2394240495031832,
+      "grad_norm": 0.6754063547956777,
+      "learning_rate": 3.338796373594346e-06,
+      "loss": 0.2846,
+      "step": 5207
+    },
+    {
+      "epoch": 1.239662045576248,
+      "grad_norm": 0.37479104141833847,
+      "learning_rate": 3.336978400815098e-06,
+      "loss": 0.3278,
+      "step": 5208
+    },
+    {
+      "epoch": 1.239900041649313,
+      "grad_norm": 0.40369056554447125,
+      "learning_rate": 3.3351606752020004e-06,
+      "loss": 0.4141,
+      "step": 5209
+    },
+    {
+      "epoch": 1.2401380377223776,
+      "grad_norm": 0.3592088638519854,
+      "learning_rate": 3.333343197025213e-06,
+      "loss": 0.3133,
+      "step": 5210
+    },
+    {
+      "epoch": 1.2403760337954424,
+      "grad_norm": 0.4150892541339133,
+      "learning_rate": 3.3315259665548554e-06,
+      "loss": 0.2712,
+      "step": 5211
+    },
+    {
+      "epoch": 1.240614029868507,
+      "grad_norm": 0.3717429857265201,
+      "learning_rate": 3.3297089840610173e-06,
+      "loss": 0.3615,
+      "step": 5212
+    },
+    {
+      "epoch": 1.2408520259415718,
+      "grad_norm": 0.37643019244681736,
+      "learning_rate": 3.3278922498137455e-06,
+      "loss": 0.376,
+      "step": 5213
+    },
+    {
+      "epoch": 1.2410900220146368,
+      "grad_norm": 0.3735812988579749,
+      "learning_rate": 3.326075764083051e-06,
+      "loss": 0.2832,
+      "step": 5214
+    },
+    {
+      "epoch": 1.2413280180877015,
+      "grad_norm": 0.4247489715555179,
+      "learning_rate": 3.3242595271389087e-06,
+      "loss": 0.3459,
+      "step": 5215
+    },
+    {
+      "epoch": 1.2415660141607663,
+      "grad_norm": 0.3782694036825223,
+      "learning_rate": 3.3224435392512565e-06,
+      "loss": 0.3574,
+      "step": 5216
+    },
+    {
+      "epoch": 1.2418040102338312,
+      "grad_norm": 0.3593379490467785,
+      "learning_rate": 3.320627800689996e-06,
+      "loss": 0.3358,
+      "step": 5217
+    },
+    {
+      "epoch": 1.242042006306896,
+      "grad_norm": 0.38439219808924074,
+      "learning_rate": 3.3188123117249884e-06,
+      "loss": 0.2689,
+      "step": 5218
+    },
+    {
+      "epoch": 1.2422800023799607,
+      "grad_norm": 0.40058679929549906,
+      "learning_rate": 3.3169970726260625e-06,
+      "loss": 0.3514,
+      "step": 5219
+    },
+    {
+      "epoch": 1.2425179984530255,
+      "grad_norm": 0.3765460507646511,
+      "learning_rate": 3.3151820836630074e-06,
+      "loss": 0.3873,
+      "step": 5220
+    },
+    {
+      "epoch": 1.2427559945260902,
+      "grad_norm": 0.4315150126863899,
+      "learning_rate": 3.313367345105575e-06,
+      "loss": 0.274,
+      "step": 5221
+    },
+    {
+      "epoch": 1.2429939905991552,
+      "grad_norm": 0.3785008936826988,
+      "learning_rate": 3.31155285722348e-06,
+      "loss": 0.3163,
+      "step": 5222
+    },
+    {
+      "epoch": 1.24323198667222,
+      "grad_norm": 0.3967888293663607,
+      "learning_rate": 3.309738620286401e-06,
+      "loss": 0.3529,
+      "step": 5223
+    },
+    {
+      "epoch": 1.2434699827452846,
+      "grad_norm": 0.3861441173261868,
+      "learning_rate": 3.307924634563978e-06,
+      "loss": 0.3399,
+      "step": 5224
+    },
+    {
+      "epoch": 1.2437079788183496,
+      "grad_norm": 0.3750655532874109,
+      "learning_rate": 3.306110900325813e-06,
+      "loss": 0.3173,
+      "step": 5225
+    },
+    {
+      "epoch": 1.2439459748914143,
+      "grad_norm": 0.3925378776235632,
+      "learning_rate": 3.304297417841474e-06,
+      "loss": 0.3065,
+      "step": 5226
+    },
+    {
+      "epoch": 1.244183970964479,
+      "grad_norm": 0.39518178594914705,
+      "learning_rate": 3.3024841873804885e-06,
+      "loss": 0.3659,
+      "step": 5227
+    },
+    {
+      "epoch": 1.2444219670375438,
+      "grad_norm": 0.37902705321028035,
+      "learning_rate": 3.300671209212347e-06,
+      "loss": 0.3182,
+      "step": 5228
+    },
+    {
+      "epoch": 1.2446599631106086,
+      "grad_norm": 0.3720334846385885,
+      "learning_rate": 3.298858483606504e-06,
+      "loss": 0.2705,
+      "step": 5229
+    },
+    {
+      "epoch": 1.2448979591836735,
+      "grad_norm": 0.3733999529266795,
+      "learning_rate": 3.2970460108323744e-06,
+      "loss": 0.3387,
+      "step": 5230
+    },
+    {
+      "epoch": 1.2451359552567383,
+      "grad_norm": 0.3772749257271583,
+      "learning_rate": 3.295233791159336e-06,
+      "loss": 0.3739,
+      "step": 5231
+    },
+    {
+      "epoch": 1.245373951329803,
+      "grad_norm": 0.3695783892022936,
+      "learning_rate": 3.2934218248567294e-06,
+      "loss": 0.283,
+      "step": 5232
+    },
+    {
+      "epoch": 1.245611947402868,
+      "grad_norm": 0.4014905086469289,
+      "learning_rate": 3.29161011219386e-06,
+      "loss": 0.3269,
+      "step": 5233
+    },
+    {
+      "epoch": 1.2458499434759327,
+      "grad_norm": 0.4628150231356265,
+      "learning_rate": 3.2897986534399908e-06,
+      "loss": 0.39,
+      "step": 5234
+    },
+    {
+      "epoch": 1.2460879395489974,
+      "grad_norm": 0.35110622672123937,
+      "learning_rate": 3.2879874488643504e-06,
+      "loss": 0.3049,
+      "step": 5235
+    },
+    {
+      "epoch": 1.2463259356220622,
+      "grad_norm": 0.39624400624138423,
+      "learning_rate": 3.2861764987361276e-06,
+      "loss": 0.2845,
+      "step": 5236
+    },
+    {
+      "epoch": 1.246563931695127,
+      "grad_norm": 0.3548580217470856,
+      "learning_rate": 3.284365803324476e-06,
+      "loss": 0.3565,
+      "step": 5237
+    },
+    {
+      "epoch": 1.2468019277681919,
+      "grad_norm": 0.3795495986882351,
+      "learning_rate": 3.2825553628985078e-06,
+      "loss": 0.3632,
+      "step": 5238
+    },
+    {
+      "epoch": 1.2470399238412566,
+      "grad_norm": 0.4043537341500232,
+      "learning_rate": 3.280745177727299e-06,
+      "loss": 0.3075,
+      "step": 5239
+    },
+    {
+      "epoch": 1.2472779199143214,
+      "grad_norm": 0.40712561384193874,
+      "learning_rate": 3.278935248079891e-06,
+      "loss": 0.2944,
+      "step": 5240
+    },
+    {
+      "epoch": 1.2475159159873863,
+      "grad_norm": 0.3557393968961097,
+      "learning_rate": 3.2771255742252817e-06,
+      "loss": 0.3703,
+      "step": 5241
+    },
+    {
+      "epoch": 1.247753912060451,
+      "grad_norm": 0.360568352135061,
+      "learning_rate": 3.2753161564324344e-06,
+      "loss": 0.3443,
+      "step": 5242
+    },
+    {
+      "epoch": 1.2479919081335158,
+      "grad_norm": 0.3773441969979045,
+      "learning_rate": 3.2735069949702723e-06,
+      "loss": 0.3008,
+      "step": 5243
+    },
+    {
+      "epoch": 1.2482299042065805,
+      "grad_norm": 0.35789698002368536,
+      "learning_rate": 3.271698090107682e-06,
+      "loss": 0.3428,
+      "step": 5244
+    },
+    {
+      "epoch": 1.2484679002796453,
+      "grad_norm": 0.3799994603077194,
+      "learning_rate": 3.269889442113512e-06,
+      "loss": 0.3857,
+      "step": 5245
+    },
+    {
+      "epoch": 1.2487058963527102,
+      "grad_norm": 0.3732392024366841,
+      "learning_rate": 3.2680810512565714e-06,
+      "loss": 0.2917,
+      "step": 5246
+    },
+    {
+      "epoch": 1.248943892425775,
+      "grad_norm": 0.40137116851946253,
+      "learning_rate": 3.266272917805633e-06,
+      "loss": 0.273,
+      "step": 5247
+    },
+    {
+      "epoch": 1.2491818884988397,
+      "grad_norm": 0.36249398408452443,
+      "learning_rate": 3.2644650420294288e-06,
+      "loss": 0.3821,
+      "step": 5248
+    },
+    {
+      "epoch": 1.2494198845719047,
+      "grad_norm": 0.38050458261938763,
+      "learning_rate": 3.2626574241966546e-06,
+      "loss": 0.3256,
+      "step": 5249
+    },
+    {
+      "epoch": 1.2496578806449694,
+      "grad_norm": 0.3726179289456404,
+      "learning_rate": 3.2608500645759673e-06,
+      "loss": 0.2807,
+      "step": 5250
+    },
+    {
+      "epoch": 1.2498958767180341,
+      "grad_norm": 0.35938712373080584,
+      "learning_rate": 3.2590429634359845e-06,
+      "loss": 0.3395,
+      "step": 5251
+    },
+    {
+      "epoch": 1.2501338727910989,
+      "grad_norm": 0.4040569210815416,
+      "learning_rate": 3.257236121045286e-06,
+      "loss": 0.386,
+      "step": 5252
+    },
+    {
+      "epoch": 1.2503718688641636,
+      "grad_norm": 0.3844777853293955,
+      "learning_rate": 3.2554295376724134e-06,
+      "loss": 0.2991,
+      "step": 5253
+    },
+    {
+      "epoch": 1.2506098649372286,
+      "grad_norm": 0.4040805562877999,
+      "learning_rate": 3.253623213585871e-06,
+      "loss": 0.2617,
+      "step": 5254
+    },
+    {
+      "epoch": 1.2508478610102933,
+      "grad_norm": 0.3728383275268288,
+      "learning_rate": 3.2518171490541222e-06,
+      "loss": 0.3587,
+      "step": 5255
+    },
+    {
+      "epoch": 1.251085857083358,
+      "grad_norm": 0.37161586537379476,
+      "learning_rate": 3.250011344345594e-06,
+      "loss": 0.351,
+      "step": 5256
+    },
+    {
+      "epoch": 1.251323853156423,
+      "grad_norm": 0.3923833529850557,
+      "learning_rate": 3.2482057997286716e-06,
+      "loss": 0.2907,
+      "step": 5257
+    },
+    {
+      "epoch": 1.2515618492294878,
+      "grad_norm": 0.46418423651277113,
+      "learning_rate": 3.2464005154717036e-06,
+      "loss": 0.3159,
+      "step": 5258
+    },
+    {
+      "epoch": 1.2517998453025525,
+      "grad_norm": 0.391141738921714,
+      "learning_rate": 3.244595491843003e-06,
+      "loss": 0.3681,
+      "step": 5259
+    },
+    {
+      "epoch": 1.2520378413756172,
+      "grad_norm": 0.3998036064404465,
+      "learning_rate": 3.2427907291108383e-06,
+      "loss": 0.3235,
+      "step": 5260
+    },
+    {
+      "epoch": 1.252275837448682,
+      "grad_norm": 0.3895934764387763,
+      "learning_rate": 3.2409862275434425e-06,
+      "loss": 0.2746,
+      "step": 5261
+    },
+    {
+      "epoch": 1.252513833521747,
+      "grad_norm": 0.3704415894740105,
+      "learning_rate": 3.239181987409009e-06,
+      "loss": 0.3457,
+      "step": 5262
+    },
+    {
+      "epoch": 1.2527518295948117,
+      "grad_norm": 0.37132868730200563,
+      "learning_rate": 3.2373780089756934e-06,
+      "loss": 0.3927,
+      "step": 5263
+    },
+    {
+      "epoch": 1.2529898256678764,
+      "grad_norm": 0.3434397334645843,
+      "learning_rate": 3.2355742925116103e-06,
+      "loss": 0.2817,
+      "step": 5264
+    },
+    {
+      "epoch": 1.2532278217409414,
+      "grad_norm": 0.4033979941592838,
+      "learning_rate": 3.233770838284837e-06,
+      "loss": 0.3391,
+      "step": 5265
+    },
+    {
+      "epoch": 1.2534658178140061,
+      "grad_norm": 0.4863421240520152,
+      "learning_rate": 3.231967646563412e-06,
+      "loss": 0.3836,
+      "step": 5266
+    },
+    {
+      "epoch": 1.2537038138870709,
+      "grad_norm": 0.3863926541257452,
+      "learning_rate": 3.230164717615331e-06,
+      "loss": 0.3292,
+      "step": 5267
+    },
+    {
+      "epoch": 1.2539418099601356,
+      "grad_norm": 0.35788876853240303,
+      "learning_rate": 3.228362051708559e-06,
+      "loss": 0.2755,
+      "step": 5268
+    },
+    {
+      "epoch": 1.2541798060332003,
+      "grad_norm": 0.3946622156577876,
+      "learning_rate": 3.2265596491110145e-06,
+      "loss": 0.3349,
+      "step": 5269
+    },
+    {
+      "epoch": 1.2544178021062653,
+      "grad_norm": 0.3862161899983044,
+      "learning_rate": 3.2247575100905794e-06,
+      "loss": 0.4002,
+      "step": 5270
+    },
+    {
+      "epoch": 1.25465579817933,
+      "grad_norm": 0.3970524252747173,
+      "learning_rate": 3.2229556349150947e-06,
+      "loss": 0.2879,
+      "step": 5271
+    },
+    {
+      "epoch": 1.2548937942523948,
+      "grad_norm": 0.38366178715626265,
+      "learning_rate": 3.221154023852364e-06,
+      "loss": 0.3337,
+      "step": 5272
+    },
+    {
+      "epoch": 1.2551317903254597,
+      "grad_norm": 0.37924724516016045,
+      "learning_rate": 3.2193526771701526e-06,
+      "loss": 0.3732,
+      "step": 5273
+    },
+    {
+      "epoch": 1.2553697863985245,
+      "grad_norm": 0.3720944559532506,
+      "learning_rate": 3.2175515951361844e-06,
+      "loss": 0.3384,
+      "step": 5274
+    },
+    {
+      "epoch": 1.2556077824715892,
+      "grad_norm": 0.37916136226673736,
+      "learning_rate": 3.2157507780181457e-06,
+      "loss": 0.2804,
+      "step": 5275
+    },
+    {
+      "epoch": 1.255845778544654,
+      "grad_norm": 0.37290254799955885,
+      "learning_rate": 3.2139502260836806e-06,
+      "loss": 0.3394,
+      "step": 5276
+    },
+    {
+      "epoch": 1.2560837746177187,
+      "grad_norm": 0.4120937160799173,
+      "learning_rate": 3.2121499396003974e-06,
+      "loss": 0.3896,
+      "step": 5277
+    },
+    {
+      "epoch": 1.2563217706907837,
+      "grad_norm": 0.3631101334404687,
+      "learning_rate": 3.2103499188358623e-06,
+      "loss": 0.3017,
+      "step": 5278
+    },
+    {
+      "epoch": 1.2565597667638484,
+      "grad_norm": 0.38219111111630644,
+      "learning_rate": 3.2085501640576035e-06,
+      "loss": 0.2705,
+      "step": 5279
+    },
+    {
+      "epoch": 1.2567977628369131,
+      "grad_norm": 0.4099100476999923,
+      "learning_rate": 3.206750675533106e-06,
+      "loss": 0.368,
+      "step": 5280
+    },
+    {
+      "epoch": 1.257035758909978,
+      "grad_norm": 0.37534175826439325,
+      "learning_rate": 3.2049514535298227e-06,
+      "loss": 0.333,
+      "step": 5281
+    },
+    {
+      "epoch": 1.2572737549830428,
+      "grad_norm": 0.3958938531308809,
+      "learning_rate": 3.2031524983151606e-06,
+      "loss": 0.3224,
+      "step": 5282
+    },
+    {
+      "epoch": 1.2575117510561076,
+      "grad_norm": 0.3755193173874838,
+      "learning_rate": 3.20135381015649e-06,
+      "loss": 0.3333,
+      "step": 5283
+    },
+    {
+      "epoch": 1.2577497471291723,
+      "grad_norm": 0.405190126827234,
+      "learning_rate": 3.1995553893211395e-06,
+      "loss": 0.3538,
+      "step": 5284
+    },
+    {
+      "epoch": 1.257987743202237,
+      "grad_norm": 0.3554947984040851,
+      "learning_rate": 3.1977572360763977e-06,
+      "loss": 0.3086,
+      "step": 5285
+    },
+    {
+      "epoch": 1.258225739275302,
+      "grad_norm": 0.36944294131448824,
+      "learning_rate": 3.195959350689517e-06,
+      "loss": 0.3169,
+      "step": 5286
+    },
+    {
+      "epoch": 1.2584637353483668,
+      "grad_norm": 0.38302961275351355,
+      "learning_rate": 3.1941617334277066e-06,
+      "loss": 0.3534,
+      "step": 5287
+    },
+    {
+      "epoch": 1.2587017314214315,
+      "grad_norm": 0.34830251003999074,
+      "learning_rate": 3.1923643845581364e-06,
+      "loss": 0.3458,
+      "step": 5288
+    },
+    {
+      "epoch": 1.2589397274944965,
+      "grad_norm": 0.4007008524746251,
+      "learning_rate": 3.190567304347938e-06,
+      "loss": 0.2737,
+      "step": 5289
+    },
+    {
+      "epoch": 1.2591777235675612,
+      "grad_norm": 0.3743627078891674,
+      "learning_rate": 3.1887704930642016e-06,
+      "loss": 0.3119,
+      "step": 5290
+    },
+    {
+      "epoch": 1.259415719640626,
+      "grad_norm": 0.39475666902200535,
+      "learning_rate": 3.1869739509739772e-06,
+      "loss": 0.3684,
+      "step": 5291
+    },
+    {
+      "epoch": 1.2596537157136907,
+      "grad_norm": 0.3507214620255833,
+      "learning_rate": 3.1851776783442754e-06,
+      "loss": 0.3523,
+      "step": 5292
+    },
+    {
+      "epoch": 1.2598917117867554,
+      "grad_norm": 0.4174143666004117,
+      "learning_rate": 3.1833816754420678e-06,
+      "loss": 0.2806,
+      "step": 5293
+    },
+    {
+      "epoch": 1.2601297078598204,
+      "grad_norm": 0.3626773824316747,
+      "learning_rate": 3.181585942534282e-06,
+      "loss": 0.3449,
+      "step": 5294
+    },
+    {
+      "epoch": 1.260367703932885,
+      "grad_norm": 0.3746333463473584,
+      "learning_rate": 3.179790479887812e-06,
+      "loss": 0.3698,
+      "step": 5295
+    },
+    {
+      "epoch": 1.2606057000059498,
+      "grad_norm": 0.39419946032729236,
+      "learning_rate": 3.1779952877695077e-06,
+      "loss": 0.2622,
+      "step": 5296
+    },
+    {
+      "epoch": 1.2608436960790148,
+      "grad_norm": 0.3665818295311167,
+      "learning_rate": 3.176200366446177e-06,
+      "loss": 0.2874,
+      "step": 5297
+    },
+    {
+      "epoch": 1.2610816921520795,
+      "grad_norm": 0.41342588700331717,
+      "learning_rate": 3.174405716184591e-06,
+      "loss": 0.3516,
+      "step": 5298
+    },
+    {
+      "epoch": 1.2613196882251443,
+      "grad_norm": 0.35359012506797194,
+      "learning_rate": 3.1726113372514767e-06,
+      "loss": 0.3293,
+      "step": 5299
+    },
+    {
+      "epoch": 1.261557684298209,
+      "grad_norm": 0.40608712887436516,
+      "learning_rate": 3.1708172299135266e-06,
+      "loss": 0.2651,
+      "step": 5300
+    },
+    {
+      "epoch": 1.2617956803712738,
+      "grad_norm": 0.39078363662670146,
+      "learning_rate": 3.1690233944373883e-06,
+      "loss": 0.3475,
+      "step": 5301
+    },
+    {
+      "epoch": 1.2620336764443387,
+      "grad_norm": 0.39442057764959976,
+      "learning_rate": 3.1672298310896693e-06,
+      "loss": 0.448,
+      "step": 5302
+    },
+    {
+      "epoch": 1.2622716725174035,
+      "grad_norm": 0.3870252703040663,
+      "learning_rate": 3.165436540136938e-06,
+      "loss": 0.3393,
+      "step": 5303
+    },
+    {
+      "epoch": 1.2625096685904682,
+      "grad_norm": 0.35385684836054004,
+      "learning_rate": 3.1636435218457216e-06,
+      "loss": 0.2895,
+      "step": 5304
+    },
+    {
+      "epoch": 1.2627476646635332,
+      "grad_norm": 0.38830448667646283,
+      "learning_rate": 3.161850776482508e-06,
+      "loss": 0.3622,
+      "step": 5305
+    },
+    {
+      "epoch": 1.262985660736598,
+      "grad_norm": 0.3724314662342735,
+      "learning_rate": 3.160058304313742e-06,
+      "loss": 0.3318,
+      "step": 5306
+    },
+    {
+      "epoch": 1.2632236568096626,
+      "grad_norm": 0.4014989473946267,
+      "learning_rate": 3.1582661056058294e-06,
+      "loss": 0.2739,
+      "step": 5307
+    },
+    {
+      "epoch": 1.2634616528827274,
+      "grad_norm": 0.37792045811506125,
+      "learning_rate": 3.156474180625134e-06,
+      "loss": 0.3098,
+      "step": 5308
+    },
+    {
+      "epoch": 1.2636996489557921,
+      "grad_norm": 0.3923195999280906,
+      "learning_rate": 3.1546825296379845e-06,
+      "loss": 0.3802,
+      "step": 5309
+    },
+    {
+      "epoch": 1.263937645028857,
+      "grad_norm": 0.36757724848018525,
+      "learning_rate": 3.152891152910662e-06,
+      "loss": 0.3026,
+      "step": 5310
+    },
+    {
+      "epoch": 1.2641756411019218,
+      "grad_norm": 0.3689543106637064,
+      "learning_rate": 3.1511000507094097e-06,
+      "loss": 0.3073,
+      "step": 5311
+    },
+    {
+      "epoch": 1.2644136371749866,
+      "grad_norm": 0.36026728365035765,
+      "learning_rate": 3.149309223300428e-06,
+      "loss": 0.3464,
+      "step": 5312
+    },
+    {
+      "epoch": 1.2646516332480515,
+      "grad_norm": 0.3666885097703029,
+      "learning_rate": 3.1475186709498806e-06,
+      "loss": 0.397,
+      "step": 5313
+    },
+    {
+      "epoch": 1.2648896293211163,
+      "grad_norm": 0.35535524090766163,
+      "learning_rate": 3.145728393923887e-06,
+      "loss": 0.2911,
+      "step": 5314
+    },
+    {
+      "epoch": 1.265127625394181,
+      "grad_norm": 0.3699552572065724,
+      "learning_rate": 3.143938392488527e-06,
+      "loss": 0.3085,
+      "step": 5315
+    },
+    {
+      "epoch": 1.2653656214672457,
+      "grad_norm": 0.371881166524415,
+      "learning_rate": 3.1421486669098376e-06,
+      "loss": 0.3925,
+      "step": 5316
+    },
+    {
+      "epoch": 1.2656036175403105,
+      "grad_norm": 0.3979148251780867,
+      "learning_rate": 3.1403592174538176e-06,
+      "loss": 0.3259,
+      "step": 5317
+    },
+    {
+      "epoch": 1.2658416136133754,
+      "grad_norm": 0.35728138026407263,
+      "learning_rate": 3.138570044386424e-06,
+      "loss": 0.2795,
+      "step": 5318
+    },
+    {
+      "epoch": 1.2660796096864402,
+      "grad_norm": 0.40000777660528697,
+      "learning_rate": 3.136781147973571e-06,
+      "loss": 0.318,
+      "step": 5319
+    },
+    {
+      "epoch": 1.266317605759505,
+      "grad_norm": 0.37847837862392103,
+      "learning_rate": 3.134992528481133e-06,
+      "loss": 0.3983,
+      "step": 5320
+    },
+    {
+      "epoch": 1.2665556018325699,
+      "grad_norm": 0.36014200615492253,
+      "learning_rate": 3.133204186174942e-06,
+      "loss": 0.3218,
+      "step": 5321
+    },
+    {
+      "epoch": 1.2667935979056346,
+      "grad_norm": 0.3823696624234763,
+      "learning_rate": 3.131416121320792e-06,
+      "loss": 0.2918,
+      "step": 5322
+    },
+    {
+      "epoch": 1.2670315939786994,
+      "grad_norm": 0.3722726581374576,
+      "learning_rate": 3.129628334184433e-06,
+      "loss": 0.3727,
+      "step": 5323
+    },
+    {
+      "epoch": 1.267269590051764,
+      "grad_norm": 0.3659346216629232,
+      "learning_rate": 3.127840825031575e-06,
+      "loss": 0.3274,
+      "step": 5324
+    },
+    {
+      "epoch": 1.2675075861248288,
+      "grad_norm": 0.38942693904105447,
+      "learning_rate": 3.126053594127885e-06,
+      "loss": 0.3065,
+      "step": 5325
+    },
+    {
+      "epoch": 1.2677455821978938,
+      "grad_norm": 0.37772681354730653,
+      "learning_rate": 3.1242666417389896e-06,
+      "loss": 0.3058,
+      "step": 5326
+    },
+    {
+      "epoch": 1.2679835782709585,
+      "grad_norm": 0.38030394224311925,
+      "learning_rate": 3.122479968130475e-06,
+      "loss": 0.3953,
+      "step": 5327
+    },
+    {
+      "epoch": 1.2682215743440233,
+      "grad_norm": 0.38391257852782784,
+      "learning_rate": 3.1206935735678855e-06,
+      "loss": 0.2879,
+      "step": 5328
+    },
+    {
+      "epoch": 1.2684595704170882,
+      "grad_norm": 0.3731989200180512,
+      "learning_rate": 3.118907458316722e-06,
+      "loss": 0.2993,
+      "step": 5329
+    },
+    {
+      "epoch": 1.268697566490153,
+      "grad_norm": 0.6492017938323883,
+      "learning_rate": 3.1171216226424466e-06,
+      "loss": 0.359,
+      "step": 5330
+    },
+    {
+      "epoch": 1.2689355625632177,
+      "grad_norm": 0.389677366563528,
+      "learning_rate": 3.115336066810478e-06,
+      "loss": 0.3899,
+      "step": 5331
+    },
+    {
+      "epoch": 1.2691735586362825,
+      "grad_norm": 0.38739411432567555,
+      "learning_rate": 3.113550791086195e-06,
+      "loss": 0.2738,
+      "step": 5332
+    },
+    {
+      "epoch": 1.2694115547093472,
+      "grad_norm": 0.3697993858168483,
+      "learning_rate": 3.1117657957349325e-06,
+      "loss": 0.3336,
+      "step": 5333
+    },
+    {
+      "epoch": 1.2696495507824122,
+      "grad_norm": 0.3851139381884033,
+      "learning_rate": 3.109981081021985e-06,
+      "loss": 0.4039,
+      "step": 5334
+    },
+    {
+      "epoch": 1.269887546855477,
+      "grad_norm": 0.3663302911358832,
+      "learning_rate": 3.108196647212605e-06,
+      "loss": 0.3209,
+      "step": 5335
+    },
+    {
+      "epoch": 1.2701255429285416,
+      "grad_norm": 0.41294325410097843,
+      "learning_rate": 3.106412494572004e-06,
+      "loss": 0.2828,
+      "step": 5336
+    },
+    {
+      "epoch": 1.2703635390016066,
+      "grad_norm": 0.35880515050491946,
+      "learning_rate": 3.104628623365352e-06,
+      "loss": 0.3626,
+      "step": 5337
+    },
+    {
+      "epoch": 1.2706015350746713,
+      "grad_norm": 0.4004043215019132,
+      "learning_rate": 3.1028450338577767e-06,
+      "loss": 0.353,
+      "step": 5338
+    },
+    {
+      "epoch": 1.270839531147736,
+      "grad_norm": 0.3743970698217683,
+      "learning_rate": 3.1010617263143616e-06,
+      "loss": 0.2735,
+      "step": 5339
+    },
+    {
+      "epoch": 1.2710775272208008,
+      "grad_norm": 0.4001724562613586,
+      "learning_rate": 3.099278701000152e-06,
+      "loss": 0.3294,
+      "step": 5340
+    },
+    {
+      "epoch": 1.2713155232938655,
+      "grad_norm": 0.3606181579759712,
+      "learning_rate": 3.097495958180149e-06,
+      "loss": 0.3885,
+      "step": 5341
+    },
+    {
+      "epoch": 1.2715535193669305,
+      "grad_norm": 0.3757180444725599,
+      "learning_rate": 3.0957134981193125e-06,
+      "loss": 0.2869,
+      "step": 5342
+    },
+    {
+      "epoch": 1.2717915154399952,
+      "grad_norm": 0.37476881387920113,
+      "learning_rate": 3.09393132108256e-06,
+      "loss": 0.2862,
+      "step": 5343
+    },
+    {
+      "epoch": 1.27202951151306,
+      "grad_norm": 0.36713464862325923,
+      "learning_rate": 3.0921494273347676e-06,
+      "loss": 0.3074,
+      "step": 5344
+    },
+    {
+      "epoch": 1.272267507586125,
+      "grad_norm": 0.3747964323798518,
+      "learning_rate": 3.090367817140767e-06,
+      "loss": 0.4127,
+      "step": 5345
+    },
+    {
+      "epoch": 1.2725055036591897,
+      "grad_norm": 0.39120167273162276,
+      "learning_rate": 3.0885864907653517e-06,
+      "loss": 0.34,
+      "step": 5346
+    },
+    {
+      "epoch": 1.2727434997322544,
+      "grad_norm": 0.3851975033032519,
+      "learning_rate": 3.0868054484732696e-06,
+      "loss": 0.3095,
+      "step": 5347
+    },
+    {
+      "epoch": 1.2729814958053192,
+      "grad_norm": 0.4511302802067385,
+      "learning_rate": 3.085024690529227e-06,
+      "loss": 0.3931,
+      "step": 5348
+    },
+    {
+      "epoch": 1.273219491878384,
+      "grad_norm": 0.41851016514587525,
+      "learning_rate": 3.0832442171978895e-06,
+      "loss": 0.3374,
+      "step": 5349
+    },
+    {
+      "epoch": 1.2734574879514489,
+      "grad_norm": 0.3763563477399065,
+      "learning_rate": 3.0814640287438813e-06,
+      "loss": 0.2577,
+      "step": 5350
+    },
+    {
+      "epoch": 1.2736954840245136,
+      "grad_norm": 0.39491317152340816,
+      "learning_rate": 3.07968412543178e-06,
+      "loss": 0.3377,
+      "step": 5351
+    },
+    {
+      "epoch": 1.2739334800975783,
+      "grad_norm": 0.42386611214893466,
+      "learning_rate": 3.0779045075261237e-06,
+      "loss": 0.3715,
+      "step": 5352
+    },
+    {
+      "epoch": 1.2741714761706433,
+      "grad_norm": 0.35087910158459273,
+      "learning_rate": 3.0761251752914077e-06,
+      "loss": 0.3094,
+      "step": 5353
+    },
+    {
+      "epoch": 1.274409472243708,
+      "grad_norm": 0.347152988488092,
+      "learning_rate": 3.074346128992086e-06,
+      "loss": 0.2898,
+      "step": 5354
+    },
+    {
+      "epoch": 1.2746474683167728,
+      "grad_norm": 0.37229430513448064,
+      "learning_rate": 3.072567368892567e-06,
+      "loss": 0.3509,
+      "step": 5355
+    },
+    {
+      "epoch": 1.2748854643898375,
+      "grad_norm": 0.38027963988847857,
+      "learning_rate": 3.0707888952572206e-06,
+      "loss": 0.3587,
+      "step": 5356
+    },
+    {
+      "epoch": 1.2751234604629023,
+      "grad_norm": 0.3518662269310278,
+      "learning_rate": 3.06901070835037e-06,
+      "loss": 0.2739,
+      "step": 5357
+    },
+    {
+      "epoch": 1.2753614565359672,
+      "grad_norm": 0.371700159550814,
+      "learning_rate": 3.067232808436299e-06,
+      "loss": 0.3261,
+      "step": 5358
+    },
+    {
+      "epoch": 1.275599452609032,
+      "grad_norm": 0.36480653010967556,
+      "learning_rate": 3.0654551957792465e-06,
+      "loss": 0.3602,
+      "step": 5359
+    },
+    {
+      "epoch": 1.2758374486820967,
+      "grad_norm": 0.3755693780387078,
+      "learning_rate": 3.0636778706434104e-06,
+      "loss": 0.3067,
+      "step": 5360
+    },
+    {
+      "epoch": 1.2760754447551617,
+      "grad_norm": 0.4104166990367289,
+      "learning_rate": 3.061900833292945e-06,
+      "loss": 0.288,
+      "step": 5361
+    },
+    {
+      "epoch": 1.2763134408282264,
+      "grad_norm": 0.3860546262996206,
+      "learning_rate": 3.060124083991961e-06,
+      "loss": 0.3287,
+      "step": 5362
+    },
+    {
+      "epoch": 1.2765514369012911,
+      "grad_norm": 0.41369962174697755,
+      "learning_rate": 3.0583476230045283e-06,
+      "loss": 0.3812,
+      "step": 5363
+    },
+    {
+      "epoch": 1.2767894329743559,
+      "grad_norm": 0.36361020540067207,
+      "learning_rate": 3.0565714505946744e-06,
+      "loss": 0.2886,
+      "step": 5364
+    },
+    {
+      "epoch": 1.2770274290474206,
+      "grad_norm": 0.39741880740743823,
+      "learning_rate": 3.0547955670263805e-06,
+      "loss": 0.301,
+      "step": 5365
+    },
+    {
+      "epoch": 1.2772654251204856,
+      "grad_norm": 0.450792782481189,
+      "learning_rate": 3.0530199725635868e-06,
+      "loss": 0.3703,
+      "step": 5366
+    },
+    {
+      "epoch": 1.2775034211935503,
+      "grad_norm": 0.3897194119813881,
+      "learning_rate": 3.0512446674701923e-06,
+      "loss": 0.3263,
+      "step": 5367
+    },
+    {
+      "epoch": 1.277741417266615,
+      "grad_norm": 0.38367874389511963,
+      "learning_rate": 3.04946965201005e-06,
+      "loss": 0.2804,
+      "step": 5368
+    },
+    {
+      "epoch": 1.27797941333968,
+      "grad_norm": 0.39417790396645735,
+      "learning_rate": 3.047694926446971e-06,
+      "loss": 0.3121,
+      "step": 5369
+    },
+    {
+      "epoch": 1.2782174094127448,
+      "grad_norm": 0.3857971707373409,
+      "learning_rate": 3.0459204910447236e-06,
+      "loss": 0.39,
+      "step": 5370
+    },
+    {
+      "epoch": 1.2784554054858095,
+      "grad_norm": 0.3471166269383708,
+      "learning_rate": 3.044146346067033e-06,
+      "loss": 0.2831,
+      "step": 5371
+    },
+    {
+      "epoch": 1.2786934015588742,
+      "grad_norm": 0.42964523510831726,
+      "learning_rate": 3.0423724917775806e-06,
+      "loss": 0.3029,
+      "step": 5372
+    },
+    {
+      "epoch": 1.278931397631939,
+      "grad_norm": 0.39624435945594666,
+      "learning_rate": 3.040598928440005e-06,
+      "loss": 0.3685,
+      "step": 5373
+    },
+    {
+      "epoch": 1.279169393705004,
+      "grad_norm": 0.400417141293227,
+      "learning_rate": 3.0388256563179024e-06,
+      "loss": 0.3483,
+      "step": 5374
+    },
+    {
+      "epoch": 1.2794073897780687,
+      "grad_norm": 0.41684013264207087,
+      "learning_rate": 3.037052675674823e-06,
+      "loss": 0.2727,
+      "step": 5375
+    },
+    {
+      "epoch": 1.2796453858511334,
+      "grad_norm": 0.3675883946338833,
+      "learning_rate": 3.0352799867742788e-06,
+      "loss": 0.3194,
+      "step": 5376
+    },
+    {
+      "epoch": 1.2798833819241984,
+      "grad_norm": 0.35438374829565183,
+      "learning_rate": 3.0335075898797315e-06,
+      "loss": 0.3911,
+      "step": 5377
+    },
+    {
+      "epoch": 1.280121377997263,
+      "grad_norm": 0.40473636844510946,
+      "learning_rate": 3.0317354852546067e-06,
+      "loss": 0.3176,
+      "step": 5378
+    },
+    {
+      "epoch": 1.2803593740703278,
+      "grad_norm": 0.3699615041682992,
+      "learning_rate": 3.02996367316228e-06,
+      "loss": 0.2816,
+      "step": 5379
+    },
+    {
+      "epoch": 1.2805973701433926,
+      "grad_norm": 0.3870198992613291,
+      "learning_rate": 3.0281921538660885e-06,
+      "loss": 0.3498,
+      "step": 5380
+    },
+    {
+      "epoch": 1.2808353662164573,
+      "grad_norm": 0.39893170072412276,
+      "learning_rate": 3.026420927629323e-06,
+      "loss": 0.3506,
+      "step": 5381
+    },
+    {
+      "epoch": 1.2810733622895223,
+      "grad_norm": 0.39285551881065733,
+      "learning_rate": 3.024649994715233e-06,
+      "loss": 0.2825,
+      "step": 5382
+    },
+    {
+      "epoch": 1.281311358362587,
+      "grad_norm": 0.4255415402851682,
+      "learning_rate": 3.0228793553870204e-06,
+      "loss": 0.3282,
+      "step": 5383
+    },
+    {
+      "epoch": 1.2815493544356518,
+      "grad_norm": 0.37137193613780684,
+      "learning_rate": 3.0211090099078475e-06,
+      "loss": 0.3586,
+      "step": 5384
+    },
+    {
+      "epoch": 1.2817873505087167,
+      "grad_norm": 0.3897265976197935,
+      "learning_rate": 3.019338958540831e-06,
+      "loss": 0.314,
+      "step": 5385
+    },
+    {
+      "epoch": 1.2820253465817815,
+      "grad_norm": 0.43026997209773205,
+      "learning_rate": 3.0175692015490443e-06,
+      "loss": 0.2804,
+      "step": 5386
+    },
+    {
+      "epoch": 1.2822633426548462,
+      "grad_norm": 0.39280510289196574,
+      "learning_rate": 3.0157997391955172e-06,
+      "loss": 0.3495,
+      "step": 5387
+    },
+    {
+      "epoch": 1.282501338727911,
+      "grad_norm": 0.3808724957489551,
+      "learning_rate": 3.014030571743236e-06,
+      "loss": 0.384,
+      "step": 5388
+    },
+    {
+      "epoch": 1.2827393348009757,
+      "grad_norm": 0.4445927409783158,
+      "learning_rate": 3.0122616994551413e-06,
+      "loss": 0.2838,
+      "step": 5389
+    },
+    {
+      "epoch": 1.2829773308740406,
+      "grad_norm": 0.3877029737610317,
+      "learning_rate": 3.0104931225941335e-06,
+      "loss": 0.3451,
+      "step": 5390
+    },
+    {
+      "epoch": 1.2832153269471054,
+      "grad_norm": 0.40529519211735904,
+      "learning_rate": 3.008724841423064e-06,
+      "loss": 0.3857,
+      "step": 5391
+    },
+    {
+      "epoch": 1.2834533230201701,
+      "grad_norm": 0.36044362418622544,
+      "learning_rate": 3.006956856204747e-06,
+      "loss": 0.2969,
+      "step": 5392
+    },
+    {
+      "epoch": 1.283691319093235,
+      "grad_norm": 0.4635471433725574,
+      "learning_rate": 3.005189167201945e-06,
+      "loss": 0.3053,
+      "step": 5393
+    },
+    {
+      "epoch": 1.2839293151662998,
+      "grad_norm": 0.3878593945568825,
+      "learning_rate": 3.003421774677383e-06,
+      "loss": 0.3243,
+      "step": 5394
+    },
+    {
+      "epoch": 1.2841673112393646,
+      "grad_norm": 0.4323624105558046,
+      "learning_rate": 3.0016546788937385e-06,
+      "loss": 0.3795,
+      "step": 5395
+    },
+    {
+      "epoch": 1.2844053073124293,
+      "grad_norm": 0.37258405194326444,
+      "learning_rate": 2.9998878801136442e-06,
+      "loss": 0.2922,
+      "step": 5396
+    },
+    {
+      "epoch": 1.284643303385494,
+      "grad_norm": 0.3984352432024578,
+      "learning_rate": 2.9981213785996925e-06,
+      "loss": 0.3179,
+      "step": 5397
+    },
+    {
+      "epoch": 1.284881299458559,
+      "grad_norm": 0.3451845968169443,
+      "learning_rate": 2.996355174614428e-06,
+      "loss": 0.3662,
+      "step": 5398
+    },
+    {
+      "epoch": 1.2851192955316237,
+      "grad_norm": 0.3786688602776222,
+      "learning_rate": 2.994589268420352e-06,
+      "loss": 0.3069,
+      "step": 5399
+    },
+    {
+      "epoch": 1.2853572916046885,
+      "grad_norm": 0.41067641786263726,
+      "learning_rate": 2.992823660279922e-06,
+      "loss": 0.2652,
+      "step": 5400
+    },
+    {
+      "epoch": 1.2855952876777534,
+      "grad_norm": 0.3922515926989098,
+      "learning_rate": 2.9910583504555516e-06,
+      "loss": 0.3191,
+      "step": 5401
+    },
+    {
+      "epoch": 1.2858332837508182,
+      "grad_norm": 0.38153653409480226,
+      "learning_rate": 2.989293339209608e-06,
+      "loss": 0.3894,
+      "step": 5402
+    },
+    {
+      "epoch": 1.286071279823883,
+      "grad_norm": 0.37037061821185474,
+      "learning_rate": 2.987528626804418e-06,
+      "loss": 0.2799,
+      "step": 5403
+    },
+    {
+      "epoch": 1.2863092758969477,
+      "grad_norm": 0.39954728304549386,
+      "learning_rate": 2.98576421350226e-06,
+      "loss": 0.2984,
+      "step": 5404
+    },
+    {
+      "epoch": 1.2865472719700124,
+      "grad_norm": 0.42121203353357217,
+      "learning_rate": 2.9840000995653684e-06,
+      "loss": 0.3576,
+      "step": 5405
+    },
+    {
+      "epoch": 1.2867852680430774,
+      "grad_norm": 0.378600084659916,
+      "learning_rate": 2.9822362852559363e-06,
+      "loss": 0.3538,
+      "step": 5406
+    },
+    {
+      "epoch": 1.287023264116142,
+      "grad_norm": 0.37558983767345905,
+      "learning_rate": 2.9804727708361094e-06,
+      "loss": 0.288,
+      "step": 5407
+    },
+    {
+      "epoch": 1.2872612601892068,
+      "grad_norm": 0.4097981060709528,
+      "learning_rate": 2.9787095565679893e-06,
+      "loss": 0.3418,
+      "step": 5408
+    },
+    {
+      "epoch": 1.2874992562622718,
+      "grad_norm": 0.3880613286304227,
+      "learning_rate": 2.976946642713634e-06,
+      "loss": 0.3919,
+      "step": 5409
+    },
+    {
+      "epoch": 1.2877372523353365,
+      "grad_norm": 0.3576965119071218,
+      "learning_rate": 2.9751840295350554e-06,
+      "loss": 0.3108,
+      "step": 5410
+    },
+    {
+      "epoch": 1.2879752484084013,
+      "grad_norm": 0.386930113400703,
+      "learning_rate": 2.9734217172942216e-06,
+      "loss": 0.2923,
+      "step": 5411
+    },
+    {
+      "epoch": 1.288213244481466,
+      "grad_norm": 0.37377948379351883,
+      "learning_rate": 2.971659706253055e-06,
+      "loss": 0.3518,
+      "step": 5412
+    },
+    {
+      "epoch": 1.2884512405545308,
+      "grad_norm": 0.3907730616845053,
+      "learning_rate": 2.9698979966734353e-06,
+      "loss": 0.3765,
+      "step": 5413
+    },
+    {
+      "epoch": 1.2886892366275957,
+      "grad_norm": 0.35233568134416837,
+      "learning_rate": 2.968136588817196e-06,
+      "loss": 0.2804,
+      "step": 5414
+    },
+    {
+      "epoch": 1.2889272327006605,
+      "grad_norm": 0.3695879561285849,
+      "learning_rate": 2.966375482946125e-06,
+      "loss": 0.3208,
+      "step": 5415
+    },
+    {
+      "epoch": 1.2891652287737252,
+      "grad_norm": 0.383905634810903,
+      "learning_rate": 2.964614679321966e-06,
+      "loss": 0.3657,
+      "step": 5416
+    },
+    {
+      "epoch": 1.2894032248467902,
+      "grad_norm": 0.36362432939576694,
+      "learning_rate": 2.962854178206419e-06,
+      "loss": 0.3283,
+      "step": 5417
+    },
+    {
+      "epoch": 1.289641220919855,
+      "grad_norm": 0.37983820596988416,
+      "learning_rate": 2.961093979861137e-06,
+      "loss": 0.2758,
+      "step": 5418
+    },
+    {
+      "epoch": 1.2898792169929196,
+      "grad_norm": 0.3719877524530933,
+      "learning_rate": 2.9593340845477315e-06,
+      "loss": 0.3312,
+      "step": 5419
+    },
+    {
+      "epoch": 1.2901172130659844,
+      "grad_norm": 0.39592619284538855,
+      "learning_rate": 2.9575744925277626e-06,
+      "loss": 0.3998,
+      "step": 5420
+    },
+    {
+      "epoch": 1.290355209139049,
+      "grad_norm": 0.36564015551941653,
+      "learning_rate": 2.955815204062753e-06,
+      "loss": 0.2775,
+      "step": 5421
+    },
+    {
+      "epoch": 1.290593205212114,
+      "grad_norm": 0.3883816093497894,
+      "learning_rate": 2.954056219414174e-06,
+      "loss": 0.3132,
+      "step": 5422
+    },
+    {
+      "epoch": 1.2908312012851788,
+      "grad_norm": 0.3830741599542019,
+      "learning_rate": 2.952297538843456e-06,
+      "loss": 0.3843,
+      "step": 5423
+    },
+    {
+      "epoch": 1.2910691973582435,
+      "grad_norm": 0.3735082644605664,
+      "learning_rate": 2.9505391626119804e-06,
+      "loss": 0.3218,
+      "step": 5424
+    },
+    {
+      "epoch": 1.2913071934313085,
+      "grad_norm": 0.39360349279204443,
+      "learning_rate": 2.9487810909810876e-06,
+      "loss": 0.2838,
+      "step": 5425
+    },
+    {
+      "epoch": 1.2915451895043732,
+      "grad_norm": 0.39199734550206006,
+      "learning_rate": 2.947023324212069e-06,
+      "loss": 0.3363,
+      "step": 5426
+    },
+    {
+      "epoch": 1.291783185577438,
+      "grad_norm": 0.3946282023306039,
+      "learning_rate": 2.945265862566172e-06,
+      "loss": 0.3958,
+      "step": 5427
+    },
+    {
+      "epoch": 1.2920211816505027,
+      "grad_norm": 0.3833972335654247,
+      "learning_rate": 2.9435087063045997e-06,
+      "loss": 0.3168,
+      "step": 5428
+    },
+    {
+      "epoch": 1.2922591777235675,
+      "grad_norm": 0.36324852808260755,
+      "learning_rate": 2.9417518556885085e-06,
+      "loss": 0.3138,
+      "step": 5429
+    },
+    {
+      "epoch": 1.2924971737966324,
+      "grad_norm": 0.4118882053851258,
+      "learning_rate": 2.9399953109790104e-06,
+      "loss": 0.364,
+      "step": 5430
+    },
+    {
+      "epoch": 1.2927351698696972,
+      "grad_norm": 0.4094387139947238,
+      "learning_rate": 2.938239072437171e-06,
+      "loss": 0.358,
+      "step": 5431
+    },
+    {
+      "epoch": 1.292973165942762,
+      "grad_norm": 0.3746366818760013,
+      "learning_rate": 2.936483140324011e-06,
+      "loss": 0.3152,
+      "step": 5432
+    },
+    {
+      "epoch": 1.2932111620158269,
+      "grad_norm": 0.3905453061521183,
+      "learning_rate": 2.9347275149005046e-06,
+      "loss": 0.3501,
+      "step": 5433
+    },
+    {
+      "epoch": 1.2934491580888916,
+      "grad_norm": 0.4541439412535773,
+      "learning_rate": 2.9329721964275827e-06,
+      "loss": 0.401,
+      "step": 5434
+    },
+    {
+      "epoch": 1.2936871541619563,
+      "grad_norm": 0.398933342343559,
+      "learning_rate": 2.9312171851661285e-06,
+      "loss": 0.3134,
+      "step": 5435
+    },
+    {
+      "epoch": 1.293925150235021,
+      "grad_norm": 0.42307597168640104,
+      "learning_rate": 2.9294624813769795e-06,
+      "loss": 0.2636,
+      "step": 5436
+    },
+    {
+      "epoch": 1.2941631463080858,
+      "grad_norm": 0.4037706037517579,
+      "learning_rate": 2.9277080853209284e-06,
+      "loss": 0.3405,
+      "step": 5437
+    },
+    {
+      "epoch": 1.2944011423811508,
+      "grad_norm": 0.3812267379791281,
+      "learning_rate": 2.9259539972587227e-06,
+      "loss": 0.382,
+      "step": 5438
+    },
+    {
+      "epoch": 1.2946391384542155,
+      "grad_norm": 0.364830365933242,
+      "learning_rate": 2.9242002174510613e-06,
+      "loss": 0.2903,
+      "step": 5439
+    },
+    {
+      "epoch": 1.2948771345272803,
+      "grad_norm": 0.3897401469920776,
+      "learning_rate": 2.922446746158601e-06,
+      "loss": 0.3097,
+      "step": 5440
+    },
+    {
+      "epoch": 1.2951151306003452,
+      "grad_norm": 0.44911236787123654,
+      "learning_rate": 2.92069358364195e-06,
+      "loss": 0.4018,
+      "step": 5441
+    },
+    {
+      "epoch": 1.29535312667341,
+      "grad_norm": 0.37167943879371046,
+      "learning_rate": 2.918940730161672e-06,
+      "loss": 0.3206,
+      "step": 5442
+    },
+    {
+      "epoch": 1.2955911227464747,
+      "grad_norm": 0.35873134087334885,
+      "learning_rate": 2.9171881859782854e-06,
+      "loss": 0.2722,
+      "step": 5443
+    },
+    {
+      "epoch": 1.2958291188195394,
+      "grad_norm": 0.37574331573192493,
+      "learning_rate": 2.91543595135226e-06,
+      "loss": 0.3285,
+      "step": 5444
+    },
+    {
+      "epoch": 1.2960671148926042,
+      "grad_norm": 0.37774497295843434,
+      "learning_rate": 2.9136840265440213e-06,
+      "loss": 0.3934,
+      "step": 5445
+    },
+    {
+      "epoch": 1.2963051109656691,
+      "grad_norm": 0.37035405909980235,
+      "learning_rate": 2.9119324118139482e-06,
+      "loss": 0.2877,
+      "step": 5446
+    },
+    {
+      "epoch": 1.2965431070387339,
+      "grad_norm": 0.3719053824299425,
+      "learning_rate": 2.9101811074223762e-06,
+      "loss": 0.3049,
+      "step": 5447
+    },
+    {
+      "epoch": 1.2967811031117986,
+      "grad_norm": 0.38418916465237035,
+      "learning_rate": 2.9084301136295922e-06,
+      "loss": 0.3589,
+      "step": 5448
+    },
+    {
+      "epoch": 1.2970190991848636,
+      "grad_norm": 0.38154835318426467,
+      "learning_rate": 2.9066794306958356e-06,
+      "loss": 0.3518,
+      "step": 5449
+    },
+    {
+      "epoch": 1.2972570952579283,
+      "grad_norm": 0.37098490527058375,
+      "learning_rate": 2.904929058881302e-06,
+      "loss": 0.2862,
+      "step": 5450
+    },
+    {
+      "epoch": 1.297495091330993,
+      "grad_norm": 0.35922712949248786,
+      "learning_rate": 2.90317899844614e-06,
+      "loss": 0.3084,
+      "step": 5451
+    },
+    {
+      "epoch": 1.2977330874040578,
+      "grad_norm": 0.3752626421355902,
+      "learning_rate": 2.9014292496504493e-06,
+      "loss": 0.3723,
+      "step": 5452
+    },
+    {
+      "epoch": 1.2979710834771225,
+      "grad_norm": 0.3660609605786775,
+      "learning_rate": 2.89967981275429e-06,
+      "loss": 0.3029,
+      "step": 5453
+    },
+    {
+      "epoch": 1.2982090795501875,
+      "grad_norm": 0.383302664535824,
+      "learning_rate": 2.8979306880176706e-06,
+      "loss": 0.3021,
+      "step": 5454
+    },
+    {
+      "epoch": 1.2984470756232522,
+      "grad_norm": 0.3994458439573466,
+      "learning_rate": 2.8961818757005533e-06,
+      "loss": 0.3402,
+      "step": 5455
+    },
+    {
+      "epoch": 1.298685071696317,
+      "grad_norm": 0.3971982374786896,
+      "learning_rate": 2.894433376062855e-06,
+      "loss": 0.3528,
+      "step": 5456
+    },
+    {
+      "epoch": 1.298923067769382,
+      "grad_norm": 0.3882804894172824,
+      "learning_rate": 2.892685189364447e-06,
+      "loss": 0.2716,
+      "step": 5457
+    },
+    {
+      "epoch": 1.2991610638424467,
+      "grad_norm": 0.3899579933994844,
+      "learning_rate": 2.8909373158651523e-06,
+      "loss": 0.3196,
+      "step": 5458
+    },
+    {
+      "epoch": 1.2993990599155114,
+      "grad_norm": 0.3855353642705331,
+      "learning_rate": 2.889189755824747e-06,
+      "loss": 0.3789,
+      "step": 5459
+    },
+    {
+      "epoch": 1.2996370559885762,
+      "grad_norm": 0.3622462412263447,
+      "learning_rate": 2.887442509502962e-06,
+      "loss": 0.2928,
+      "step": 5460
+    },
+    {
+      "epoch": 1.299875052061641,
+      "grad_norm": 0.38805537438834,
+      "learning_rate": 2.885695577159484e-06,
+      "loss": 0.2817,
+      "step": 5461
+    },
+    {
+      "epoch": 1.3001130481347059,
+      "grad_norm": 0.4137674453573019,
+      "learning_rate": 2.8839489590539482e-06,
+      "loss": 0.3426,
+      "step": 5462
+    },
+    {
+      "epoch": 1.3003510442077706,
+      "grad_norm": 0.37778518714576936,
+      "learning_rate": 2.882202655445946e-06,
+      "loss": 0.3454,
+      "step": 5463
+    },
+    {
+      "epoch": 1.3005890402808353,
+      "grad_norm": 0.34689911133155854,
+      "learning_rate": 2.8804566665950207e-06,
+      "loss": 0.2948,
+      "step": 5464
+    },
+    {
+      "epoch": 1.3008270363539003,
+      "grad_norm": 0.3881888255918,
+      "learning_rate": 2.8787109927606704e-06,
+      "loss": 0.3178,
+      "step": 5465
+    },
+    {
+      "epoch": 1.301065032426965,
+      "grad_norm": 0.3689773814104203,
+      "learning_rate": 2.8769656342023445e-06,
+      "loss": 0.3571,
+      "step": 5466
+    },
+    {
+      "epoch": 1.3013030285000298,
+      "grad_norm": 0.37850711600818543,
+      "learning_rate": 2.8752205911794463e-06,
+      "loss": 0.3288,
+      "step": 5467
+    },
+    {
+      "epoch": 1.3015410245730945,
+      "grad_norm": 0.3848381260854274,
+      "learning_rate": 2.8734758639513327e-06,
+      "loss": 0.2831,
+      "step": 5468
+    },
+    {
+      "epoch": 1.3017790206461592,
+      "grad_norm": 0.3517844887890128,
+      "learning_rate": 2.8717314527773134e-06,
+      "loss": 0.3123,
+      "step": 5469
+    },
+    {
+      "epoch": 1.3020170167192242,
+      "grad_norm": 0.380584564601741,
+      "learning_rate": 2.8699873579166517e-06,
+      "loss": 0.3761,
+      "step": 5470
+    },
+    {
+      "epoch": 1.302255012792289,
+      "grad_norm": 0.3577811945107307,
+      "learning_rate": 2.8682435796285617e-06,
+      "loss": 0.285,
+      "step": 5471
+    },
+    {
+      "epoch": 1.3024930088653537,
+      "grad_norm": 0.3819440878979895,
+      "learning_rate": 2.8665001181722134e-06,
+      "loss": 0.2917,
+      "step": 5472
+    },
+    {
+      "epoch": 1.3027310049384186,
+      "grad_norm": 0.38014911424877107,
+      "learning_rate": 2.8647569738067262e-06,
+      "loss": 0.3719,
+      "step": 5473
+    },
+    {
+      "epoch": 1.3029690010114834,
+      "grad_norm": 0.38123121571081403,
+      "learning_rate": 2.8630141467911777e-06,
+      "loss": 0.3411,
+      "step": 5474
+    },
+    {
+      "epoch": 1.3032069970845481,
+      "grad_norm": 0.3930220532783937,
+      "learning_rate": 2.8612716373845927e-06,
+      "loss": 0.2891,
+      "step": 5475
+    },
+    {
+      "epoch": 1.3034449931576129,
+      "grad_norm": 0.4003883784524193,
+      "learning_rate": 2.859529445845953e-06,
+      "loss": 0.3191,
+      "step": 5476
+    },
+    {
+      "epoch": 1.3036829892306776,
+      "grad_norm": 0.39728799061286724,
+      "learning_rate": 2.8577875724341897e-06,
+      "loss": 0.3656,
+      "step": 5477
+    },
+    {
+      "epoch": 1.3039209853037426,
+      "grad_norm": 0.36276882877846156,
+      "learning_rate": 2.8560460174081896e-06,
+      "loss": 0.3009,
+      "step": 5478
+    },
+    {
+      "epoch": 1.3041589813768073,
+      "grad_norm": 0.34834226783278643,
+      "learning_rate": 2.8543047810267876e-06,
+      "loss": 0.2964,
+      "step": 5479
+    },
+    {
+      "epoch": 1.304396977449872,
+      "grad_norm": 0.37989807168764456,
+      "learning_rate": 2.852563863548779e-06,
+      "loss": 0.3486,
+      "step": 5480
+    },
+    {
+      "epoch": 1.304634973522937,
+      "grad_norm": 0.36552520765488317,
+      "learning_rate": 2.850823265232906e-06,
+      "loss": 0.3672,
+      "step": 5481
+    },
+    {
+      "epoch": 1.3048729695960017,
+      "grad_norm": 0.36751382218934403,
+      "learning_rate": 2.849082986337863e-06,
+      "loss": 0.251,
+      "step": 5482
+    },
+    {
+      "epoch": 1.3051109656690665,
+      "grad_norm": 0.40093768891686205,
+      "learning_rate": 2.8473430271222994e-06,
+      "loss": 0.3348,
+      "step": 5483
+    },
+    {
+      "epoch": 1.3053489617421312,
+      "grad_norm": 0.3780458249964564,
+      "learning_rate": 2.845603387844817e-06,
+      "loss": 0.3953,
+      "step": 5484
+    },
+    {
+      "epoch": 1.305586957815196,
+      "grad_norm": 0.3807478074730611,
+      "learning_rate": 2.8438640687639675e-06,
+      "loss": 0.3076,
+      "step": 5485
+    },
+    {
+      "epoch": 1.305824953888261,
+      "grad_norm": 0.3703942594376492,
+      "learning_rate": 2.842125070138258e-06,
+      "loss": 0.2822,
+      "step": 5486
+    },
+    {
+      "epoch": 1.3060629499613257,
+      "grad_norm": 0.3600318688567636,
+      "learning_rate": 2.8403863922261444e-06,
+      "loss": 0.3293,
+      "step": 5487
+    },
+    {
+      "epoch": 1.3063009460343904,
+      "grad_norm": 0.3609650481015529,
+      "learning_rate": 2.8386480352860414e-06,
+      "loss": 0.3507,
+      "step": 5488
+    },
+    {
+      "epoch": 1.3065389421074554,
+      "grad_norm": 0.36174141337750676,
+      "learning_rate": 2.8369099995763088e-06,
+      "loss": 0.2888,
+      "step": 5489
+    },
+    {
+      "epoch": 1.30677693818052,
+      "grad_norm": 0.3641547780000098,
+      "learning_rate": 2.835172285355263e-06,
+      "loss": 0.2959,
+      "step": 5490
+    },
+    {
+      "epoch": 1.3070149342535848,
+      "grad_norm": 0.35270388989031864,
+      "learning_rate": 2.833434892881171e-06,
+      "loss": 0.3901,
+      "step": 5491
+    },
+    {
+      "epoch": 1.3072529303266496,
+      "grad_norm": 0.41231868386069426,
+      "learning_rate": 2.831697822412252e-06,
+      "loss": 0.3232,
+      "step": 5492
+    },
+    {
+      "epoch": 1.3074909263997143,
+      "grad_norm": 0.4084581682764854,
+      "learning_rate": 2.8299610742066778e-06,
+      "loss": 0.2764,
+      "step": 5493
+    },
+    {
+      "epoch": 1.3077289224727793,
+      "grad_norm": 0.391038761585091,
+      "learning_rate": 2.8282246485225722e-06,
+      "loss": 0.3169,
+      "step": 5494
+    },
+    {
+      "epoch": 1.307966918545844,
+      "grad_norm": 0.392419286404586,
+      "learning_rate": 2.826488545618011e-06,
+      "loss": 0.41,
+      "step": 5495
+    },
+    {
+      "epoch": 1.3082049146189088,
+      "grad_norm": 0.36111528650542957,
+      "learning_rate": 2.824752765751022e-06,
+      "loss": 0.2922,
+      "step": 5496
+    },
+    {
+      "epoch": 1.3084429106919737,
+      "grad_norm": 0.40433967227880035,
+      "learning_rate": 2.8230173091795853e-06,
+      "loss": 0.312,
+      "step": 5497
+    },
+    {
+      "epoch": 1.3086809067650385,
+      "grad_norm": 0.4150889724294196,
+      "learning_rate": 2.8212821761616314e-06,
+      "loss": 0.3416,
+      "step": 5498
+    },
+    {
+      "epoch": 1.3089189028381032,
+      "grad_norm": 0.3784389771780397,
+      "learning_rate": 2.819547366955046e-06,
+      "loss": 0.3283,
+      "step": 5499
+    },
+    {
+      "epoch": 1.309156898911168,
+      "grad_norm": 0.39020777536186346,
+      "learning_rate": 2.817812881817663e-06,
+      "loss": 0.3039,
+      "step": 5500
+    },
+    {
+      "epoch": 1.3093948949842327,
+      "grad_norm": 0.38412700599020266,
+      "learning_rate": 2.8160787210072695e-06,
+      "loss": 0.335,
+      "step": 5501
+    },
+    {
+      "epoch": 1.3096328910572976,
+      "grad_norm": 0.37696789257027463,
+      "learning_rate": 2.814344884781607e-06,
+      "loss": 0.3842,
+      "step": 5502
+    },
+    {
+      "epoch": 1.3098708871303624,
+      "grad_norm": 0.3592980808212725,
+      "learning_rate": 2.812611373398365e-06,
+      "loss": 0.2864,
+      "step": 5503
+    },
+    {
+      "epoch": 1.310108883203427,
+      "grad_norm": 0.3932282988991156,
+      "learning_rate": 2.8108781871151866e-06,
+      "loss": 0.2911,
+      "step": 5504
+    },
+    {
+      "epoch": 1.310346879276492,
+      "grad_norm": 0.366771028303434,
+      "learning_rate": 2.8091453261896657e-06,
+      "loss": 0.3637,
+      "step": 5505
+    },
+    {
+      "epoch": 1.3105848753495568,
+      "grad_norm": 0.3612811617097267,
+      "learning_rate": 2.8074127908793464e-06,
+      "loss": 0.3796,
+      "step": 5506
+    },
+    {
+      "epoch": 1.3108228714226215,
+      "grad_norm": 0.3522436714920477,
+      "learning_rate": 2.8056805814417305e-06,
+      "loss": 0.2732,
+      "step": 5507
+    },
+    {
+      "epoch": 1.3110608674956863,
+      "grad_norm": 0.40918249461660394,
+      "learning_rate": 2.8039486981342647e-06,
+      "loss": 0.3357,
+      "step": 5508
+    },
+    {
+      "epoch": 1.311298863568751,
+      "grad_norm": 0.4104488620009276,
+      "learning_rate": 2.8022171412143504e-06,
+      "loss": 0.3646,
+      "step": 5509
+    },
+    {
+      "epoch": 1.311536859641816,
+      "grad_norm": 0.3643401976041939,
+      "learning_rate": 2.800485910939339e-06,
+      "loss": 0.3133,
+      "step": 5510
+    },
+    {
+      "epoch": 1.3117748557148807,
+      "grad_norm": 0.3894095595689637,
+      "learning_rate": 2.7987550075665356e-06,
+      "loss": 0.2957,
+      "step": 5511
+    },
+    {
+      "epoch": 1.3120128517879455,
+      "grad_norm": 0.38028414239770225,
+      "learning_rate": 2.7970244313531935e-06,
+      "loss": 0.3483,
+      "step": 5512
+    },
+    {
+      "epoch": 1.3122508478610104,
+      "grad_norm": 0.383503362557972,
+      "learning_rate": 2.7952941825565193e-06,
+      "loss": 0.3826,
+      "step": 5513
+    },
+    {
+      "epoch": 1.3124888439340752,
+      "grad_norm": 0.3753460629109302,
+      "learning_rate": 2.793564261433672e-06,
+      "loss": 0.2777,
+      "step": 5514
+    },
+    {
+      "epoch": 1.31272684000714,
+      "grad_norm": 0.38982091819158365,
+      "learning_rate": 2.7918346682417585e-06,
+      "loss": 0.3217,
+      "step": 5515
+    },
+    {
+      "epoch": 1.3129648360802046,
+      "grad_norm": 0.39383363481041805,
+      "learning_rate": 2.7901054032378426e-06,
+      "loss": 0.3899,
+      "step": 5516
+    },
+    {
+      "epoch": 1.3132028321532694,
+      "grad_norm": 0.43566741874589754,
+      "learning_rate": 2.7883764666789336e-06,
+      "loss": 0.2971,
+      "step": 5517
+    },
+    {
+      "epoch": 1.3134408282263343,
+      "grad_norm": 0.38349493003039725,
+      "learning_rate": 2.7866478588219945e-06,
+      "loss": 0.2766,
+      "step": 5518
+    },
+    {
+      "epoch": 1.313678824299399,
+      "grad_norm": 0.3767513973073563,
+      "learning_rate": 2.784919579923939e-06,
+      "loss": 0.3182,
+      "step": 5519
+    },
+    {
+      "epoch": 1.3139168203724638,
+      "grad_norm": 0.37130947209320847,
+      "learning_rate": 2.783191630241633e-06,
+      "loss": 0.3982,
+      "step": 5520
+    },
+    {
+      "epoch": 1.3141548164455288,
+      "grad_norm": 0.36376573221358194,
+      "learning_rate": 2.7814640100318917e-06,
+      "loss": 0.2772,
+      "step": 5521
+    },
+    {
+      "epoch": 1.3143928125185935,
+      "grad_norm": 0.504440580572112,
+      "learning_rate": 2.7797367195514825e-06,
+      "loss": 0.3145,
+      "step": 5522
+    },
+    {
+      "epoch": 1.3146308085916583,
+      "grad_norm": 0.3834097822594875,
+      "learning_rate": 2.778009759057123e-06,
+      "loss": 0.3448,
+      "step": 5523
+    },
+    {
+      "epoch": 1.314868804664723,
+      "grad_norm": 0.3670665934349199,
+      "learning_rate": 2.7762831288054836e-06,
+      "loss": 0.3409,
+      "step": 5524
+    },
+    {
+      "epoch": 1.3151068007377877,
+      "grad_norm": 0.37382482389619465,
+      "learning_rate": 2.7745568290531827e-06,
+      "loss": 0.2827,
+      "step": 5525
+    },
+    {
+      "epoch": 1.3153447968108527,
+      "grad_norm": 0.3685314429800074,
+      "learning_rate": 2.772830860056792e-06,
+      "loss": 0.3201,
+      "step": 5526
+    },
+    {
+      "epoch": 1.3155827928839174,
+      "grad_norm": 0.36043420452079505,
+      "learning_rate": 2.771105222072833e-06,
+      "loss": 0.426,
+      "step": 5527
+    },
+    {
+      "epoch": 1.3158207889569822,
+      "grad_norm": 0.36717830372844196,
+      "learning_rate": 2.769379915357776e-06,
+      "loss": 0.2985,
+      "step": 5528
+    },
+    {
+      "epoch": 1.3160587850300471,
+      "grad_norm": 0.37554243694911554,
+      "learning_rate": 2.7676549401680486e-06,
+      "loss": 0.2885,
+      "step": 5529
+    },
+    {
+      "epoch": 1.3162967811031119,
+      "grad_norm": 0.3741544659657053,
+      "learning_rate": 2.7659302967600226e-06,
+      "loss": 0.3648,
+      "step": 5530
+    },
+    {
+      "epoch": 1.3165347771761766,
+      "grad_norm": 0.38217417099471307,
+      "learning_rate": 2.764205985390023e-06,
+      "loss": 0.3696,
+      "step": 5531
+    },
+    {
+      "epoch": 1.3167727732492414,
+      "grad_norm": 0.3587235824649817,
+      "learning_rate": 2.762482006314324e-06,
+      "loss": 0.2703,
+      "step": 5532
+    },
+    {
+      "epoch": 1.317010769322306,
+      "grad_norm": 0.37109675762063177,
+      "learning_rate": 2.760758359789151e-06,
+      "loss": 0.3151,
+      "step": 5533
+    },
+    {
+      "epoch": 1.317248765395371,
+      "grad_norm": 0.4298879106634171,
+      "learning_rate": 2.7590350460706845e-06,
+      "loss": 0.3901,
+      "step": 5534
+    },
+    {
+      "epoch": 1.3174867614684358,
+      "grad_norm": 0.3555539814561159,
+      "learning_rate": 2.757312065415048e-06,
+      "loss": 0.2981,
+      "step": 5535
+    },
+    {
+      "epoch": 1.3177247575415005,
+      "grad_norm": 0.3717826964329484,
+      "learning_rate": 2.7555894180783203e-06,
+      "loss": 0.3002,
+      "step": 5536
+    },
+    {
+      "epoch": 1.3179627536145655,
+      "grad_norm": 0.41601130773589784,
+      "learning_rate": 2.7538671043165295e-06,
+      "loss": 0.3482,
+      "step": 5537
+    },
+    {
+      "epoch": 1.3182007496876302,
+      "grad_norm": 0.3901057217046057,
+      "learning_rate": 2.7521451243856533e-06,
+      "loss": 0.3883,
+      "step": 5538
+    },
+    {
+      "epoch": 1.318438745760695,
+      "grad_norm": 0.4228120609797883,
+      "learning_rate": 2.7504234785416206e-06,
+      "loss": 0.3078,
+      "step": 5539
+    },
+    {
+      "epoch": 1.3186767418337597,
+      "grad_norm": 0.3842260920241478,
+      "learning_rate": 2.7487021670403115e-06,
+      "loss": 0.315,
+      "step": 5540
+    },
+    {
+      "epoch": 1.3189147379068245,
+      "grad_norm": 0.41311984790557893,
+      "learning_rate": 2.746981190137554e-06,
+      "loss": 0.3725,
+      "step": 5541
+    },
+    {
+      "epoch": 1.3191527339798894,
+      "grad_norm": 0.3577865374099851,
+      "learning_rate": 2.7452605480891276e-06,
+      "loss": 0.3359,
+      "step": 5542
+    },
+    {
+      "epoch": 1.3193907300529542,
+      "grad_norm": 0.3845593269549925,
+      "learning_rate": 2.743540241150765e-06,
+      "loss": 0.2928,
+      "step": 5543
+    },
+    {
+      "epoch": 1.319628726126019,
+      "grad_norm": 0.3988246990305431,
+      "learning_rate": 2.7418202695781443e-06,
+      "loss": 0.3563,
+      "step": 5544
+    },
+    {
+      "epoch": 1.3198667221990839,
+      "grad_norm": 0.3926174352866639,
+      "learning_rate": 2.7401006336268966e-06,
+      "loss": 0.3319,
+      "step": 5545
+    },
+    {
+      "epoch": 1.3201047182721486,
+      "grad_norm": 0.3877729503505814,
+      "learning_rate": 2.738381333552601e-06,
+      "loss": 0.2998,
+      "step": 5546
+    },
+    {
+      "epoch": 1.3203427143452133,
+      "grad_norm": 0.40291838283798237,
+      "learning_rate": 2.73666236961079e-06,
+      "loss": 0.3096,
+      "step": 5547
+    },
+    {
+      "epoch": 1.320580710418278,
+      "grad_norm": 0.36068628191130936,
+      "learning_rate": 2.734943742056943e-06,
+      "loss": 0.3678,
+      "step": 5548
+    },
+    {
+      "epoch": 1.3208187064913428,
+      "grad_norm": 0.3709054992297036,
+      "learning_rate": 2.7332254511464906e-06,
+      "loss": 0.3333,
+      "step": 5549
+    },
+    {
+      "epoch": 1.3210567025644078,
+      "grad_norm": 0.3673136445373127,
+      "learning_rate": 2.7315074971348133e-06,
+      "loss": 0.303,
+      "step": 5550
+    },
+    {
+      "epoch": 1.3212946986374725,
+      "grad_norm": 0.3626370946230582,
+      "learning_rate": 2.729789880277242e-06,
+      "loss": 0.3252,
+      "step": 5551
+    },
+    {
+      "epoch": 1.3215326947105372,
+      "grad_norm": 0.3761315640459379,
+      "learning_rate": 2.728072600829057e-06,
+      "loss": 0.4076,
+      "step": 5552
+    },
+    {
+      "epoch": 1.3217706907836022,
+      "grad_norm": 0.38806642072912284,
+      "learning_rate": 2.7263556590454874e-06,
+      "loss": 0.3095,
+      "step": 5553
+    },
+    {
+      "epoch": 1.322008686856667,
+      "grad_norm": 0.3827363695116688,
+      "learning_rate": 2.724639055181715e-06,
+      "loss": 0.3012,
+      "step": 5554
+    },
+    {
+      "epoch": 1.3222466829297317,
+      "grad_norm": 0.406266926754827,
+      "learning_rate": 2.7229227894928666e-06,
+      "loss": 0.3563,
+      "step": 5555
+    },
+    {
+      "epoch": 1.3224846790027964,
+      "grad_norm": 0.3601509616480053,
+      "learning_rate": 2.721206862234026e-06,
+      "loss": 0.3359,
+      "step": 5556
+    },
+    {
+      "epoch": 1.3227226750758612,
+      "grad_norm": 0.39622194453014564,
+      "learning_rate": 2.71949127366022e-06,
+      "loss": 0.3009,
+      "step": 5557
+    },
+    {
+      "epoch": 1.322960671148926,
+      "grad_norm": 0.40428113767002444,
+      "learning_rate": 2.7177760240264273e-06,
+      "loss": 0.3141,
+      "step": 5558
+    },
+    {
+      "epoch": 1.3231986672219909,
+      "grad_norm": 0.39692065016659434,
+      "learning_rate": 2.7160611135875774e-06,
+      "loss": 0.3779,
+      "step": 5559
+    },
+    {
+      "epoch": 1.3234366632950556,
+      "grad_norm": 0.3546448081276393,
+      "learning_rate": 2.714346542598546e-06,
+      "loss": 0.3072,
+      "step": 5560
+    },
+    {
+      "epoch": 1.3236746593681203,
+      "grad_norm": 0.40729641787879445,
+      "learning_rate": 2.712632311314165e-06,
+      "loss": 0.2691,
+      "step": 5561
+    },
+    {
+      "epoch": 1.3239126554411853,
+      "grad_norm": 0.374238994477962,
+      "learning_rate": 2.7109184199892093e-06,
+      "loss": 0.3453,
+      "step": 5562
+    },
+    {
+      "epoch": 1.32415065151425,
+      "grad_norm": 0.3684026547826237,
+      "learning_rate": 2.7092048688784046e-06,
+      "loss": 0.3535,
+      "step": 5563
+    },
+    {
+      "epoch": 1.3243886475873148,
+      "grad_norm": 0.3957287753476546,
+      "learning_rate": 2.7074916582364284e-06,
+      "loss": 0.2692,
+      "step": 5564
+    },
+    {
+      "epoch": 1.3246266436603795,
+      "grad_norm": 0.3807784927555232,
+      "learning_rate": 2.705778788317906e-06,
+      "loss": 0.3471,
+      "step": 5565
+    },
+    {
+      "epoch": 1.3248646397334443,
+      "grad_norm": 0.36315758859429115,
+      "learning_rate": 2.7040662593774114e-06,
+      "loss": 0.384,
+      "step": 5566
+    },
+    {
+      "epoch": 1.3251026358065092,
+      "grad_norm": 0.3570108097887906,
+      "learning_rate": 2.70235407166947e-06,
+      "loss": 0.3291,
+      "step": 5567
+    },
+    {
+      "epoch": 1.325340631879574,
+      "grad_norm": 0.38719520188943835,
+      "learning_rate": 2.700642225448554e-06,
+      "loss": 0.3024,
+      "step": 5568
+    },
+    {
+      "epoch": 1.3255786279526387,
+      "grad_norm": 0.3896705832913739,
+      "learning_rate": 2.698930720969087e-06,
+      "loss": 0.3169,
+      "step": 5569
+    },
+    {
+      "epoch": 1.3258166240257037,
+      "grad_norm": 0.3888507607782142,
+      "learning_rate": 2.697219558485439e-06,
+      "loss": 0.3841,
+      "step": 5570
+    },
+    {
+      "epoch": 1.3260546200987684,
+      "grad_norm": 0.3670049338221428,
+      "learning_rate": 2.695508738251934e-06,
+      "loss": 0.2712,
+      "step": 5571
+    },
+    {
+      "epoch": 1.3262926161718331,
+      "grad_norm": 0.39842737745653384,
+      "learning_rate": 2.693798260522841e-06,
+      "loss": 0.2985,
+      "step": 5572
+    },
+    {
+      "epoch": 1.3265306122448979,
+      "grad_norm": 0.37727527288531104,
+      "learning_rate": 2.692088125552379e-06,
+      "loss": 0.3754,
+      "step": 5573
+    },
+    {
+      "epoch": 1.3267686083179626,
+      "grad_norm": 0.43871454005089827,
+      "learning_rate": 2.690378333594717e-06,
+      "loss": 0.3219,
+      "step": 5574
+    },
+    {
+      "epoch": 1.3270066043910276,
+      "grad_norm": 0.39318467207824387,
+      "learning_rate": 2.6886688849039717e-06,
+      "loss": 0.2756,
+      "step": 5575
+    },
+    {
+      "epoch": 1.3272446004640923,
+      "grad_norm": 0.3994623481607067,
+      "learning_rate": 2.6869597797342096e-06,
+      "loss": 0.3249,
+      "step": 5576
+    },
+    {
+      "epoch": 1.327482596537157,
+      "grad_norm": 0.44155749097845365,
+      "learning_rate": 2.685251018339446e-06,
+      "loss": 0.3763,
+      "step": 5577
+    },
+    {
+      "epoch": 1.327720592610222,
+      "grad_norm": 0.3665656601981612,
+      "learning_rate": 2.6835426009736455e-06,
+      "loss": 0.3239,
+      "step": 5578
+    },
+    {
+      "epoch": 1.3279585886832868,
+      "grad_norm": 0.4022642181119508,
+      "learning_rate": 2.681834527890721e-06,
+      "loss": 0.2761,
+      "step": 5579
+    },
+    {
+      "epoch": 1.3281965847563515,
+      "grad_norm": 0.3788048689911649,
+      "learning_rate": 2.6801267993445335e-06,
+      "loss": 0.3626,
+      "step": 5580
+    },
+    {
+      "epoch": 1.3284345808294162,
+      "grad_norm": 0.399964703157908,
+      "learning_rate": 2.6784194155888953e-06,
+      "loss": 0.364,
+      "step": 5581
+    },
+    {
+      "epoch": 1.328672576902481,
+      "grad_norm": 0.46248131169930756,
+      "learning_rate": 2.6767123768775627e-06,
+      "loss": 0.3001,
+      "step": 5582
+    },
+    {
+      "epoch": 1.328910572975546,
+      "grad_norm": 0.3790430115317474,
+      "learning_rate": 2.675005683464248e-06,
+      "loss": 0.3368,
+      "step": 5583
+    },
+    {
+      "epoch": 1.3291485690486107,
+      "grad_norm": 0.42296015749472615,
+      "learning_rate": 2.6732993356026073e-06,
+      "loss": 0.3964,
+      "step": 5584
+    },
+    {
+      "epoch": 1.3293865651216754,
+      "grad_norm": 0.37767108037602987,
+      "learning_rate": 2.671593333546244e-06,
+      "loss": 0.3186,
+      "step": 5585
+    },
+    {
+      "epoch": 1.3296245611947404,
+      "grad_norm": 0.3556314352660897,
+      "learning_rate": 2.669887677548712e-06,
+      "loss": 0.2928,
+      "step": 5586
+    },
+    {
+      "epoch": 1.3298625572678051,
+      "grad_norm": 0.3646511968328883,
+      "learning_rate": 2.6681823678635177e-06,
+      "loss": 0.3345,
+      "step": 5587
+    },
+    {
+      "epoch": 1.3301005533408699,
+      "grad_norm": 0.38411250506226,
+      "learning_rate": 2.6664774047441087e-06,
+      "loss": 0.3821,
+      "step": 5588
+    },
+    {
+      "epoch": 1.3303385494139346,
+      "grad_norm": 0.3780727074870984,
+      "learning_rate": 2.6647727884438866e-06,
+      "loss": 0.287,
+      "step": 5589
+    },
+    {
+      "epoch": 1.3305765454869993,
+      "grad_norm": 0.3595186487413398,
+      "learning_rate": 2.6630685192161995e-06,
+      "loss": 0.3187,
+      "step": 5590
+    },
+    {
+      "epoch": 1.3308145415600643,
+      "grad_norm": 0.352373082985822,
+      "learning_rate": 2.6613645973143427e-06,
+      "loss": 0.3637,
+      "step": 5591
+    },
+    {
+      "epoch": 1.331052537633129,
+      "grad_norm": 0.3611789272199898,
+      "learning_rate": 2.659661022991562e-06,
+      "loss": 0.3405,
+      "step": 5592
+    },
+    {
+      "epoch": 1.3312905337061938,
+      "grad_norm": 0.4065306276115284,
+      "learning_rate": 2.65795779650105e-06,
+      "loss": 0.293,
+      "step": 5593
+    },
+    {
+      "epoch": 1.3315285297792587,
+      "grad_norm": 0.3987698770035065,
+      "learning_rate": 2.656254918095949e-06,
+      "loss": 0.3428,
+      "step": 5594
+    },
+    {
+      "epoch": 1.3317665258523235,
+      "grad_norm": 0.3943243833309216,
+      "learning_rate": 2.654552388029349e-06,
+      "loss": 0.3822,
+      "step": 5595
+    },
+    {
+      "epoch": 1.3320045219253882,
+      "grad_norm": 0.36509611859681235,
+      "learning_rate": 2.652850206554287e-06,
+      "loss": 0.2925,
+      "step": 5596
+    },
+    {
+      "epoch": 1.332242517998453,
+      "grad_norm": 0.38035780331790836,
+      "learning_rate": 2.6511483739237508e-06,
+      "loss": 0.3156,
+      "step": 5597
+    },
+    {
+      "epoch": 1.3324805140715177,
+      "grad_norm": 0.37332659121072254,
+      "learning_rate": 2.649446890390671e-06,
+      "loss": 0.3465,
+      "step": 5598
+    },
+    {
+      "epoch": 1.3327185101445826,
+      "grad_norm": 0.3665673689467114,
+      "learning_rate": 2.647745756207937e-06,
+      "loss": 0.3658,
+      "step": 5599
+    },
+    {
+      "epoch": 1.3329565062176474,
+      "grad_norm": 0.3820284830332031,
+      "learning_rate": 2.646044971628374e-06,
+      "loss": 0.2949,
+      "step": 5600
+    },
+    {
+      "epoch": 1.3331945022907121,
+      "grad_norm": 0.37188294924637955,
+      "learning_rate": 2.6443445369047625e-06,
+      "loss": 0.3224,
+      "step": 5601
+    },
+    {
+      "epoch": 1.333432498363777,
+      "grad_norm": 0.3735185472412557,
+      "learning_rate": 2.6426444522898286e-06,
+      "loss": 0.3703,
+      "step": 5602
+    },
+    {
+      "epoch": 1.3336704944368418,
+      "grad_norm": 0.35518226222360766,
+      "learning_rate": 2.640944718036248e-06,
+      "loss": 0.3015,
+      "step": 5603
+    },
+    {
+      "epoch": 1.3339084905099066,
+      "grad_norm": 0.389450698522127,
+      "learning_rate": 2.6392453343966422e-06,
+      "loss": 0.2587,
+      "step": 5604
+    },
+    {
+      "epoch": 1.3341464865829713,
+      "grad_norm": 0.3971444862417287,
+      "learning_rate": 2.6375463016235826e-06,
+      "loss": 0.3954,
+      "step": 5605
+    },
+    {
+      "epoch": 1.334384482656036,
+      "grad_norm": 0.3502692387454883,
+      "learning_rate": 2.6358476199695858e-06,
+      "loss": 0.3411,
+      "step": 5606
+    },
+    {
+      "epoch": 1.334622478729101,
+      "grad_norm": 0.36914607170334995,
+      "learning_rate": 2.6341492896871198e-06,
+      "loss": 0.2936,
+      "step": 5607
+    },
+    {
+      "epoch": 1.3348604748021657,
+      "grad_norm": 0.4088428743135435,
+      "learning_rate": 2.632451311028598e-06,
+      "loss": 0.3102,
+      "step": 5608
+    },
+    {
+      "epoch": 1.3350984708752305,
+      "grad_norm": 0.37074985008059036,
+      "learning_rate": 2.630753684246378e-06,
+      "loss": 0.3687,
+      "step": 5609
+    },
+    {
+      "epoch": 1.3353364669482954,
+      "grad_norm": 0.35775847771800834,
+      "learning_rate": 2.6290564095927762e-06,
+      "loss": 0.3196,
+      "step": 5610
+    },
+    {
+      "epoch": 1.3355744630213602,
+      "grad_norm": 0.3834131260856632,
+      "learning_rate": 2.627359487320046e-06,
+      "loss": 0.2913,
+      "step": 5611
+    },
+    {
+      "epoch": 1.335812459094425,
+      "grad_norm": 0.37855664847349857,
+      "learning_rate": 2.6256629176803925e-06,
+      "loss": 0.3398,
+      "step": 5612
+    },
+    {
+      "epoch": 1.3360504551674897,
+      "grad_norm": 0.3915920080407146,
+      "learning_rate": 2.623966700925965e-06,
+      "loss": 0.3756,
+      "step": 5613
+    },
+    {
+      "epoch": 1.3362884512405544,
+      "grad_norm": 0.3936719541417686,
+      "learning_rate": 2.622270837308869e-06,
+      "loss": 0.2686,
+      "step": 5614
+    },
+    {
+      "epoch": 1.3365264473136194,
+      "grad_norm": 0.37541566356978545,
+      "learning_rate": 2.620575327081148e-06,
+      "loss": 0.3409,
+      "step": 5615
+    },
+    {
+      "epoch": 1.336764443386684,
+      "grad_norm": 0.39242038775201626,
+      "learning_rate": 2.6188801704947976e-06,
+      "loss": 0.4255,
+      "step": 5616
+    },
+    {
+      "epoch": 1.3370024394597488,
+      "grad_norm": 0.3738300495004616,
+      "learning_rate": 2.61718536780176e-06,
+      "loss": 0.3243,
+      "step": 5617
+    },
+    {
+      "epoch": 1.3372404355328138,
+      "grad_norm": 0.37526710784159806,
+      "learning_rate": 2.6154909192539248e-06,
+      "loss": 0.2954,
+      "step": 5618
+    },
+    {
+      "epoch": 1.3374784316058785,
+      "grad_norm": 0.3831635215027392,
+      "learning_rate": 2.613796825103129e-06,
+      "loss": 0.3692,
+      "step": 5619
+    },
+    {
+      "epoch": 1.3377164276789433,
+      "grad_norm": 0.3928912025057844,
+      "learning_rate": 2.6121030856011562e-06,
+      "loss": 0.4076,
+      "step": 5620
+    },
+    {
+      "epoch": 1.337954423752008,
+      "grad_norm": 0.3747902265725176,
+      "learning_rate": 2.61040970099974e-06,
+      "loss": 0.3122,
+      "step": 5621
+    },
+    {
+      "epoch": 1.3381924198250728,
+      "grad_norm": 0.3852536780569142,
+      "learning_rate": 2.6087166715505563e-06,
+      "loss": 0.3281,
+      "step": 5622
+    },
+    {
+      "epoch": 1.3384304158981377,
+      "grad_norm": 0.4203309500020824,
+      "learning_rate": 2.6070239975052334e-06,
+      "loss": 0.3748,
+      "step": 5623
+    },
+    {
+      "epoch": 1.3386684119712025,
+      "grad_norm": 0.37552479294518815,
+      "learning_rate": 2.605331679115344e-06,
+      "loss": 0.3544,
+      "step": 5624
+    },
+    {
+      "epoch": 1.3389064080442672,
+      "grad_norm": 0.3796552948357275,
+      "learning_rate": 2.6036397166324062e-06,
+      "loss": 0.257,
+      "step": 5625
+    },
+    {
+      "epoch": 1.3391444041173322,
+      "grad_norm": 0.36585880934173215,
+      "learning_rate": 2.6019481103078912e-06,
+      "loss": 0.3224,
+      "step": 5626
+    },
+    {
+      "epoch": 1.339382400190397,
+      "grad_norm": 0.3784318209352287,
+      "learning_rate": 2.6002568603932127e-06,
+      "loss": 0.4231,
+      "step": 5627
+    },
+    {
+      "epoch": 1.3396203962634616,
+      "grad_norm": 0.3581778615226102,
+      "learning_rate": 2.598565967139731e-06,
+      "loss": 0.269,
+      "step": 5628
+    },
+    {
+      "epoch": 1.3398583923365264,
+      "grad_norm": 0.3613572419065405,
+      "learning_rate": 2.5968754307987556e-06,
+      "loss": 0.3059,
+      "step": 5629
+    },
+    {
+      "epoch": 1.340096388409591,
+      "grad_norm": 0.41685880452742957,
+      "learning_rate": 2.5951852516215415e-06,
+      "loss": 0.3547,
+      "step": 5630
+    },
+    {
+      "epoch": 1.340334384482656,
+      "grad_norm": 0.3476538727617446,
+      "learning_rate": 2.593495429859291e-06,
+      "loss": 0.3522,
+      "step": 5631
+    },
+    {
+      "epoch": 1.3405723805557208,
+      "grad_norm": 0.38372719904096886,
+      "learning_rate": 2.5918059657631532e-06,
+      "loss": 0.2839,
+      "step": 5632
+    },
+    {
+      "epoch": 1.3408103766287855,
+      "grad_norm": 0.4168082237276136,
+      "learning_rate": 2.5901168595842256e-06,
+      "loss": 0.3166,
+      "step": 5633
+    },
+    {
+      "epoch": 1.3410483727018505,
+      "grad_norm": 0.3990911280449388,
+      "learning_rate": 2.58842811157355e-06,
+      "loss": 0.3948,
+      "step": 5634
+    },
+    {
+      "epoch": 1.3412863687749152,
+      "grad_norm": 0.3619981264622829,
+      "learning_rate": 2.5867397219821166e-06,
+      "loss": 0.3115,
+      "step": 5635
+    },
+    {
+      "epoch": 1.34152436484798,
+      "grad_norm": 0.37183438540217123,
+      "learning_rate": 2.5850516910608596e-06,
+      "loss": 0.299,
+      "step": 5636
+    },
+    {
+      "epoch": 1.3417623609210447,
+      "grad_norm": 0.3821320851640882,
+      "learning_rate": 2.5833640190606663e-06,
+      "loss": 0.3677,
+      "step": 5637
+    },
+    {
+      "epoch": 1.3420003569941095,
+      "grad_norm": 0.4340280479794182,
+      "learning_rate": 2.5816767062323646e-06,
+      "loss": 0.3653,
+      "step": 5638
+    },
+    {
+      "epoch": 1.3422383530671744,
+      "grad_norm": 0.3935957665861045,
+      "learning_rate": 2.5799897528267304e-06,
+      "loss": 0.3125,
+      "step": 5639
+    },
+    {
+      "epoch": 1.3424763491402392,
+      "grad_norm": 0.3551306900169191,
+      "learning_rate": 2.5783031590944853e-06,
+      "loss": 0.3321,
+      "step": 5640
+    },
+    {
+      "epoch": 1.342714345213304,
+      "grad_norm": 0.39027530241074593,
+      "learning_rate": 2.5766169252863026e-06,
+      "loss": 0.3719,
+      "step": 5641
+    },
+    {
+      "epoch": 1.3429523412863689,
+      "grad_norm": 0.3669775255066402,
+      "learning_rate": 2.574931051652796e-06,
+      "loss": 0.3106,
+      "step": 5642
+    },
+    {
+      "epoch": 1.3431903373594336,
+      "grad_norm": 0.36467746278063334,
+      "learning_rate": 2.573245538444529e-06,
+      "loss": 0.2812,
+      "step": 5643
+    },
+    {
+      "epoch": 1.3434283334324983,
+      "grad_norm": 0.44665190837644525,
+      "learning_rate": 2.5715603859120095e-06,
+      "loss": 0.3282,
+      "step": 5644
+    },
+    {
+      "epoch": 1.343666329505563,
+      "grad_norm": 0.36150659846601496,
+      "learning_rate": 2.569875594305694e-06,
+      "loss": 0.3898,
+      "step": 5645
+    },
+    {
+      "epoch": 1.3439043255786278,
+      "grad_norm": 0.36749221425362993,
+      "learning_rate": 2.5681911638759837e-06,
+      "loss": 0.2971,
+      "step": 5646
+    },
+    {
+      "epoch": 1.3441423216516928,
+      "grad_norm": 0.37962067008941414,
+      "learning_rate": 2.5665070948732258e-06,
+      "loss": 0.2935,
+      "step": 5647
+    },
+    {
+      "epoch": 1.3443803177247575,
+      "grad_norm": 0.4350753894556899,
+      "learning_rate": 2.564823387547716e-06,
+      "loss": 0.3688,
+      "step": 5648
+    },
+    {
+      "epoch": 1.3446183137978223,
+      "grad_norm": 0.36978176336633634,
+      "learning_rate": 2.5631400421496934e-06,
+      "loss": 0.3408,
+      "step": 5649
+    },
+    {
+      "epoch": 1.3448563098708872,
+      "grad_norm": 0.3672540601703578,
+      "learning_rate": 2.5614570589293457e-06,
+      "loss": 0.3011,
+      "step": 5650
+    },
+    {
+      "epoch": 1.345094305943952,
+      "grad_norm": 0.3767220367127715,
+      "learning_rate": 2.5597744381368063e-06,
+      "loss": 0.3395,
+      "step": 5651
+    },
+    {
+      "epoch": 1.3453323020170167,
+      "grad_norm": 0.38145663607283764,
+      "learning_rate": 2.558092180022153e-06,
+      "loss": 0.3896,
+      "step": 5652
+    },
+    {
+      "epoch": 1.3455702980900814,
+      "grad_norm": 0.36950737323434774,
+      "learning_rate": 2.5564102848354098e-06,
+      "loss": 0.3138,
+      "step": 5653
+    },
+    {
+      "epoch": 1.3458082941631462,
+      "grad_norm": 0.3764771683321259,
+      "learning_rate": 2.5547287528265517e-06,
+      "loss": 0.282,
+      "step": 5654
+    },
+    {
+      "epoch": 1.3460462902362111,
+      "grad_norm": 0.4283336061272312,
+      "learning_rate": 2.553047584245495e-06,
+      "loss": 0.372,
+      "step": 5655
+    },
+    {
+      "epoch": 1.3462842863092759,
+      "grad_norm": 0.3676206486750502,
+      "learning_rate": 2.5513667793421002e-06,
+      "loss": 0.3684,
+      "step": 5656
+    },
+    {
+      "epoch": 1.3465222823823406,
+      "grad_norm": 0.40021565693988703,
+      "learning_rate": 2.5496863383661797e-06,
+      "loss": 0.2643,
+      "step": 5657
+    },
+    {
+      "epoch": 1.3467602784554056,
+      "grad_norm": 0.3777923710780302,
+      "learning_rate": 2.548006261567487e-06,
+      "loss": 0.3046,
+      "step": 5658
+    },
+    {
+      "epoch": 1.3469982745284703,
+      "grad_norm": 0.38725995517731304,
+      "learning_rate": 2.5463265491957224e-06,
+      "loss": 0.3831,
+      "step": 5659
+    },
+    {
+      "epoch": 1.347236270601535,
+      "grad_norm": 0.3576244257854606,
+      "learning_rate": 2.544647201500534e-06,
+      "loss": 0.3197,
+      "step": 5660
+    },
+    {
+      "epoch": 1.3474742666745998,
+      "grad_norm": 0.36464513110800595,
+      "learning_rate": 2.542968218731514e-06,
+      "loss": 0.2718,
+      "step": 5661
+    },
+    {
+      "epoch": 1.3477122627476645,
+      "grad_norm": 0.3904440981846269,
+      "learning_rate": 2.541289601138201e-06,
+      "loss": 0.3488,
+      "step": 5662
+    },
+    {
+      "epoch": 1.3479502588207295,
+      "grad_norm": 0.3894407299674647,
+      "learning_rate": 2.539611348970077e-06,
+      "loss": 0.3785,
+      "step": 5663
+    },
+    {
+      "epoch": 1.3481882548937942,
+      "grad_norm": 0.36018047764775457,
+      "learning_rate": 2.537933462476575e-06,
+      "loss": 0.294,
+      "step": 5664
+    },
+    {
+      "epoch": 1.348426250966859,
+      "grad_norm": 0.45032023637446106,
+      "learning_rate": 2.5362559419070693e-06,
+      "loss": 0.3058,
+      "step": 5665
+    },
+    {
+      "epoch": 1.348664247039924,
+      "grad_norm": 0.3520211254562473,
+      "learning_rate": 2.534578787510881e-06,
+      "loss": 0.3629,
+      "step": 5666
+    },
+    {
+      "epoch": 1.3489022431129887,
+      "grad_norm": 0.35778323923398264,
+      "learning_rate": 2.532901999537274e-06,
+      "loss": 0.3121,
+      "step": 5667
+    },
+    {
+      "epoch": 1.3491402391860534,
+      "grad_norm": 0.40768614895424304,
+      "learning_rate": 2.531225578235465e-06,
+      "loss": 0.2942,
+      "step": 5668
+    },
+    {
+      "epoch": 1.3493782352591182,
+      "grad_norm": 0.40587873605164554,
+      "learning_rate": 2.52954952385461e-06,
+      "loss": 0.3128,
+      "step": 5669
+    },
+    {
+      "epoch": 1.349616231332183,
+      "grad_norm": 0.3665695548104563,
+      "learning_rate": 2.527873836643811e-06,
+      "loss": 0.3682,
+      "step": 5670
+    },
+    {
+      "epoch": 1.3498542274052479,
+      "grad_norm": 0.3757527838138373,
+      "learning_rate": 2.5261985168521174e-06,
+      "loss": 0.2842,
+      "step": 5671
+    },
+    {
+      "epoch": 1.3500922234783126,
+      "grad_norm": 0.38913613595462837,
+      "learning_rate": 2.5245235647285238e-06,
+      "loss": 0.3045,
+      "step": 5672
+    },
+    {
+      "epoch": 1.3503302195513773,
+      "grad_norm": 0.39284781780088013,
+      "learning_rate": 2.5228489805219684e-06,
+      "loss": 0.3733,
+      "step": 5673
+    },
+    {
+      "epoch": 1.3505682156244423,
+      "grad_norm": 0.39264836647519236,
+      "learning_rate": 2.5211747644813367e-06,
+      "loss": 0.323,
+      "step": 5674
+    },
+    {
+      "epoch": 1.350806211697507,
+      "grad_norm": 0.38369466339855723,
+      "learning_rate": 2.5195009168554572e-06,
+      "loss": 0.2919,
+      "step": 5675
+    },
+    {
+      "epoch": 1.3510442077705718,
+      "grad_norm": 0.37067590315834653,
+      "learning_rate": 2.517827437893107e-06,
+      "loss": 0.2875,
+      "step": 5676
+    },
+    {
+      "epoch": 1.3512822038436365,
+      "grad_norm": 0.38302813218504755,
+      "learning_rate": 2.5161543278430055e-06,
+      "loss": 0.3684,
+      "step": 5677
+    },
+    {
+      "epoch": 1.3515201999167012,
+      "grad_norm": 0.3605795300765205,
+      "learning_rate": 2.5144815869538177e-06,
+      "loss": 0.2855,
+      "step": 5678
+    },
+    {
+      "epoch": 1.3517581959897662,
+      "grad_norm": 0.36372350908656326,
+      "learning_rate": 2.512809215474155e-06,
+      "loss": 0.2996,
+      "step": 5679
+    },
+    {
+      "epoch": 1.351996192062831,
+      "grad_norm": 0.36974222068701523,
+      "learning_rate": 2.5111372136525713e-06,
+      "loss": 0.3311,
+      "step": 5680
+    },
+    {
+      "epoch": 1.3522341881358957,
+      "grad_norm": 0.36626523624003543,
+      "learning_rate": 2.509465581737571e-06,
+      "loss": 0.3287,
+      "step": 5681
+    },
+    {
+      "epoch": 1.3524721842089606,
+      "grad_norm": 0.3940460629565776,
+      "learning_rate": 2.5077943199775978e-06,
+      "loss": 0.29,
+      "step": 5682
+    },
+    {
+      "epoch": 1.3527101802820254,
+      "grad_norm": 0.38718589993001984,
+      "learning_rate": 2.5061234286210436e-06,
+      "loss": 0.2943,
+      "step": 5683
+    },
+    {
+      "epoch": 1.3529481763550901,
+      "grad_norm": 0.4250135569344329,
+      "learning_rate": 2.5044529079162426e-06,
+      "loss": 0.3941,
+      "step": 5684
+    },
+    {
+      "epoch": 1.3531861724281549,
+      "grad_norm": 0.395297953846377,
+      "learning_rate": 2.502782758111477e-06,
+      "loss": 0.2984,
+      "step": 5685
+    },
+    {
+      "epoch": 1.3534241685012196,
+      "grad_norm": 0.3758290085462567,
+      "learning_rate": 2.5011129794549717e-06,
+      "loss": 0.258,
+      "step": 5686
+    },
+    {
+      "epoch": 1.3536621645742846,
+      "grad_norm": 0.3900626080187409,
+      "learning_rate": 2.4994435721948966e-06,
+      "loss": 0.3284,
+      "step": 5687
+    },
+    {
+      "epoch": 1.3539001606473493,
+      "grad_norm": 0.38881872385050364,
+      "learning_rate": 2.4977745365793676e-06,
+      "loss": 0.3925,
+      "step": 5688
+    },
+    {
+      "epoch": 1.354138156720414,
+      "grad_norm": 0.37508295198986963,
+      "learning_rate": 2.4961058728564446e-06,
+      "loss": 0.2848,
+      "step": 5689
+    },
+    {
+      "epoch": 1.354376152793479,
+      "grad_norm": 0.4969361288451411,
+      "learning_rate": 2.4944375812741304e-06,
+      "loss": 0.3224,
+      "step": 5690
+    },
+    {
+      "epoch": 1.3546141488665437,
+      "grad_norm": 0.4191683283997192,
+      "learning_rate": 2.4927696620803783e-06,
+      "loss": 0.3844,
+      "step": 5691
+    },
+    {
+      "epoch": 1.3548521449396085,
+      "grad_norm": 0.3632365871635604,
+      "learning_rate": 2.4911021155230804e-06,
+      "loss": 0.3277,
+      "step": 5692
+    },
+    {
+      "epoch": 1.3550901410126732,
+      "grad_norm": 0.36797859774918557,
+      "learning_rate": 2.4894349418500745e-06,
+      "loss": 0.2917,
+      "step": 5693
+    },
+    {
+      "epoch": 1.355328137085738,
+      "grad_norm": 0.36534359001275923,
+      "learning_rate": 2.487768141309144e-06,
+      "loss": 0.3252,
+      "step": 5694
+    },
+    {
+      "epoch": 1.355566133158803,
+      "grad_norm": 0.41390949501152025,
+      "learning_rate": 2.486101714148018e-06,
+      "loss": 0.3776,
+      "step": 5695
+    },
+    {
+      "epoch": 1.3558041292318677,
+      "grad_norm": 0.3596335904706074,
+      "learning_rate": 2.4844356606143687e-06,
+      "loss": 0.3014,
+      "step": 5696
+    },
+    {
+      "epoch": 1.3560421253049324,
+      "grad_norm": 0.3738157482380636,
+      "learning_rate": 2.482769980955812e-06,
+      "loss": 0.3108,
+      "step": 5697
+    },
+    {
+      "epoch": 1.3562801213779974,
+      "grad_norm": 0.3712839660896256,
+      "learning_rate": 2.4811046754199098e-06,
+      "loss": 0.3732,
+      "step": 5698
+    },
+    {
+      "epoch": 1.356518117451062,
+      "grad_norm": 0.3879484832908067,
+      "learning_rate": 2.479439744254167e-06,
+      "loss": 0.3083,
+      "step": 5699
+    },
+    {
+      "epoch": 1.3567561135241268,
+      "grad_norm": 0.4129860781327451,
+      "learning_rate": 2.4777751877060343e-06,
+      "loss": 0.2727,
+      "step": 5700
+    },
+    {
+      "epoch": 1.3569941095971916,
+      "grad_norm": 0.3695216250582914,
+      "learning_rate": 2.476111006022905e-06,
+      "loss": 0.3,
+      "step": 5701
+    },
+    {
+      "epoch": 1.3572321056702563,
+      "grad_norm": 0.38079989838460043,
+      "learning_rate": 2.4744471994521184e-06,
+      "loss": 0.387,
+      "step": 5702
+    },
+    {
+      "epoch": 1.3574701017433213,
+      "grad_norm": 0.3851371719901647,
+      "learning_rate": 2.4727837682409574e-06,
+      "loss": 0.2918,
+      "step": 5703
+    },
+    {
+      "epoch": 1.357708097816386,
+      "grad_norm": 0.4030470108804071,
+      "learning_rate": 2.4711207126366483e-06,
+      "loss": 0.2774,
+      "step": 5704
+    },
+    {
+      "epoch": 1.3579460938894508,
+      "grad_norm": 0.4043361359872621,
+      "learning_rate": 2.4694580328863633e-06,
+      "loss": 0.3715,
+      "step": 5705
+    },
+    {
+      "epoch": 1.3581840899625157,
+      "grad_norm": 0.36653113787008357,
+      "learning_rate": 2.4677957292372166e-06,
+      "loss": 0.3627,
+      "step": 5706
+    },
+    {
+      "epoch": 1.3584220860355805,
+      "grad_norm": 0.3850700049312921,
+      "learning_rate": 2.4661338019362684e-06,
+      "loss": 0.2887,
+      "step": 5707
+    },
+    {
+      "epoch": 1.3586600821086452,
+      "grad_norm": 0.38782380526495214,
+      "learning_rate": 2.46447225123052e-06,
+      "loss": 0.35,
+      "step": 5708
+    },
+    {
+      "epoch": 1.35889807818171,
+      "grad_norm": 0.3875168580573579,
+      "learning_rate": 2.4628110773669235e-06,
+      "loss": 0.3909,
+      "step": 5709
+    },
+    {
+      "epoch": 1.3591360742547747,
+      "grad_norm": 0.3706961560048635,
+      "learning_rate": 2.4611502805923677e-06,
+      "loss": 0.3001,
+      "step": 5710
+    },
+    {
+      "epoch": 1.3593740703278396,
+      "grad_norm": 0.41022795283685426,
+      "learning_rate": 2.459489861153688e-06,
+      "loss": 0.2858,
+      "step": 5711
+    },
+    {
+      "epoch": 1.3596120664009044,
+      "grad_norm": 0.3507488121743938,
+      "learning_rate": 2.4578298192976646e-06,
+      "loss": 0.3481,
+      "step": 5712
+    },
+    {
+      "epoch": 1.3598500624739691,
+      "grad_norm": 0.3882156595432714,
+      "learning_rate": 2.4561701552710198e-06,
+      "loss": 0.3727,
+      "step": 5713
+    },
+    {
+      "epoch": 1.360088058547034,
+      "grad_norm": 0.41147742183986585,
+      "learning_rate": 2.454510869320422e-06,
+      "loss": 0.2848,
+      "step": 5714
+    },
+    {
+      "epoch": 1.3603260546200988,
+      "grad_norm": 0.3932347054914899,
+      "learning_rate": 2.4528519616924807e-06,
+      "loss": 0.319,
+      "step": 5715
+    },
+    {
+      "epoch": 1.3605640506931636,
+      "grad_norm": 0.3962467933758855,
+      "learning_rate": 2.4511934326337516e-06,
+      "loss": 0.3792,
+      "step": 5716
+    },
+    {
+      "epoch": 1.3608020467662283,
+      "grad_norm": 0.38954161787628644,
+      "learning_rate": 2.449535282390731e-06,
+      "loss": 0.3375,
+      "step": 5717
+    },
+    {
+      "epoch": 1.361040042839293,
+      "grad_norm": 0.3926513205788529,
+      "learning_rate": 2.4478775112098644e-06,
+      "loss": 0.271,
+      "step": 5718
+    },
+    {
+      "epoch": 1.361278038912358,
+      "grad_norm": 0.4009062810034733,
+      "learning_rate": 2.446220119337536e-06,
+      "loss": 0.3599,
+      "step": 5719
+    },
+    {
+      "epoch": 1.3615160349854227,
+      "grad_norm": 0.367410786702826,
+      "learning_rate": 2.444563107020076e-06,
+      "loss": 0.3562,
+      "step": 5720
+    },
+    {
+      "epoch": 1.3617540310584875,
+      "grad_norm": 0.37374300698130064,
+      "learning_rate": 2.4429064745037562e-06,
+      "loss": 0.2799,
+      "step": 5721
+    },
+    {
+      "epoch": 1.3619920271315524,
+      "grad_norm": 0.39466179262858975,
+      "learning_rate": 2.441250222034792e-06,
+      "loss": 0.3144,
+      "step": 5722
+    },
+    {
+      "epoch": 1.3622300232046172,
+      "grad_norm": 0.4045572620870956,
+      "learning_rate": 2.4395943498593476e-06,
+      "loss": 0.3638,
+      "step": 5723
+    },
+    {
+      "epoch": 1.362468019277682,
+      "grad_norm": 0.35753443357854653,
+      "learning_rate": 2.4379388582235236e-06,
+      "loss": 0.3324,
+      "step": 5724
+    },
+    {
+      "epoch": 1.3627060153507466,
+      "grad_norm": 0.3827497837460359,
+      "learning_rate": 2.436283747373368e-06,
+      "loss": 0.2728,
+      "step": 5725
+    },
+    {
+      "epoch": 1.3629440114238114,
+      "grad_norm": 0.369331720445923,
+      "learning_rate": 2.43462901755487e-06,
+      "loss": 0.3226,
+      "step": 5726
+    },
+    {
+      "epoch": 1.3631820074968763,
+      "grad_norm": 0.3868589649921183,
+      "learning_rate": 2.4329746690139656e-06,
+      "loss": 0.4188,
+      "step": 5727
+    },
+    {
+      "epoch": 1.363420003569941,
+      "grad_norm": 0.5212121670082667,
+      "learning_rate": 2.4313207019965295e-06,
+      "loss": 0.3194,
+      "step": 5728
+    },
+    {
+      "epoch": 1.3636579996430058,
+      "grad_norm": 0.380202951444516,
+      "learning_rate": 2.429667116748383e-06,
+      "loss": 0.2932,
+      "step": 5729
+    },
+    {
+      "epoch": 1.3638959957160708,
+      "grad_norm": 0.3965822543317094,
+      "learning_rate": 2.4280139135152906e-06,
+      "loss": 0.3701,
+      "step": 5730
+    },
+    {
+      "epoch": 1.3641339917891355,
+      "grad_norm": 0.37353099551658364,
+      "learning_rate": 2.4263610925429588e-06,
+      "loss": 0.3374,
+      "step": 5731
+    },
+    {
+      "epoch": 1.3643719878622003,
+      "grad_norm": 0.3661423191951257,
+      "learning_rate": 2.4247086540770365e-06,
+      "loss": 0.2938,
+      "step": 5732
+    },
+    {
+      "epoch": 1.364609983935265,
+      "grad_norm": 0.3647519555796,
+      "learning_rate": 2.4230565983631184e-06,
+      "loss": 0.3204,
+      "step": 5733
+    },
+    {
+      "epoch": 1.3648479800083297,
+      "grad_norm": 0.43532671681837853,
+      "learning_rate": 2.42140492564674e-06,
+      "loss": 0.3777,
+      "step": 5734
+    },
+    {
+      "epoch": 1.3650859760813947,
+      "grad_norm": 0.38492642963237705,
+      "learning_rate": 2.4197536361733792e-06,
+      "loss": 0.3075,
+      "step": 5735
+    },
+    {
+      "epoch": 1.3653239721544594,
+      "grad_norm": 0.3814772631917648,
+      "learning_rate": 2.418102730188462e-06,
+      "loss": 0.2729,
+      "step": 5736
+    },
+    {
+      "epoch": 1.3655619682275242,
+      "grad_norm": 0.39209652486984,
+      "learning_rate": 2.4164522079373525e-06,
+      "loss": 0.344,
+      "step": 5737
+    },
+    {
+      "epoch": 1.3657999643005891,
+      "grad_norm": 0.3664720899560753,
+      "learning_rate": 2.4148020696653583e-06,
+      "loss": 0.3635,
+      "step": 5738
+    },
+    {
+      "epoch": 1.3660379603736539,
+      "grad_norm": 0.3534142149903722,
+      "learning_rate": 2.413152315617732e-06,
+      "loss": 0.2905,
+      "step": 5739
+    },
+    {
+      "epoch": 1.3662759564467186,
+      "grad_norm": 0.3734213259184978,
+      "learning_rate": 2.4115029460396666e-06,
+      "loss": 0.2853,
+      "step": 5740
+    },
+    {
+      "epoch": 1.3665139525197834,
+      "grad_norm": 0.45252181922728846,
+      "learning_rate": 2.4098539611763e-06,
+      "loss": 0.3878,
+      "step": 5741
+    },
+    {
+      "epoch": 1.366751948592848,
+      "grad_norm": 0.4147075627953894,
+      "learning_rate": 2.408205361272712e-06,
+      "loss": 0.3316,
+      "step": 5742
+    },
+    {
+      "epoch": 1.366989944665913,
+      "grad_norm": 0.355835782980996,
+      "learning_rate": 2.4065571465739247e-06,
+      "loss": 0.2902,
+      "step": 5743
+    },
+    {
+      "epoch": 1.3672279407389778,
+      "grad_norm": 0.40862200982116076,
+      "learning_rate": 2.4049093173249026e-06,
+      "loss": 0.3258,
+      "step": 5744
+    },
+    {
+      "epoch": 1.3674659368120425,
+      "grad_norm": 0.3792126875199552,
+      "learning_rate": 2.403261873770557e-06,
+      "loss": 0.3966,
+      "step": 5745
+    },
+    {
+      "epoch": 1.3677039328851075,
+      "grad_norm": 0.39370562835081924,
+      "learning_rate": 2.401614816155737e-06,
+      "loss": 0.2929,
+      "step": 5746
+    },
+    {
+      "epoch": 1.3679419289581722,
+      "grad_norm": 0.3985119119940615,
+      "learning_rate": 2.399968144725237e-06,
+      "loss": 0.3155,
+      "step": 5747
+    },
+    {
+      "epoch": 1.368179925031237,
+      "grad_norm": 0.4030281878578675,
+      "learning_rate": 2.3983218597237924e-06,
+      "loss": 0.3834,
+      "step": 5748
+    },
+    {
+      "epoch": 1.3684179211043017,
+      "grad_norm": 0.3690031365226529,
+      "learning_rate": 2.3966759613960796e-06,
+      "loss": 0.3083,
+      "step": 5749
+    },
+    {
+      "epoch": 1.3686559171773665,
+      "grad_norm": 0.38052926649872104,
+      "learning_rate": 2.3950304499867237e-06,
+      "loss": 0.2933,
+      "step": 5750
+    },
+    {
+      "epoch": 1.3688939132504314,
+      "grad_norm": 0.3970644105901736,
+      "learning_rate": 2.393385325740287e-06,
+      "loss": 0.3469,
+      "step": 5751
+    },
+    {
+      "epoch": 1.3691319093234962,
+      "grad_norm": 0.38735698530577334,
+      "learning_rate": 2.391740588901276e-06,
+      "loss": 0.4109,
+      "step": 5752
+    },
+    {
+      "epoch": 1.369369905396561,
+      "grad_norm": 0.37818783783560916,
+      "learning_rate": 2.3900962397141375e-06,
+      "loss": 0.2838,
+      "step": 5753
+    },
+    {
+      "epoch": 1.3696079014696259,
+      "grad_norm": 0.3963292077588658,
+      "learning_rate": 2.3884522784232645e-06,
+      "loss": 0.2984,
+      "step": 5754
+    },
+    {
+      "epoch": 1.3698458975426906,
+      "grad_norm": 0.3783349832246415,
+      "learning_rate": 2.3868087052729893e-06,
+      "loss": 0.3633,
+      "step": 5755
+    },
+    {
+      "epoch": 1.3700838936157553,
+      "grad_norm": 0.3663177261232768,
+      "learning_rate": 2.385165520507588e-06,
+      "loss": 0.357,
+      "step": 5756
+    },
+    {
+      "epoch": 1.37032188968882,
+      "grad_norm": 0.40376407001395825,
+      "learning_rate": 2.383522724371278e-06,
+      "loss": 0.3001,
+      "step": 5757
+    },
+    {
+      "epoch": 1.3705598857618848,
+      "grad_norm": 0.3888540867491295,
+      "learning_rate": 2.38188031710822e-06,
+      "loss": 0.3109,
+      "step": 5758
+    },
+    {
+      "epoch": 1.3707978818349498,
+      "grad_norm": 0.3823516378443356,
+      "learning_rate": 2.3802382989625156e-06,
+      "loss": 0.374,
+      "step": 5759
+    },
+    {
+      "epoch": 1.3710358779080145,
+      "grad_norm": 0.38773823298396787,
+      "learning_rate": 2.3785966701782098e-06,
+      "loss": 0.3037,
+      "step": 5760
+    },
+    {
+      "epoch": 1.3712738739810792,
+      "grad_norm": 0.4285193328439413,
+      "learning_rate": 2.3769554309992894e-06,
+      "loss": 0.2735,
+      "step": 5761
+    },
+    {
+      "epoch": 1.3715118700541442,
+      "grad_norm": 0.36656643054155086,
+      "learning_rate": 2.3753145816696827e-06,
+      "loss": 0.3558,
+      "step": 5762
+    },
+    {
+      "epoch": 1.371749866127209,
+      "grad_norm": 0.3731902128515065,
+      "learning_rate": 2.3736741224332592e-06,
+      "loss": 0.3605,
+      "step": 5763
+    },
+    {
+      "epoch": 1.3719878622002737,
+      "grad_norm": 0.3743570952391234,
+      "learning_rate": 2.372034053533835e-06,
+      "loss": 0.279,
+      "step": 5764
+    },
+    {
+      "epoch": 1.3722258582733384,
+      "grad_norm": 0.36798469228190406,
+      "learning_rate": 2.370394375215163e-06,
+      "loss": 0.3228,
+      "step": 5765
+    },
+    {
+      "epoch": 1.3724638543464032,
+      "grad_norm": 0.40727564860454385,
+      "learning_rate": 2.36875508772094e-06,
+      "loss": 0.3691,
+      "step": 5766
+    },
+    {
+      "epoch": 1.3727018504194681,
+      "grad_norm": 0.379396729169711,
+      "learning_rate": 2.367116191294806e-06,
+      "loss": 0.3249,
+      "step": 5767
+    },
+    {
+      "epoch": 1.3729398464925329,
+      "grad_norm": 0.4026927710670588,
+      "learning_rate": 2.365477686180339e-06,
+      "loss": 0.2866,
+      "step": 5768
+    },
+    {
+      "epoch": 1.3731778425655976,
+      "grad_norm": 0.37305525839630027,
+      "learning_rate": 2.3638395726210633e-06,
+      "loss": 0.3318,
+      "step": 5769
+    },
+    {
+      "epoch": 1.3734158386386626,
+      "grad_norm": 0.39186058761949977,
+      "learning_rate": 2.362201850860443e-06,
+      "loss": 0.3883,
+      "step": 5770
+    },
+    {
+      "epoch": 1.3736538347117273,
+      "grad_norm": 0.3880251678929161,
+      "learning_rate": 2.3605645211418817e-06,
+      "loss": 0.3171,
+      "step": 5771
+    },
+    {
+      "epoch": 1.373891830784792,
+      "grad_norm": 0.38767893993417474,
+      "learning_rate": 2.35892758370873e-06,
+      "loss": 0.3205,
+      "step": 5772
+    },
+    {
+      "epoch": 1.3741298268578568,
+      "grad_norm": 0.40297474576829545,
+      "learning_rate": 2.357291038804277e-06,
+      "loss": 0.3814,
+      "step": 5773
+    },
+    {
+      "epoch": 1.3743678229309215,
+      "grad_norm": 0.3831088465518714,
+      "learning_rate": 2.3556548866717532e-06,
+      "loss": 0.3225,
+      "step": 5774
+    },
+    {
+      "epoch": 1.3746058190039865,
+      "grad_norm": 0.4134284378393506,
+      "learning_rate": 2.3540191275543313e-06,
+      "loss": 0.2801,
+      "step": 5775
+    },
+    {
+      "epoch": 1.3748438150770512,
+      "grad_norm": 0.42059665325132306,
+      "learning_rate": 2.352383761695125e-06,
+      "loss": 0.3214,
+      "step": 5776
+    },
+    {
+      "epoch": 1.375081811150116,
+      "grad_norm": 0.39143587678806474,
+      "learning_rate": 2.350748789337189e-06,
+      "loss": 0.3738,
+      "step": 5777
+    },
+    {
+      "epoch": 1.375319807223181,
+      "grad_norm": 0.349803339437346,
+      "learning_rate": 2.349114210723524e-06,
+      "loss": 0.2689,
+      "step": 5778
+    },
+    {
+      "epoch": 1.3755578032962457,
+      "grad_norm": 0.3951646071549478,
+      "learning_rate": 2.3474800260970663e-06,
+      "loss": 0.3007,
+      "step": 5779
+    },
+    {
+      "epoch": 1.3757957993693104,
+      "grad_norm": 0.3898501872812015,
+      "learning_rate": 2.345846235700698e-06,
+      "loss": 0.3639,
+      "step": 5780
+    },
+    {
+      "epoch": 1.3760337954423751,
+      "grad_norm": 0.38912512748823874,
+      "learning_rate": 2.3442128397772396e-06,
+      "loss": 0.3618,
+      "step": 5781
+    },
+    {
+      "epoch": 1.3762717915154399,
+      "grad_norm": 0.38556905651285667,
+      "learning_rate": 2.3425798385694536e-06,
+      "loss": 0.26,
+      "step": 5782
+    },
+    {
+      "epoch": 1.3765097875885048,
+      "grad_norm": 0.4281982976966381,
+      "learning_rate": 2.3409472323200456e-06,
+      "loss": 0.3449,
+      "step": 5783
+    },
+    {
+      "epoch": 1.3767477836615696,
+      "grad_norm": 0.39046675152439464,
+      "learning_rate": 2.3393150212716604e-06,
+      "loss": 0.3875,
+      "step": 5784
+    },
+    {
+      "epoch": 1.3769857797346343,
+      "grad_norm": 0.3544336457158023,
+      "learning_rate": 2.337683205666885e-06,
+      "loss": 0.3233,
+      "step": 5785
+    },
+    {
+      "epoch": 1.3772237758076993,
+      "grad_norm": 0.36474388790592926,
+      "learning_rate": 2.336051785748248e-06,
+      "loss": 0.285,
+      "step": 5786
+    },
+    {
+      "epoch": 1.377461771880764,
+      "grad_norm": 0.3585335598025468,
+      "learning_rate": 2.334420761758219e-06,
+      "loss": 0.3508,
+      "step": 5787
+    },
+    {
+      "epoch": 1.3776997679538288,
+      "grad_norm": 0.395098339023238,
+      "learning_rate": 2.332790133939207e-06,
+      "loss": 0.4056,
+      "step": 5788
+    },
+    {
+      "epoch": 1.3779377640268935,
+      "grad_norm": 0.3916655159612242,
+      "learning_rate": 2.3311599025335654e-06,
+      "loss": 0.3117,
+      "step": 5789
+    },
+    {
+      "epoch": 1.3781757600999582,
+      "grad_norm": 0.39352579920052977,
+      "learning_rate": 2.3295300677835857e-06,
+      "loss": 0.3329,
+      "step": 5790
+    },
+    {
+      "epoch": 1.3784137561730232,
+      "grad_norm": 0.3949421237210347,
+      "learning_rate": 2.327900629931501e-06,
+      "loss": 0.4006,
+      "step": 5791
+    },
+    {
+      "epoch": 1.378651752246088,
+      "grad_norm": 0.3930410571926244,
+      "learning_rate": 2.3262715892194885e-06,
+      "loss": 0.3191,
+      "step": 5792
+    },
+    {
+      "epoch": 1.3788897483191527,
+      "grad_norm": 0.3767521685835201,
+      "learning_rate": 2.3246429458896637e-06,
+      "loss": 0.2756,
+      "step": 5793
+    },
+    {
+      "epoch": 1.3791277443922176,
+      "grad_norm": 0.4252278529678899,
+      "learning_rate": 2.3230147001840814e-06,
+      "loss": 0.3496,
+      "step": 5794
+    },
+    {
+      "epoch": 1.3793657404652824,
+      "grad_norm": 0.40177327803867513,
+      "learning_rate": 2.3213868523447404e-06,
+      "loss": 0.3888,
+      "step": 5795
+    },
+    {
+      "epoch": 1.3796037365383471,
+      "grad_norm": 0.3991978634595426,
+      "learning_rate": 2.3197594026135785e-06,
+      "loss": 0.2992,
+      "step": 5796
+    },
+    {
+      "epoch": 1.3798417326114119,
+      "grad_norm": 0.3820122275069646,
+      "learning_rate": 2.3181323512324754e-06,
+      "loss": 0.3062,
+      "step": 5797
+    },
+    {
+      "epoch": 1.3800797286844766,
+      "grad_norm": 0.394013127202058,
+      "learning_rate": 2.3165056984432493e-06,
+      "loss": 0.3725,
+      "step": 5798
+    },
+    {
+      "epoch": 1.3803177247575416,
+      "grad_norm": 0.3865299916518418,
+      "learning_rate": 2.314879444487665e-06,
+      "loss": 0.308,
+      "step": 5799
+    },
+    {
+      "epoch": 1.3805557208306063,
+      "grad_norm": 0.3606035448903917,
+      "learning_rate": 2.3132535896074213e-06,
+      "loss": 0.2737,
+      "step": 5800
+    },
+    {
+      "epoch": 1.380793716903671,
+      "grad_norm": 0.39435874636986457,
+      "learning_rate": 2.3116281340441616e-06,
+      "loss": 0.3243,
+      "step": 5801
+    },
+    {
+      "epoch": 1.381031712976736,
+      "grad_norm": 0.3944881659871505,
+      "learning_rate": 2.310003078039468e-06,
+      "loss": 0.3845,
+      "step": 5802
+    },
+    {
+      "epoch": 1.3812697090498007,
+      "grad_norm": 0.3739664004902333,
+      "learning_rate": 2.3083784218348646e-06,
+      "loss": 0.3068,
+      "step": 5803
+    },
+    {
+      "epoch": 1.3815077051228655,
+      "grad_norm": 0.3784435268416916,
+      "learning_rate": 2.3067541656718133e-06,
+      "loss": 0.2978,
+      "step": 5804
+    },
+    {
+      "epoch": 1.3817457011959302,
+      "grad_norm": 0.35412487368158974,
+      "learning_rate": 2.305130309791723e-06,
+      "loss": 0.3379,
+      "step": 5805
+    },
+    {
+      "epoch": 1.381983697268995,
+      "grad_norm": 0.408945279084268,
+      "learning_rate": 2.303506854435936e-06,
+      "loss": 0.3552,
+      "step": 5806
+    },
+    {
+      "epoch": 1.38222169334206,
+      "grad_norm": 0.3753005006752039,
+      "learning_rate": 2.3018837998457384e-06,
+      "loss": 0.2929,
+      "step": 5807
+    },
+    {
+      "epoch": 1.3824596894151246,
+      "grad_norm": 0.36459640658547277,
+      "learning_rate": 2.3002611462623563e-06,
+      "loss": 0.3338,
+      "step": 5808
+    },
+    {
+      "epoch": 1.3826976854881894,
+      "grad_norm": 0.3665121988763295,
+      "learning_rate": 2.2986388939269567e-06,
+      "loss": 0.3665,
+      "step": 5809
+    },
+    {
+      "epoch": 1.3829356815612543,
+      "grad_norm": 0.3517995162869186,
+      "learning_rate": 2.2970170430806455e-06,
+      "loss": 0.2952,
+      "step": 5810
+    },
+    {
+      "epoch": 1.383173677634319,
+      "grad_norm": 0.39818503253324855,
+      "learning_rate": 2.29539559396447e-06,
+      "loss": 0.2816,
+      "step": 5811
+    },
+    {
+      "epoch": 1.3834116737073838,
+      "grad_norm": 0.3883965107678621,
+      "learning_rate": 2.2937745468194186e-06,
+      "loss": 0.3767,
+      "step": 5812
+    },
+    {
+      "epoch": 1.3836496697804486,
+      "grad_norm": 0.34808683822000047,
+      "learning_rate": 2.2921539018864177e-06,
+      "loss": 0.3601,
+      "step": 5813
+    },
+    {
+      "epoch": 1.3838876658535133,
+      "grad_norm": 0.3785593295383857,
+      "learning_rate": 2.290533659406336e-06,
+      "loss": 0.3008,
+      "step": 5814
+    },
+    {
+      "epoch": 1.3841256619265783,
+      "grad_norm": 0.38995875252493734,
+      "learning_rate": 2.2889138196199816e-06,
+      "loss": 0.3139,
+      "step": 5815
+    },
+    {
+      "epoch": 1.384363657999643,
+      "grad_norm": 0.37852442117564816,
+      "learning_rate": 2.287294382768103e-06,
+      "loss": 0.3529,
+      "step": 5816
+    },
+    {
+      "epoch": 1.3846016540727077,
+      "grad_norm": 0.37788122292949944,
+      "learning_rate": 2.285675349091388e-06,
+      "loss": 0.3376,
+      "step": 5817
+    },
+    {
+      "epoch": 1.3848396501457727,
+      "grad_norm": 0.42130823797471484,
+      "learning_rate": 2.284056718830463e-06,
+      "loss": 0.2815,
+      "step": 5818
+    },
+    {
+      "epoch": 1.3850776462188374,
+      "grad_norm": 0.4103925080704719,
+      "learning_rate": 2.2824384922259005e-06,
+      "loss": 0.3086,
+      "step": 5819
+    },
+    {
+      "epoch": 1.3853156422919022,
+      "grad_norm": 0.42671753513437,
+      "learning_rate": 2.280820669518208e-06,
+      "loss": 0.4045,
+      "step": 5820
+    },
+    {
+      "epoch": 1.385553638364967,
+      "grad_norm": 0.39262412594080603,
+      "learning_rate": 2.2792032509478335e-06,
+      "loss": 0.2985,
+      "step": 5821
+    },
+    {
+      "epoch": 1.3857916344380317,
+      "grad_norm": 0.3889466132697115,
+      "learning_rate": 2.2775862367551642e-06,
+      "loss": 0.3125,
+      "step": 5822
+    },
+    {
+      "epoch": 1.3860296305110966,
+      "grad_norm": 0.3873197747905789,
+      "learning_rate": 2.2759696271805298e-06,
+      "loss": 0.3508,
+      "step": 5823
+    },
+    {
+      "epoch": 1.3862676265841614,
+      "grad_norm": 0.38959234217638355,
+      "learning_rate": 2.274353422464198e-06,
+      "loss": 0.327,
+      "step": 5824
+    },
+    {
+      "epoch": 1.386505622657226,
+      "grad_norm": 0.49119856662683686,
+      "learning_rate": 2.2727376228463754e-06,
+      "loss": 0.2518,
+      "step": 5825
+    },
+    {
+      "epoch": 1.386743618730291,
+      "grad_norm": 0.3592727494766306,
+      "learning_rate": 2.271122228567213e-06,
+      "loss": 0.3103,
+      "step": 5826
+    },
+    {
+      "epoch": 1.3869816148033558,
+      "grad_norm": 0.40339073061591146,
+      "learning_rate": 2.2695072398667965e-06,
+      "loss": 0.4171,
+      "step": 5827
+    },
+    {
+      "epoch": 1.3872196108764205,
+      "grad_norm": 0.35219893549707026,
+      "learning_rate": 2.2678926569851535e-06,
+      "loss": 0.3219,
+      "step": 5828
+    },
+    {
+      "epoch": 1.3874576069494853,
+      "grad_norm": 0.3964161658508986,
+      "learning_rate": 2.26627848016225e-06,
+      "loss": 0.2777,
+      "step": 5829
+    },
+    {
+      "epoch": 1.38769560302255,
+      "grad_norm": 0.38942352677332387,
+      "learning_rate": 2.2646647096379944e-06,
+      "loss": 0.3602,
+      "step": 5830
+    },
+    {
+      "epoch": 1.387933599095615,
+      "grad_norm": 0.42284500435613087,
+      "learning_rate": 2.2630513456522315e-06,
+      "loss": 0.3794,
+      "step": 5831
+    },
+    {
+      "epoch": 1.3881715951686797,
+      "grad_norm": 0.3975269422746937,
+      "learning_rate": 2.261438388444745e-06,
+      "loss": 0.2931,
+      "step": 5832
+    },
+    {
+      "epoch": 1.3884095912417445,
+      "grad_norm": 0.3911429557797404,
+      "learning_rate": 2.259825838255265e-06,
+      "loss": 0.3317,
+      "step": 5833
+    },
+    {
+      "epoch": 1.3886475873148094,
+      "grad_norm": 0.40487348869777584,
+      "learning_rate": 2.2582136953234544e-06,
+      "loss": 0.3936,
+      "step": 5834
+    },
+    {
+      "epoch": 1.3888855833878742,
+      "grad_norm": 0.3675010804576965,
+      "learning_rate": 2.2566019598889168e-06,
+      "loss": 0.301,
+      "step": 5835
+    },
+    {
+      "epoch": 1.389123579460939,
+      "grad_norm": 0.3755478510541879,
+      "learning_rate": 2.254990632191197e-06,
+      "loss": 0.2646,
+      "step": 5836
+    },
+    {
+      "epoch": 1.3893615755340036,
+      "grad_norm": 0.3717516657625129,
+      "learning_rate": 2.2533797124697763e-06,
+      "loss": 0.3402,
+      "step": 5837
+    },
+    {
+      "epoch": 1.3895995716070684,
+      "grad_norm": 0.37873677832015795,
+      "learning_rate": 2.2517692009640796e-06,
+      "loss": 0.3532,
+      "step": 5838
+    },
+    {
+      "epoch": 1.3898375676801333,
+      "grad_norm": 0.3812765678125732,
+      "learning_rate": 2.250159097913467e-06,
+      "loss": 0.2933,
+      "step": 5839
+    },
+    {
+      "epoch": 1.390075563753198,
+      "grad_norm": 0.3833578003169235,
+      "learning_rate": 2.2485494035572404e-06,
+      "loss": 0.2927,
+      "step": 5840
+    },
+    {
+      "epoch": 1.3903135598262628,
+      "grad_norm": 0.37497476870990365,
+      "learning_rate": 2.24694011813464e-06,
+      "loss": 0.3708,
+      "step": 5841
+    },
+    {
+      "epoch": 1.3905515558993278,
+      "grad_norm": 0.35418256159670797,
+      "learning_rate": 2.2453312418848454e-06,
+      "loss": 0.318,
+      "step": 5842
+    },
+    {
+      "epoch": 1.3907895519723925,
+      "grad_norm": 0.3892901098428216,
+      "learning_rate": 2.243722775046976e-06,
+      "loss": 0.3176,
+      "step": 5843
+    },
+    {
+      "epoch": 1.3910275480454573,
+      "grad_norm": 0.37876686309642,
+      "learning_rate": 2.2421147178600883e-06,
+      "loss": 0.2907,
+      "step": 5844
+    },
+    {
+      "epoch": 1.391265544118522,
+      "grad_norm": 0.3847753143788955,
+      "learning_rate": 2.2405070705631808e-06,
+      "loss": 0.3906,
+      "step": 5845
+    },
+    {
+      "epoch": 1.3915035401915867,
+      "grad_norm": 0.3638169205713526,
+      "learning_rate": 2.238899833395188e-06,
+      "loss": 0.2749,
+      "step": 5846
+    },
+    {
+      "epoch": 1.3917415362646517,
+      "grad_norm": 0.40407484545756533,
+      "learning_rate": 2.2372930065949873e-06,
+      "loss": 0.3018,
+      "step": 5847
+    },
+    {
+      "epoch": 1.3919795323377164,
+      "grad_norm": 0.37766478316846785,
+      "learning_rate": 2.235686590401392e-06,
+      "loss": 0.3687,
+      "step": 5848
+    },
+    {
+      "epoch": 1.3922175284107812,
+      "grad_norm": 0.34719174148467286,
+      "learning_rate": 2.234080585053155e-06,
+      "loss": 0.3441,
+      "step": 5849
+    },
+    {
+      "epoch": 1.3924555244838461,
+      "grad_norm": 0.37713121646339276,
+      "learning_rate": 2.232474990788969e-06,
+      "loss": 0.2669,
+      "step": 5850
+    },
+    {
+      "epoch": 1.3926935205569109,
+      "grad_norm": 0.40478233582795153,
+      "learning_rate": 2.2308698078474645e-06,
+      "loss": 0.3362,
+      "step": 5851
+    },
+    {
+      "epoch": 1.3929315166299756,
+      "grad_norm": 0.4118362847556917,
+      "learning_rate": 2.2292650364672096e-06,
+      "loss": 0.3946,
+      "step": 5852
+    },
+    {
+      "epoch": 1.3931695127030403,
+      "grad_norm": 0.3558802597442529,
+      "learning_rate": 2.2276606768867172e-06,
+      "loss": 0.2805,
+      "step": 5853
+    },
+    {
+      "epoch": 1.393407508776105,
+      "grad_norm": 0.4691025727643871,
+      "learning_rate": 2.226056729344432e-06,
+      "loss": 0.2816,
+      "step": 5854
+    },
+    {
+      "epoch": 1.39364550484917,
+      "grad_norm": 0.4030556981141061,
+      "learning_rate": 2.2244531940787413e-06,
+      "loss": 0.3574,
+      "step": 5855
+    },
+    {
+      "epoch": 1.3938835009222348,
+      "grad_norm": 0.3809802135059961,
+      "learning_rate": 2.22285007132797e-06,
+      "loss": 0.3576,
+      "step": 5856
+    },
+    {
+      "epoch": 1.3941214969952995,
+      "grad_norm": 0.35544990461765996,
+      "learning_rate": 2.2212473613303807e-06,
+      "loss": 0.2642,
+      "step": 5857
+    },
+    {
+      "epoch": 1.3943594930683645,
+      "grad_norm": 0.3692152888228544,
+      "learning_rate": 2.2196450643241768e-06,
+      "loss": 0.3212,
+      "step": 5858
+    },
+    {
+      "epoch": 1.3945974891414292,
+      "grad_norm": 0.4034006295018707,
+      "learning_rate": 2.218043180547499e-06,
+      "loss": 0.3548,
+      "step": 5859
+    },
+    {
+      "epoch": 1.394835485214494,
+      "grad_norm": 0.3750782883132117,
+      "learning_rate": 2.216441710238425e-06,
+      "loss": 0.2913,
+      "step": 5860
+    },
+    {
+      "epoch": 1.3950734812875587,
+      "grad_norm": 0.4040307357444328,
+      "learning_rate": 2.214840653634977e-06,
+      "loss": 0.2816,
+      "step": 5861
+    },
+    {
+      "epoch": 1.3953114773606234,
+      "grad_norm": 0.3673705781593482,
+      "learning_rate": 2.213240010975109e-06,
+      "loss": 0.3495,
+      "step": 5862
+    },
+    {
+      "epoch": 1.3955494734336884,
+      "grad_norm": 0.3704834222648198,
+      "learning_rate": 2.211639782496717e-06,
+      "loss": 0.3757,
+      "step": 5863
+    },
+    {
+      "epoch": 1.3957874695067531,
+      "grad_norm": 0.37384507283033885,
+      "learning_rate": 2.2100399684376333e-06,
+      "loss": 0.2976,
+      "step": 5864
+    },
+    {
+      "epoch": 1.3960254655798179,
+      "grad_norm": 0.38977052889591346,
+      "learning_rate": 2.208440569035631e-06,
+      "loss": 0.3131,
+      "step": 5865
+    },
+    {
+      "epoch": 1.3962634616528828,
+      "grad_norm": 0.3803082135668662,
+      "learning_rate": 2.2068415845284197e-06,
+      "loss": 0.3629,
+      "step": 5866
+    },
+    {
+      "epoch": 1.3965014577259476,
+      "grad_norm": 0.3428155698284938,
+      "learning_rate": 2.2052430151536488e-06,
+      "loss": 0.306,
+      "step": 5867
+    },
+    {
+      "epoch": 1.3967394537990123,
+      "grad_norm": 0.4200376862637538,
+      "learning_rate": 2.203644861148904e-06,
+      "loss": 0.3033,
+      "step": 5868
+    },
+    {
+      "epoch": 1.396977449872077,
+      "grad_norm": 0.41707618321368245,
+      "learning_rate": 2.202047122751712e-06,
+      "loss": 0.3557,
+      "step": 5869
+    },
+    {
+      "epoch": 1.3972154459451418,
+      "grad_norm": 0.39243821595567624,
+      "learning_rate": 2.2004498001995355e-06,
+      "loss": 0.3608,
+      "step": 5870
+    },
+    {
+      "epoch": 1.3974534420182068,
+      "grad_norm": 0.3533215441822008,
+      "learning_rate": 2.1988528937297764e-06,
+      "loss": 0.2832,
+      "step": 5871
+    },
+    {
+      "epoch": 1.3976914380912715,
+      "grad_norm": 0.37292525643353835,
+      "learning_rate": 2.1972564035797738e-06,
+      "loss": 0.313,
+      "step": 5872
+    },
+    {
+      "epoch": 1.3979294341643362,
+      "grad_norm": 0.41160282830855566,
+      "learning_rate": 2.1956603299868052e-06,
+      "loss": 0.3649,
+      "step": 5873
+    },
+    {
+      "epoch": 1.3981674302374012,
+      "grad_norm": 0.3789548998028757,
+      "learning_rate": 2.1940646731880887e-06,
+      "loss": 0.3277,
+      "step": 5874
+    },
+    {
+      "epoch": 1.398405426310466,
+      "grad_norm": 0.40108158442899056,
+      "learning_rate": 2.1924694334207773e-06,
+      "loss": 0.2789,
+      "step": 5875
+    },
+    {
+      "epoch": 1.3986434223835307,
+      "grad_norm": 0.38479323451240566,
+      "learning_rate": 2.1908746109219633e-06,
+      "loss": 0.3232,
+      "step": 5876
+    },
+    {
+      "epoch": 1.3988814184565954,
+      "grad_norm": 0.39277475179443505,
+      "learning_rate": 2.189280205928676e-06,
+      "loss": 0.4036,
+      "step": 5877
+    },
+    {
+      "epoch": 1.3991194145296602,
+      "grad_norm": 0.3629626558888327,
+      "learning_rate": 2.1876862186778847e-06,
+      "loss": 0.292,
+      "step": 5878
+    },
+    {
+      "epoch": 1.3993574106027251,
+      "grad_norm": 0.39386803443255086,
+      "learning_rate": 2.186092649406492e-06,
+      "loss": 0.2813,
+      "step": 5879
+    },
+    {
+      "epoch": 1.3995954066757899,
+      "grad_norm": 0.4045307535442236,
+      "learning_rate": 2.184499498351347e-06,
+      "loss": 0.3579,
+      "step": 5880
+    },
+    {
+      "epoch": 1.3998334027488546,
+      "grad_norm": 0.38562044486044106,
+      "learning_rate": 2.182906765749228e-06,
+      "loss": 0.3259,
+      "step": 5881
+    },
+    {
+      "epoch": 1.4000713988219196,
+      "grad_norm": 0.3722054840630449,
+      "learning_rate": 2.1813144518368556e-06,
+      "loss": 0.2682,
+      "step": 5882
+    },
+    {
+      "epoch": 1.4003093948949843,
+      "grad_norm": 0.4853537819335595,
+      "learning_rate": 2.1797225568508863e-06,
+      "loss": 0.3586,
+      "step": 5883
+    },
+    {
+      "epoch": 1.400547390968049,
+      "grad_norm": 0.3917377497499103,
+      "learning_rate": 2.1781310810279156e-06,
+      "loss": 0.3889,
+      "step": 5884
+    },
+    {
+      "epoch": 1.4007853870411138,
+      "grad_norm": 0.3820969906527712,
+      "learning_rate": 2.1765400246044755e-06,
+      "loss": 0.286,
+      "step": 5885
+    },
+    {
+      "epoch": 1.4010233831141785,
+      "grad_norm": 0.3901672408734566,
+      "learning_rate": 2.1749493878170368e-06,
+      "loss": 0.268,
+      "step": 5886
+    },
+    {
+      "epoch": 1.4012613791872435,
+      "grad_norm": 0.5012229103908235,
+      "learning_rate": 2.173359170902006e-06,
+      "loss": 0.3502,
+      "step": 5887
+    },
+    {
+      "epoch": 1.4014993752603082,
+      "grad_norm": 0.4163979879875083,
+      "learning_rate": 2.171769374095732e-06,
+      "loss": 0.3584,
+      "step": 5888
+    },
+    {
+      "epoch": 1.401737371333373,
+      "grad_norm": 0.3825230668134233,
+      "learning_rate": 2.1701799976344956e-06,
+      "loss": 0.2952,
+      "step": 5889
+    },
+    {
+      "epoch": 1.401975367406438,
+      "grad_norm": 0.3692912522268186,
+      "learning_rate": 2.168591041754518e-06,
+      "loss": 0.3206,
+      "step": 5890
+    },
+    {
+      "epoch": 1.4022133634795027,
+      "grad_norm": 0.4020957675464499,
+      "learning_rate": 2.1670025066919575e-06,
+      "loss": 0.3645,
+      "step": 5891
+    },
+    {
+      "epoch": 1.4024513595525674,
+      "grad_norm": 0.3652682950372471,
+      "learning_rate": 2.1654143926829095e-06,
+      "loss": 0.3088,
+      "step": 5892
+    },
+    {
+      "epoch": 1.4026893556256321,
+      "grad_norm": 0.39357600720054553,
+      "learning_rate": 2.163826699963407e-06,
+      "loss": 0.3002,
+      "step": 5893
+    },
+    {
+      "epoch": 1.4029273516986969,
+      "grad_norm": 0.39176835973179347,
+      "learning_rate": 2.1622394287694203e-06,
+      "loss": 0.3524,
+      "step": 5894
+    },
+    {
+      "epoch": 1.4031653477717618,
+      "grad_norm": 0.4030630624539821,
+      "learning_rate": 2.1606525793368578e-06,
+      "loss": 0.4,
+      "step": 5895
+    },
+    {
+      "epoch": 1.4034033438448266,
+      "grad_norm": 0.3581093728731858,
+      "learning_rate": 2.159066151901563e-06,
+      "loss": 0.3042,
+      "step": 5896
+    },
+    {
+      "epoch": 1.4036413399178913,
+      "grad_norm": 0.3958734156159978,
+      "learning_rate": 2.1574801466993204e-06,
+      "loss": 0.2925,
+      "step": 5897
+    },
+    {
+      "epoch": 1.4038793359909563,
+      "grad_norm": 0.44017456212239137,
+      "learning_rate": 2.155894563965848e-06,
+      "loss": 0.3305,
+      "step": 5898
+    },
+    {
+      "epoch": 1.404117332064021,
+      "grad_norm": 0.36546554037443685,
+      "learning_rate": 2.1543094039368034e-06,
+      "loss": 0.3465,
+      "step": 5899
+    },
+    {
+      "epoch": 1.4043553281370857,
+      "grad_norm": 0.4261135558150708,
+      "learning_rate": 2.15272466684778e-06,
+      "loss": 0.2733,
+      "step": 5900
+    },
+    {
+      "epoch": 1.4045933242101505,
+      "grad_norm": 0.42031143455545217,
+      "learning_rate": 2.151140352934308e-06,
+      "loss": 0.3398,
+      "step": 5901
+    },
+    {
+      "epoch": 1.4048313202832152,
+      "grad_norm": 0.3751387002028948,
+      "learning_rate": 2.149556462431859e-06,
+      "loss": 0.3863,
+      "step": 5902
+    },
+    {
+      "epoch": 1.4050693163562802,
+      "grad_norm": 0.3438466812731875,
+      "learning_rate": 2.1479729955758354e-06,
+      "loss": 0.3185,
+      "step": 5903
+    },
+    {
+      "epoch": 1.405307312429345,
+      "grad_norm": 0.3693116279880859,
+      "learning_rate": 2.146389952601581e-06,
+      "loss": 0.2928,
+      "step": 5904
+    },
+    {
+      "epoch": 1.4055453085024097,
+      "grad_norm": 0.39195086077989794,
+      "learning_rate": 2.1448073337443743e-06,
+      "loss": 0.3798,
+      "step": 5905
+    },
+    {
+      "epoch": 1.4057833045754746,
+      "grad_norm": 0.379901101442375,
+      "learning_rate": 2.1432251392394303e-06,
+      "loss": 0.3315,
+      "step": 5906
+    },
+    {
+      "epoch": 1.4060213006485394,
+      "grad_norm": 0.3827424931752997,
+      "learning_rate": 2.141643369321905e-06,
+      "loss": 0.2723,
+      "step": 5907
+    },
+    {
+      "epoch": 1.406259296721604,
+      "grad_norm": 0.3816989292493151,
+      "learning_rate": 2.1400620242268883e-06,
+      "loss": 0.3374,
+      "step": 5908
+    },
+    {
+      "epoch": 1.4064972927946688,
+      "grad_norm": 0.38230399719450614,
+      "learning_rate": 2.1384811041894055e-06,
+      "loss": 0.387,
+      "step": 5909
+    },
+    {
+      "epoch": 1.4067352888677336,
+      "grad_norm": 0.3725751003550838,
+      "learning_rate": 2.1369006094444215e-06,
+      "loss": 0.2997,
+      "step": 5910
+    },
+    {
+      "epoch": 1.4069732849407985,
+      "grad_norm": 0.385430704556038,
+      "learning_rate": 2.1353205402268368e-06,
+      "loss": 0.3085,
+      "step": 5911
+    },
+    {
+      "epoch": 1.4072112810138633,
+      "grad_norm": 0.376139109287831,
+      "learning_rate": 2.1337408967714883e-06,
+      "loss": 0.3395,
+      "step": 5912
+    },
+    {
+      "epoch": 1.407449277086928,
+      "grad_norm": 0.3715517206789897,
+      "learning_rate": 2.1321616793131507e-06,
+      "loss": 0.3763,
+      "step": 5913
+    },
+    {
+      "epoch": 1.407687273159993,
+      "grad_norm": 0.39429770665412,
+      "learning_rate": 2.130582888086534e-06,
+      "loss": 0.2859,
+      "step": 5914
+    },
+    {
+      "epoch": 1.4079252692330577,
+      "grad_norm": 0.3733749256636019,
+      "learning_rate": 2.129004523326284e-06,
+      "loss": 0.2936,
+      "step": 5915
+    },
+    {
+      "epoch": 1.4081632653061225,
+      "grad_norm": 0.8974813867363939,
+      "learning_rate": 2.1274265852669894e-06,
+      "loss": 0.3692,
+      "step": 5916
+    },
+    {
+      "epoch": 1.4084012613791872,
+      "grad_norm": 0.3681033154441353,
+      "learning_rate": 2.125849074143168e-06,
+      "loss": 0.3132,
+      "step": 5917
+    },
+    {
+      "epoch": 1.408639257452252,
+      "grad_norm": 0.3871816903644566,
+      "learning_rate": 2.124271990189277e-06,
+      "loss": 0.2816,
+      "step": 5918
+    },
+    {
+      "epoch": 1.408877253525317,
+      "grad_norm": 0.3733497998587013,
+      "learning_rate": 2.1226953336397105e-06,
+      "loss": 0.3519,
+      "step": 5919
+    },
+    {
+      "epoch": 1.4091152495983816,
+      "grad_norm": 0.3581335746536029,
+      "learning_rate": 2.1211191047287988e-06,
+      "loss": 0.3848,
+      "step": 5920
+    },
+    {
+      "epoch": 1.4093532456714464,
+      "grad_norm": 0.35647096310686627,
+      "learning_rate": 2.119543303690808e-06,
+      "loss": 0.2912,
+      "step": 5921
+    },
+    {
+      "epoch": 1.4095912417445113,
+      "grad_norm": 0.3783590852785142,
+      "learning_rate": 2.117967930759941e-06,
+      "loss": 0.3103,
+      "step": 5922
+    },
+    {
+      "epoch": 1.409829237817576,
+      "grad_norm": 0.39200620885506376,
+      "learning_rate": 2.1163929861703383e-06,
+      "loss": 0.3547,
+      "step": 5923
+    },
+    {
+      "epoch": 1.4100672338906408,
+      "grad_norm": 0.3607566196572999,
+      "learning_rate": 2.1148184701560742e-06,
+      "loss": 0.3402,
+      "step": 5924
+    },
+    {
+      "epoch": 1.4103052299637056,
+      "grad_norm": 0.36895223669749566,
+      "learning_rate": 2.113244382951162e-06,
+      "loss": 0.275,
+      "step": 5925
+    },
+    {
+      "epoch": 1.4105432260367703,
+      "grad_norm": 0.3459412598350246,
+      "learning_rate": 2.1116707247895484e-06,
+      "loss": 0.3434,
+      "step": 5926
+    },
+    {
+      "epoch": 1.4107812221098353,
+      "grad_norm": 0.3767275620938957,
+      "learning_rate": 2.1100974959051198e-06,
+      "loss": 0.38,
+      "step": 5927
+    },
+    {
+      "epoch": 1.4110192181829,
+      "grad_norm": 0.39657367634889606,
+      "learning_rate": 2.1085246965316936e-06,
+      "loss": 0.3327,
+      "step": 5928
+    },
+    {
+      "epoch": 1.4112572142559647,
+      "grad_norm": 0.39804610232153587,
+      "learning_rate": 2.106952326903031e-06,
+      "loss": 0.2669,
+      "step": 5929
+    },
+    {
+      "epoch": 1.4114952103290297,
+      "grad_norm": 0.42512925166801646,
+      "learning_rate": 2.105380387252824e-06,
+      "loss": 0.3799,
+      "step": 5930
+    },
+    {
+      "epoch": 1.4117332064020944,
+      "grad_norm": 0.367576966472368,
+      "learning_rate": 2.1038088778147004e-06,
+      "loss": 0.3648,
+      "step": 5931
+    },
+    {
+      "epoch": 1.4119712024751592,
+      "grad_norm": 0.4071023704661706,
+      "learning_rate": 2.1022377988222255e-06,
+      "loss": 0.2493,
+      "step": 5932
+    },
+    {
+      "epoch": 1.412209198548224,
+      "grad_norm": 0.3517027129740413,
+      "learning_rate": 2.100667150508899e-06,
+      "loss": 0.3263,
+      "step": 5933
+    },
+    {
+      "epoch": 1.4124471946212886,
+      "grad_norm": 0.3630304666957127,
+      "learning_rate": 2.099096933108163e-06,
+      "loss": 0.3633,
+      "step": 5934
+    },
+    {
+      "epoch": 1.4126851906943536,
+      "grad_norm": 0.36827148716226255,
+      "learning_rate": 2.0975271468533864e-06,
+      "loss": 0.3073,
+      "step": 5935
+    },
+    {
+      "epoch": 1.4129231867674183,
+      "grad_norm": 0.39577401190423284,
+      "learning_rate": 2.0959577919778803e-06,
+      "loss": 0.2861,
+      "step": 5936
+    },
+    {
+      "epoch": 1.413161182840483,
+      "grad_norm": 0.40173519198012886,
+      "learning_rate": 2.0943888687148883e-06,
+      "loss": 0.3431,
+      "step": 5937
+    },
+    {
+      "epoch": 1.413399178913548,
+      "grad_norm": 0.3992484721549741,
+      "learning_rate": 2.0928203772975917e-06,
+      "loss": 0.3809,
+      "step": 5938
+    },
+    {
+      "epoch": 1.4136371749866128,
+      "grad_norm": 0.39599927216720754,
+      "learning_rate": 2.0912523179591076e-06,
+      "loss": 0.3008,
+      "step": 5939
+    },
+    {
+      "epoch": 1.4138751710596775,
+      "grad_norm": 0.4242280603241625,
+      "learning_rate": 2.0896846909324874e-06,
+      "loss": 0.3229,
+      "step": 5940
+    },
+    {
+      "epoch": 1.4141131671327423,
+      "grad_norm": 0.3664045587336667,
+      "learning_rate": 2.0881174964507205e-06,
+      "loss": 0.4031,
+      "step": 5941
+    },
+    {
+      "epoch": 1.414351163205807,
+      "grad_norm": 0.36607982694161056,
+      "learning_rate": 2.086550734746728e-06,
+      "loss": 0.3328,
+      "step": 5942
+    },
+    {
+      "epoch": 1.414589159278872,
+      "grad_norm": 0.3924516864004525,
+      "learning_rate": 2.0849844060533736e-06,
+      "loss": 0.2976,
+      "step": 5943
+    },
+    {
+      "epoch": 1.4148271553519367,
+      "grad_norm": 0.4057578371930518,
+      "learning_rate": 2.0834185106034503e-06,
+      "loss": 0.3015,
+      "step": 5944
+    },
+    {
+      "epoch": 1.4150651514250014,
+      "grad_norm": 0.3805032627968153,
+      "learning_rate": 2.081853048629689e-06,
+      "loss": 0.3876,
+      "step": 5945
+    },
+    {
+      "epoch": 1.4153031474980664,
+      "grad_norm": 0.3698394601043139,
+      "learning_rate": 2.0802880203647565e-06,
+      "loss": 0.2778,
+      "step": 5946
+    },
+    {
+      "epoch": 1.4155411435711311,
+      "grad_norm": 0.3828459136639372,
+      "learning_rate": 2.078723426041254e-06,
+      "loss": 0.3119,
+      "step": 5947
+    },
+    {
+      "epoch": 1.4157791396441959,
+      "grad_norm": 0.38957935834327895,
+      "learning_rate": 2.0771592658917196e-06,
+      "loss": 0.3794,
+      "step": 5948
+    },
+    {
+      "epoch": 1.4160171357172606,
+      "grad_norm": 0.3851845261585521,
+      "learning_rate": 2.0755955401486255e-06,
+      "loss": 0.3443,
+      "step": 5949
+    },
+    {
+      "epoch": 1.4162551317903254,
+      "grad_norm": 0.3546131940340063,
+      "learning_rate": 2.0740322490443802e-06,
+      "loss": 0.3024,
+      "step": 5950
+    },
+    {
+      "epoch": 1.4164931278633903,
+      "grad_norm": 0.4166580686412631,
+      "learning_rate": 2.072469392811329e-06,
+      "loss": 0.3346,
+      "step": 5951
+    },
+    {
+      "epoch": 1.416731123936455,
+      "grad_norm": 0.40477461909600554,
+      "learning_rate": 2.070906971681748e-06,
+      "loss": 0.3986,
+      "step": 5952
+    },
+    {
+      "epoch": 1.4169691200095198,
+      "grad_norm": 0.3863431875261796,
+      "learning_rate": 2.0693449858878543e-06,
+      "loss": 0.2625,
+      "step": 5953
+    },
+    {
+      "epoch": 1.4172071160825848,
+      "grad_norm": 0.42378362283530085,
+      "learning_rate": 2.0677834356617967e-06,
+      "loss": 0.2718,
+      "step": 5954
+    },
+    {
+      "epoch": 1.4174451121556495,
+      "grad_norm": 0.37534140974542646,
+      "learning_rate": 2.066222321235659e-06,
+      "loss": 0.3537,
+      "step": 5955
+    },
+    {
+      "epoch": 1.4176831082287142,
+      "grad_norm": 0.3421669011162348,
+      "learning_rate": 2.064661642841462e-06,
+      "loss": 0.3523,
+      "step": 5956
+    },
+    {
+      "epoch": 1.417921104301779,
+      "grad_norm": 0.37943922157081356,
+      "learning_rate": 2.0631014007111627e-06,
+      "loss": 0.2818,
+      "step": 5957
+    },
+    {
+      "epoch": 1.4181591003748437,
+      "grad_norm": 0.37246301648402214,
+      "learning_rate": 2.0615415950766504e-06,
+      "loss": 0.3343,
+      "step": 5958
+    },
+    {
+      "epoch": 1.4183970964479087,
+      "grad_norm": 0.38400127878264756,
+      "learning_rate": 2.0599822261697516e-06,
+      "loss": 0.3702,
+      "step": 5959
+    },
+    {
+      "epoch": 1.4186350925209734,
+      "grad_norm": 0.39219554424930037,
+      "learning_rate": 2.0584232942222247e-06,
+      "loss": 0.2835,
+      "step": 5960
+    },
+    {
+      "epoch": 1.4188730885940382,
+      "grad_norm": 0.39117355067041304,
+      "learning_rate": 2.056864799465769e-06,
+      "loss": 0.2966,
+      "step": 5961
+    },
+    {
+      "epoch": 1.4191110846671031,
+      "grad_norm": 0.39451400558685296,
+      "learning_rate": 2.055306742132014e-06,
+      "loss": 0.337,
+      "step": 5962
+    },
+    {
+      "epoch": 1.4193490807401679,
+      "grad_norm": 0.36635780928223005,
+      "learning_rate": 2.053749122452525e-06,
+      "loss": 0.362,
+      "step": 5963
+    },
+    {
+      "epoch": 1.4195870768132326,
+      "grad_norm": 0.3873092005359913,
+      "learning_rate": 2.052191940658803e-06,
+      "loss": 0.3079,
+      "step": 5964
+    },
+    {
+      "epoch": 1.4198250728862973,
+      "grad_norm": 0.3836880319475697,
+      "learning_rate": 2.050635196982284e-06,
+      "loss": 0.3435,
+      "step": 5965
+    },
+    {
+      "epoch": 1.420063068959362,
+      "grad_norm": 0.44296005657411364,
+      "learning_rate": 2.049078891654339e-06,
+      "loss": 0.37,
+      "step": 5966
+    },
+    {
+      "epoch": 1.420301065032427,
+      "grad_norm": 0.35016323703149027,
+      "learning_rate": 2.0475230249062727e-06,
+      "loss": 0.315,
+      "step": 5967
+    },
+    {
+      "epoch": 1.4205390611054918,
+      "grad_norm": 0.3950620054962618,
+      "learning_rate": 2.0459675969693256e-06,
+      "loss": 0.3173,
+      "step": 5968
+    },
+    {
+      "epoch": 1.4207770571785565,
+      "grad_norm": 0.3952176517038452,
+      "learning_rate": 2.044412608074672e-06,
+      "loss": 0.3234,
+      "step": 5969
+    },
+    {
+      "epoch": 1.4210150532516215,
+      "grad_norm": 0.38307975634819236,
+      "learning_rate": 2.042858058453422e-06,
+      "loss": 0.405,
+      "step": 5970
+    },
+    {
+      "epoch": 1.4212530493246862,
+      "grad_norm": 0.3364899399185918,
+      "learning_rate": 2.041303948336622e-06,
+      "loss": 0.3013,
+      "step": 5971
+    },
+    {
+      "epoch": 1.421491045397751,
+      "grad_norm": 0.38894109353307277,
+      "learning_rate": 2.0397502779552498e-06,
+      "loss": 0.2979,
+      "step": 5972
+    },
+    {
+      "epoch": 1.4217290414708157,
+      "grad_norm": 0.3510083872107239,
+      "learning_rate": 2.0381970475402196e-06,
+      "loss": 0.3435,
+      "step": 5973
+    },
+    {
+      "epoch": 1.4219670375438804,
+      "grad_norm": 0.3947818997277785,
+      "learning_rate": 2.0366442573223795e-06,
+      "loss": 0.3896,
+      "step": 5974
+    },
+    {
+      "epoch": 1.4222050336169454,
+      "grad_norm": 0.39395505014453025,
+      "learning_rate": 2.0350919075325124e-06,
+      "loss": 0.2891,
+      "step": 5975
+    },
+    {
+      "epoch": 1.4224430296900101,
+      "grad_norm": 0.3978783535558053,
+      "learning_rate": 2.0335399984013366e-06,
+      "loss": 0.3043,
+      "step": 5976
+    },
+    {
+      "epoch": 1.4226810257630749,
+      "grad_norm": 0.3682896039882474,
+      "learning_rate": 2.0319885301595034e-06,
+      "loss": 0.4002,
+      "step": 5977
+    },
+    {
+      "epoch": 1.4229190218361398,
+      "grad_norm": 0.37550375575321043,
+      "learning_rate": 2.0304375030375996e-06,
+      "loss": 0.2694,
+      "step": 5978
+    },
+    {
+      "epoch": 1.4231570179092046,
+      "grad_norm": 0.38790715040650325,
+      "learning_rate": 2.0288869172661463e-06,
+      "loss": 0.3062,
+      "step": 5979
+    },
+    {
+      "epoch": 1.4233950139822693,
+      "grad_norm": 0.39968408903014346,
+      "learning_rate": 2.0273367730755993e-06,
+      "loss": 0.3375,
+      "step": 5980
+    },
+    {
+      "epoch": 1.423633010055334,
+      "grad_norm": 0.3878895341440517,
+      "learning_rate": 2.025787070696348e-06,
+      "loss": 0.3548,
+      "step": 5981
+    },
+    {
+      "epoch": 1.4238710061283988,
+      "grad_norm": 0.39123351222051045,
+      "learning_rate": 2.0242378103587157e-06,
+      "loss": 0.2858,
+      "step": 5982
+    },
+    {
+      "epoch": 1.4241090022014637,
+      "grad_norm": 0.37665471025750874,
+      "learning_rate": 2.0226889922929603e-06,
+      "loss": 0.3093,
+      "step": 5983
+    },
+    {
+      "epoch": 1.4243469982745285,
+      "grad_norm": 0.39559868934355824,
+      "learning_rate": 2.0211406167292775e-06,
+      "loss": 0.3772,
+      "step": 5984
+    },
+    {
+      "epoch": 1.4245849943475932,
+      "grad_norm": 0.3851654558218671,
+      "learning_rate": 2.0195926838977926e-06,
+      "loss": 0.2878,
+      "step": 5985
+    },
+    {
+      "epoch": 1.4248229904206582,
+      "grad_norm": 0.37919297006292657,
+      "learning_rate": 2.018045194028567e-06,
+      "loss": 0.2562,
+      "step": 5986
+    },
+    {
+      "epoch": 1.425060986493723,
+      "grad_norm": 0.40761397021940626,
+      "learning_rate": 2.0164981473515926e-06,
+      "loss": 0.3191,
+      "step": 5987
+    },
+    {
+      "epoch": 1.4252989825667877,
+      "grad_norm": 0.4197443032408188,
+      "learning_rate": 2.014951544096804e-06,
+      "loss": 0.3657,
+      "step": 5988
+    },
+    {
+      "epoch": 1.4255369786398524,
+      "grad_norm": 0.37473340700791113,
+      "learning_rate": 2.013405384494063e-06,
+      "loss": 0.2811,
+      "step": 5989
+    },
+    {
+      "epoch": 1.4257749747129171,
+      "grad_norm": 0.40715204280059647,
+      "learning_rate": 2.0118596687731666e-06,
+      "loss": 0.3151,
+      "step": 5990
+    },
+    {
+      "epoch": 1.426012970785982,
+      "grad_norm": 0.3737319670986343,
+      "learning_rate": 2.0103143971638463e-06,
+      "loss": 0.3727,
+      "step": 5991
+    },
+    {
+      "epoch": 1.4262509668590468,
+      "grad_norm": 0.3698219377663196,
+      "learning_rate": 2.0087695698957676e-06,
+      "loss": 0.2946,
+      "step": 5992
+    },
+    {
+      "epoch": 1.4264889629321116,
+      "grad_norm": 0.37300313864826995,
+      "learning_rate": 2.0072251871985306e-06,
+      "loss": 0.2735,
+      "step": 5993
+    },
+    {
+      "epoch": 1.4267269590051765,
+      "grad_norm": 0.3944808909836751,
+      "learning_rate": 2.0056812493016684e-06,
+      "loss": 0.3603,
+      "step": 5994
+    },
+    {
+      "epoch": 1.4269649550782413,
+      "grad_norm": 0.3741150627911366,
+      "learning_rate": 2.0041377564346484e-06,
+      "loss": 0.3969,
+      "step": 5995
+    },
+    {
+      "epoch": 1.427202951151306,
+      "grad_norm": 0.36618783349147516,
+      "learning_rate": 2.0025947088268714e-06,
+      "loss": 0.3102,
+      "step": 5996
+    },
+    {
+      "epoch": 1.4274409472243708,
+      "grad_norm": 0.4085121504761507,
+      "learning_rate": 2.001052106707672e-06,
+      "loss": 0.3215,
+      "step": 5997
+    },
+    {
+      "epoch": 1.4276789432974355,
+      "grad_norm": 0.42374164985590546,
+      "learning_rate": 1.9995099503063214e-06,
+      "loss": 0.361,
+      "step": 5998
+    },
+    {
+      "epoch": 1.4279169393705005,
+      "grad_norm": 0.36596013003548744,
+      "learning_rate": 1.9979682398520205e-06,
+      "loss": 0.3197,
+      "step": 5999
+    },
+    {
+      "epoch": 1.4281549354435652,
+      "grad_norm": 0.38401848047343784,
+      "learning_rate": 1.9964269755739057e-06,
+      "loss": 0.2791,
+      "step": 6000
+    },
+    {
+      "epoch": 1.42839293151663,
+      "grad_norm": 0.3988267552673252,
+      "learning_rate": 1.9948861577010475e-06,
+      "loss": 0.313,
+      "step": 6001
+    },
+    {
+      "epoch": 1.428630927589695,
+      "grad_norm": 0.38693890781588514,
+      "learning_rate": 1.993345786462449e-06,
+      "loss": 0.391,
+      "step": 6002
+    },
+    {
+      "epoch": 1.4288689236627596,
+      "grad_norm": 0.36685062755023384,
+      "learning_rate": 1.991805862087048e-06,
+      "loss": 0.2948,
+      "step": 6003
+    },
+    {
+      "epoch": 1.4291069197358244,
+      "grad_norm": 0.5848360061934482,
+      "learning_rate": 1.9902663848037147e-06,
+      "loss": 0.278,
+      "step": 6004
+    },
+    {
+      "epoch": 1.4293449158088891,
+      "grad_norm": 0.405283771817613,
+      "learning_rate": 1.988727354841254e-06,
+      "loss": 0.3247,
+      "step": 6005
+    },
+    {
+      "epoch": 1.4295829118819539,
+      "grad_norm": 0.38737497620202277,
+      "learning_rate": 1.987188772428403e-06,
+      "loss": 0.3388,
+      "step": 6006
+    },
+    {
+      "epoch": 1.4298209079550188,
+      "grad_norm": 0.3778534408788696,
+      "learning_rate": 1.985650637793835e-06,
+      "loss": 0.2796,
+      "step": 6007
+    },
+    {
+      "epoch": 1.4300589040280836,
+      "grad_norm": 0.3917808141958219,
+      "learning_rate": 1.9841129511661526e-06,
+      "loss": 0.3428,
+      "step": 6008
+    },
+    {
+      "epoch": 1.4302969001011483,
+      "grad_norm": 0.39975809188737405,
+      "learning_rate": 1.9825757127738957e-06,
+      "loss": 0.396,
+      "step": 6009
+    },
+    {
+      "epoch": 1.4305348961742133,
+      "grad_norm": 0.37502506922397805,
+      "learning_rate": 1.9810389228455334e-06,
+      "loss": 0.2975,
+      "step": 6010
+    },
+    {
+      "epoch": 1.430772892247278,
+      "grad_norm": 0.38726845931918774,
+      "learning_rate": 1.9795025816094747e-06,
+      "loss": 0.273,
+      "step": 6011
+    },
+    {
+      "epoch": 1.4310108883203427,
+      "grad_norm": 0.3632876751872661,
+      "learning_rate": 1.9779666892940557e-06,
+      "loss": 0.3556,
+      "step": 6012
+    },
+    {
+      "epoch": 1.4312488843934075,
+      "grad_norm": 0.38274179138386594,
+      "learning_rate": 1.9764312461275482e-06,
+      "loss": 0.3783,
+      "step": 6013
+    },
+    {
+      "epoch": 1.4314868804664722,
+      "grad_norm": 0.3696767725810876,
+      "learning_rate": 1.974896252338155e-06,
+      "loss": 0.277,
+      "step": 6014
+    },
+    {
+      "epoch": 1.4317248765395372,
+      "grad_norm": 0.35996212200903566,
+      "learning_rate": 1.973361708154018e-06,
+      "loss": 0.3241,
+      "step": 6015
+    },
+    {
+      "epoch": 1.431962872612602,
+      "grad_norm": 0.3746809277181633,
+      "learning_rate": 1.9718276138032066e-06,
+      "loss": 0.4065,
+      "step": 6016
+    },
+    {
+      "epoch": 1.4322008686856667,
+      "grad_norm": 0.3662515570773327,
+      "learning_rate": 1.970293969513725e-06,
+      "loss": 0.3321,
+      "step": 6017
+    },
+    {
+      "epoch": 1.4324388647587316,
+      "grad_norm": 0.3893301791415084,
+      "learning_rate": 1.9687607755135114e-06,
+      "loss": 0.269,
+      "step": 6018
+    },
+    {
+      "epoch": 1.4326768608317964,
+      "grad_norm": 0.38494260892368,
+      "learning_rate": 1.9672280320304356e-06,
+      "loss": 0.3051,
+      "step": 6019
+    },
+    {
+      "epoch": 1.432914856904861,
+      "grad_norm": 0.38863096717866963,
+      "learning_rate": 1.965695739292301e-06,
+      "loss": 0.4005,
+      "step": 6020
+    },
+    {
+      "epoch": 1.4331528529779258,
+      "grad_norm": 0.3906075944195601,
+      "learning_rate": 1.964163897526845e-06,
+      "loss": 0.3081,
+      "step": 6021
+    },
+    {
+      "epoch": 1.4333908490509906,
+      "grad_norm": 0.4134133532934077,
+      "learning_rate": 1.9626325069617365e-06,
+      "loss": 0.3055,
+      "step": 6022
+    },
+    {
+      "epoch": 1.4336288451240555,
+      "grad_norm": 0.4136513446069149,
+      "learning_rate": 1.9611015678245786e-06,
+      "loss": 0.3594,
+      "step": 6023
+    },
+    {
+      "epoch": 1.4338668411971203,
+      "grad_norm": 0.3878063222204458,
+      "learning_rate": 1.9595710803429064e-06,
+      "loss": 0.3438,
+      "step": 6024
+    },
+    {
+      "epoch": 1.434104837270185,
+      "grad_norm": 0.3940209179695724,
+      "learning_rate": 1.958041044744186e-06,
+      "loss": 0.2871,
+      "step": 6025
+    },
+    {
+      "epoch": 1.43434283334325,
+      "grad_norm": 0.3824890901587333,
+      "learning_rate": 1.9565114612558232e-06,
+      "loss": 0.34,
+      "step": 6026
+    },
+    {
+      "epoch": 1.4345808294163147,
+      "grad_norm": 0.38272339908598974,
+      "learning_rate": 1.95498233010515e-06,
+      "loss": 0.3798,
+      "step": 6027
+    },
+    {
+      "epoch": 1.4348188254893794,
+      "grad_norm": 0.3610519966627758,
+      "learning_rate": 1.9534536515194312e-06,
+      "loss": 0.3026,
+      "step": 6028
+    },
+    {
+      "epoch": 1.4350568215624442,
+      "grad_norm": 0.3922292942167091,
+      "learning_rate": 1.9519254257258684e-06,
+      "loss": 0.2929,
+      "step": 6029
+    },
+    {
+      "epoch": 1.435294817635509,
+      "grad_norm": 0.4055572921879517,
+      "learning_rate": 1.950397652951593e-06,
+      "loss": 0.3556,
+      "step": 6030
+    },
+    {
+      "epoch": 1.4355328137085739,
+      "grad_norm": 0.41972072883290057,
+      "learning_rate": 1.94887033342367e-06,
+      "loss": 0.347,
+      "step": 6031
+    },
+    {
+      "epoch": 1.4357708097816386,
+      "grad_norm": 0.3783613808548642,
+      "learning_rate": 1.9473434673690974e-06,
+      "loss": 0.2675,
+      "step": 6032
+    },
+    {
+      "epoch": 1.4360088058547034,
+      "grad_norm": 0.374012703303672,
+      "learning_rate": 1.945817055014804e-06,
+      "loss": 0.3121,
+      "step": 6033
+    },
+    {
+      "epoch": 1.4362468019277683,
+      "grad_norm": 0.37927442883741,
+      "learning_rate": 1.9442910965876533e-06,
+      "loss": 0.3758,
+      "step": 6034
+    },
+    {
+      "epoch": 1.436484798000833,
+      "grad_norm": 0.38190378180904344,
+      "learning_rate": 1.94276559231444e-06,
+      "loss": 0.292,
+      "step": 6035
+    },
+    {
+      "epoch": 1.4367227940738978,
+      "grad_norm": 0.37836022258000657,
+      "learning_rate": 1.9412405424218915e-06,
+      "loss": 0.2698,
+      "step": 6036
+    },
+    {
+      "epoch": 1.4369607901469625,
+      "grad_norm": 0.38901824469541524,
+      "learning_rate": 1.9397159471366677e-06,
+      "loss": 0.3496,
+      "step": 6037
+    },
+    {
+      "epoch": 1.4371987862200273,
+      "grad_norm": 0.37161186422002496,
+      "learning_rate": 1.9381918066853632e-06,
+      "loss": 0.3693,
+      "step": 6038
+    },
+    {
+      "epoch": 1.4374367822930922,
+      "grad_norm": 0.38662177675097864,
+      "learning_rate": 1.9366681212945014e-06,
+      "loss": 0.2883,
+      "step": 6039
+    },
+    {
+      "epoch": 1.437674778366157,
+      "grad_norm": 0.3647394695355797,
+      "learning_rate": 1.9351448911905407e-06,
+      "loss": 0.2887,
+      "step": 6040
+    },
+    {
+      "epoch": 1.4379127744392217,
+      "grad_norm": 0.364210263276713,
+      "learning_rate": 1.933622116599868e-06,
+      "loss": 0.3901,
+      "step": 6041
+    },
+    {
+      "epoch": 1.4381507705122867,
+      "grad_norm": 0.4151429490488636,
+      "learning_rate": 1.9320997977488086e-06,
+      "loss": 0.3212,
+      "step": 6042
+    },
+    {
+      "epoch": 1.4383887665853514,
+      "grad_norm": 0.4095987714243739,
+      "learning_rate": 1.930577934863616e-06,
+      "loss": 0.2889,
+      "step": 6043
+    },
+    {
+      "epoch": 1.4386267626584162,
+      "grad_norm": 0.36732368664961057,
+      "learning_rate": 1.929056528170476e-06,
+      "loss": 0.324,
+      "step": 6044
+    },
+    {
+      "epoch": 1.438864758731481,
+      "grad_norm": 0.36769386073709703,
+      "learning_rate": 1.9275355778955073e-06,
+      "loss": 0.3886,
+      "step": 6045
+    },
+    {
+      "epoch": 1.4391027548045456,
+      "grad_norm": 0.3914105235529593,
+      "learning_rate": 1.926015084264761e-06,
+      "loss": 0.27,
+      "step": 6046
+    },
+    {
+      "epoch": 1.4393407508776106,
+      "grad_norm": 0.35201976589044254,
+      "learning_rate": 1.9244950475042195e-06,
+      "loss": 0.3152,
+      "step": 6047
+    },
+    {
+      "epoch": 1.4395787469506753,
+      "grad_norm": 0.37403165542953687,
+      "learning_rate": 1.922975467839799e-06,
+      "loss": 0.3402,
+      "step": 6048
+    },
+    {
+      "epoch": 1.43981674302374,
+      "grad_norm": 0.3704566985118877,
+      "learning_rate": 1.9214563454973452e-06,
+      "loss": 0.3232,
+      "step": 6049
+    },
+    {
+      "epoch": 1.440054739096805,
+      "grad_norm": 0.3591762003021762,
+      "learning_rate": 1.9199376807026383e-06,
+      "loss": 0.3001,
+      "step": 6050
+    },
+    {
+      "epoch": 1.4402927351698698,
+      "grad_norm": 0.3710069533511562,
+      "learning_rate": 1.918419473681389e-06,
+      "loss": 0.3465,
+      "step": 6051
+    },
+    {
+      "epoch": 1.4405307312429345,
+      "grad_norm": 0.40999024447887217,
+      "learning_rate": 1.9169017246592404e-06,
+      "loss": 0.4,
+      "step": 6052
+    },
+    {
+      "epoch": 1.4407687273159993,
+      "grad_norm": 0.3683061673886724,
+      "learning_rate": 1.915384433861766e-06,
+      "loss": 0.3081,
+      "step": 6053
+    },
+    {
+      "epoch": 1.441006723389064,
+      "grad_norm": 0.3822641283219383,
+      "learning_rate": 1.9138676015144765e-06,
+      "loss": 0.2872,
+      "step": 6054
+    },
+    {
+      "epoch": 1.441244719462129,
+      "grad_norm": 0.46115476067766986,
+      "learning_rate": 1.912351227842808e-06,
+      "loss": 0.3579,
+      "step": 6055
+    },
+    {
+      "epoch": 1.4414827155351937,
+      "grad_norm": 0.3663936685499609,
+      "learning_rate": 1.9108353130721326e-06,
+      "loss": 0.3425,
+      "step": 6056
+    },
+    {
+      "epoch": 1.4417207116082584,
+      "grad_norm": 0.3549319540090481,
+      "learning_rate": 1.9093198574277516e-06,
+      "loss": 0.2929,
+      "step": 6057
+    },
+    {
+      "epoch": 1.4419587076813234,
+      "grad_norm": 0.42037911174742915,
+      "learning_rate": 1.9078048611348992e-06,
+      "loss": 0.3327,
+      "step": 6058
+    },
+    {
+      "epoch": 1.4421967037543881,
+      "grad_norm": 0.3914790764090726,
+      "learning_rate": 1.9062903244187419e-06,
+      "loss": 0.3751,
+      "step": 6059
+    },
+    {
+      "epoch": 1.4424346998274529,
+      "grad_norm": 0.39191069864519656,
+      "learning_rate": 1.9047762475043775e-06,
+      "loss": 0.2849,
+      "step": 6060
+    },
+    {
+      "epoch": 1.4426726959005176,
+      "grad_norm": 0.5035295970708908,
+      "learning_rate": 1.9032626306168344e-06,
+      "loss": 0.2822,
+      "step": 6061
+    },
+    {
+      "epoch": 1.4429106919735823,
+      "grad_norm": 0.3617034512387851,
+      "learning_rate": 1.9017494739810737e-06,
+      "loss": 0.3608,
+      "step": 6062
+    },
+    {
+      "epoch": 1.4431486880466473,
+      "grad_norm": 0.3801209303019163,
+      "learning_rate": 1.9002367778219889e-06,
+      "loss": 0.3888,
+      "step": 6063
+    },
+    {
+      "epoch": 1.443386684119712,
+      "grad_norm": 0.38415720552229266,
+      "learning_rate": 1.8987245423644012e-06,
+      "loss": 0.3067,
+      "step": 6064
+    },
+    {
+      "epoch": 1.4436246801927768,
+      "grad_norm": 0.35618939435471303,
+      "learning_rate": 1.8972127678330703e-06,
+      "loss": 0.3021,
+      "step": 6065
+    },
+    {
+      "epoch": 1.4438626762658417,
+      "grad_norm": 0.39214870024334303,
+      "learning_rate": 1.8957014544526808e-06,
+      "loss": 0.3576,
+      "step": 6066
+    },
+    {
+      "epoch": 1.4441006723389065,
+      "grad_norm": 0.36937785124750994,
+      "learning_rate": 1.8941906024478524e-06,
+      "loss": 0.315,
+      "step": 6067
+    },
+    {
+      "epoch": 1.4443386684119712,
+      "grad_norm": 0.37155671677042273,
+      "learning_rate": 1.8926802120431325e-06,
+      "loss": 0.3286,
+      "step": 6068
+    },
+    {
+      "epoch": 1.444576664485036,
+      "grad_norm": 0.4021046642008438,
+      "learning_rate": 1.8911702834630063e-06,
+      "loss": 0.3203,
+      "step": 6069
+    },
+    {
+      "epoch": 1.4448146605581007,
+      "grad_norm": 0.4017487540968996,
+      "learning_rate": 1.8896608169318847e-06,
+      "loss": 0.375,
+      "step": 6070
+    },
+    {
+      "epoch": 1.4450526566311657,
+      "grad_norm": 0.36752083749151027,
+      "learning_rate": 1.8881518126741121e-06,
+      "loss": 0.2926,
+      "step": 6071
+    },
+    {
+      "epoch": 1.4452906527042304,
+      "grad_norm": 0.36577529364212397,
+      "learning_rate": 1.886643270913963e-06,
+      "loss": 0.2998,
+      "step": 6072
+    },
+    {
+      "epoch": 1.4455286487772951,
+      "grad_norm": 0.38891808139085265,
+      "learning_rate": 1.885135191875645e-06,
+      "loss": 0.3527,
+      "step": 6073
+    },
+    {
+      "epoch": 1.44576664485036,
+      "grad_norm": 0.3858655286851012,
+      "learning_rate": 1.8836275757832957e-06,
+      "loss": 0.3269,
+      "step": 6074
+    },
+    {
+      "epoch": 1.4460046409234248,
+      "grad_norm": 0.4342531875991616,
+      "learning_rate": 1.8821204228609835e-06,
+      "loss": 0.2787,
+      "step": 6075
+    },
+    {
+      "epoch": 1.4462426369964896,
+      "grad_norm": 0.38696089076302526,
+      "learning_rate": 1.880613733332709e-06,
+      "loss": 0.3401,
+      "step": 6076
+    },
+    {
+      "epoch": 1.4464806330695543,
+      "grad_norm": 0.3949331561518244,
+      "learning_rate": 1.879107507422404e-06,
+      "loss": 0.3802,
+      "step": 6077
+    },
+    {
+      "epoch": 1.446718629142619,
+      "grad_norm": 0.405406854410338,
+      "learning_rate": 1.8776017453539307e-06,
+      "loss": 0.3093,
+      "step": 6078
+    },
+    {
+      "epoch": 1.446956625215684,
+      "grad_norm": 0.37618600362179,
+      "learning_rate": 1.8760964473510823e-06,
+      "loss": 0.2896,
+      "step": 6079
+    },
+    {
+      "epoch": 1.4471946212887488,
+      "grad_norm": 0.4006013598116971,
+      "learning_rate": 1.8745916136375814e-06,
+      "loss": 0.3919,
+      "step": 6080
+    },
+    {
+      "epoch": 1.4474326173618135,
+      "grad_norm": 0.38408656575398153,
+      "learning_rate": 1.8730872444370874e-06,
+      "loss": 0.366,
+      "step": 6081
+    },
+    {
+      "epoch": 1.4476706134348785,
+      "grad_norm": 0.3948058469025689,
+      "learning_rate": 1.8715833399731854e-06,
+      "loss": 0.2908,
+      "step": 6082
+    },
+    {
+      "epoch": 1.4479086095079432,
+      "grad_norm": 0.3849530328785906,
+      "learning_rate": 1.870079900469392e-06,
+      "loss": 0.2974,
+      "step": 6083
+    },
+    {
+      "epoch": 1.448146605581008,
+      "grad_norm": 0.4378511368770222,
+      "learning_rate": 1.868576926149156e-06,
+      "loss": 0.3846,
+      "step": 6084
+    },
+    {
+      "epoch": 1.4483846016540727,
+      "grad_norm": 0.3862944035305655,
+      "learning_rate": 1.8670744172358563e-06,
+      "loss": 0.3343,
+      "step": 6085
+    },
+    {
+      "epoch": 1.4486225977271374,
+      "grad_norm": 0.4279048030239408,
+      "learning_rate": 1.865572373952803e-06,
+      "loss": 0.3044,
+      "step": 6086
+    },
+    {
+      "epoch": 1.4488605938002024,
+      "grad_norm": 0.45420594269270115,
+      "learning_rate": 1.8640707965232375e-06,
+      "loss": 0.337,
+      "step": 6087
+    },
+    {
+      "epoch": 1.4490985898732671,
+      "grad_norm": 0.38069002954177594,
+      "learning_rate": 1.86256968517033e-06,
+      "loss": 0.3705,
+      "step": 6088
+    },
+    {
+      "epoch": 1.4493365859463319,
+      "grad_norm": 0.41715068301428293,
+      "learning_rate": 1.8610690401171845e-06,
+      "loss": 0.2696,
+      "step": 6089
+    },
+    {
+      "epoch": 1.4495745820193968,
+      "grad_norm": 0.41094165450892295,
+      "learning_rate": 1.8595688615868324e-06,
+      "loss": 0.3237,
+      "step": 6090
+    },
+    {
+      "epoch": 1.4498125780924616,
+      "grad_norm": 0.40909914025067795,
+      "learning_rate": 1.8580691498022364e-06,
+      "loss": 0.3643,
+      "step": 6091
+    },
+    {
+      "epoch": 1.4500505741655263,
+      "grad_norm": 0.39406224435785253,
+      "learning_rate": 1.8565699049862939e-06,
+      "loss": 0.3093,
+      "step": 6092
+    },
+    {
+      "epoch": 1.450288570238591,
+      "grad_norm": 0.4265470858706377,
+      "learning_rate": 1.8550711273618283e-06,
+      "loss": 0.2583,
+      "step": 6093
+    },
+    {
+      "epoch": 1.4505265663116558,
+      "grad_norm": 0.4557324973621654,
+      "learning_rate": 1.8535728171515949e-06,
+      "loss": 0.3407,
+      "step": 6094
+    },
+    {
+      "epoch": 1.4507645623847207,
+      "grad_norm": 0.39977705869048036,
+      "learning_rate": 1.8520749745782784e-06,
+      "loss": 0.3649,
+      "step": 6095
+    },
+    {
+      "epoch": 1.4510025584577855,
+      "grad_norm": 0.37280847284038304,
+      "learning_rate": 1.8505775998644982e-06,
+      "loss": 0.3052,
+      "step": 6096
+    },
+    {
+      "epoch": 1.4512405545308502,
+      "grad_norm": 0.39792431548064744,
+      "learning_rate": 1.8490806932327993e-06,
+      "loss": 0.3123,
+      "step": 6097
+    },
+    {
+      "epoch": 1.4514785506039152,
+      "grad_norm": 0.42229729780036634,
+      "learning_rate": 1.8475842549056594e-06,
+      "loss": 0.3582,
+      "step": 6098
+    },
+    {
+      "epoch": 1.45171654667698,
+      "grad_norm": 0.3864779005126171,
+      "learning_rate": 1.8460882851054867e-06,
+      "loss": 0.3315,
+      "step": 6099
+    },
+    {
+      "epoch": 1.4519545427500447,
+      "grad_norm": 0.3953598919130047,
+      "learning_rate": 1.8445927840546186e-06,
+      "loss": 0.2967,
+      "step": 6100
+    },
+    {
+      "epoch": 1.4521925388231094,
+      "grad_norm": 0.38578775492268436,
+      "learning_rate": 1.8430977519753235e-06,
+      "loss": 0.3217,
+      "step": 6101
+    },
+    {
+      "epoch": 1.4524305348961741,
+      "grad_norm": 0.44987910181185625,
+      "learning_rate": 1.8416031890898006e-06,
+      "loss": 0.3781,
+      "step": 6102
+    },
+    {
+      "epoch": 1.4526685309692389,
+      "grad_norm": 0.3912127016231833,
+      "learning_rate": 1.840109095620179e-06,
+      "loss": 0.2981,
+      "step": 6103
+    },
+    {
+      "epoch": 1.4529065270423038,
+      "grad_norm": 0.4302943166340117,
+      "learning_rate": 1.838615471788518e-06,
+      "loss": 0.2725,
+      "step": 6104
+    },
+    {
+      "epoch": 1.4531445231153686,
+      "grad_norm": 0.43080639650484565,
+      "learning_rate": 1.8371223178168063e-06,
+      "loss": 0.3699,
+      "step": 6105
+    },
+    {
+      "epoch": 1.4533825191884333,
+      "grad_norm": 0.4082319577973362,
+      "learning_rate": 1.8356296339269635e-06,
+      "loss": 0.3588,
+      "step": 6106
+    },
+    {
+      "epoch": 1.4536205152614983,
+      "grad_norm": 0.36934438218455085,
+      "learning_rate": 1.8341374203408407e-06,
+      "loss": 0.2699,
+      "step": 6107
+    },
+    {
+      "epoch": 1.453858511334563,
+      "grad_norm": 0.3802597728212302,
+      "learning_rate": 1.8326456772802148e-06,
+      "loss": 0.3429,
+      "step": 6108
+    },
+    {
+      "epoch": 1.4540965074076277,
+      "grad_norm": 0.3935734960848936,
+      "learning_rate": 1.8311544049668001e-06,
+      "loss": 0.4122,
+      "step": 6109
+    },
+    {
+      "epoch": 1.4543345034806925,
+      "grad_norm": 0.4192793931557721,
+      "learning_rate": 1.8296636036222338e-06,
+      "loss": 0.2778,
+      "step": 6110
+    },
+    {
+      "epoch": 1.4545724995537572,
+      "grad_norm": 0.383099877153263,
+      "learning_rate": 1.8281732734680863e-06,
+      "loss": 0.2909,
+      "step": 6111
+    },
+    {
+      "epoch": 1.4548104956268222,
+      "grad_norm": 0.5516280401082402,
+      "learning_rate": 1.8266834147258577e-06,
+      "loss": 0.3406,
+      "step": 6112
+    },
+    {
+      "epoch": 1.455048491699887,
+      "grad_norm": 0.3873783424168311,
+      "learning_rate": 1.8251940276169777e-06,
+      "loss": 0.3955,
+      "step": 6113
+    },
+    {
+      "epoch": 1.4552864877729517,
+      "grad_norm": 0.38380201175264,
+      "learning_rate": 1.8237051123628057e-06,
+      "loss": 0.2968,
+      "step": 6114
+    },
+    {
+      "epoch": 1.4555244838460166,
+      "grad_norm": 0.39275514580382104,
+      "learning_rate": 1.8222166691846321e-06,
+      "loss": 0.2943,
+      "step": 6115
+    },
+    {
+      "epoch": 1.4557624799190814,
+      "grad_norm": 0.40228420689226446,
+      "learning_rate": 1.8207286983036765e-06,
+      "loss": 0.3527,
+      "step": 6116
+    },
+    {
+      "epoch": 1.456000475992146,
+      "grad_norm": 0.37618752702614194,
+      "learning_rate": 1.819241199941087e-06,
+      "loss": 0.3262,
+      "step": 6117
+    },
+    {
+      "epoch": 1.4562384720652108,
+      "grad_norm": 0.37547398305167284,
+      "learning_rate": 1.8177541743179423e-06,
+      "loss": 0.2852,
+      "step": 6118
+    },
+    {
+      "epoch": 1.4564764681382756,
+      "grad_norm": 0.41032362127204985,
+      "learning_rate": 1.8162676216552533e-06,
+      "loss": 0.3307,
+      "step": 6119
+    },
+    {
+      "epoch": 1.4567144642113405,
+      "grad_norm": 0.4347197449287449,
+      "learning_rate": 1.8147815421739578e-06,
+      "loss": 0.3799,
+      "step": 6120
+    },
+    {
+      "epoch": 1.4569524602844053,
+      "grad_norm": 0.3563768966230964,
+      "learning_rate": 1.8132959360949237e-06,
+      "loss": 0.2712,
+      "step": 6121
+    },
+    {
+      "epoch": 1.45719045635747,
+      "grad_norm": 0.38101919344310853,
+      "learning_rate": 1.811810803638947e-06,
+      "loss": 0.279,
+      "step": 6122
+    },
+    {
+      "epoch": 1.457428452430535,
+      "grad_norm": 0.418475439520036,
+      "learning_rate": 1.810326145026759e-06,
+      "loss": 0.3644,
+      "step": 6123
+    },
+    {
+      "epoch": 1.4576664485035997,
+      "grad_norm": 0.34723819719324317,
+      "learning_rate": 1.8088419604790135e-06,
+      "loss": 0.313,
+      "step": 6124
+    },
+    {
+      "epoch": 1.4579044445766645,
+      "grad_norm": 0.40905964751759843,
+      "learning_rate": 1.807358250216299e-06,
+      "loss": 0.2901,
+      "step": 6125
+    },
+    {
+      "epoch": 1.4581424406497292,
+      "grad_norm": 0.39190988886703376,
+      "learning_rate": 1.8058750144591308e-06,
+      "loss": 0.3643,
+      "step": 6126
+    },
+    {
+      "epoch": 1.458380436722794,
+      "grad_norm": 0.39412625081899383,
+      "learning_rate": 1.804392253427954e-06,
+      "loss": 0.4072,
+      "step": 6127
+    },
+    {
+      "epoch": 1.458618432795859,
+      "grad_norm": 0.38625582558289245,
+      "learning_rate": 1.8029099673431438e-06,
+      "loss": 0.3247,
+      "step": 6128
+    },
+    {
+      "epoch": 1.4588564288689236,
+      "grad_norm": 0.37374977195081105,
+      "learning_rate": 1.8014281564250046e-06,
+      "loss": 0.2899,
+      "step": 6129
+    },
+    {
+      "epoch": 1.4590944249419884,
+      "grad_norm": 0.4022256905018728,
+      "learning_rate": 1.7999468208937698e-06,
+      "loss": 0.3349,
+      "step": 6130
+    },
+    {
+      "epoch": 1.4593324210150533,
+      "grad_norm": 0.3782717385013846,
+      "learning_rate": 1.7984659609696037e-06,
+      "loss": 0.3659,
+      "step": 6131
+    },
+    {
+      "epoch": 1.459570417088118,
+      "grad_norm": 0.42371774399832857,
+      "learning_rate": 1.7969855768725973e-06,
+      "loss": 0.2867,
+      "step": 6132
+    },
+    {
+      "epoch": 1.4598084131611828,
+      "grad_norm": 0.4828644226623123,
+      "learning_rate": 1.7955056688227735e-06,
+      "loss": 0.3071,
+      "step": 6133
+    },
+    {
+      "epoch": 1.4600464092342476,
+      "grad_norm": 0.3711693938332572,
+      "learning_rate": 1.7940262370400823e-06,
+      "loss": 0.3749,
+      "step": 6134
+    },
+    {
+      "epoch": 1.4602844053073123,
+      "grad_norm": 0.38061515908607924,
+      "learning_rate": 1.792547281744403e-06,
+      "loss": 0.2909,
+      "step": 6135
+    },
+    {
+      "epoch": 1.4605224013803773,
+      "grad_norm": 0.39151517351766074,
+      "learning_rate": 1.7910688031555473e-06,
+      "loss": 0.2842,
+      "step": 6136
+    },
+    {
+      "epoch": 1.460760397453442,
+      "grad_norm": 0.41546568095239644,
+      "learning_rate": 1.7895908014932529e-06,
+      "loss": 0.3413,
+      "step": 6137
+    },
+    {
+      "epoch": 1.4609983935265067,
+      "grad_norm": 0.40057642081942424,
+      "learning_rate": 1.788113276977187e-06,
+      "loss": 0.3942,
+      "step": 6138
+    },
+    {
+      "epoch": 1.4612363895995717,
+      "grad_norm": 0.4025786155182079,
+      "learning_rate": 1.7866362298269468e-06,
+      "loss": 0.287,
+      "step": 6139
+    },
+    {
+      "epoch": 1.4614743856726364,
+      "grad_norm": 0.36206986714191153,
+      "learning_rate": 1.7851596602620568e-06,
+      "loss": 0.3139,
+      "step": 6140
+    },
+    {
+      "epoch": 1.4617123817457012,
+      "grad_norm": 0.4078012441791886,
+      "learning_rate": 1.7836835685019732e-06,
+      "loss": 0.37,
+      "step": 6141
+    },
+    {
+      "epoch": 1.461950377818766,
+      "grad_norm": 0.36946996756792294,
+      "learning_rate": 1.7822079547660792e-06,
+      "loss": 0.3086,
+      "step": 6142
+    },
+    {
+      "epoch": 1.4621883738918307,
+      "grad_norm": 0.3817427648802179,
+      "learning_rate": 1.7807328192736872e-06,
+      "loss": 0.2633,
+      "step": 6143
+    },
+    {
+      "epoch": 1.4624263699648956,
+      "grad_norm": 0.38265305398727073,
+      "learning_rate": 1.7792581622440392e-06,
+      "loss": 0.3307,
+      "step": 6144
+    },
+    {
+      "epoch": 1.4626643660379604,
+      "grad_norm": 0.5748853333539239,
+      "learning_rate": 1.777783983896304e-06,
+      "loss": 0.3755,
+      "step": 6145
+    },
+    {
+      "epoch": 1.462902362111025,
+      "grad_norm": 0.39571210226781933,
+      "learning_rate": 1.7763102844495838e-06,
+      "loss": 0.3066,
+      "step": 6146
+    },
+    {
+      "epoch": 1.46314035818409,
+      "grad_norm": 0.39702192397444136,
+      "learning_rate": 1.7748370641229063e-06,
+      "loss": 0.2969,
+      "step": 6147
+    },
+    {
+      "epoch": 1.4633783542571548,
+      "grad_norm": 0.3736958018420523,
+      "learning_rate": 1.773364323135227e-06,
+      "loss": 0.3624,
+      "step": 6148
+    },
+    {
+      "epoch": 1.4636163503302195,
+      "grad_norm": 0.3853561768534024,
+      "learning_rate": 1.7718920617054313e-06,
+      "loss": 0.359,
+      "step": 6149
+    },
+    {
+      "epoch": 1.4638543464032843,
+      "grad_norm": 0.3882989341472717,
+      "learning_rate": 1.7704202800523362e-06,
+      "loss": 0.267,
+      "step": 6150
+    },
+    {
+      "epoch": 1.464092342476349,
+      "grad_norm": 0.40235971375885304,
+      "learning_rate": 1.768948978394684e-06,
+      "loss": 0.3269,
+      "step": 6151
+    },
+    {
+      "epoch": 1.464330338549414,
+      "grad_norm": 0.41400386307178666,
+      "learning_rate": 1.7674781569511451e-06,
+      "loss": 0.3767,
+      "step": 6152
+    },
+    {
+      "epoch": 1.4645683346224787,
+      "grad_norm": 0.4015051161422788,
+      "learning_rate": 1.766007815940321e-06,
+      "loss": 0.2967,
+      "step": 6153
+    },
+    {
+      "epoch": 1.4648063306955434,
+      "grad_norm": 0.39178112671184995,
+      "learning_rate": 1.7645379555807408e-06,
+      "loss": 0.276,
+      "step": 6154
+    },
+    {
+      "epoch": 1.4650443267686084,
+      "grad_norm": 0.4115792252224875,
+      "learning_rate": 1.7630685760908623e-06,
+      "loss": 0.3404,
+      "step": 6155
+    },
+    {
+      "epoch": 1.4652823228416731,
+      "grad_norm": 0.38556483981697237,
+      "learning_rate": 1.7615996776890704e-06,
+      "loss": 0.3528,
+      "step": 6156
+    },
+    {
+      "epoch": 1.4655203189147379,
+      "grad_norm": 0.37944105872349776,
+      "learning_rate": 1.760131260593681e-06,
+      "loss": 0.2946,
+      "step": 6157
+    },
+    {
+      "epoch": 1.4657583149878026,
+      "grad_norm": 0.3902900142096629,
+      "learning_rate": 1.7586633250229368e-06,
+      "loss": 0.3307,
+      "step": 6158
+    },
+    {
+      "epoch": 1.4659963110608674,
+      "grad_norm": 0.3814146994358725,
+      "learning_rate": 1.7571958711950088e-06,
+      "loss": 0.3736,
+      "step": 6159
+    },
+    {
+      "epoch": 1.4662343071339323,
+      "grad_norm": 0.37047847418626345,
+      "learning_rate": 1.7557288993279981e-06,
+      "loss": 0.3168,
+      "step": 6160
+    },
+    {
+      "epoch": 1.466472303206997,
+      "grad_norm": 0.38122549662040817,
+      "learning_rate": 1.754262409639932e-06,
+      "loss": 0.2881,
+      "step": 6161
+    },
+    {
+      "epoch": 1.4667102992800618,
+      "grad_norm": 0.3933736686295024,
+      "learning_rate": 1.7527964023487676e-06,
+      "loss": 0.3072,
+      "step": 6162
+    },
+    {
+      "epoch": 1.4669482953531268,
+      "grad_norm": 0.4116114307272517,
+      "learning_rate": 1.751330877672388e-06,
+      "loss": 0.3761,
+      "step": 6163
+    },
+    {
+      "epoch": 1.4671862914261915,
+      "grad_norm": 0.36687857623003073,
+      "learning_rate": 1.7498658358286098e-06,
+      "loss": 0.285,
+      "step": 6164
+    },
+    {
+      "epoch": 1.4674242874992562,
+      "grad_norm": 0.35502119768687834,
+      "learning_rate": 1.7484012770351732e-06,
+      "loss": 0.327,
+      "step": 6165
+    },
+    {
+      "epoch": 1.467662283572321,
+      "grad_norm": 0.38055825453365044,
+      "learning_rate": 1.7469372015097469e-06,
+      "loss": 0.4053,
+      "step": 6166
+    },
+    {
+      "epoch": 1.4679002796453857,
+      "grad_norm": 0.37110944722737155,
+      "learning_rate": 1.7454736094699298e-06,
+      "loss": 0.2759,
+      "step": 6167
+    },
+    {
+      "epoch": 1.4681382757184507,
+      "grad_norm": 0.4369470705323925,
+      "learning_rate": 1.7440105011332476e-06,
+      "loss": 0.2671,
+      "step": 6168
+    },
+    {
+      "epoch": 1.4683762717915154,
+      "grad_norm": 0.5041602732328805,
+      "learning_rate": 1.7425478767171539e-06,
+      "loss": 0.3163,
+      "step": 6169
+    },
+    {
+      "epoch": 1.4686142678645802,
+      "grad_norm": 0.38125323959910473,
+      "learning_rate": 1.741085736439031e-06,
+      "loss": 0.3758,
+      "step": 6170
+    },
+    {
+      "epoch": 1.4688522639376451,
+      "grad_norm": 0.4598903890600633,
+      "learning_rate": 1.7396240805161896e-06,
+      "loss": 0.3144,
+      "step": 6171
+    },
+    {
+      "epoch": 1.4690902600107099,
+      "grad_norm": 0.38337527005033567,
+      "learning_rate": 1.7381629091658664e-06,
+      "loss": 0.288,
+      "step": 6172
+    },
+    {
+      "epoch": 1.4693282560837746,
+      "grad_norm": 0.37602121389981175,
+      "learning_rate": 1.7367022226052299e-06,
+      "loss": 0.3685,
+      "step": 6173
+    },
+    {
+      "epoch": 1.4695662521568393,
+      "grad_norm": 0.3515901765862798,
+      "learning_rate": 1.7352420210513732e-06,
+      "loss": 0.3525,
+      "step": 6174
+    },
+    {
+      "epoch": 1.469804248229904,
+      "grad_norm": 0.37679635447891247,
+      "learning_rate": 1.7337823047213186e-06,
+      "loss": 0.269,
+      "step": 6175
+    },
+    {
+      "epoch": 1.470042244302969,
+      "grad_norm": 0.421007003539755,
+      "learning_rate": 1.7323230738320162e-06,
+      "loss": 0.3184,
+      "step": 6176
+    },
+    {
+      "epoch": 1.4702802403760338,
+      "grad_norm": 0.3953495895644738,
+      "learning_rate": 1.7308643286003412e-06,
+      "loss": 0.4181,
+      "step": 6177
+    },
+    {
+      "epoch": 1.4705182364490985,
+      "grad_norm": 0.3904183739455049,
+      "learning_rate": 1.7294060692431035e-06,
+      "loss": 0.3224,
+      "step": 6178
+    },
+    {
+      "epoch": 1.4707562325221635,
+      "grad_norm": 0.3958574641622456,
+      "learning_rate": 1.7279482959770345e-06,
+      "loss": 0.281,
+      "step": 6179
+    },
+    {
+      "epoch": 1.4709942285952282,
+      "grad_norm": 0.3657347135001612,
+      "learning_rate": 1.7264910090187952e-06,
+      "loss": 0.3256,
+      "step": 6180
+    },
+    {
+      "epoch": 1.471232224668293,
+      "grad_norm": 0.36459178634963607,
+      "learning_rate": 1.7250342085849747e-06,
+      "loss": 0.3699,
+      "step": 6181
+    },
+    {
+      "epoch": 1.4714702207413577,
+      "grad_norm": 0.35909114336445935,
+      "learning_rate": 1.72357789489209e-06,
+      "loss": 0.2993,
+      "step": 6182
+    },
+    {
+      "epoch": 1.4717082168144224,
+      "grad_norm": 0.35376404383488347,
+      "learning_rate": 1.7221220681565842e-06,
+      "loss": 0.3323,
+      "step": 6183
+    },
+    {
+      "epoch": 1.4719462128874874,
+      "grad_norm": 0.39465896136944995,
+      "learning_rate": 1.7206667285948303e-06,
+      "loss": 0.3739,
+      "step": 6184
+    },
+    {
+      "epoch": 1.4721842089605521,
+      "grad_norm": 0.365373583943615,
+      "learning_rate": 1.7192118764231276e-06,
+      "loss": 0.2824,
+      "step": 6185
+    },
+    {
+      "epoch": 1.4724222050336169,
+      "grad_norm": 0.3970412374830905,
+      "learning_rate": 1.7177575118577022e-06,
+      "loss": 0.2851,
+      "step": 6186
+    },
+    {
+      "epoch": 1.4726602011066818,
+      "grad_norm": 0.38867388441314077,
+      "learning_rate": 1.7163036351147094e-06,
+      "loss": 0.368,
+      "step": 6187
+    },
+    {
+      "epoch": 1.4728981971797466,
+      "grad_norm": 0.3726608425142561,
+      "learning_rate": 1.7148502464102312e-06,
+      "loss": 0.3686,
+      "step": 6188
+    },
+    {
+      "epoch": 1.4731361932528113,
+      "grad_norm": 0.3895210104408985,
+      "learning_rate": 1.7133973459602776e-06,
+      "loss": 0.2775,
+      "step": 6189
+    },
+    {
+      "epoch": 1.473374189325876,
+      "grad_norm": 0.4279705014335874,
+      "learning_rate": 1.7119449339807825e-06,
+      "loss": 0.3314,
+      "step": 6190
+    },
+    {
+      "epoch": 1.4736121853989408,
+      "grad_norm": 0.376456866460005,
+      "learning_rate": 1.710493010687615e-06,
+      "loss": 0.3556,
+      "step": 6191
+    },
+    {
+      "epoch": 1.4738501814720057,
+      "grad_norm": 0.3605199356919443,
+      "learning_rate": 1.7090415762965646e-06,
+      "loss": 0.3045,
+      "step": 6192
+    },
+    {
+      "epoch": 1.4740881775450705,
+      "grad_norm": 0.44055911897196376,
+      "learning_rate": 1.7075906310233503e-06,
+      "loss": 0.292,
+      "step": 6193
+    },
+    {
+      "epoch": 1.4743261736181352,
+      "grad_norm": 0.3901396740042231,
+      "learning_rate": 1.7061401750836182e-06,
+      "loss": 0.3442,
+      "step": 6194
+    },
+    {
+      "epoch": 1.4745641696912002,
+      "grad_norm": 0.3987599004644487,
+      "learning_rate": 1.7046902086929428e-06,
+      "loss": 0.4148,
+      "step": 6195
+    },
+    {
+      "epoch": 1.474802165764265,
+      "grad_norm": 0.37091763674215245,
+      "learning_rate": 1.7032407320668243e-06,
+      "loss": 0.2732,
+      "step": 6196
+    },
+    {
+      "epoch": 1.4750401618373297,
+      "grad_norm": 0.347095443573286,
+      "learning_rate": 1.7017917454206905e-06,
+      "loss": 0.2876,
+      "step": 6197
+    },
+    {
+      "epoch": 1.4752781579103944,
+      "grad_norm": 0.3759182793149618,
+      "learning_rate": 1.7003432489698974e-06,
+      "loss": 0.3734,
+      "step": 6198
+    },
+    {
+      "epoch": 1.4755161539834591,
+      "grad_norm": 0.4063498607269111,
+      "learning_rate": 1.698895242929725e-06,
+      "loss": 0.3454,
+      "step": 6199
+    },
+    {
+      "epoch": 1.475754150056524,
+      "grad_norm": 0.3813713594888871,
+      "learning_rate": 1.6974477275153872e-06,
+      "loss": 0.2854,
+      "step": 6200
+    },
+    {
+      "epoch": 1.4759921461295888,
+      "grad_norm": 0.35289582689259275,
+      "learning_rate": 1.696000702942018e-06,
+      "loss": 0.3151,
+      "step": 6201
+    },
+    {
+      "epoch": 1.4762301422026536,
+      "grad_norm": 0.38603206935555157,
+      "learning_rate": 1.6945541694246809e-06,
+      "loss": 0.3949,
+      "step": 6202
+    },
+    {
+      "epoch": 1.4764681382757185,
+      "grad_norm": 0.35892990933497454,
+      "learning_rate": 1.6931081271783679e-06,
+      "loss": 0.2862,
+      "step": 6203
+    },
+    {
+      "epoch": 1.4767061343487833,
+      "grad_norm": 0.3831559472250615,
+      "learning_rate": 1.6916625764179934e-06,
+      "loss": 0.2722,
+      "step": 6204
+    },
+    {
+      "epoch": 1.476944130421848,
+      "grad_norm": 0.40027720342137013,
+      "learning_rate": 1.6902175173584062e-06,
+      "loss": 0.3442,
+      "step": 6205
+    },
+    {
+      "epoch": 1.4771821264949128,
+      "grad_norm": 0.383928566968621,
+      "learning_rate": 1.6887729502143762e-06,
+      "loss": 0.3227,
+      "step": 6206
+    },
+    {
+      "epoch": 1.4774201225679775,
+      "grad_norm": 0.4035104381552861,
+      "learning_rate": 1.6873288752006013e-06,
+      "loss": 0.2994,
+      "step": 6207
+    },
+    {
+      "epoch": 1.4776581186410425,
+      "grad_norm": 0.37119371517468214,
+      "learning_rate": 1.685885292531707e-06,
+      "loss": 0.333,
+      "step": 6208
+    },
+    {
+      "epoch": 1.4778961147141072,
+      "grad_norm": 0.3897446214278183,
+      "learning_rate": 1.6844422024222462e-06,
+      "loss": 0.3858,
+      "step": 6209
+    },
+    {
+      "epoch": 1.478134110787172,
+      "grad_norm": 0.36090692256055146,
+      "learning_rate": 1.6829996050866965e-06,
+      "loss": 0.2898,
+      "step": 6210
+    },
+    {
+      "epoch": 1.478372106860237,
+      "grad_norm": 0.3903554712089631,
+      "learning_rate": 1.6815575007394641e-06,
+      "loss": 0.2851,
+      "step": 6211
+    },
+    {
+      "epoch": 1.4786101029333016,
+      "grad_norm": 0.38668090625918494,
+      "learning_rate": 1.6801158895948816e-06,
+      "loss": 0.3617,
+      "step": 6212
+    },
+    {
+      "epoch": 1.4788480990063664,
+      "grad_norm": 0.3934627308345188,
+      "learning_rate": 1.6786747718672076e-06,
+      "loss": 0.3588,
+      "step": 6213
+    },
+    {
+      "epoch": 1.4790860950794311,
+      "grad_norm": 0.43569448317897996,
+      "learning_rate": 1.6772341477706284e-06,
+      "loss": 0.2777,
+      "step": 6214
+    },
+    {
+      "epoch": 1.4793240911524959,
+      "grad_norm": 0.4089467347728445,
+      "learning_rate": 1.675794017519256e-06,
+      "loss": 0.3008,
+      "step": 6215
+    },
+    {
+      "epoch": 1.4795620872255608,
+      "grad_norm": 0.3882878487464565,
+      "learning_rate": 1.6743543813271296e-06,
+      "loss": 0.3329,
+      "step": 6216
+    },
+    {
+      "epoch": 1.4798000832986256,
+      "grad_norm": 0.37303839591336113,
+      "learning_rate": 1.6729152394082144e-06,
+      "loss": 0.302,
+      "step": 6217
+    },
+    {
+      "epoch": 1.4800380793716903,
+      "grad_norm": 0.38100751920843007,
+      "learning_rate": 1.6714765919764015e-06,
+      "loss": 0.2819,
+      "step": 6218
+    },
+    {
+      "epoch": 1.4802760754447553,
+      "grad_norm": 0.3845807803382066,
+      "learning_rate": 1.6700384392455122e-06,
+      "loss": 0.3177,
+      "step": 6219
+    },
+    {
+      "epoch": 1.48051407151782,
+      "grad_norm": 0.34937828068409094,
+      "learning_rate": 1.6686007814292898e-06,
+      "loss": 0.371,
+      "step": 6220
+    },
+    {
+      "epoch": 1.4807520675908847,
+      "grad_norm": 0.3465525601796606,
+      "learning_rate": 1.6671636187414065e-06,
+      "loss": 0.2809,
+      "step": 6221
+    },
+    {
+      "epoch": 1.4809900636639495,
+      "grad_norm": 0.37505512040887995,
+      "learning_rate": 1.66572695139546e-06,
+      "loss": 0.3043,
+      "step": 6222
+    },
+    {
+      "epoch": 1.4812280597370142,
+      "grad_norm": 0.3849866579968192,
+      "learning_rate": 1.664290779604974e-06,
+      "loss": 0.3555,
+      "step": 6223
+    },
+    {
+      "epoch": 1.4814660558100792,
+      "grad_norm": 0.38249173168240747,
+      "learning_rate": 1.6628551035833995e-06,
+      "loss": 0.3256,
+      "step": 6224
+    },
+    {
+      "epoch": 1.481704051883144,
+      "grad_norm": 0.4133019140360971,
+      "learning_rate": 1.6614199235441141e-06,
+      "loss": 0.2854,
+      "step": 6225
+    },
+    {
+      "epoch": 1.4819420479562087,
+      "grad_norm": 0.3804351372256457,
+      "learning_rate": 1.6599852397004184e-06,
+      "loss": 0.3361,
+      "step": 6226
+    },
+    {
+      "epoch": 1.4821800440292736,
+      "grad_norm": 0.38603923989851835,
+      "learning_rate": 1.6585510522655463e-06,
+      "loss": 0.4,
+      "step": 6227
+    },
+    {
+      "epoch": 1.4824180401023384,
+      "grad_norm": 0.3518414478466677,
+      "learning_rate": 1.657117361452651e-06,
+      "loss": 0.2999,
+      "step": 6228
+    },
+    {
+      "epoch": 1.482656036175403,
+      "grad_norm": 0.4067824455417079,
+      "learning_rate": 1.6556841674748148e-06,
+      "loss": 0.2751,
+      "step": 6229
+    },
+    {
+      "epoch": 1.4828940322484678,
+      "grad_norm": 0.3888377616545728,
+      "learning_rate": 1.6542514705450453e-06,
+      "loss": 0.3871,
+      "step": 6230
+    },
+    {
+      "epoch": 1.4831320283215326,
+      "grad_norm": 0.37104379881621813,
+      "learning_rate": 1.6528192708762775e-06,
+      "loss": 0.3358,
+      "step": 6231
+    },
+    {
+      "epoch": 1.4833700243945975,
+      "grad_norm": 0.37321076058841374,
+      "learning_rate": 1.6513875686813696e-06,
+      "loss": 0.2893,
+      "step": 6232
+    },
+    {
+      "epoch": 1.4836080204676623,
+      "grad_norm": 0.40249229900762473,
+      "learning_rate": 1.6499563641731115e-06,
+      "loss": 0.3353,
+      "step": 6233
+    },
+    {
+      "epoch": 1.483846016540727,
+      "grad_norm": 0.39795621117297714,
+      "learning_rate": 1.6485256575642133e-06,
+      "loss": 0.4186,
+      "step": 6234
+    },
+    {
+      "epoch": 1.484084012613792,
+      "grad_norm": 0.3719720131348448,
+      "learning_rate": 1.6470954490673141e-06,
+      "loss": 0.3079,
+      "step": 6235
+    },
+    {
+      "epoch": 1.4843220086868567,
+      "grad_norm": 0.4065939918654111,
+      "learning_rate": 1.6456657388949782e-06,
+      "loss": 0.2705,
+      "step": 6236
+    },
+    {
+      "epoch": 1.4845600047599214,
+      "grad_norm": 0.38543792010700656,
+      "learning_rate": 1.6442365272596955e-06,
+      "loss": 0.3343,
+      "step": 6237
+    },
+    {
+      "epoch": 1.4847980008329862,
+      "grad_norm": 0.364965468199743,
+      "learning_rate": 1.6428078143738828e-06,
+      "loss": 0.3702,
+      "step": 6238
+    },
+    {
+      "epoch": 1.485035996906051,
+      "grad_norm": 0.39025088197310315,
+      "learning_rate": 1.6413796004498816e-06,
+      "loss": 0.2871,
+      "step": 6239
+    },
+    {
+      "epoch": 1.4852739929791159,
+      "grad_norm": 0.38934407019438205,
+      "learning_rate": 1.6399518856999597e-06,
+      "loss": 0.3311,
+      "step": 6240
+    },
+    {
+      "epoch": 1.4855119890521806,
+      "grad_norm": 0.4045158386339914,
+      "learning_rate": 1.6385246703363117e-06,
+      "loss": 0.3784,
+      "step": 6241
+    },
+    {
+      "epoch": 1.4857499851252454,
+      "grad_norm": 0.36882815626153603,
+      "learning_rate": 1.6370979545710564e-06,
+      "loss": 0.3222,
+      "step": 6242
+    },
+    {
+      "epoch": 1.4859879811983103,
+      "grad_norm": 0.38031964694752574,
+      "learning_rate": 1.6356717386162392e-06,
+      "loss": 0.297,
+      "step": 6243
+    },
+    {
+      "epoch": 1.486225977271375,
+      "grad_norm": 0.3535105134819767,
+      "learning_rate": 1.634246022683831e-06,
+      "loss": 0.3139,
+      "step": 6244
+    },
+    {
+      "epoch": 1.4864639733444398,
+      "grad_norm": 0.45655045384221404,
+      "learning_rate": 1.6328208069857288e-06,
+      "loss": 0.4003,
+      "step": 6245
+    },
+    {
+      "epoch": 1.4867019694175045,
+      "grad_norm": 0.358001681330835,
+      "learning_rate": 1.631396091733753e-06,
+      "loss": 0.2827,
+      "step": 6246
+    },
+    {
+      "epoch": 1.4869399654905693,
+      "grad_norm": 0.39972769951910697,
+      "learning_rate": 1.6299718771396544e-06,
+      "loss": 0.32,
+      "step": 6247
+    },
+    {
+      "epoch": 1.4871779615636342,
+      "grad_norm": 0.37405419443502563,
+      "learning_rate": 1.6285481634151057e-06,
+      "loss": 0.3557,
+      "step": 6248
+    },
+    {
+      "epoch": 1.487415957636699,
+      "grad_norm": 0.36391998904722717,
+      "learning_rate": 1.6271249507717058e-06,
+      "loss": 0.3482,
+      "step": 6249
+    },
+    {
+      "epoch": 1.4876539537097637,
+      "grad_norm": 0.36597440907805356,
+      "learning_rate": 1.6257022394209787e-06,
+      "loss": 0.2612,
+      "step": 6250
+    },
+    {
+      "epoch": 1.4878919497828287,
+      "grad_norm": 0.37262756229567356,
+      "learning_rate": 1.6242800295743755e-06,
+      "loss": 0.3512,
+      "step": 6251
+    },
+    {
+      "epoch": 1.4881299458558934,
+      "grad_norm": 0.393961403942417,
+      "learning_rate": 1.6228583214432708e-06,
+      "loss": 0.3568,
+      "step": 6252
+    },
+    {
+      "epoch": 1.4883679419289582,
+      "grad_norm": 0.3800387502035991,
+      "learning_rate": 1.6214371152389646e-06,
+      "loss": 0.2948,
+      "step": 6253
+    },
+    {
+      "epoch": 1.488605938002023,
+      "grad_norm": 0.4078216630673394,
+      "learning_rate": 1.6200164111726857e-06,
+      "loss": 0.2829,
+      "step": 6254
+    },
+    {
+      "epoch": 1.4888439340750876,
+      "grad_norm": 0.4060042960242336,
+      "learning_rate": 1.6185962094555857e-06,
+      "loss": 0.335,
+      "step": 6255
+    },
+    {
+      "epoch": 1.4890819301481526,
+      "grad_norm": 0.37150604082715344,
+      "learning_rate": 1.6171765102987401e-06,
+      "loss": 0.3452,
+      "step": 6256
+    },
+    {
+      "epoch": 1.4893199262212173,
+      "grad_norm": 0.4109956321194422,
+      "learning_rate": 1.6157573139131527e-06,
+      "loss": 0.2697,
+      "step": 6257
+    },
+    {
+      "epoch": 1.489557922294282,
+      "grad_norm": 0.3579171265280181,
+      "learning_rate": 1.6143386205097506e-06,
+      "loss": 0.3006,
+      "step": 6258
+    },
+    {
+      "epoch": 1.489795918367347,
+      "grad_norm": 0.4053809117913336,
+      "learning_rate": 1.6129204302993845e-06,
+      "loss": 0.3799,
+      "step": 6259
+    },
+    {
+      "epoch": 1.4900339144404118,
+      "grad_norm": 0.3824402887929443,
+      "learning_rate": 1.611502743492837e-06,
+      "loss": 0.3014,
+      "step": 6260
+    },
+    {
+      "epoch": 1.4902719105134765,
+      "grad_norm": 0.37518205183540226,
+      "learning_rate": 1.6100855603008087e-06,
+      "loss": 0.3104,
+      "step": 6261
+    },
+    {
+      "epoch": 1.4905099065865413,
+      "grad_norm": 0.4420001474069447,
+      "learning_rate": 1.6086688809339291e-06,
+      "loss": 0.3232,
+      "step": 6262
+    },
+    {
+      "epoch": 1.490747902659606,
+      "grad_norm": 0.3790341363906722,
+      "learning_rate": 1.6072527056027509e-06,
+      "loss": 0.4002,
+      "step": 6263
+    },
+    {
+      "epoch": 1.490985898732671,
+      "grad_norm": 0.3871942676533004,
+      "learning_rate": 1.6058370345177531e-06,
+      "loss": 0.2823,
+      "step": 6264
+    },
+    {
+      "epoch": 1.4912238948057357,
+      "grad_norm": 0.39208286818203664,
+      "learning_rate": 1.6044218678893398e-06,
+      "loss": 0.3084,
+      "step": 6265
+    },
+    {
+      "epoch": 1.4914618908788004,
+      "grad_norm": 0.4084685412861594,
+      "learning_rate": 1.6030072059278396e-06,
+      "loss": 0.382,
+      "step": 6266
+    },
+    {
+      "epoch": 1.4916998869518654,
+      "grad_norm": 0.399630334749065,
+      "learning_rate": 1.6015930488435055e-06,
+      "loss": 0.3122,
+      "step": 6267
+    },
+    {
+      "epoch": 1.4919378830249301,
+      "grad_norm": 0.3744742890214211,
+      "learning_rate": 1.6001793968465173e-06,
+      "loss": 0.2994,
+      "step": 6268
+    },
+    {
+      "epoch": 1.4921758790979949,
+      "grad_norm": 0.4074445031836467,
+      "learning_rate": 1.5987662501469787e-06,
+      "loss": 0.3392,
+      "step": 6269
+    },
+    {
+      "epoch": 1.4924138751710596,
+      "grad_norm": 0.3701010877824389,
+      "learning_rate": 1.5973536089549174e-06,
+      "loss": 0.3836,
+      "step": 6270
+    },
+    {
+      "epoch": 1.4926518712441244,
+      "grad_norm": 0.34247020644927706,
+      "learning_rate": 1.595941473480287e-06,
+      "loss": 0.2838,
+      "step": 6271
+    },
+    {
+      "epoch": 1.4928898673171893,
+      "grad_norm": 0.4149376790144497,
+      "learning_rate": 1.594529843932966e-06,
+      "loss": 0.2893,
+      "step": 6272
+    },
+    {
+      "epoch": 1.493127863390254,
+      "grad_norm": 0.37778290292253097,
+      "learning_rate": 1.593118720522756e-06,
+      "loss": 0.3544,
+      "step": 6273
+    },
+    {
+      "epoch": 1.4933658594633188,
+      "grad_norm": 0.38515242574627634,
+      "learning_rate": 1.591708103459388e-06,
+      "loss": 0.3389,
+      "step": 6274
+    },
+    {
+      "epoch": 1.4936038555363838,
+      "grad_norm": 0.392490049369985,
+      "learning_rate": 1.590297992952513e-06,
+      "loss": 0.2979,
+      "step": 6275
+    },
+    {
+      "epoch": 1.4938418516094485,
+      "grad_norm": 0.4109028404933468,
+      "learning_rate": 1.588888389211708e-06,
+      "loss": 0.3001,
+      "step": 6276
+    },
+    {
+      "epoch": 1.4940798476825132,
+      "grad_norm": 0.37566084684050377,
+      "learning_rate": 1.587479292446475e-06,
+      "loss": 0.3806,
+      "step": 6277
+    },
+    {
+      "epoch": 1.494317843755578,
+      "grad_norm": 0.3770457435748563,
+      "learning_rate": 1.5860707028662415e-06,
+      "loss": 0.2925,
+      "step": 6278
+    },
+    {
+      "epoch": 1.4945558398286427,
+      "grad_norm": 0.40003813495891827,
+      "learning_rate": 1.5846626206803572e-06,
+      "loss": 0.317,
+      "step": 6279
+    },
+    {
+      "epoch": 1.4947938359017077,
+      "grad_norm": 0.38923435537963624,
+      "learning_rate": 1.5832550460980978e-06,
+      "loss": 0.3441,
+      "step": 6280
+    },
+    {
+      "epoch": 1.4950318319747724,
+      "grad_norm": 0.3630845762326767,
+      "learning_rate": 1.5818479793286663e-06,
+      "loss": 0.3697,
+      "step": 6281
+    },
+    {
+      "epoch": 1.4952698280478371,
+      "grad_norm": 0.38811524683296966,
+      "learning_rate": 1.5804414205811864e-06,
+      "loss": 0.2647,
+      "step": 6282
+    },
+    {
+      "epoch": 1.495507824120902,
+      "grad_norm": 0.36922187210672575,
+      "learning_rate": 1.5790353700647066e-06,
+      "loss": 0.3284,
+      "step": 6283
+    },
+    {
+      "epoch": 1.4957458201939668,
+      "grad_norm": 0.400379309159919,
+      "learning_rate": 1.5776298279882018e-06,
+      "loss": 0.3806,
+      "step": 6284
+    },
+    {
+      "epoch": 1.4959838162670316,
+      "grad_norm": 0.38009298047601064,
+      "learning_rate": 1.5762247945605696e-06,
+      "loss": 0.2978,
+      "step": 6285
+    },
+    {
+      "epoch": 1.4962218123400963,
+      "grad_norm": 0.3884850826229188,
+      "learning_rate": 1.5748202699906335e-06,
+      "loss": 0.2626,
+      "step": 6286
+    },
+    {
+      "epoch": 1.496459808413161,
+      "grad_norm": 0.43122694564171327,
+      "learning_rate": 1.5734162544871379e-06,
+      "loss": 0.349,
+      "step": 6287
+    },
+    {
+      "epoch": 1.496697804486226,
+      "grad_norm": 0.37117494480049085,
+      "learning_rate": 1.5720127482587582e-06,
+      "loss": 0.3717,
+      "step": 6288
+    },
+    {
+      "epoch": 1.4969358005592908,
+      "grad_norm": 0.6961660701599369,
+      "learning_rate": 1.5706097515140888e-06,
+      "loss": 0.2725,
+      "step": 6289
+    },
+    {
+      "epoch": 1.4971737966323555,
+      "grad_norm": 0.39992955711015266,
+      "learning_rate": 1.5692072644616497e-06,
+      "loss": 0.3149,
+      "step": 6290
+    },
+    {
+      "epoch": 1.4974117927054205,
+      "grad_norm": 0.3844509561461545,
+      "learning_rate": 1.5678052873098843e-06,
+      "loss": 0.3416,
+      "step": 6291
+    },
+    {
+      "epoch": 1.4976497887784852,
+      "grad_norm": 0.4114588280934326,
+      "learning_rate": 1.5664038202671616e-06,
+      "loss": 0.2867,
+      "step": 6292
+    },
+    {
+      "epoch": 1.49788778485155,
+      "grad_norm": 0.3838890205158945,
+      "learning_rate": 1.565002863541774e-06,
+      "loss": 0.2827,
+      "step": 6293
+    },
+    {
+      "epoch": 1.4981257809246147,
+      "grad_norm": 0.4048312301464947,
+      "learning_rate": 1.5636024173419389e-06,
+      "loss": 0.3504,
+      "step": 6294
+    },
+    {
+      "epoch": 1.4983637769976794,
+      "grad_norm": 0.3751861323780018,
+      "learning_rate": 1.562202481875797e-06,
+      "loss": 0.362,
+      "step": 6295
+    },
+    {
+      "epoch": 1.4986017730707444,
+      "grad_norm": 0.37261348585719195,
+      "learning_rate": 1.5608030573514131e-06,
+      "loss": 0.2872,
+      "step": 6296
+    },
+    {
+      "epoch": 1.4988397691438091,
+      "grad_norm": 0.6348899760877189,
+      "learning_rate": 1.5594041439767772e-06,
+      "loss": 0.2771,
+      "step": 6297
+    },
+    {
+      "epoch": 1.4990777652168739,
+      "grad_norm": 0.38641372515323824,
+      "learning_rate": 1.5580057419598011e-06,
+      "loss": 0.352,
+      "step": 6298
+    },
+    {
+      "epoch": 1.4993157612899388,
+      "grad_norm": 0.39465392512377706,
+      "learning_rate": 1.5566078515083227e-06,
+      "loss": 0.3409,
+      "step": 6299
+    },
+    {
+      "epoch": 1.4995537573630036,
+      "grad_norm": 0.3627033946670975,
+      "learning_rate": 1.5552104728301031e-06,
+      "loss": 0.2711,
+      "step": 6300
+    },
+    {
+      "epoch": 1.4997917534360683,
+      "grad_norm": 0.35357386216875875,
+      "learning_rate": 1.5538136061328256e-06,
+      "loss": 0.3098,
+      "step": 6301
+    },
+    {
+      "epoch": 1.500029749509133,
+      "grad_norm": 0.401206280542586,
+      "learning_rate": 1.552417251624102e-06,
+      "loss": 0.3832,
+      "step": 6302
+    },
+    {
+      "epoch": 1.5002677455821978,
+      "grad_norm": 0.35052024645889873,
+      "learning_rate": 1.5510214095114644e-06,
+      "loss": 0.2957,
+      "step": 6303
+    },
+    {
+      "epoch": 1.5005057416552627,
+      "grad_norm": 0.4270002745679526,
+      "learning_rate": 1.5496260800023688e-06,
+      "loss": 0.3093,
+      "step": 6304
+    },
+    {
+      "epoch": 1.5007437377283275,
+      "grad_norm": 0.39472985597101196,
+      "learning_rate": 1.5482312633041956e-06,
+      "loss": 0.3537,
+      "step": 6305
+    },
+    {
+      "epoch": 1.5009817338013924,
+      "grad_norm": 0.3586093492162438,
+      "learning_rate": 1.54683695962425e-06,
+      "loss": 0.3202,
+      "step": 6306
+    },
+    {
+      "epoch": 1.5012197298744572,
+      "grad_norm": 0.4005505431292118,
+      "learning_rate": 1.5454431691697575e-06,
+      "loss": 0.2976,
+      "step": 6307
+    },
+    {
+      "epoch": 1.501457725947522,
+      "grad_norm": 0.35924720446659664,
+      "learning_rate": 1.5440498921478726e-06,
+      "loss": 0.2957,
+      "step": 6308
+    },
+    {
+      "epoch": 1.5016957220205867,
+      "grad_norm": 0.396240213306404,
+      "learning_rate": 1.5426571287656705e-06,
+      "loss": 0.3937,
+      "step": 6309
+    },
+    {
+      "epoch": 1.5019337180936514,
+      "grad_norm": 0.44063755653514214,
+      "learning_rate": 1.5412648792301494e-06,
+      "loss": 0.3058,
+      "step": 6310
+    },
+    {
+      "epoch": 1.5021717141667161,
+      "grad_norm": 0.39238660452322954,
+      "learning_rate": 1.5398731437482322e-06,
+      "loss": 0.283,
+      "step": 6311
+    },
+    {
+      "epoch": 1.502409710239781,
+      "grad_norm": 0.3882700159407384,
+      "learning_rate": 1.538481922526765e-06,
+      "loss": 0.3765,
+      "step": 6312
+    },
+    {
+      "epoch": 1.5026477063128458,
+      "grad_norm": 0.39130343929171285,
+      "learning_rate": 1.5370912157725183e-06,
+      "loss": 0.3601,
+      "step": 6313
+    },
+    {
+      "epoch": 1.5028857023859108,
+      "grad_norm": 0.3727457835348889,
+      "learning_rate": 1.5357010236921853e-06,
+      "loss": 0.308,
+      "step": 6314
+    },
+    {
+      "epoch": 1.5031236984589755,
+      "grad_norm": 0.38525499619958226,
+      "learning_rate": 1.5343113464923808e-06,
+      "loss": 0.3076,
+      "step": 6315
+    },
+    {
+      "epoch": 1.5033616945320403,
+      "grad_norm": 0.3731034146144273,
+      "learning_rate": 1.5329221843796492e-06,
+      "loss": 0.3616,
+      "step": 6316
+    },
+    {
+      "epoch": 1.503599690605105,
+      "grad_norm": 0.38290221481731973,
+      "learning_rate": 1.5315335375604523e-06,
+      "loss": 0.2957,
+      "step": 6317
+    },
+    {
+      "epoch": 1.5038376866781697,
+      "grad_norm": 0.3851034270150811,
+      "learning_rate": 1.5301454062411769e-06,
+      "loss": 0.3072,
+      "step": 6318
+    },
+    {
+      "epoch": 1.5040756827512345,
+      "grad_norm": 0.4023568178915069,
+      "learning_rate": 1.5287577906281347e-06,
+      "loss": 0.3343,
+      "step": 6319
+    },
+    {
+      "epoch": 1.5043136788242994,
+      "grad_norm": 0.47865734964101286,
+      "learning_rate": 1.5273706909275593e-06,
+      "loss": 0.3684,
+      "step": 6320
+    },
+    {
+      "epoch": 1.5045516748973642,
+      "grad_norm": 0.36513450339658127,
+      "learning_rate": 1.5259841073456078e-06,
+      "loss": 0.3044,
+      "step": 6321
+    },
+    {
+      "epoch": 1.5047896709704291,
+      "grad_norm": 0.37595196535771713,
+      "learning_rate": 1.5245980400883609e-06,
+      "loss": 0.3099,
+      "step": 6322
+    },
+    {
+      "epoch": 1.5050276670434939,
+      "grad_norm": 0.386737994428304,
+      "learning_rate": 1.5232124893618228e-06,
+      "loss": 0.3566,
+      "step": 6323
+    },
+    {
+      "epoch": 1.5052656631165586,
+      "grad_norm": 0.341571020509951,
+      "learning_rate": 1.5218274553719198e-06,
+      "loss": 0.3182,
+      "step": 6324
+    },
+    {
+      "epoch": 1.5055036591896234,
+      "grad_norm": 0.3790719754813889,
+      "learning_rate": 1.5204429383245034e-06,
+      "loss": 0.2723,
+      "step": 6325
+    },
+    {
+      "epoch": 1.505741655262688,
+      "grad_norm": 0.3997178109826318,
+      "learning_rate": 1.5190589384253458e-06,
+      "loss": 0.2983,
+      "step": 6326
+    },
+    {
+      "epoch": 1.5059796513357528,
+      "grad_norm": 0.36641685229169446,
+      "learning_rate": 1.5176754558801448e-06,
+      "loss": 0.4068,
+      "step": 6327
+    },
+    {
+      "epoch": 1.5062176474088178,
+      "grad_norm": 0.34991820225784404,
+      "learning_rate": 1.516292490894518e-06,
+      "loss": 0.2743,
+      "step": 6328
+    },
+    {
+      "epoch": 1.5064556434818825,
+      "grad_norm": 0.35331131328827264,
+      "learning_rate": 1.514910043674011e-06,
+      "loss": 0.2919,
+      "step": 6329
+    },
+    {
+      "epoch": 1.5066936395549475,
+      "grad_norm": 0.37730212152272385,
+      "learning_rate": 1.5135281144240888e-06,
+      "loss": 0.3632,
+      "step": 6330
+    },
+    {
+      "epoch": 1.5069316356280122,
+      "grad_norm": 0.3634842374888263,
+      "learning_rate": 1.5121467033501403e-06,
+      "loss": 0.3557,
+      "step": 6331
+    },
+    {
+      "epoch": 1.507169631701077,
+      "grad_norm": 0.37734949621666297,
+      "learning_rate": 1.5107658106574764e-06,
+      "loss": 0.2589,
+      "step": 6332
+    },
+    {
+      "epoch": 1.5074076277741417,
+      "grad_norm": 0.3625755897347857,
+      "learning_rate": 1.5093854365513327e-06,
+      "loss": 0.3306,
+      "step": 6333
+    },
+    {
+      "epoch": 1.5076456238472065,
+      "grad_norm": 0.3709257192300279,
+      "learning_rate": 1.5080055812368654e-06,
+      "loss": 0.3503,
+      "step": 6334
+    },
+    {
+      "epoch": 1.5078836199202712,
+      "grad_norm": 0.3697093486108935,
+      "learning_rate": 1.5066262449191576e-06,
+      "loss": 0.2987,
+      "step": 6335
+    },
+    {
+      "epoch": 1.5081216159933362,
+      "grad_norm": 0.3975396214243995,
+      "learning_rate": 1.5052474278032125e-06,
+      "loss": 0.2942,
+      "step": 6336
+    },
+    {
+      "epoch": 1.508359612066401,
+      "grad_norm": 0.36248809899098255,
+      "learning_rate": 1.5038691300939552e-06,
+      "loss": 0.3452,
+      "step": 6337
+    },
+    {
+      "epoch": 1.5085976081394659,
+      "grad_norm": 0.37957957692819194,
+      "learning_rate": 1.5024913519962353e-06,
+      "loss": 0.38,
+      "step": 6338
+    },
+    {
+      "epoch": 1.5088356042125306,
+      "grad_norm": 0.36619032508426047,
+      "learning_rate": 1.5011140937148243e-06,
+      "loss": 0.2854,
+      "step": 6339
+    },
+    {
+      "epoch": 1.5090736002855953,
+      "grad_norm": 0.36939842494901753,
+      "learning_rate": 1.4997373554544176e-06,
+      "loss": 0.2817,
+      "step": 6340
+    },
+    {
+      "epoch": 1.50931159635866,
+      "grad_norm": 0.3996275043396052,
+      "learning_rate": 1.4983611374196323e-06,
+      "loss": 0.3731,
+      "step": 6341
+    },
+    {
+      "epoch": 1.5095495924317248,
+      "grad_norm": 0.37179348365865195,
+      "learning_rate": 1.4969854398150069e-06,
+      "loss": 0.3166,
+      "step": 6342
+    },
+    {
+      "epoch": 1.5097875885047896,
+      "grad_norm": 0.4072043189526779,
+      "learning_rate": 1.4956102628450065e-06,
+      "loss": 0.2694,
+      "step": 6343
+    },
+    {
+      "epoch": 1.5100255845778545,
+      "grad_norm": 0.41059367411920217,
+      "learning_rate": 1.4942356067140162e-06,
+      "loss": 0.327,
+      "step": 6344
+    },
+    {
+      "epoch": 1.5102635806509193,
+      "grad_norm": 0.3933303176194039,
+      "learning_rate": 1.4928614716263428e-06,
+      "loss": 0.3873,
+      "step": 6345
+    },
+    {
+      "epoch": 1.5105015767239842,
+      "grad_norm": 0.38198672911717163,
+      "learning_rate": 1.4914878577862168e-06,
+      "loss": 0.2796,
+      "step": 6346
+    },
+    {
+      "epoch": 1.510739572797049,
+      "grad_norm": 0.5557986276562655,
+      "learning_rate": 1.490114765397792e-06,
+      "loss": 0.3061,
+      "step": 6347
+    },
+    {
+      "epoch": 1.5109775688701137,
+      "grad_norm": 0.40633469354717716,
+      "learning_rate": 1.4887421946651436e-06,
+      "loss": 0.3823,
+      "step": 6348
+    },
+    {
+      "epoch": 1.5112155649431784,
+      "grad_norm": 0.40308923354303,
+      "learning_rate": 1.487370145792269e-06,
+      "loss": 0.3266,
+      "step": 6349
+    },
+    {
+      "epoch": 1.5114535610162432,
+      "grad_norm": 0.38544791649854926,
+      "learning_rate": 1.4859986189830894e-06,
+      "loss": 0.2822,
+      "step": 6350
+    },
+    {
+      "epoch": 1.511691557089308,
+      "grad_norm": 0.3987203797284687,
+      "learning_rate": 1.4846276144414468e-06,
+      "loss": 0.3075,
+      "step": 6351
+    },
+    {
+      "epoch": 1.5119295531623729,
+      "grad_norm": 0.6071459510037844,
+      "learning_rate": 1.483257132371107e-06,
+      "loss": 0.4008,
+      "step": 6352
+    },
+    {
+      "epoch": 1.5121675492354376,
+      "grad_norm": 0.3813619772349361,
+      "learning_rate": 1.4818871729757572e-06,
+      "loss": 0.277,
+      "step": 6353
+    },
+    {
+      "epoch": 1.5124055453085026,
+      "grad_norm": 0.39614277089555516,
+      "learning_rate": 1.4805177364590078e-06,
+      "loss": 0.3156,
+      "step": 6354
+    },
+    {
+      "epoch": 1.5126435413815673,
+      "grad_norm": 0.360231458449255,
+      "learning_rate": 1.4791488230243907e-06,
+      "loss": 0.3506,
+      "step": 6355
+    },
+    {
+      "epoch": 1.512881537454632,
+      "grad_norm": 0.3694141952246499,
+      "learning_rate": 1.4777804328753582e-06,
+      "loss": 0.3909,
+      "step": 6356
+    },
+    {
+      "epoch": 1.5131195335276968,
+      "grad_norm": 0.3556081977766239,
+      "learning_rate": 1.47641256621529e-06,
+      "loss": 0.2827,
+      "step": 6357
+    },
+    {
+      "epoch": 1.5133575296007615,
+      "grad_norm": 0.38860260107924,
+      "learning_rate": 1.4750452232474843e-06,
+      "loss": 0.3373,
+      "step": 6358
+    },
+    {
+      "epoch": 1.5135955256738263,
+      "grad_norm": 0.38185389209509235,
+      "learning_rate": 1.4736784041751617e-06,
+      "loss": 0.4104,
+      "step": 6359
+    },
+    {
+      "epoch": 1.5138335217468912,
+      "grad_norm": 0.35687668113180804,
+      "learning_rate": 1.4723121092014654e-06,
+      "loss": 0.2966,
+      "step": 6360
+    },
+    {
+      "epoch": 1.514071517819956,
+      "grad_norm": 0.3771926399784858,
+      "learning_rate": 1.4709463385294586e-06,
+      "loss": 0.294,
+      "step": 6361
+    },
+    {
+      "epoch": 1.514309513893021,
+      "grad_norm": 0.39400534017538535,
+      "learning_rate": 1.4695810923621323e-06,
+      "loss": 0.3273,
+      "step": 6362
+    },
+    {
+      "epoch": 1.5145475099660857,
+      "grad_norm": 0.36830931162825825,
+      "learning_rate": 1.4682163709023934e-06,
+      "loss": 0.389,
+      "step": 6363
+    },
+    {
+      "epoch": 1.5147855060391504,
+      "grad_norm": 0.3829512181500474,
+      "learning_rate": 1.4668521743530745e-06,
+      "loss": 0.3118,
+      "step": 6364
+    },
+    {
+      "epoch": 1.5150235021122151,
+      "grad_norm": 0.38703933279454056,
+      "learning_rate": 1.465488502916928e-06,
+      "loss": 0.3094,
+      "step": 6365
+    },
+    {
+      "epoch": 1.5152614981852799,
+      "grad_norm": 0.3730048469300558,
+      "learning_rate": 1.4641253567966302e-06,
+      "loss": 0.38,
+      "step": 6366
+    },
+    {
+      "epoch": 1.5154994942583446,
+      "grad_norm": 0.3774200036872977,
+      "learning_rate": 1.462762736194777e-06,
+      "loss": 0.3355,
+      "step": 6367
+    },
+    {
+      "epoch": 1.5157374903314096,
+      "grad_norm": 0.4263601312122989,
+      "learning_rate": 1.4614006413138882e-06,
+      "loss": 0.2627,
+      "step": 6368
+    },
+    {
+      "epoch": 1.5159754864044743,
+      "grad_norm": 0.45738040178098543,
+      "learning_rate": 1.460039072356405e-06,
+      "loss": 0.3157,
+      "step": 6369
+    },
+    {
+      "epoch": 1.516213482477539,
+      "grad_norm": 0.3827099888179909,
+      "learning_rate": 1.4586780295246888e-06,
+      "loss": 0.3668,
+      "step": 6370
+    },
+    {
+      "epoch": 1.516451478550604,
+      "grad_norm": 0.3839304092663823,
+      "learning_rate": 1.457317513021027e-06,
+      "loss": 0.2815,
+      "step": 6371
+    },
+    {
+      "epoch": 1.5166894746236688,
+      "grad_norm": 0.3874574711568108,
+      "learning_rate": 1.455957523047624e-06,
+      "loss": 0.292,
+      "step": 6372
+    },
+    {
+      "epoch": 1.5169274706967335,
+      "grad_norm": 0.365985707188615,
+      "learning_rate": 1.454598059806609e-06,
+      "loss": 0.368,
+      "step": 6373
+    },
+    {
+      "epoch": 1.5171654667697982,
+      "grad_norm": 0.3664109124443364,
+      "learning_rate": 1.4532391235000316e-06,
+      "loss": 0.339,
+      "step": 6374
+    },
+    {
+      "epoch": 1.517403462842863,
+      "grad_norm": 0.3873829121823865,
+      "learning_rate": 1.4518807143298625e-06,
+      "loss": 0.2742,
+      "step": 6375
+    },
+    {
+      "epoch": 1.5176414589159277,
+      "grad_norm": 0.36086287426960706,
+      "learning_rate": 1.4505228324979954e-06,
+      "loss": 0.3181,
+      "step": 6376
+    },
+    {
+      "epoch": 1.5178794549889927,
+      "grad_norm": 0.3800615640914873,
+      "learning_rate": 1.449165478206246e-06,
+      "loss": 0.3744,
+      "step": 6377
+    },
+    {
+      "epoch": 1.5181174510620574,
+      "grad_norm": 0.34268205116914047,
+      "learning_rate": 1.447808651656349e-06,
+      "loss": 0.3002,
+      "step": 6378
+    },
+    {
+      "epoch": 1.5183554471351224,
+      "grad_norm": 0.4034392600733182,
+      "learning_rate": 1.4464523530499636e-06,
+      "loss": 0.2793,
+      "step": 6379
+    },
+    {
+      "epoch": 1.5185934432081871,
+      "grad_norm": 0.3778878775280351,
+      "learning_rate": 1.4450965825886693e-06,
+      "loss": 0.3479,
+      "step": 6380
+    },
+    {
+      "epoch": 1.5188314392812519,
+      "grad_norm": 0.3622997863722037,
+      "learning_rate": 1.4437413404739669e-06,
+      "loss": 0.3375,
+      "step": 6381
+    },
+    {
+      "epoch": 1.5190694353543166,
+      "grad_norm": 0.3806285094389011,
+      "learning_rate": 1.442386626907279e-06,
+      "loss": 0.2532,
+      "step": 6382
+    },
+    {
+      "epoch": 1.5193074314273813,
+      "grad_norm": 0.3702090170524379,
+      "learning_rate": 1.4410324420899475e-06,
+      "loss": 0.3195,
+      "step": 6383
+    },
+    {
+      "epoch": 1.519545427500446,
+      "grad_norm": 0.3715020539789907,
+      "learning_rate": 1.4396787862232413e-06,
+      "loss": 0.3655,
+      "step": 6384
+    },
+    {
+      "epoch": 1.519783423573511,
+      "grad_norm": 0.3683352719161142,
+      "learning_rate": 1.438325659508346e-06,
+      "loss": 0.2953,
+      "step": 6385
+    },
+    {
+      "epoch": 1.5200214196465758,
+      "grad_norm": 0.3796246049843194,
+      "learning_rate": 1.436973062146369e-06,
+      "loss": 0.2857,
+      "step": 6386
+    },
+    {
+      "epoch": 1.5202594157196407,
+      "grad_norm": 0.37688288858107755,
+      "learning_rate": 1.4356209943383386e-06,
+      "loss": 0.3469,
+      "step": 6387
+    },
+    {
+      "epoch": 1.5204974117927055,
+      "grad_norm": 0.37028336062287065,
+      "learning_rate": 1.4342694562852084e-06,
+      "loss": 0.3181,
+      "step": 6388
+    },
+    {
+      "epoch": 1.5207354078657702,
+      "grad_norm": 0.39109691442662825,
+      "learning_rate": 1.432918448187849e-06,
+      "loss": 0.2942,
+      "step": 6389
+    },
+    {
+      "epoch": 1.520973403938835,
+      "grad_norm": 0.44854038551840963,
+      "learning_rate": 1.4315679702470537e-06,
+      "loss": 0.3434,
+      "step": 6390
+    },
+    {
+      "epoch": 1.5212114000118997,
+      "grad_norm": 0.41884702455758605,
+      "learning_rate": 1.430218022663536e-06,
+      "loss": 0.3763,
+      "step": 6391
+    },
+    {
+      "epoch": 1.5214493960849644,
+      "grad_norm": 0.3594762787773168,
+      "learning_rate": 1.428868605637933e-06,
+      "loss": 0.325,
+      "step": 6392
+    },
+    {
+      "epoch": 1.5216873921580294,
+      "grad_norm": 0.4335927443090939,
+      "learning_rate": 1.4275197193708007e-06,
+      "loss": 0.2827,
+      "step": 6393
+    },
+    {
+      "epoch": 1.5219253882310941,
+      "grad_norm": 0.39709453363546243,
+      "learning_rate": 1.4261713640626168e-06,
+      "loss": 0.3365,
+      "step": 6394
+    },
+    {
+      "epoch": 1.522163384304159,
+      "grad_norm": 0.3747147101012874,
+      "learning_rate": 1.4248235399137805e-06,
+      "loss": 0.3943,
+      "step": 6395
+    },
+    {
+      "epoch": 1.5224013803772238,
+      "grad_norm": 0.48060494905748885,
+      "learning_rate": 1.4234762471246116e-06,
+      "loss": 0.3053,
+      "step": 6396
+    },
+    {
+      "epoch": 1.5226393764502886,
+      "grad_norm": 0.3908845745235385,
+      "learning_rate": 1.4221294858953499e-06,
+      "loss": 0.3409,
+      "step": 6397
+    },
+    {
+      "epoch": 1.5228773725233533,
+      "grad_norm": 0.3658107584700326,
+      "learning_rate": 1.4207832564261603e-06,
+      "loss": 0.3652,
+      "step": 6398
+    },
+    {
+      "epoch": 1.523115368596418,
+      "grad_norm": 0.3781266672514177,
+      "learning_rate": 1.419437558917125e-06,
+      "loss": 0.334,
+      "step": 6399
+    },
+    {
+      "epoch": 1.5233533646694828,
+      "grad_norm": 0.3934768910170049,
+      "learning_rate": 1.4180923935682467e-06,
+      "loss": 0.2721,
+      "step": 6400
+    },
+    {
+      "epoch": 1.5235913607425478,
+      "grad_norm": 0.42820638390694693,
+      "learning_rate": 1.4167477605794505e-06,
+      "loss": 0.2985,
+      "step": 6401
+    },
+    {
+      "epoch": 1.5238293568156125,
+      "grad_norm": 0.40845089003089985,
+      "learning_rate": 1.4154036601505834e-06,
+      "loss": 0.3593,
+      "step": 6402
+    },
+    {
+      "epoch": 1.5240673528886775,
+      "grad_norm": 0.39510953166289925,
+      "learning_rate": 1.4140600924814101e-06,
+      "loss": 0.3068,
+      "step": 6403
+    },
+    {
+      "epoch": 1.5243053489617422,
+      "grad_norm": 0.39306339830594605,
+      "learning_rate": 1.4127170577716193e-06,
+      "loss": 0.2853,
+      "step": 6404
+    },
+    {
+      "epoch": 1.524543345034807,
+      "grad_norm": 0.3633843389560684,
+      "learning_rate": 1.4113745562208191e-06,
+      "loss": 0.368,
+      "step": 6405
+    },
+    {
+      "epoch": 1.5247813411078717,
+      "grad_norm": 0.3681166528899588,
+      "learning_rate": 1.4100325880285381e-06,
+      "loss": 0.3689,
+      "step": 6406
+    },
+    {
+      "epoch": 1.5250193371809364,
+      "grad_norm": 0.3597587077669697,
+      "learning_rate": 1.4086911533942254e-06,
+      "loss": 0.2705,
+      "step": 6407
+    },
+    {
+      "epoch": 1.5252573332540011,
+      "grad_norm": 0.3678376723681916,
+      "learning_rate": 1.4073502525172528e-06,
+      "loss": 0.302,
+      "step": 6408
+    },
+    {
+      "epoch": 1.525495329327066,
+      "grad_norm": 0.3967644755270021,
+      "learning_rate": 1.4060098855969102e-06,
+      "loss": 0.3791,
+      "step": 6409
+    },
+    {
+      "epoch": 1.5257333254001308,
+      "grad_norm": 0.3765325293672744,
+      "learning_rate": 1.4046700528324082e-06,
+      "loss": 0.2981,
+      "step": 6410
+    },
+    {
+      "epoch": 1.5259713214731958,
+      "grad_norm": 0.36491847328820715,
+      "learning_rate": 1.403330754422882e-06,
+      "loss": 0.2766,
+      "step": 6411
+    },
+    {
+      "epoch": 1.5262093175462605,
+      "grad_norm": 0.38299479011732535,
+      "learning_rate": 1.401991990567383e-06,
+      "loss": 0.3489,
+      "step": 6412
+    },
+    {
+      "epoch": 1.5264473136193253,
+      "grad_norm": 0.38079523523420805,
+      "learning_rate": 1.4006537614648846e-06,
+      "loss": 0.3457,
+      "step": 6413
+    },
+    {
+      "epoch": 1.52668530969239,
+      "grad_norm": 0.3893545607621366,
+      "learning_rate": 1.3993160673142791e-06,
+      "loss": 0.2929,
+      "step": 6414
+    },
+    {
+      "epoch": 1.5269233057654548,
+      "grad_norm": 0.3650551166262176,
+      "learning_rate": 1.3979789083143847e-06,
+      "loss": 0.3031,
+      "step": 6415
+    },
+    {
+      "epoch": 1.5271613018385195,
+      "grad_norm": 0.370501273331385,
+      "learning_rate": 1.3966422846639338e-06,
+      "loss": 0.3539,
+      "step": 6416
+    },
+    {
+      "epoch": 1.5273992979115845,
+      "grad_norm": 0.37724908299456045,
+      "learning_rate": 1.3953061965615822e-06,
+      "loss": 0.3347,
+      "step": 6417
+    },
+    {
+      "epoch": 1.5276372939846492,
+      "grad_norm": 0.4060598296506883,
+      "learning_rate": 1.3939706442059054e-06,
+      "loss": 0.253,
+      "step": 6418
+    },
+    {
+      "epoch": 1.5278752900577142,
+      "grad_norm": 0.435691084317172,
+      "learning_rate": 1.3926356277954001e-06,
+      "loss": 0.3471,
+      "step": 6419
+    },
+    {
+      "epoch": 1.528113286130779,
+      "grad_norm": 0.38356363994640225,
+      "learning_rate": 1.391301147528482e-06,
+      "loss": 0.3906,
+      "step": 6420
+    },
+    {
+      "epoch": 1.5283512822038436,
+      "grad_norm": 0.3809047541613839,
+      "learning_rate": 1.389967203603489e-06,
+      "loss": 0.3001,
+      "step": 6421
+    },
+    {
+      "epoch": 1.5285892782769084,
+      "grad_norm": 0.40431301095479877,
+      "learning_rate": 1.388633796218677e-06,
+      "loss": 0.2865,
+      "step": 6422
+    },
+    {
+      "epoch": 1.5288272743499731,
+      "grad_norm": 0.3637907972443383,
+      "learning_rate": 1.3873009255722236e-06,
+      "loss": 0.3513,
+      "step": 6423
+    },
+    {
+      "epoch": 1.5290652704230379,
+      "grad_norm": 0.36132364029306596,
+      "learning_rate": 1.3859685918622269e-06,
+      "loss": 0.3441,
+      "step": 6424
+    },
+    {
+      "epoch": 1.5293032664961028,
+      "grad_norm": 0.3913993625800555,
+      "learning_rate": 1.3846367952867025e-06,
+      "loss": 0.2711,
+      "step": 6425
+    },
+    {
+      "epoch": 1.5295412625691676,
+      "grad_norm": 0.3707148982517722,
+      "learning_rate": 1.3833055360435916e-06,
+      "loss": 0.308,
+      "step": 6426
+    },
+    {
+      "epoch": 1.5297792586422325,
+      "grad_norm": 0.3729561418753125,
+      "learning_rate": 1.3819748143307509e-06,
+      "loss": 0.3893,
+      "step": 6427
+    },
+    {
+      "epoch": 1.5300172547152973,
+      "grad_norm": 0.4422339804101922,
+      "learning_rate": 1.3806446303459586e-06,
+      "loss": 0.2895,
+      "step": 6428
+    },
+    {
+      "epoch": 1.530255250788362,
+      "grad_norm": 0.3578722524768168,
+      "learning_rate": 1.3793149842869125e-06,
+      "loss": 0.3015,
+      "step": 6429
+    },
+    {
+      "epoch": 1.5304932468614267,
+      "grad_norm": 0.39833175086792916,
+      "learning_rate": 1.377985876351231e-06,
+      "loss": 0.3476,
+      "step": 6430
+    },
+    {
+      "epoch": 1.5307312429344915,
+      "grad_norm": 0.3670466020585641,
+      "learning_rate": 1.376657306736453e-06,
+      "loss": 0.3009,
+      "step": 6431
+    },
+    {
+      "epoch": 1.5309692390075562,
+      "grad_norm": 0.3668993481646055,
+      "learning_rate": 1.375329275640036e-06,
+      "loss": 0.2879,
+      "step": 6432
+    },
+    {
+      "epoch": 1.5312072350806212,
+      "grad_norm": 0.3760947174886567,
+      "learning_rate": 1.3740017832593588e-06,
+      "loss": 0.3278,
+      "step": 6433
+    },
+    {
+      "epoch": 1.531445231153686,
+      "grad_norm": 0.42040674402359185,
+      "learning_rate": 1.3726748297917196e-06,
+      "loss": 0.4024,
+      "step": 6434
+    },
+    {
+      "epoch": 1.5316832272267509,
+      "grad_norm": 0.35338872181053815,
+      "learning_rate": 1.3713484154343366e-06,
+      "loss": 0.3243,
+      "step": 6435
+    },
+    {
+      "epoch": 1.5319212232998156,
+      "grad_norm": 0.36677960202123283,
+      "learning_rate": 1.370022540384347e-06,
+      "loss": 0.2743,
+      "step": 6436
+    },
+    {
+      "epoch": 1.5321592193728804,
+      "grad_norm": 0.39245535596015746,
+      "learning_rate": 1.368697204838808e-06,
+      "loss": 0.3666,
+      "step": 6437
+    },
+    {
+      "epoch": 1.532397215445945,
+      "grad_norm": 0.4062599071898869,
+      "learning_rate": 1.3673724089947005e-06,
+      "loss": 0.3457,
+      "step": 6438
+    },
+    {
+      "epoch": 1.5326352115190098,
+      "grad_norm": 0.406451461182181,
+      "learning_rate": 1.366048153048919e-06,
+      "loss": 0.2789,
+      "step": 6439
+    },
+    {
+      "epoch": 1.5328732075920746,
+      "grad_norm": 0.41876146445384105,
+      "learning_rate": 1.364724437198282e-06,
+      "loss": 0.2907,
+      "step": 6440
+    },
+    {
+      "epoch": 1.5331112036651395,
+      "grad_norm": 0.3781185900759409,
+      "learning_rate": 1.3634012616395249e-06,
+      "loss": 0.3733,
+      "step": 6441
+    },
+    {
+      "epoch": 1.5333491997382043,
+      "grad_norm": 0.355167285944409,
+      "learning_rate": 1.3620786265693065e-06,
+      "loss": 0.2928,
+      "step": 6442
+    },
+    {
+      "epoch": 1.5335871958112692,
+      "grad_norm": 0.37322147215613916,
+      "learning_rate": 1.360756532184202e-06,
+      "loss": 0.2858,
+      "step": 6443
+    },
+    {
+      "epoch": 1.533825191884334,
+      "grad_norm": 0.3630088334381094,
+      "learning_rate": 1.3594349786807075e-06,
+      "loss": 0.3354,
+      "step": 6444
+    },
+    {
+      "epoch": 1.5340631879573987,
+      "grad_norm": 0.37666334694230696,
+      "learning_rate": 1.3581139662552384e-06,
+      "loss": 0.3768,
+      "step": 6445
+    },
+    {
+      "epoch": 1.5343011840304634,
+      "grad_norm": 0.3646570924309373,
+      "learning_rate": 1.3567934951041295e-06,
+      "loss": 0.3052,
+      "step": 6446
+    },
+    {
+      "epoch": 1.5345391801035282,
+      "grad_norm": 0.40326392321455273,
+      "learning_rate": 1.3554735654236362e-06,
+      "loss": 0.3308,
+      "step": 6447
+    },
+    {
+      "epoch": 1.534777176176593,
+      "grad_norm": 0.3811017852408969,
+      "learning_rate": 1.354154177409932e-06,
+      "loss": 0.3621,
+      "step": 6448
+    },
+    {
+      "epoch": 1.5350151722496579,
+      "grad_norm": 0.3847249344728611,
+      "learning_rate": 1.3528353312591113e-06,
+      "loss": 0.3481,
+      "step": 6449
+    },
+    {
+      "epoch": 1.5352531683227226,
+      "grad_norm": 0.40310178526788526,
+      "learning_rate": 1.3515170271671869e-06,
+      "loss": 0.2535,
+      "step": 6450
+    },
+    {
+      "epoch": 1.5354911643957876,
+      "grad_norm": 0.4142458710891034,
+      "learning_rate": 1.350199265330091e-06,
+      "loss": 0.3009,
+      "step": 6451
+    },
+    {
+      "epoch": 1.5357291604688523,
+      "grad_norm": 0.41762456749772203,
+      "learning_rate": 1.3488820459436746e-06,
+      "loss": 0.3848,
+      "step": 6452
+    },
+    {
+      "epoch": 1.535967156541917,
+      "grad_norm": 0.3546073693357267,
+      "learning_rate": 1.3475653692037121e-06,
+      "loss": 0.2921,
+      "step": 6453
+    },
+    {
+      "epoch": 1.5362051526149818,
+      "grad_norm": 0.40900895467204246,
+      "learning_rate": 1.3462492353058925e-06,
+      "loss": 0.2866,
+      "step": 6454
+    },
+    {
+      "epoch": 1.5364431486880465,
+      "grad_norm": 0.3839566221518419,
+      "learning_rate": 1.3449336444458262e-06,
+      "loss": 0.351,
+      "step": 6455
+    },
+    {
+      "epoch": 1.5366811447611113,
+      "grad_norm": 0.39130970196546844,
+      "learning_rate": 1.3436185968190424e-06,
+      "loss": 0.3163,
+      "step": 6456
+    },
+    {
+      "epoch": 1.5369191408341762,
+      "grad_norm": 0.3741648167638271,
+      "learning_rate": 1.3423040926209891e-06,
+      "loss": 0.2968,
+      "step": 6457
+    },
+    {
+      "epoch": 1.537157136907241,
+      "grad_norm": 0.37882647569434985,
+      "learning_rate": 1.3409901320470353e-06,
+      "loss": 0.3218,
+      "step": 6458
+    },
+    {
+      "epoch": 1.537395132980306,
+      "grad_norm": 0.40274124899921193,
+      "learning_rate": 1.3396767152924667e-06,
+      "loss": 0.3784,
+      "step": 6459
+    },
+    {
+      "epoch": 1.5376331290533707,
+      "grad_norm": 0.3978813236044458,
+      "learning_rate": 1.3383638425524909e-06,
+      "loss": 0.3023,
+      "step": 6460
+    },
+    {
+      "epoch": 1.5378711251264354,
+      "grad_norm": 0.38122434614979217,
+      "learning_rate": 1.3370515140222322e-06,
+      "loss": 0.2824,
+      "step": 6461
+    },
+    {
+      "epoch": 1.5381091211995002,
+      "grad_norm": 0.3763755588799363,
+      "learning_rate": 1.3357397298967356e-06,
+      "loss": 0.358,
+      "step": 6462
+    },
+    {
+      "epoch": 1.538347117272565,
+      "grad_norm": 0.37502057107131376,
+      "learning_rate": 1.3344284903709647e-06,
+      "loss": 0.3488,
+      "step": 6463
+    },
+    {
+      "epoch": 1.5385851133456296,
+      "grad_norm": 0.3851990727799265,
+      "learning_rate": 1.3331177956398007e-06,
+      "loss": 0.2791,
+      "step": 6464
+    },
+    {
+      "epoch": 1.5388231094186946,
+      "grad_norm": 0.4611773475919031,
+      "learning_rate": 1.3318076458980472e-06,
+      "loss": 0.3228,
+      "step": 6465
+    },
+    {
+      "epoch": 1.5390611054917593,
+      "grad_norm": 0.4023609502927905,
+      "learning_rate": 1.330498041340425e-06,
+      "loss": 0.3756,
+      "step": 6466
+    },
+    {
+      "epoch": 1.5392991015648243,
+      "grad_norm": 0.36953198203324916,
+      "learning_rate": 1.3291889821615728e-06,
+      "loss": 0.3158,
+      "step": 6467
+    },
+    {
+      "epoch": 1.539537097637889,
+      "grad_norm": 0.39443173542625404,
+      "learning_rate": 1.3278804685560476e-06,
+      "loss": 0.2908,
+      "step": 6468
+    },
+    {
+      "epoch": 1.5397750937109538,
+      "grad_norm": 0.3713126115765714,
+      "learning_rate": 1.3265725007183306e-06,
+      "loss": 0.3272,
+      "step": 6469
+    },
+    {
+      "epoch": 1.5400130897840185,
+      "grad_norm": 0.38677506574041237,
+      "learning_rate": 1.3252650788428156e-06,
+      "loss": 0.3867,
+      "step": 6470
+    },
+    {
+      "epoch": 1.5402510858570833,
+      "grad_norm": 0.3478181826041587,
+      "learning_rate": 1.3239582031238191e-06,
+      "loss": 0.2743,
+      "step": 6471
+    },
+    {
+      "epoch": 1.540489081930148,
+      "grad_norm": 0.3406479266488313,
+      "learning_rate": 1.3226518737555744e-06,
+      "loss": 0.3027,
+      "step": 6472
+    },
+    {
+      "epoch": 1.540727078003213,
+      "grad_norm": 0.39762042117594537,
+      "learning_rate": 1.321346090932234e-06,
+      "loss": 0.3722,
+      "step": 6473
+    },
+    {
+      "epoch": 1.5409650740762777,
+      "grad_norm": 0.37089894103258814,
+      "learning_rate": 1.3200408548478705e-06,
+      "loss": 0.3375,
+      "step": 6474
+    },
+    {
+      "epoch": 1.5412030701493427,
+      "grad_norm": 0.3965123393190166,
+      "learning_rate": 1.318736165696473e-06,
+      "loss": 0.2923,
+      "step": 6475
+    },
+    {
+      "epoch": 1.5414410662224074,
+      "grad_norm": 0.39001895537688247,
+      "learning_rate": 1.3174320236719524e-06,
+      "loss": 0.3209,
+      "step": 6476
+    },
+    {
+      "epoch": 1.5416790622954721,
+      "grad_norm": 0.39041593273975445,
+      "learning_rate": 1.3161284289681347e-06,
+      "loss": 0.4009,
+      "step": 6477
+    },
+    {
+      "epoch": 1.5419170583685369,
+      "grad_norm": 0.3757834666815788,
+      "learning_rate": 1.3148253817787671e-06,
+      "loss": 0.3065,
+      "step": 6478
+    },
+    {
+      "epoch": 1.5421550544416016,
+      "grad_norm": 0.3850973543387022,
+      "learning_rate": 1.3135228822975148e-06,
+      "loss": 0.2928,
+      "step": 6479
+    },
+    {
+      "epoch": 1.5423930505146664,
+      "grad_norm": 0.38795859626153806,
+      "learning_rate": 1.3122209307179602e-06,
+      "loss": 0.3493,
+      "step": 6480
+    },
+    {
+      "epoch": 1.5426310465877313,
+      "grad_norm": 0.3781000254192901,
+      "learning_rate": 1.3109195272336073e-06,
+      "loss": 0.3487,
+      "step": 6481
+    },
+    {
+      "epoch": 1.542869042660796,
+      "grad_norm": 0.42706004699561867,
+      "learning_rate": 1.3096186720378762e-06,
+      "loss": 0.2718,
+      "step": 6482
+    },
+    {
+      "epoch": 1.543107038733861,
+      "grad_norm": 0.4346631528907676,
+      "learning_rate": 1.308318365324106e-06,
+      "loss": 0.3182,
+      "step": 6483
+    },
+    {
+      "epoch": 1.5433450348069258,
+      "grad_norm": 0.3671857753557718,
+      "learning_rate": 1.3070186072855551e-06,
+      "loss": 0.3902,
+      "step": 6484
+    },
+    {
+      "epoch": 1.5435830308799905,
+      "grad_norm": 0.3756868316351957,
+      "learning_rate": 1.3057193981153986e-06,
+      "loss": 0.3181,
+      "step": 6485
+    },
+    {
+      "epoch": 1.5438210269530552,
+      "grad_norm": 0.3921792506523292,
+      "learning_rate": 1.304420738006732e-06,
+      "loss": 0.2774,
+      "step": 6486
+    },
+    {
+      "epoch": 1.54405902302612,
+      "grad_norm": 0.3998042980004656,
+      "learning_rate": 1.3031226271525677e-06,
+      "loss": 0.3231,
+      "step": 6487
+    },
+    {
+      "epoch": 1.5442970190991847,
+      "grad_norm": 0.40991845946955335,
+      "learning_rate": 1.3018250657458377e-06,
+      "loss": 0.3567,
+      "step": 6488
+    },
+    {
+      "epoch": 1.5445350151722497,
+      "grad_norm": 0.3688970315279264,
+      "learning_rate": 1.3005280539793908e-06,
+      "loss": 0.2933,
+      "step": 6489
+    },
+    {
+      "epoch": 1.5447730112453144,
+      "grad_norm": 0.37617632189473976,
+      "learning_rate": 1.2992315920459958e-06,
+      "loss": 0.3308,
+      "step": 6490
+    },
+    {
+      "epoch": 1.5450110073183794,
+      "grad_norm": 0.3884799402239865,
+      "learning_rate": 1.2979356801383376e-06,
+      "loss": 0.3745,
+      "step": 6491
+    },
+    {
+      "epoch": 1.545249003391444,
+      "grad_norm": 0.3545462251747847,
+      "learning_rate": 1.296640318449023e-06,
+      "loss": 0.3218,
+      "step": 6492
+    },
+    {
+      "epoch": 1.5454869994645088,
+      "grad_norm": 0.3910225423977597,
+      "learning_rate": 1.295345507170574e-06,
+      "loss": 0.2844,
+      "step": 6493
+    },
+    {
+      "epoch": 1.5457249955375736,
+      "grad_norm": 0.39920924863182033,
+      "learning_rate": 1.2940512464954313e-06,
+      "loss": 0.3485,
+      "step": 6494
+    },
+    {
+      "epoch": 1.5459629916106383,
+      "grad_norm": 0.3725385615151038,
+      "learning_rate": 1.2927575366159529e-06,
+      "loss": 0.3894,
+      "step": 6495
+    },
+    {
+      "epoch": 1.546200987683703,
+      "grad_norm": 0.35822929977999596,
+      "learning_rate": 1.291464377724419e-06,
+      "loss": 0.3083,
+      "step": 6496
+    },
+    {
+      "epoch": 1.546438983756768,
+      "grad_norm": 0.394890044687265,
+      "learning_rate": 1.290171770013023e-06,
+      "loss": 0.3128,
+      "step": 6497
+    },
+    {
+      "epoch": 1.5466769798298328,
+      "grad_norm": 0.36206101450795275,
+      "learning_rate": 1.2888797136738785e-06,
+      "loss": 0.3694,
+      "step": 6498
+    },
+    {
+      "epoch": 1.5469149759028977,
+      "grad_norm": 0.35891273995138384,
+      "learning_rate": 1.2875882088990172e-06,
+      "loss": 0.3243,
+      "step": 6499
+    },
+    {
+      "epoch": 1.5471529719759625,
+      "grad_norm": 0.39198989856107563,
+      "learning_rate": 1.2862972558803893e-06,
+      "loss": 0.2723,
+      "step": 6500
+    },
+    {
+      "epoch": 1.5473909680490272,
+      "grad_norm": 0.4058267955634325,
+      "learning_rate": 1.285006854809861e-06,
+      "loss": 0.3337,
+      "step": 6501
+    },
+    {
+      "epoch": 1.547628964122092,
+      "grad_norm": 0.38126495691581086,
+      "learning_rate": 1.2837170058792192e-06,
+      "loss": 0.374,
+      "step": 6502
+    },
+    {
+      "epoch": 1.5478669601951567,
+      "grad_norm": 0.37306157896006814,
+      "learning_rate": 1.2824277092801663e-06,
+      "loss": 0.2942,
+      "step": 6503
+    },
+    {
+      "epoch": 1.5481049562682214,
+      "grad_norm": 0.3821763777566332,
+      "learning_rate": 1.2811389652043244e-06,
+      "loss": 0.2898,
+      "step": 6504
+    },
+    {
+      "epoch": 1.5483429523412864,
+      "grad_norm": 0.38236792478815307,
+      "learning_rate": 1.2798507738432326e-06,
+      "loss": 0.3272,
+      "step": 6505
+    },
+    {
+      "epoch": 1.5485809484143511,
+      "grad_norm": 0.3844001205504733,
+      "learning_rate": 1.2785631353883476e-06,
+      "loss": 0.3261,
+      "step": 6506
+    },
+    {
+      "epoch": 1.548818944487416,
+      "grad_norm": 0.3847625058188862,
+      "learning_rate": 1.2772760500310444e-06,
+      "loss": 0.3074,
+      "step": 6507
+    },
+    {
+      "epoch": 1.5490569405604808,
+      "grad_norm": 0.4101461881227883,
+      "learning_rate": 1.2759895179626147e-06,
+      "loss": 0.3132,
+      "step": 6508
+    },
+    {
+      "epoch": 1.5492949366335456,
+      "grad_norm": 0.3736053815442064,
+      "learning_rate": 1.2747035393742718e-06,
+      "loss": 0.3962,
+      "step": 6509
+    },
+    {
+      "epoch": 1.5495329327066103,
+      "grad_norm": 0.36963876676347174,
+      "learning_rate": 1.2734181144571422e-06,
+      "loss": 0.2847,
+      "step": 6510
+    },
+    {
+      "epoch": 1.549770928779675,
+      "grad_norm": 0.42277471628828484,
+      "learning_rate": 1.272133243402272e-06,
+      "loss": 0.2858,
+      "step": 6511
+    },
+    {
+      "epoch": 1.5500089248527398,
+      "grad_norm": 0.35688249609379674,
+      "learning_rate": 1.2708489264006246e-06,
+      "loss": 0.3236,
+      "step": 6512
+    },
+    {
+      "epoch": 1.5502469209258047,
+      "grad_norm": 0.3855474832519827,
+      "learning_rate": 1.2695651636430816e-06,
+      "loss": 0.3387,
+      "step": 6513
+    },
+    {
+      "epoch": 1.5504849169988695,
+      "grad_norm": 0.36968276951753154,
+      "learning_rate": 1.2682819553204424e-06,
+      "loss": 0.2947,
+      "step": 6514
+    },
+    {
+      "epoch": 1.5507229130719344,
+      "grad_norm": 0.399145007942629,
+      "learning_rate": 1.2669993016234222e-06,
+      "loss": 0.3258,
+      "step": 6515
+    },
+    {
+      "epoch": 1.5509609091449992,
+      "grad_norm": 0.39491923210379837,
+      "learning_rate": 1.2657172027426556e-06,
+      "loss": 0.3859,
+      "step": 6516
+    },
+    {
+      "epoch": 1.551198905218064,
+      "grad_norm": 0.3614434517449473,
+      "learning_rate": 1.264435658868695e-06,
+      "loss": 0.2889,
+      "step": 6517
+    },
+    {
+      "epoch": 1.5514369012911287,
+      "grad_norm": 0.3977718991319386,
+      "learning_rate": 1.2631546701920073e-06,
+      "loss": 0.2863,
+      "step": 6518
+    },
+    {
+      "epoch": 1.5516748973641934,
+      "grad_norm": 0.493812725827731,
+      "learning_rate": 1.2618742369029819e-06,
+      "loss": 0.3471,
+      "step": 6519
+    },
+    {
+      "epoch": 1.5519128934372581,
+      "grad_norm": 0.396858346322872,
+      "learning_rate": 1.2605943591919218e-06,
+      "loss": 0.3643,
+      "step": 6520
+    },
+    {
+      "epoch": 1.552150889510323,
+      "grad_norm": 0.4006368823356171,
+      "learning_rate": 1.2593150372490482e-06,
+      "loss": 0.2958,
+      "step": 6521
+    },
+    {
+      "epoch": 1.5523888855833878,
+      "grad_norm": 0.3651666797447085,
+      "learning_rate": 1.258036271264499e-06,
+      "loss": 0.2965,
+      "step": 6522
+    },
+    {
+      "epoch": 1.5526268816564528,
+      "grad_norm": 0.3609444928045324,
+      "learning_rate": 1.2567580614283331e-06,
+      "loss": 0.3453,
+      "step": 6523
+    },
+    {
+      "epoch": 1.5528648777295175,
+      "grad_norm": 0.35866895527736053,
+      "learning_rate": 1.255480407930522e-06,
+      "loss": 0.3647,
+      "step": 6524
+    },
+    {
+      "epoch": 1.5531028738025823,
+      "grad_norm": 0.3792035303974508,
+      "learning_rate": 1.254203310960958e-06,
+      "loss": 0.3127,
+      "step": 6525
+    },
+    {
+      "epoch": 1.553340869875647,
+      "grad_norm": 0.38922874090947,
+      "learning_rate": 1.2529267707094483e-06,
+      "loss": 0.3213,
+      "step": 6526
+    },
+    {
+      "epoch": 1.5535788659487118,
+      "grad_norm": 0.38138547100943027,
+      "learning_rate": 1.2516507873657184e-06,
+      "loss": 0.3742,
+      "step": 6527
+    },
+    {
+      "epoch": 1.5538168620217765,
+      "grad_norm": 0.3818945794940767,
+      "learning_rate": 1.2503753611194113e-06,
+      "loss": 0.2942,
+      "step": 6528
+    },
+    {
+      "epoch": 1.5540548580948415,
+      "grad_norm": 0.3825385830982494,
+      "learning_rate": 1.2491004921600868e-06,
+      "loss": 0.2755,
+      "step": 6529
+    },
+    {
+      "epoch": 1.5542928541679062,
+      "grad_norm": 0.3998069859296653,
+      "learning_rate": 1.247826180677222e-06,
+      "loss": 0.3644,
+      "step": 6530
+    },
+    {
+      "epoch": 1.5545308502409712,
+      "grad_norm": 0.35741830458922036,
+      "learning_rate": 1.2465524268602107e-06,
+      "loss": 0.3762,
+      "step": 6531
+    },
+    {
+      "epoch": 1.554768846314036,
+      "grad_norm": 0.3777877889719805,
+      "learning_rate": 1.2452792308983646e-06,
+      "loss": 0.2921,
+      "step": 6532
+    },
+    {
+      "epoch": 1.5550068423871006,
+      "grad_norm": 0.3803043899122809,
+      "learning_rate": 1.2440065929809125e-06,
+      "loss": 0.3158,
+      "step": 6533
+    },
+    {
+      "epoch": 1.5552448384601654,
+      "grad_norm": 0.36524440591889934,
+      "learning_rate": 1.2427345132969997e-06,
+      "loss": 0.3631,
+      "step": 6534
+    },
+    {
+      "epoch": 1.55548283453323,
+      "grad_norm": 0.38177834729615784,
+      "learning_rate": 1.2414629920356862e-06,
+      "loss": 0.2983,
+      "step": 6535
+    },
+    {
+      "epoch": 1.5557208306062948,
+      "grad_norm": 0.3903386365033605,
+      "learning_rate": 1.2401920293859558e-06,
+      "loss": 0.2908,
+      "step": 6536
+    },
+    {
+      "epoch": 1.5559588266793598,
+      "grad_norm": 0.37546580053400613,
+      "learning_rate": 1.238921625536703e-06,
+      "loss": 0.3295,
+      "step": 6537
+    },
+    {
+      "epoch": 1.5561968227524245,
+      "grad_norm": 0.37598661605642897,
+      "learning_rate": 1.2376517806767412e-06,
+      "loss": 0.377,
+      "step": 6538
+    },
+    {
+      "epoch": 1.5564348188254895,
+      "grad_norm": 0.38383478247802943,
+      "learning_rate": 1.2363824949948012e-06,
+      "loss": 0.2775,
+      "step": 6539
+    },
+    {
+      "epoch": 1.5566728148985542,
+      "grad_norm": 0.3763981357885576,
+      "learning_rate": 1.2351137686795294e-06,
+      "loss": 0.3129,
+      "step": 6540
+    },
+    {
+      "epoch": 1.556910810971619,
+      "grad_norm": 0.40852446358280264,
+      "learning_rate": 1.2338456019194906e-06,
+      "loss": 0.3695,
+      "step": 6541
+    },
+    {
+      "epoch": 1.5571488070446837,
+      "grad_norm": 0.3695119241962966,
+      "learning_rate": 1.2325779949031653e-06,
+      "loss": 0.3202,
+      "step": 6542
+    },
+    {
+      "epoch": 1.5573868031177485,
+      "grad_norm": 0.44568416123855525,
+      "learning_rate": 1.231310947818951e-06,
+      "loss": 0.2751,
+      "step": 6543
+    },
+    {
+      "epoch": 1.5576247991908132,
+      "grad_norm": 0.39409919509614555,
+      "learning_rate": 1.2300444608551636e-06,
+      "loss": 0.3013,
+      "step": 6544
+    },
+    {
+      "epoch": 1.5578627952638782,
+      "grad_norm": 0.3742902470173008,
+      "learning_rate": 1.228778534200032e-06,
+      "loss": 0.381,
+      "step": 6545
+    },
+    {
+      "epoch": 1.558100791336943,
+      "grad_norm": 0.32924222634354267,
+      "learning_rate": 1.2275131680417069e-06,
+      "loss": 0.3009,
+      "step": 6546
+    },
+    {
+      "epoch": 1.5583387874100079,
+      "grad_norm": 0.3667969007920045,
+      "learning_rate": 1.2262483625682514e-06,
+      "loss": 0.2832,
+      "step": 6547
+    },
+    {
+      "epoch": 1.5585767834830726,
+      "grad_norm": 0.3894529324837924,
+      "learning_rate": 1.2249841179676481e-06,
+      "loss": 0.4056,
+      "step": 6548
+    },
+    {
+      "epoch": 1.5588147795561373,
+      "grad_norm": 0.36782423847690526,
+      "learning_rate": 1.2237204344277925e-06,
+      "loss": 0.3222,
+      "step": 6549
+    },
+    {
+      "epoch": 1.559052775629202,
+      "grad_norm": 0.3724914266382319,
+      "learning_rate": 1.2224573121365024e-06,
+      "loss": 0.2706,
+      "step": 6550
+    },
+    {
+      "epoch": 1.5592907717022668,
+      "grad_norm": 0.418089388727327,
+      "learning_rate": 1.2211947512815076e-06,
+      "loss": 0.3172,
+      "step": 6551
+    },
+    {
+      "epoch": 1.5595287677753316,
+      "grad_norm": 0.402369301328807,
+      "learning_rate": 1.2199327520504562e-06,
+      "loss": 0.378,
+      "step": 6552
+    },
+    {
+      "epoch": 1.5597667638483965,
+      "grad_norm": 0.42301913940674296,
+      "learning_rate": 1.218671314630912e-06,
+      "loss": 0.3409,
+      "step": 6553
+    },
+    {
+      "epoch": 1.5600047599214613,
+      "grad_norm": 0.38961962135190426,
+      "learning_rate": 1.2174104392103558e-06,
+      "loss": 0.2681,
+      "step": 6554
+    },
+    {
+      "epoch": 1.5602427559945262,
+      "grad_norm": 0.4054494324727994,
+      "learning_rate": 1.2161501259761855e-06,
+      "loss": 0.3628,
+      "step": 6555
+    },
+    {
+      "epoch": 1.560480752067591,
+      "grad_norm": 0.3373988273851839,
+      "learning_rate": 1.2148903751157144e-06,
+      "loss": 0.3361,
+      "step": 6556
+    },
+    {
+      "epoch": 1.5607187481406557,
+      "grad_norm": 0.40537913952657784,
+      "learning_rate": 1.213631186816172e-06,
+      "loss": 0.2743,
+      "step": 6557
+    },
+    {
+      "epoch": 1.5609567442137204,
+      "grad_norm": 0.38794190280853064,
+      "learning_rate": 1.2123725612647064e-06,
+      "loss": 0.327,
+      "step": 6558
+    },
+    {
+      "epoch": 1.5611947402867852,
+      "grad_norm": 0.38511311663261943,
+      "learning_rate": 1.2111144986483792e-06,
+      "loss": 0.3744,
+      "step": 6559
+    },
+    {
+      "epoch": 1.56143273635985,
+      "grad_norm": 0.3674215063731254,
+      "learning_rate": 1.2098569991541698e-06,
+      "loss": 0.3067,
+      "step": 6560
+    },
+    {
+      "epoch": 1.5616707324329149,
+      "grad_norm": 0.40312804407214897,
+      "learning_rate": 1.2086000629689743e-06,
+      "loss": 0.3093,
+      "step": 6561
+    },
+    {
+      "epoch": 1.5619087285059796,
+      "grad_norm": 0.3762916570398983,
+      "learning_rate": 1.2073436902796038e-06,
+      "loss": 0.3063,
+      "step": 6562
+    },
+    {
+      "epoch": 1.5621467245790446,
+      "grad_norm": 0.36072105739082216,
+      "learning_rate": 1.2060878812727854e-06,
+      "loss": 0.3568,
+      "step": 6563
+    },
+    {
+      "epoch": 1.5623847206521093,
+      "grad_norm": 0.354835431200888,
+      "learning_rate": 1.2048326361351658e-06,
+      "loss": 0.2666,
+      "step": 6564
+    },
+    {
+      "epoch": 1.562622716725174,
+      "grad_norm": 0.4104035256897569,
+      "learning_rate": 1.203577955053305e-06,
+      "loss": 0.3111,
+      "step": 6565
+    },
+    {
+      "epoch": 1.5628607127982388,
+      "grad_norm": 0.3512367007914632,
+      "learning_rate": 1.2023238382136787e-06,
+      "loss": 0.3713,
+      "step": 6566
+    },
+    {
+      "epoch": 1.5630987088713035,
+      "grad_norm": 0.3646761606995802,
+      "learning_rate": 1.20107028580268e-06,
+      "loss": 0.2961,
+      "step": 6567
+    },
+    {
+      "epoch": 1.5633367049443683,
+      "grad_norm": 0.35406751256411023,
+      "learning_rate": 1.199817298006618e-06,
+      "loss": 0.2711,
+      "step": 6568
+    },
+    {
+      "epoch": 1.5635747010174332,
+      "grad_norm": 0.4043510614910653,
+      "learning_rate": 1.1985648750117179e-06,
+      "loss": 0.3151,
+      "step": 6569
+    },
+    {
+      "epoch": 1.563812697090498,
+      "grad_norm": 0.3791673629759354,
+      "learning_rate": 1.1973130170041198e-06,
+      "loss": 0.4031,
+      "step": 6570
+    },
+    {
+      "epoch": 1.564050693163563,
+      "grad_norm": 0.3842915787993255,
+      "learning_rate": 1.1960617241698813e-06,
+      "loss": 0.2776,
+      "step": 6571
+    },
+    {
+      "epoch": 1.5642886892366277,
+      "grad_norm": 0.4119663617052321,
+      "learning_rate": 1.1948109966949745e-06,
+      "loss": 0.3098,
+      "step": 6572
+    },
+    {
+      "epoch": 1.5645266853096924,
+      "grad_norm": 0.3592669784242863,
+      "learning_rate": 1.1935608347652906e-06,
+      "loss": 0.3729,
+      "step": 6573
+    },
+    {
+      "epoch": 1.5647646813827571,
+      "grad_norm": 0.373131383360226,
+      "learning_rate": 1.192311238566633e-06,
+      "loss": 0.3484,
+      "step": 6574
+    },
+    {
+      "epoch": 1.5650026774558219,
+      "grad_norm": 0.3833981711887363,
+      "learning_rate": 1.1910622082847234e-06,
+      "loss": 0.2392,
+      "step": 6575
+    },
+    {
+      "epoch": 1.5652406735288866,
+      "grad_norm": 0.34671002508415816,
+      "learning_rate": 1.1898137441051982e-06,
+      "loss": 0.3245,
+      "step": 6576
+    },
+    {
+      "epoch": 1.5654786696019516,
+      "grad_norm": 0.3632990316751966,
+      "learning_rate": 1.1885658462136085e-06,
+      "loss": 0.4118,
+      "step": 6577
+    },
+    {
+      "epoch": 1.5657166656750163,
+      "grad_norm": 0.35796363702341216,
+      "learning_rate": 1.1873185147954263e-06,
+      "loss": 0.2738,
+      "step": 6578
+    },
+    {
+      "epoch": 1.5659546617480813,
+      "grad_norm": 0.3527329899780401,
+      "learning_rate": 1.1860717500360336e-06,
+      "loss": 0.2714,
+      "step": 6579
+    },
+    {
+      "epoch": 1.566192657821146,
+      "grad_norm": 0.41834053159356094,
+      "learning_rate": 1.1848255521207308e-06,
+      "loss": 0.3636,
+      "step": 6580
+    },
+    {
+      "epoch": 1.5664306538942108,
+      "grad_norm": 0.39616933163013157,
+      "learning_rate": 1.1835799212347343e-06,
+      "loss": 0.3605,
+      "step": 6581
+    },
+    {
+      "epoch": 1.5666686499672755,
+      "grad_norm": 0.37466230103769954,
+      "learning_rate": 1.1823348575631749e-06,
+      "loss": 0.2658,
+      "step": 6582
+    },
+    {
+      "epoch": 1.5669066460403402,
+      "grad_norm": 0.39963898406246245,
+      "learning_rate": 1.1810903612911001e-06,
+      "loss": 0.3117,
+      "step": 6583
+    },
+    {
+      "epoch": 1.567144642113405,
+      "grad_norm": 0.3910180293713378,
+      "learning_rate": 1.179846432603473e-06,
+      "loss": 0.3908,
+      "step": 6584
+    },
+    {
+      "epoch": 1.56738263818647,
+      "grad_norm": 0.3978347412470743,
+      "learning_rate": 1.178603071685172e-06,
+      "loss": 0.3303,
+      "step": 6585
+    },
+    {
+      "epoch": 1.5676206342595347,
+      "grad_norm": 0.4194169551330718,
+      "learning_rate": 1.1773602787209915e-06,
+      "loss": 0.2969,
+      "step": 6586
+    },
+    {
+      "epoch": 1.5678586303325996,
+      "grad_norm": 0.4026400405007115,
+      "learning_rate": 1.1761180538956408e-06,
+      "loss": 0.3348,
+      "step": 6587
+    },
+    {
+      "epoch": 1.5680966264056644,
+      "grad_norm": 0.3774639635057926,
+      "learning_rate": 1.1748763973937454e-06,
+      "loss": 0.3725,
+      "step": 6588
+    },
+    {
+      "epoch": 1.5683346224787291,
+      "grad_norm": 0.37626958019160905,
+      "learning_rate": 1.1736353093998471e-06,
+      "loss": 0.3056,
+      "step": 6589
+    },
+    {
+      "epoch": 1.5685726185517939,
+      "grad_norm": 0.3938142965328492,
+      "learning_rate": 1.1723947900983995e-06,
+      "loss": 0.2885,
+      "step": 6590
+    },
+    {
+      "epoch": 1.5688106146248586,
+      "grad_norm": 0.3988797825778564,
+      "learning_rate": 1.1711548396737782e-06,
+      "loss": 0.3832,
+      "step": 6591
+    },
+    {
+      "epoch": 1.5690486106979233,
+      "grad_norm": 0.3627863918712503,
+      "learning_rate": 1.1699154583102684e-06,
+      "loss": 0.2937,
+      "step": 6592
+    },
+    {
+      "epoch": 1.5692866067709883,
+      "grad_norm": 0.37559495224162986,
+      "learning_rate": 1.1686766461920735e-06,
+      "loss": 0.3193,
+      "step": 6593
+    },
+    {
+      "epoch": 1.569524602844053,
+      "grad_norm": 0.3864117859142998,
+      "learning_rate": 1.1674384035033115e-06,
+      "loss": 0.3263,
+      "step": 6594
+    },
+    {
+      "epoch": 1.569762598917118,
+      "grad_norm": 0.3796716007438366,
+      "learning_rate": 1.1662007304280153e-06,
+      "loss": 0.3871,
+      "step": 6595
+    },
+    {
+      "epoch": 1.5700005949901827,
+      "grad_norm": 0.3728167400610051,
+      "learning_rate": 1.1649636271501346e-06,
+      "loss": 0.2965,
+      "step": 6596
+    },
+    {
+      "epoch": 1.5702385910632475,
+      "grad_norm": 0.3719186929240788,
+      "learning_rate": 1.1637270938535334e-06,
+      "loss": 0.298,
+      "step": 6597
+    },
+    {
+      "epoch": 1.5704765871363122,
+      "grad_norm": 0.37701153994508685,
+      "learning_rate": 1.1624911307219904e-06,
+      "loss": 0.353,
+      "step": 6598
+    },
+    {
+      "epoch": 1.570714583209377,
+      "grad_norm": 0.37159186789859183,
+      "learning_rate": 1.1612557379391993e-06,
+      "loss": 0.3372,
+      "step": 6599
+    },
+    {
+      "epoch": 1.5709525792824417,
+      "grad_norm": 0.3868110826031627,
+      "learning_rate": 1.1600209156887732e-06,
+      "loss": 0.2635,
+      "step": 6600
+    },
+    {
+      "epoch": 1.5711905753555067,
+      "grad_norm": 0.3807624942737973,
+      "learning_rate": 1.158786664154235e-06,
+      "loss": 0.3153,
+      "step": 6601
+    },
+    {
+      "epoch": 1.5714285714285714,
+      "grad_norm": 0.409345535932182,
+      "learning_rate": 1.1575529835190264e-06,
+      "loss": 0.3898,
+      "step": 6602
+    },
+    {
+      "epoch": 1.5716665675016364,
+      "grad_norm": 0.3879086278028145,
+      "learning_rate": 1.1563198739665017e-06,
+      "loss": 0.2732,
+      "step": 6603
+    },
+    {
+      "epoch": 1.571904563574701,
+      "grad_norm": 0.40901261697805064,
+      "learning_rate": 1.1550873356799297e-06,
+      "loss": 0.2728,
+      "step": 6604
+    },
+    {
+      "epoch": 1.5721425596477658,
+      "grad_norm": 0.3791446730046189,
+      "learning_rate": 1.1538553688425002e-06,
+      "loss": 0.365,
+      "step": 6605
+    },
+    {
+      "epoch": 1.5723805557208306,
+      "grad_norm": 0.3662473834687694,
+      "learning_rate": 1.1526239736373118e-06,
+      "loss": 0.3693,
+      "step": 6606
+    },
+    {
+      "epoch": 1.5726185517938953,
+      "grad_norm": 0.3745667664354613,
+      "learning_rate": 1.1513931502473807e-06,
+      "loss": 0.2745,
+      "step": 6607
+    },
+    {
+      "epoch": 1.57285654786696,
+      "grad_norm": 0.3758851474104438,
+      "learning_rate": 1.1501628988556368e-06,
+      "loss": 0.3148,
+      "step": 6608
+    },
+    {
+      "epoch": 1.573094543940025,
+      "grad_norm": 0.40537121108015156,
+      "learning_rate": 1.1489332196449271e-06,
+      "loss": 0.3888,
+      "step": 6609
+    },
+    {
+      "epoch": 1.5733325400130898,
+      "grad_norm": 0.3957681771676492,
+      "learning_rate": 1.1477041127980115e-06,
+      "loss": 0.3328,
+      "step": 6610
+    },
+    {
+      "epoch": 1.5735705360861547,
+      "grad_norm": 0.4123620794488522,
+      "learning_rate": 1.146475578497566e-06,
+      "loss": 0.2634,
+      "step": 6611
+    },
+    {
+      "epoch": 1.5738085321592195,
+      "grad_norm": 0.4057487023110117,
+      "learning_rate": 1.1452476169261812e-06,
+      "loss": 0.341,
+      "step": 6612
+    },
+    {
+      "epoch": 1.5740465282322842,
+      "grad_norm": 0.39959292930319834,
+      "learning_rate": 1.144020228266362e-06,
+      "loss": 0.3469,
+      "step": 6613
+    },
+    {
+      "epoch": 1.574284524305349,
+      "grad_norm": 0.34870011465820333,
+      "learning_rate": 1.1427934127005296e-06,
+      "loss": 0.284,
+      "step": 6614
+    },
+    {
+      "epoch": 1.5745225203784137,
+      "grad_norm": 0.41846723355624454,
+      "learning_rate": 1.1415671704110193e-06,
+      "loss": 0.2993,
+      "step": 6615
+    },
+    {
+      "epoch": 1.5747605164514784,
+      "grad_norm": 0.4035291904243755,
+      "learning_rate": 1.1403415015800801e-06,
+      "loss": 0.3623,
+      "step": 6616
+    },
+    {
+      "epoch": 1.5749985125245434,
+      "grad_norm": 0.3751084543135697,
+      "learning_rate": 1.1391164063898768e-06,
+      "loss": 0.307,
+      "step": 6617
+    },
+    {
+      "epoch": 1.575236508597608,
+      "grad_norm": 0.3848206656891993,
+      "learning_rate": 1.137891885022488e-06,
+      "loss": 0.2872,
+      "step": 6618
+    },
+    {
+      "epoch": 1.575474504670673,
+      "grad_norm": 0.36097442997024354,
+      "learning_rate": 1.1366679376599104e-06,
+      "loss": 0.3278,
+      "step": 6619
+    },
+    {
+      "epoch": 1.5757125007437378,
+      "grad_norm": 0.3676488968525861,
+      "learning_rate": 1.1354445644840516e-06,
+      "loss": 0.3784,
+      "step": 6620
+    },
+    {
+      "epoch": 1.5759504968168025,
+      "grad_norm": 0.3757067541931903,
+      "learning_rate": 1.1342217656767352e-06,
+      "loss": 0.294,
+      "step": 6621
+    },
+    {
+      "epoch": 1.5761884928898673,
+      "grad_norm": 0.4248595314868793,
+      "learning_rate": 1.1329995414196986e-06,
+      "loss": 0.3104,
+      "step": 6622
+    },
+    {
+      "epoch": 1.576426488962932,
+      "grad_norm": 0.39022645581178395,
+      "learning_rate": 1.1317778918945954e-06,
+      "loss": 0.3596,
+      "step": 6623
+    },
+    {
+      "epoch": 1.5766644850359968,
+      "grad_norm": 0.37146786462426146,
+      "learning_rate": 1.1305568172829928e-06,
+      "loss": 0.3432,
+      "step": 6624
+    },
+    {
+      "epoch": 1.5769024811090617,
+      "grad_norm": 0.3957415682588482,
+      "learning_rate": 1.1293363177663724e-06,
+      "loss": 0.2897,
+      "step": 6625
+    },
+    {
+      "epoch": 1.5771404771821265,
+      "grad_norm": 0.37407424529688554,
+      "learning_rate": 1.1281163935261298e-06,
+      "loss": 0.3319,
+      "step": 6626
+    },
+    {
+      "epoch": 1.5773784732551914,
+      "grad_norm": 0.5040169043587555,
+      "learning_rate": 1.1268970447435772e-06,
+      "loss": 0.4073,
+      "step": 6627
+    },
+    {
+      "epoch": 1.5776164693282562,
+      "grad_norm": 0.3494340725959216,
+      "learning_rate": 1.1256782715999409e-06,
+      "loss": 0.2889,
+      "step": 6628
+    },
+    {
+      "epoch": 1.577854465401321,
+      "grad_norm": 0.3592534966406634,
+      "learning_rate": 1.124460074276359e-06,
+      "loss": 0.2832,
+      "step": 6629
+    },
+    {
+      "epoch": 1.5780924614743856,
+      "grad_norm": 0.4051953408007675,
+      "learning_rate": 1.1232424529538865e-06,
+      "loss": 0.3587,
+      "step": 6630
+    },
+    {
+      "epoch": 1.5783304575474504,
+      "grad_norm": 0.3796932548423937,
+      "learning_rate": 1.1220254078134919e-06,
+      "loss": 0.3441,
+      "step": 6631
+    },
+    {
+      "epoch": 1.5785684536205151,
+      "grad_norm": 0.38402120878401813,
+      "learning_rate": 1.1208089390360566e-06,
+      "loss": 0.2712,
+      "step": 6632
+    },
+    {
+      "epoch": 1.57880644969358,
+      "grad_norm": 0.3928063655919691,
+      "learning_rate": 1.1195930468023808e-06,
+      "loss": 0.3301,
+      "step": 6633
+    },
+    {
+      "epoch": 1.5790444457666448,
+      "grad_norm": 0.3791650075605935,
+      "learning_rate": 1.1183777312931748e-06,
+      "loss": 0.3712,
+      "step": 6634
+    },
+    {
+      "epoch": 1.5792824418397098,
+      "grad_norm": 0.38088863826024594,
+      "learning_rate": 1.1171629926890648e-06,
+      "loss": 0.3067,
+      "step": 6635
+    },
+    {
+      "epoch": 1.5795204379127745,
+      "grad_norm": 0.4035315968898168,
+      "learning_rate": 1.1159488311705914e-06,
+      "loss": 0.2967,
+      "step": 6636
+    },
+    {
+      "epoch": 1.5797584339858393,
+      "grad_norm": 0.39150229184884205,
+      "learning_rate": 1.1147352469182077e-06,
+      "loss": 0.3563,
+      "step": 6637
+    },
+    {
+      "epoch": 1.579996430058904,
+      "grad_norm": 0.3945787408533381,
+      "learning_rate": 1.1135222401122835e-06,
+      "loss": 0.3706,
+      "step": 6638
+    },
+    {
+      "epoch": 1.5802344261319687,
+      "grad_norm": 0.3824380546742477,
+      "learning_rate": 1.112309810933101e-06,
+      "loss": 0.3121,
+      "step": 6639
+    },
+    {
+      "epoch": 1.5804724222050335,
+      "grad_norm": 0.38647053605937204,
+      "learning_rate": 1.111097959560858e-06,
+      "loss": 0.2656,
+      "step": 6640
+    },
+    {
+      "epoch": 1.5807104182780984,
+      "grad_norm": 0.39387059548636993,
+      "learning_rate": 1.109886686175664e-06,
+      "loss": 0.3632,
+      "step": 6641
+    },
+    {
+      "epoch": 1.5809484143511632,
+      "grad_norm": 0.38920881826537457,
+      "learning_rate": 1.1086759909575462e-06,
+      "loss": 0.2883,
+      "step": 6642
+    },
+    {
+      "epoch": 1.5811864104242281,
+      "grad_norm": 0.39014824674716264,
+      "learning_rate": 1.1074658740864425e-06,
+      "loss": 0.2812,
+      "step": 6643
+    },
+    {
+      "epoch": 1.5814244064972929,
+      "grad_norm": 0.4100290719394353,
+      "learning_rate": 1.1062563357422063e-06,
+      "loss": 0.3405,
+      "step": 6644
+    },
+    {
+      "epoch": 1.5816624025703576,
+      "grad_norm": 0.379281769403233,
+      "learning_rate": 1.1050473761046038e-06,
+      "loss": 0.3978,
+      "step": 6645
+    },
+    {
+      "epoch": 1.5819003986434224,
+      "grad_norm": 0.3843586382669758,
+      "learning_rate": 1.1038389953533197e-06,
+      "loss": 0.2815,
+      "step": 6646
+    },
+    {
+      "epoch": 1.582138394716487,
+      "grad_norm": 0.37029427889320665,
+      "learning_rate": 1.102631193667947e-06,
+      "loss": 0.2894,
+      "step": 6647
+    },
+    {
+      "epoch": 1.5823763907895518,
+      "grad_norm": 0.363151439990877,
+      "learning_rate": 1.1014239712279946e-06,
+      "loss": 0.3463,
+      "step": 6648
+    },
+    {
+      "epoch": 1.5826143868626168,
+      "grad_norm": 0.3499094972795553,
+      "learning_rate": 1.100217328212887e-06,
+      "loss": 0.3472,
+      "step": 6649
+    },
+    {
+      "epoch": 1.5828523829356815,
+      "grad_norm": 0.37693181662588054,
+      "learning_rate": 1.0990112648019595e-06,
+      "loss": 0.2838,
+      "step": 6650
+    },
+    {
+      "epoch": 1.5830903790087465,
+      "grad_norm": 0.4029662389391535,
+      "learning_rate": 1.0978057811744646e-06,
+      "loss": 0.3276,
+      "step": 6651
+    },
+    {
+      "epoch": 1.5833283750818112,
+      "grad_norm": 0.4147641724374556,
+      "learning_rate": 1.096600877509566e-06,
+      "loss": 0.3633,
+      "step": 6652
+    },
+    {
+      "epoch": 1.583566371154876,
+      "grad_norm": 0.3672722330153522,
+      "learning_rate": 1.0953965539863409e-06,
+      "loss": 0.2932,
+      "step": 6653
+    },
+    {
+      "epoch": 1.5838043672279407,
+      "grad_norm": 0.3893345967844479,
+      "learning_rate": 1.0941928107837852e-06,
+      "loss": 0.2635,
+      "step": 6654
+    },
+    {
+      "epoch": 1.5840423633010055,
+      "grad_norm": 0.39379673310513574,
+      "learning_rate": 1.0929896480808022e-06,
+      "loss": 0.372,
+      "step": 6655
+    },
+    {
+      "epoch": 1.5842803593740702,
+      "grad_norm": 0.3763842161032918,
+      "learning_rate": 1.0917870660562125e-06,
+      "loss": 0.3338,
+      "step": 6656
+    },
+    {
+      "epoch": 1.5845183554471352,
+      "grad_norm": 0.42973871300326605,
+      "learning_rate": 1.090585064888749e-06,
+      "loss": 0.2549,
+      "step": 6657
+    },
+    {
+      "epoch": 1.5847563515202,
+      "grad_norm": 0.3692281668495606,
+      "learning_rate": 1.0893836447570595e-06,
+      "loss": 0.3091,
+      "step": 6658
+    },
+    {
+      "epoch": 1.5849943475932649,
+      "grad_norm": 0.4105528991354411,
+      "learning_rate": 1.0881828058397025e-06,
+      "loss": 0.3591,
+      "step": 6659
+    },
+    {
+      "epoch": 1.5852323436663296,
+      "grad_norm": 0.3506145732459315,
+      "learning_rate": 1.0869825483151563e-06,
+      "loss": 0.3442,
+      "step": 6660
+    },
+    {
+      "epoch": 1.5854703397393943,
+      "grad_norm": 0.38609627382705203,
+      "learning_rate": 1.0857828723618063e-06,
+      "loss": 0.2698,
+      "step": 6661
+    },
+    {
+      "epoch": 1.585708335812459,
+      "grad_norm": 0.39582571031621394,
+      "learning_rate": 1.084583778157955e-06,
+      "loss": 0.3305,
+      "step": 6662
+    },
+    {
+      "epoch": 1.5859463318855238,
+      "grad_norm": 0.4773499221532177,
+      "learning_rate": 1.0833852658818167e-06,
+      "loss": 0.3659,
+      "step": 6663
+    },
+    {
+      "epoch": 1.5861843279585885,
+      "grad_norm": 0.387549231721086,
+      "learning_rate": 1.0821873357115203e-06,
+      "loss": 0.2837,
+      "step": 6664
+    },
+    {
+      "epoch": 1.5864223240316535,
+      "grad_norm": 0.3684761502468937,
+      "learning_rate": 1.0809899878251078e-06,
+      "loss": 0.3226,
+      "step": 6665
+    },
+    {
+      "epoch": 1.5866603201047182,
+      "grad_norm": 0.4189733561178494,
+      "learning_rate": 1.0797932224005348e-06,
+      "loss": 0.3902,
+      "step": 6666
+    },
+    {
+      "epoch": 1.5868983161777832,
+      "grad_norm": 0.3603259475366153,
+      "learning_rate": 1.0785970396156698e-06,
+      "loss": 0.3178,
+      "step": 6667
+    },
+    {
+      "epoch": 1.587136312250848,
+      "grad_norm": 0.37689639588625584,
+      "learning_rate": 1.0774014396482962e-06,
+      "loss": 0.2916,
+      "step": 6668
+    },
+    {
+      "epoch": 1.5873743083239127,
+      "grad_norm": 0.38804261264274786,
+      "learning_rate": 1.0762064226761087e-06,
+      "loss": 0.3446,
+      "step": 6669
+    },
+    {
+      "epoch": 1.5876123043969774,
+      "grad_norm": 0.36867385287936777,
+      "learning_rate": 1.075011988876717e-06,
+      "loss": 0.3833,
+      "step": 6670
+    },
+    {
+      "epoch": 1.5878503004700422,
+      "grad_norm": 0.7076069958920685,
+      "learning_rate": 1.073818138427643e-06,
+      "loss": 0.3034,
+      "step": 6671
+    },
+    {
+      "epoch": 1.588088296543107,
+      "grad_norm": 0.37858638589930405,
+      "learning_rate": 1.0726248715063226e-06,
+      "loss": 0.2957,
+      "step": 6672
+    },
+    {
+      "epoch": 1.5883262926161719,
+      "grad_norm": 0.3866414028696196,
+      "learning_rate": 1.0714321882901036e-06,
+      "loss": 0.3701,
+      "step": 6673
+    },
+    {
+      "epoch": 1.5885642886892366,
+      "grad_norm": 0.39497519093426026,
+      "learning_rate": 1.070240088956251e-06,
+      "loss": 0.3026,
+      "step": 6674
+    },
+    {
+      "epoch": 1.5888022847623016,
+      "grad_norm": 0.37372560276384903,
+      "learning_rate": 1.069048573681939e-06,
+      "loss": 0.2698,
+      "step": 6675
+    },
+    {
+      "epoch": 1.5890402808353663,
+      "grad_norm": 0.4009991162457296,
+      "learning_rate": 1.0678576426442554e-06,
+      "loss": 0.3123,
+      "step": 6676
+    },
+    {
+      "epoch": 1.589278276908431,
+      "grad_norm": 0.39506909405987567,
+      "learning_rate": 1.0666672960202028e-06,
+      "loss": 0.3798,
+      "step": 6677
+    },
+    {
+      "epoch": 1.5895162729814958,
+      "grad_norm": 0.3524389433062458,
+      "learning_rate": 1.0654775339866963e-06,
+      "loss": 0.2702,
+      "step": 6678
+    },
+    {
+      "epoch": 1.5897542690545605,
+      "grad_norm": 0.38272669069986726,
+      "learning_rate": 1.0642883567205635e-06,
+      "loss": 0.28,
+      "step": 6679
+    },
+    {
+      "epoch": 1.5899922651276253,
+      "grad_norm": 0.40293966435907647,
+      "learning_rate": 1.0630997643985447e-06,
+      "loss": 0.3393,
+      "step": 6680
+    },
+    {
+      "epoch": 1.5902302612006902,
+      "grad_norm": 0.3587703813764963,
+      "learning_rate": 1.061911757197296e-06,
+      "loss": 0.3498,
+      "step": 6681
+    },
+    {
+      "epoch": 1.590468257273755,
+      "grad_norm": 0.38543802046712716,
+      "learning_rate": 1.0607243352933838e-06,
+      "loss": 0.2903,
+      "step": 6682
+    },
+    {
+      "epoch": 1.59070625334682,
+      "grad_norm": 0.4301083842782692,
+      "learning_rate": 1.0595374988632884e-06,
+      "loss": 0.3182,
+      "step": 6683
+    },
+    {
+      "epoch": 1.5909442494198847,
+      "grad_norm": 0.40034262172534774,
+      "learning_rate": 1.058351248083403e-06,
+      "loss": 0.3613,
+      "step": 6684
+    },
+    {
+      "epoch": 1.5911822454929494,
+      "grad_norm": 0.382673941851363,
+      "learning_rate": 1.0571655831300342e-06,
+      "loss": 0.326,
+      "step": 6685
+    },
+    {
+      "epoch": 1.5914202415660141,
+      "grad_norm": 0.41204904076387927,
+      "learning_rate": 1.0559805041794002e-06,
+      "loss": 0.2805,
+      "step": 6686
+    },
+    {
+      "epoch": 1.5916582376390789,
+      "grad_norm": 0.38921959287793867,
+      "learning_rate": 1.0547960114076328e-06,
+      "loss": 0.3514,
+      "step": 6687
+    },
+    {
+      "epoch": 1.5918962337121436,
+      "grad_norm": 0.40306688375436,
+      "learning_rate": 1.053612104990779e-06,
+      "loss": 0.3668,
+      "step": 6688
+    },
+    {
+      "epoch": 1.5921342297852086,
+      "grad_norm": 0.4133630297518753,
+      "learning_rate": 1.052428785104795e-06,
+      "loss": 0.2802,
+      "step": 6689
+    },
+    {
+      "epoch": 1.5923722258582733,
+      "grad_norm": 0.3962023255094279,
+      "learning_rate": 1.051246051925552e-06,
+      "loss": 0.316,
+      "step": 6690
+    },
+    {
+      "epoch": 1.5926102219313383,
+      "grad_norm": 0.41861934380314547,
+      "learning_rate": 1.0500639056288325e-06,
+      "loss": 0.3799,
+      "step": 6691
+    },
+    {
+      "epoch": 1.592848218004403,
+      "grad_norm": 0.3766382477058585,
+      "learning_rate": 1.0488823463903341e-06,
+      "loss": 0.3612,
+      "step": 6692
+    },
+    {
+      "epoch": 1.5930862140774678,
+      "grad_norm": 0.3881147115564832,
+      "learning_rate": 1.0477013743856652e-06,
+      "loss": 0.274,
+      "step": 6693
+    },
+    {
+      "epoch": 1.5933242101505325,
+      "grad_norm": 0.3621383112707997,
+      "learning_rate": 1.0465209897903466e-06,
+      "loss": 0.3322,
+      "step": 6694
+    },
+    {
+      "epoch": 1.5935622062235972,
+      "grad_norm": 0.3636848700106123,
+      "learning_rate": 1.0453411927798136e-06,
+      "loss": 0.3889,
+      "step": 6695
+    },
+    {
+      "epoch": 1.593800202296662,
+      "grad_norm": 0.37473055964276525,
+      "learning_rate": 1.0441619835294125e-06,
+      "loss": 0.2783,
+      "step": 6696
+    },
+    {
+      "epoch": 1.594038198369727,
+      "grad_norm": 0.37206515088751896,
+      "learning_rate": 1.0429833622144037e-06,
+      "loss": 0.2903,
+      "step": 6697
+    },
+    {
+      "epoch": 1.5942761944427917,
+      "grad_norm": 0.39577591456358313,
+      "learning_rate": 1.0418053290099589e-06,
+      "loss": 0.3918,
+      "step": 6698
+    },
+    {
+      "epoch": 1.5945141905158566,
+      "grad_norm": 0.37165220358822043,
+      "learning_rate": 1.0406278840911632e-06,
+      "loss": 0.3808,
+      "step": 6699
+    },
+    {
+      "epoch": 1.5947521865889214,
+      "grad_norm": 0.3930220555144249,
+      "learning_rate": 1.0394510276330145e-06,
+      "loss": 0.2849,
+      "step": 6700
+    },
+    {
+      "epoch": 1.5949901826619861,
+      "grad_norm": 0.4069004575367408,
+      "learning_rate": 1.038274759810421e-06,
+      "loss": 0.3189,
+      "step": 6701
+    },
+    {
+      "epoch": 1.5952281787350509,
+      "grad_norm": 0.39687917691298413,
+      "learning_rate": 1.0370990807982073e-06,
+      "loss": 0.3705,
+      "step": 6702
+    },
+    {
+      "epoch": 1.5954661748081156,
+      "grad_norm": 0.3910720271352218,
+      "learning_rate": 1.0359239907711082e-06,
+      "loss": 0.2893,
+      "step": 6703
+    },
+    {
+      "epoch": 1.5957041708811803,
+      "grad_norm": 0.37980335830006934,
+      "learning_rate": 1.0347494899037703e-06,
+      "loss": 0.2815,
+      "step": 6704
+    },
+    {
+      "epoch": 1.5959421669542453,
+      "grad_norm": 0.39726445474059047,
+      "learning_rate": 1.033575578370754e-06,
+      "loss": 0.3508,
+      "step": 6705
+    },
+    {
+      "epoch": 1.59618016302731,
+      "grad_norm": 0.37676377876897194,
+      "learning_rate": 1.0324022563465307e-06,
+      "loss": 0.3562,
+      "step": 6706
+    },
+    {
+      "epoch": 1.596418159100375,
+      "grad_norm": 0.3593315132381354,
+      "learning_rate": 1.0312295240054853e-06,
+      "loss": 0.2756,
+      "step": 6707
+    },
+    {
+      "epoch": 1.5966561551734397,
+      "grad_norm": 0.37080438768334384,
+      "learning_rate": 1.030057381521916e-06,
+      "loss": 0.3206,
+      "step": 6708
+    },
+    {
+      "epoch": 1.5968941512465045,
+      "grad_norm": 0.37873106808738827,
+      "learning_rate": 1.028885829070032e-06,
+      "loss": 0.3971,
+      "step": 6709
+    },
+    {
+      "epoch": 1.5971321473195692,
+      "grad_norm": 0.3659434667845778,
+      "learning_rate": 1.027714866823954e-06,
+      "loss": 0.283,
+      "step": 6710
+    },
+    {
+      "epoch": 1.597370143392634,
+      "grad_norm": 0.3867687273653711,
+      "learning_rate": 1.0265444949577163e-06,
+      "loss": 0.2797,
+      "step": 6711
+    },
+    {
+      "epoch": 1.5976081394656987,
+      "grad_norm": 0.4072148084054436,
+      "learning_rate": 1.0253747136452657e-06,
+      "loss": 0.3506,
+      "step": 6712
+    },
+    {
+      "epoch": 1.5978461355387636,
+      "grad_norm": 0.3901721585652958,
+      "learning_rate": 1.0242055230604598e-06,
+      "loss": 0.3852,
+      "step": 6713
+    },
+    {
+      "epoch": 1.5980841316118284,
+      "grad_norm": 0.38077201173399056,
+      "learning_rate": 1.0230369233770688e-06,
+      "loss": 0.2825,
+      "step": 6714
+    },
+    {
+      "epoch": 1.5983221276848933,
+      "grad_norm": 0.34750351350090636,
+      "learning_rate": 1.0218689147687766e-06,
+      "loss": 0.318,
+      "step": 6715
+    },
+    {
+      "epoch": 1.598560123757958,
+      "grad_norm": 0.37793674402799227,
+      "learning_rate": 1.0207014974091788e-06,
+      "loss": 0.373,
+      "step": 6716
+    },
+    {
+      "epoch": 1.5987981198310228,
+      "grad_norm": 0.37210043250443864,
+      "learning_rate": 1.0195346714717813e-06,
+      "loss": 0.3131,
+      "step": 6717
+    },
+    {
+      "epoch": 1.5990361159040876,
+      "grad_norm": 0.397386731456454,
+      "learning_rate": 1.0183684371300035e-06,
+      "loss": 0.2892,
+      "step": 6718
+    },
+    {
+      "epoch": 1.5992741119771523,
+      "grad_norm": 0.3857199576264192,
+      "learning_rate": 1.0172027945571765e-06,
+      "loss": 0.3093,
+      "step": 6719
+    },
+    {
+      "epoch": 1.599512108050217,
+      "grad_norm": 0.3652664433334774,
+      "learning_rate": 1.0160377439265445e-06,
+      "loss": 0.3803,
+      "step": 6720
+    },
+    {
+      "epoch": 1.599750104123282,
+      "grad_norm": 0.39011567367020755,
+      "learning_rate": 1.014873285411262e-06,
+      "loss": 0.2906,
+      "step": 6721
+    },
+    {
+      "epoch": 1.5999881001963467,
+      "grad_norm": 0.4067204738621554,
+      "learning_rate": 1.013709419184396e-06,
+      "loss": 0.2948,
+      "step": 6722
+    },
+    {
+      "epoch": 1.6002260962694117,
+      "grad_norm": 0.38555422364472247,
+      "learning_rate": 1.0125461454189273e-06,
+      "loss": 0.3671,
+      "step": 6723
+    },
+    {
+      "epoch": 1.6004640923424764,
+      "grad_norm": 0.38611908239425363,
+      "learning_rate": 1.0113834642877457e-06,
+      "loss": 0.3296,
+      "step": 6724
+    },
+    {
+      "epoch": 1.6007020884155412,
+      "grad_norm": 0.4058794539643264,
+      "learning_rate": 1.0102213759636548e-06,
+      "loss": 0.2817,
+      "step": 6725
+    },
+    {
+      "epoch": 1.600940084488606,
+      "grad_norm": 0.38734559534856855,
+      "learning_rate": 1.0090598806193701e-06,
+      "loss": 0.33,
+      "step": 6726
+    },
+    {
+      "epoch": 1.6011780805616707,
+      "grad_norm": 0.393666212501604,
+      "learning_rate": 1.0078989784275183e-06,
+      "loss": 0.3874,
+      "step": 6727
+    },
+    {
+      "epoch": 1.6014160766347354,
+      "grad_norm": 0.36559390636596556,
+      "learning_rate": 1.0067386695606367e-06,
+      "loss": 0.27,
+      "step": 6728
+    },
+    {
+      "epoch": 1.6016540727078004,
+      "grad_norm": 0.409681509162357,
+      "learning_rate": 1.0055789541911788e-06,
+      "loss": 0.275,
+      "step": 6729
+    },
+    {
+      "epoch": 1.601892068780865,
+      "grad_norm": 0.4138944613457098,
+      "learning_rate": 1.004419832491505e-06,
+      "loss": 0.343,
+      "step": 6730
+    },
+    {
+      "epoch": 1.60213006485393,
+      "grad_norm": 0.3955901179511804,
+      "learning_rate": 1.0032613046338907e-06,
+      "loss": 0.3837,
+      "step": 6731
+    },
+    {
+      "epoch": 1.6023680609269948,
+      "grad_norm": 0.3840912144760111,
+      "learning_rate": 1.0021033707905202e-06,
+      "loss": 0.2799,
+      "step": 6732
+    },
+    {
+      "epoch": 1.6026060570000595,
+      "grad_norm": 0.3907610707868771,
+      "learning_rate": 1.0009460311334928e-06,
+      "loss": 0.319,
+      "step": 6733
+    },
+    {
+      "epoch": 1.6028440530731243,
+      "grad_norm": 0.38100921827655265,
+      "learning_rate": 9.99789285834815e-07,
+      "loss": 0.3672,
+      "step": 6734
+    },
+    {
+      "epoch": 1.603082049146189,
+      "grad_norm": 0.4088419649010157,
+      "learning_rate": 9.98633135066412e-07,
+      "loss": 0.3222,
+      "step": 6735
+    },
+    {
+      "epoch": 1.6033200452192538,
+      "grad_norm": 0.3801150461937333,
+      "learning_rate": 9.974775790001134e-07,
+      "loss": 0.2863,
+      "step": 6736
+    },
+    {
+      "epoch": 1.6035580412923187,
+      "grad_norm": 0.3695771436542104,
+      "learning_rate": 9.963226178076646e-07,
+      "loss": 0.3493,
+      "step": 6737
+    },
+    {
+      "epoch": 1.6037960373653835,
+      "grad_norm": 0.4646018757207895,
+      "learning_rate": 9.951682516607208e-07,
+      "loss": 0.3526,
+      "step": 6738
+    },
+    {
+      "epoch": 1.6040340334384484,
+      "grad_norm": 0.3739693404848989,
+      "learning_rate": 9.940144807308494e-07,
+      "loss": 0.2835,
+      "step": 6739
+    },
+    {
+      "epoch": 1.6042720295115132,
+      "grad_norm": 0.38989411473136254,
+      "learning_rate": 9.928613051895297e-07,
+      "loss": 0.3219,
+      "step": 6740
+    },
+    {
+      "epoch": 1.604510025584578,
+      "grad_norm": 0.39146856372760747,
+      "learning_rate": 9.91708725208152e-07,
+      "loss": 0.3897,
+      "step": 6741
+    },
+    {
+      "epoch": 1.6047480216576426,
+      "grad_norm": 0.35756410647769543,
+      "learning_rate": 9.90556740958017e-07,
+      "loss": 0.3037,
+      "step": 6742
+    },
+    {
+      "epoch": 1.6049860177307074,
+      "grad_norm": 0.41503456281085804,
+      "learning_rate": 9.894053526103397e-07,
+      "loss": 0.2837,
+      "step": 6743
+    },
+    {
+      "epoch": 1.605224013803772,
+      "grad_norm": 0.37911206000367126,
+      "learning_rate": 9.882545603362448e-07,
+      "loss": 0.3649,
+      "step": 6744
+    },
+    {
+      "epoch": 1.605462009876837,
+      "grad_norm": 0.39287333843325145,
+      "learning_rate": 9.87104364306768e-07,
+      "loss": 0.3827,
+      "step": 6745
+    },
+    {
+      "epoch": 1.6057000059499018,
+      "grad_norm": 0.3623004181552304,
+      "learning_rate": 9.859547646928568e-07,
+      "loss": 0.2571,
+      "step": 6746
+    },
+    {
+      "epoch": 1.6059380020229668,
+      "grad_norm": 0.3729145089679708,
+      "learning_rate": 9.848057616653705e-07,
+      "loss": 0.3106,
+      "step": 6747
+    },
+    {
+      "epoch": 1.6061759980960315,
+      "grad_norm": 0.39121154395488567,
+      "learning_rate": 9.83657355395079e-07,
+      "loss": 0.3548,
+      "step": 6748
+    },
+    {
+      "epoch": 1.6064139941690962,
+      "grad_norm": 0.3649057165068434,
+      "learning_rate": 9.82509546052664e-07,
+      "loss": 0.3453,
+      "step": 6749
+    },
+    {
+      "epoch": 1.606651990242161,
+      "grad_norm": 0.38056432571485566,
+      "learning_rate": 9.813623338087181e-07,
+      "loss": 0.2907,
+      "step": 6750
+    },
+    {
+      "epoch": 1.6068899863152257,
+      "grad_norm": 0.3773297281289237,
+      "learning_rate": 9.802157188337464e-07,
+      "loss": 0.3433,
+      "step": 6751
+    },
+    {
+      "epoch": 1.6071279823882905,
+      "grad_norm": 0.37897532559459307,
+      "learning_rate": 9.790697012981632e-07,
+      "loss": 0.3886,
+      "step": 6752
+    },
+    {
+      "epoch": 1.6073659784613554,
+      "grad_norm": 0.36839416215381326,
+      "learning_rate": 9.77924281372295e-07,
+      "loss": 0.2888,
+      "step": 6753
+    },
+    {
+      "epoch": 1.6076039745344202,
+      "grad_norm": 0.40572797254422144,
+      "learning_rate": 9.767794592263802e-07,
+      "loss": 0.2758,
+      "step": 6754
+    },
+    {
+      "epoch": 1.6078419706074851,
+      "grad_norm": 0.38690003338447304,
+      "learning_rate": 9.756352350305676e-07,
+      "loss": 0.3507,
+      "step": 6755
+    },
+    {
+      "epoch": 1.6080799666805499,
+      "grad_norm": 0.3633536684182977,
+      "learning_rate": 9.744916089549155e-07,
+      "loss": 0.3725,
+      "step": 6756
+    },
+    {
+      "epoch": 1.6083179627536146,
+      "grad_norm": 0.3430540268517774,
+      "learning_rate": 9.733485811693983e-07,
+      "loss": 0.2831,
+      "step": 6757
+    },
+    {
+      "epoch": 1.6085559588266793,
+      "grad_norm": 0.3967941777054526,
+      "learning_rate": 9.722061518438963e-07,
+      "loss": 0.3197,
+      "step": 6758
+    },
+    {
+      "epoch": 1.608793954899744,
+      "grad_norm": 0.3769154445985261,
+      "learning_rate": 9.71064321148203e-07,
+      "loss": 0.3843,
+      "step": 6759
+    },
+    {
+      "epoch": 1.6090319509728088,
+      "grad_norm": 0.36745266921746866,
+      "learning_rate": 9.699230892520222e-07,
+      "loss": 0.2952,
+      "step": 6760
+    },
+    {
+      "epoch": 1.6092699470458738,
+      "grad_norm": 0.3719289532826721,
+      "learning_rate": 9.687824563249687e-07,
+      "loss": 0.2952,
+      "step": 6761
+    },
+    {
+      "epoch": 1.6095079431189385,
+      "grad_norm": 0.3850579375106196,
+      "learning_rate": 9.67642422536571e-07,
+      "loss": 0.3314,
+      "step": 6762
+    },
+    {
+      "epoch": 1.6097459391920035,
+      "grad_norm": 0.36942355814013683,
+      "learning_rate": 9.665029880562655e-07,
+      "loss": 0.355,
+      "step": 6763
+    },
+    {
+      "epoch": 1.6099839352650682,
+      "grad_norm": 0.37525710772280524,
+      "learning_rate": 9.65364153053399e-07,
+      "loss": 0.2903,
+      "step": 6764
+    },
+    {
+      "epoch": 1.610221931338133,
+      "grad_norm": 0.37197742941684636,
+      "learning_rate": 9.64225917697232e-07,
+      "loss": 0.3024,
+      "step": 6765
+    },
+    {
+      "epoch": 1.6104599274111977,
+      "grad_norm": 0.3768595495711356,
+      "learning_rate": 9.630882821569338e-07,
+      "loss": 0.3852,
+      "step": 6766
+    },
+    {
+      "epoch": 1.6106979234842624,
+      "grad_norm": 0.35867128131233944,
+      "learning_rate": 9.61951246601585e-07,
+      "loss": 0.3093,
+      "step": 6767
+    },
+    {
+      "epoch": 1.6109359195573272,
+      "grad_norm": 0.40435567311806875,
+      "learning_rate": 9.608148112001774e-07,
+      "loss": 0.262,
+      "step": 6768
+    },
+    {
+      "epoch": 1.6111739156303921,
+      "grad_norm": 0.3786650690585147,
+      "learning_rate": 9.59678976121614e-07,
+      "loss": 0.319,
+      "step": 6769
+    },
+    {
+      "epoch": 1.6114119117034569,
+      "grad_norm": 0.3873536911631197,
+      "learning_rate": 9.58543741534706e-07,
+      "loss": 0.367,
+      "step": 6770
+    },
+    {
+      "epoch": 1.6116499077765218,
+      "grad_norm": 0.36824419357607824,
+      "learning_rate": 9.574091076081799e-07,
+      "loss": 0.2774,
+      "step": 6771
+    },
+    {
+      "epoch": 1.6118879038495866,
+      "grad_norm": 0.3891611443567881,
+      "learning_rate": 9.562750745106697e-07,
+      "loss": 0.2913,
+      "step": 6772
+    },
+    {
+      "epoch": 1.6121258999226513,
+      "grad_norm": 0.3995297635418241,
+      "learning_rate": 9.551416424107202e-07,
+      "loss": 0.3454,
+      "step": 6773
+    },
+    {
+      "epoch": 1.612363895995716,
+      "grad_norm": 0.41453803457475424,
+      "learning_rate": 9.54008811476787e-07,
+      "loss": 0.3433,
+      "step": 6774
+    },
+    {
+      "epoch": 1.6126018920687808,
+      "grad_norm": 0.39579544768901054,
+      "learning_rate": 9.528765818772379e-07,
+      "loss": 0.2877,
+      "step": 6775
+    },
+    {
+      "epoch": 1.6128398881418455,
+      "grad_norm": 0.3488915743792545,
+      "learning_rate": 9.517449537803497e-07,
+      "loss": 0.3048,
+      "step": 6776
+    },
+    {
+      "epoch": 1.6130778842149105,
+      "grad_norm": 0.38162237869505217,
+      "learning_rate": 9.506139273543108e-07,
+      "loss": 0.3919,
+      "step": 6777
+    },
+    {
+      "epoch": 1.6133158802879752,
+      "grad_norm": 0.40658809756542225,
+      "learning_rate": 9.494835027672189e-07,
+      "loss": 0.2859,
+      "step": 6778
+    },
+    {
+      "epoch": 1.6135538763610402,
+      "grad_norm": 0.4090597053962956,
+      "learning_rate": 9.483536801870835e-07,
+      "loss": 0.2859,
+      "step": 6779
+    },
+    {
+      "epoch": 1.613791872434105,
+      "grad_norm": 0.4162769342871607,
+      "learning_rate": 9.472244597818236e-07,
+      "loss": 0.3559,
+      "step": 6780
+    },
+    {
+      "epoch": 1.6140298685071697,
+      "grad_norm": 0.4127306882903097,
+      "learning_rate": 9.460958417192706e-07,
+      "loss": 0.3618,
+      "step": 6781
+    },
+    {
+      "epoch": 1.6142678645802344,
+      "grad_norm": 0.3705494652288955,
+      "learning_rate": 9.449678261671636e-07,
+      "loss": 0.2495,
+      "step": 6782
+    },
+    {
+      "epoch": 1.6145058606532992,
+      "grad_norm": 0.38615110793723045,
+      "learning_rate": 9.438404132931534e-07,
+      "loss": 0.3189,
+      "step": 6783
+    },
+    {
+      "epoch": 1.614743856726364,
+      "grad_norm": 0.4247897491555896,
+      "learning_rate": 9.427136032648038e-07,
+      "loss": 0.3723,
+      "step": 6784
+    },
+    {
+      "epoch": 1.6149818527994289,
+      "grad_norm": 0.38613015342079937,
+      "learning_rate": 9.415873962495847e-07,
+      "loss": 0.3214,
+      "step": 6785
+    },
+    {
+      "epoch": 1.6152198488724936,
+      "grad_norm": 0.38838589089115877,
+      "learning_rate": 9.40461792414879e-07,
+      "loss": 0.3093,
+      "step": 6786
+    },
+    {
+      "epoch": 1.6154578449455586,
+      "grad_norm": 0.4134809114242341,
+      "learning_rate": 9.393367919279794e-07,
+      "loss": 0.3258,
+      "step": 6787
+    },
+    {
+      "epoch": 1.6156958410186233,
+      "grad_norm": 0.3657710453341731,
+      "learning_rate": 9.382123949560868e-07,
+      "loss": 0.3751,
+      "step": 6788
+    },
+    {
+      "epoch": 1.615933837091688,
+      "grad_norm": 0.367777502716404,
+      "learning_rate": 9.370886016663178e-07,
+      "loss": 0.284,
+      "step": 6789
+    },
+    {
+      "epoch": 1.6161718331647528,
+      "grad_norm": 0.37224518047407673,
+      "learning_rate": 9.35965412225694e-07,
+      "loss": 0.3233,
+      "step": 6790
+    },
+    {
+      "epoch": 1.6164098292378175,
+      "grad_norm": 0.3679948065415524,
+      "learning_rate": 9.348428268011489e-07,
+      "loss": 0.38,
+      "step": 6791
+    },
+    {
+      "epoch": 1.6166478253108822,
+      "grad_norm": 0.368226197018974,
+      "learning_rate": 9.337208455595271e-07,
+      "loss": 0.3483,
+      "step": 6792
+    },
+    {
+      "epoch": 1.6168858213839472,
+      "grad_norm": 0.38742952029982053,
+      "learning_rate": 9.325994686675826e-07,
+      "loss": 0.2773,
+      "step": 6793
+    },
+    {
+      "epoch": 1.617123817457012,
+      "grad_norm": 0.4010018284261614,
+      "learning_rate": 9.314786962919798e-07,
+      "loss": 0.3711,
+      "step": 6794
+    },
+    {
+      "epoch": 1.617361813530077,
+      "grad_norm": 0.39828012262925994,
+      "learning_rate": 9.303585285992933e-07,
+      "loss": 0.3747,
+      "step": 6795
+    },
+    {
+      "epoch": 1.6175998096031416,
+      "grad_norm": 0.41099074455886697,
+      "learning_rate": 9.292389657560069e-07,
+      "loss": 0.2781,
+      "step": 6796
+    },
+    {
+      "epoch": 1.6178378056762064,
+      "grad_norm": 0.3833507644875006,
+      "learning_rate": 9.281200079285152e-07,
+      "loss": 0.2849,
+      "step": 6797
+    },
+    {
+      "epoch": 1.6180758017492711,
+      "grad_norm": 0.38415300774108235,
+      "learning_rate": 9.270016552831252e-07,
+      "loss": 0.3648,
+      "step": 6798
+    },
+    {
+      "epoch": 1.6183137978223359,
+      "grad_norm": 0.37049113228296326,
+      "learning_rate": 9.258839079860499e-07,
+      "loss": 0.3186,
+      "step": 6799
+    },
+    {
+      "epoch": 1.6185517938954006,
+      "grad_norm": 0.3725759851299388,
+      "learning_rate": 9.247667662034149e-07,
+      "loss": 0.2857,
+      "step": 6800
+    },
+    {
+      "epoch": 1.6187897899684656,
+      "grad_norm": 0.3974032467656143,
+      "learning_rate": 9.236502301012546e-07,
+      "loss": 0.3328,
+      "step": 6801
+    },
+    {
+      "epoch": 1.6190277860415303,
+      "grad_norm": 0.37422414801531034,
+      "learning_rate": 9.225342998455145e-07,
+      "loss": 0.3822,
+      "step": 6802
+    },
+    {
+      "epoch": 1.6192657821145953,
+      "grad_norm": 0.34722153079424223,
+      "learning_rate": 9.214189756020487e-07,
+      "loss": 0.2668,
+      "step": 6803
+    },
+    {
+      "epoch": 1.61950377818766,
+      "grad_norm": 0.3888234856151576,
+      "learning_rate": 9.203042575366228e-07,
+      "loss": 0.2795,
+      "step": 6804
+    },
+    {
+      "epoch": 1.6197417742607247,
+      "grad_norm": 0.40283307764457343,
+      "learning_rate": 9.191901458149106e-07,
+      "loss": 0.3836,
+      "step": 6805
+    },
+    {
+      "epoch": 1.6199797703337895,
+      "grad_norm": 0.38472144664022295,
+      "learning_rate": 9.180766406024971e-07,
+      "loss": 0.3505,
+      "step": 6806
+    },
+    {
+      "epoch": 1.6202177664068542,
+      "grad_norm": 0.37583986626555843,
+      "learning_rate": 9.16963742064877e-07,
+      "loss": 0.2628,
+      "step": 6807
+    },
+    {
+      "epoch": 1.620455762479919,
+      "grad_norm": 0.362264134804524,
+      "learning_rate": 9.158514503674543e-07,
+      "loss": 0.3411,
+      "step": 6808
+    },
+    {
+      "epoch": 1.620693758552984,
+      "grad_norm": 0.3769911946362758,
+      "learning_rate": 9.14739765675543e-07,
+      "loss": 0.371,
+      "step": 6809
+    },
+    {
+      "epoch": 1.6209317546260487,
+      "grad_norm": 0.3853498348483161,
+      "learning_rate": 9.136286881543666e-07,
+      "loss": 0.314,
+      "step": 6810
+    },
+    {
+      "epoch": 1.6211697506991136,
+      "grad_norm": 0.3983780077597233,
+      "learning_rate": 9.125182179690584e-07,
+      "loss": 0.2673,
+      "step": 6811
+    },
+    {
+      "epoch": 1.6214077467721784,
+      "grad_norm": 0.37571986287733705,
+      "learning_rate": 9.114083552846636e-07,
+      "loss": 0.3361,
+      "step": 6812
+    },
+    {
+      "epoch": 1.621645742845243,
+      "grad_norm": 0.3904265606928051,
+      "learning_rate": 9.102991002661337e-07,
+      "loss": 0.367,
+      "step": 6813
+    },
+    {
+      "epoch": 1.6218837389183078,
+      "grad_norm": 0.4223581365461809,
+      "learning_rate": 9.091904530783319e-07,
+      "loss": 0.3114,
+      "step": 6814
+    },
+    {
+      "epoch": 1.6221217349913726,
+      "grad_norm": 0.39388763487496037,
+      "learning_rate": 9.080824138860295e-07,
+      "loss": 0.3303,
+      "step": 6815
+    },
+    {
+      "epoch": 1.6223597310644373,
+      "grad_norm": 0.39770354399393776,
+      "learning_rate": 9.069749828539109e-07,
+      "loss": 0.3909,
+      "step": 6816
+    },
+    {
+      "epoch": 1.6225977271375023,
+      "grad_norm": 0.37675287316781486,
+      "learning_rate": 9.058681601465663e-07,
+      "loss": 0.3139,
+      "step": 6817
+    },
+    {
+      "epoch": 1.622835723210567,
+      "grad_norm": 0.359202858527697,
+      "learning_rate": 9.047619459284968e-07,
+      "loss": 0.2548,
+      "step": 6818
+    },
+    {
+      "epoch": 1.623073719283632,
+      "grad_norm": 0.3666753439305403,
+      "learning_rate": 9.036563403641136e-07,
+      "loss": 0.2944,
+      "step": 6819
+    },
+    {
+      "epoch": 1.6233117153566967,
+      "grad_norm": 0.48209117748060715,
+      "learning_rate": 9.025513436177368e-07,
+      "loss": 0.3864,
+      "step": 6820
+    },
+    {
+      "epoch": 1.6235497114297615,
+      "grad_norm": 0.3719293692788079,
+      "learning_rate": 9.014469558535965e-07,
+      "loss": 0.278,
+      "step": 6821
+    },
+    {
+      "epoch": 1.6237877075028262,
+      "grad_norm": 0.3969227125064937,
+      "learning_rate": 9.003431772358323e-07,
+      "loss": 0.306,
+      "step": 6822
+    },
+    {
+      "epoch": 1.624025703575891,
+      "grad_norm": 0.3951142656857,
+      "learning_rate": 8.992400079284919e-07,
+      "loss": 0.3504,
+      "step": 6823
+    },
+    {
+      "epoch": 1.6242636996489557,
+      "grad_norm": 0.3481746671653846,
+      "learning_rate": 8.981374480955347e-07,
+      "loss": 0.3283,
+      "step": 6824
+    },
+    {
+      "epoch": 1.6245016957220206,
+      "grad_norm": 0.3891762848601158,
+      "learning_rate": 8.970354979008261e-07,
+      "loss": 0.281,
+      "step": 6825
+    },
+    {
+      "epoch": 1.6247396917950854,
+      "grad_norm": 0.43069665710742855,
+      "learning_rate": 8.959341575081465e-07,
+      "loss": 0.3199,
+      "step": 6826
+    },
+    {
+      "epoch": 1.6249776878681503,
+      "grad_norm": 0.37051390449080696,
+      "learning_rate": 8.948334270811809e-07,
+      "loss": 0.3871,
+      "step": 6827
+    },
+    {
+      "epoch": 1.625215683941215,
+      "grad_norm": 0.3880703377585865,
+      "learning_rate": 8.937333067835247e-07,
+      "loss": 0.2861,
+      "step": 6828
+    },
+    {
+      "epoch": 1.6254536800142798,
+      "grad_norm": 0.384396753893017,
+      "learning_rate": 8.926337967786836e-07,
+      "loss": 0.2741,
+      "step": 6829
+    },
+    {
+      "epoch": 1.6256916760873446,
+      "grad_norm": 0.3569238940042476,
+      "learning_rate": 8.915348972300713e-07,
+      "loss": 0.357,
+      "step": 6830
+    },
+    {
+      "epoch": 1.6259296721604093,
+      "grad_norm": 0.3499062660519206,
+      "learning_rate": 8.904366083010119e-07,
+      "loss": 0.3204,
+      "step": 6831
+    },
+    {
+      "epoch": 1.626167668233474,
+      "grad_norm": 0.38069461747113986,
+      "learning_rate": 8.893389301547384e-07,
+      "loss": 0.2716,
+      "step": 6832
+    },
+    {
+      "epoch": 1.626405664306539,
+      "grad_norm": 0.35807608092297644,
+      "learning_rate": 8.882418629543926e-07,
+      "loss": 0.3028,
+      "step": 6833
+    },
+    {
+      "epoch": 1.6266436603796037,
+      "grad_norm": 0.38443325525551314,
+      "learning_rate": 8.871454068630259e-07,
+      "loss": 0.3946,
+      "step": 6834
+    },
+    {
+      "epoch": 1.6268816564526687,
+      "grad_norm": 0.357934918939066,
+      "learning_rate": 8.860495620435989e-07,
+      "loss": 0.3355,
+      "step": 6835
+    },
+    {
+      "epoch": 1.6271196525257334,
+      "grad_norm": 0.36332940789321744,
+      "learning_rate": 8.849543286589813e-07,
+      "loss": 0.297,
+      "step": 6836
+    },
+    {
+      "epoch": 1.6273576485987982,
+      "grad_norm": 0.39032305116436367,
+      "learning_rate": 8.838597068719518e-07,
+      "loss": 0.3483,
+      "step": 6837
+    },
+    {
+      "epoch": 1.627595644671863,
+      "grad_norm": 0.37557326175473943,
+      "learning_rate": 8.82765696845197e-07,
+      "loss": 0.3611,
+      "step": 6838
+    },
+    {
+      "epoch": 1.6278336407449276,
+      "grad_norm": 0.3693511552822045,
+      "learning_rate": 8.816722987413162e-07,
+      "loss": 0.2734,
+      "step": 6839
+    },
+    {
+      "epoch": 1.6280716368179924,
+      "grad_norm": 0.3718337287364868,
+      "learning_rate": 8.805795127228145e-07,
+      "loss": 0.305,
+      "step": 6840
+    },
+    {
+      "epoch": 1.6283096328910573,
+      "grad_norm": 0.3904012424314939,
+      "learning_rate": 8.794873389521069e-07,
+      "loss": 0.3579,
+      "step": 6841
+    },
+    {
+      "epoch": 1.628547628964122,
+      "grad_norm": 0.3643275958018295,
+      "learning_rate": 8.783957775915159e-07,
+      "loss": 0.3106,
+      "step": 6842
+    },
+    {
+      "epoch": 1.628785625037187,
+      "grad_norm": 0.38787892776494476,
+      "learning_rate": 8.77304828803277e-07,
+      "loss": 0.2696,
+      "step": 6843
+    },
+    {
+      "epoch": 1.6290236211102518,
+      "grad_norm": 0.3660591486708151,
+      "learning_rate": 8.762144927495309e-07,
+      "loss": 0.3095,
+      "step": 6844
+    },
+    {
+      "epoch": 1.6292616171833165,
+      "grad_norm": 0.39222908624194175,
+      "learning_rate": 8.751247695923292e-07,
+      "loss": 0.4224,
+      "step": 6845
+    },
+    {
+      "epoch": 1.6294996132563813,
+      "grad_norm": 0.42296916733578066,
+      "learning_rate": 8.740356594936311e-07,
+      "loss": 0.2812,
+      "step": 6846
+    },
+    {
+      "epoch": 1.629737609329446,
+      "grad_norm": 0.40203065491885415,
+      "learning_rate": 8.729471626153052e-07,
+      "loss": 0.345,
+      "step": 6847
+    },
+    {
+      "epoch": 1.6299756054025107,
+      "grad_norm": 0.3806358325222802,
+      "learning_rate": 8.7185927911913e-07,
+      "loss": 0.3598,
+      "step": 6848
+    },
+    {
+      "epoch": 1.6302136014755757,
+      "grad_norm": 0.3826490418036862,
+      "learning_rate": 8.707720091667904e-07,
+      "loss": 0.3363,
+      "step": 6849
+    },
+    {
+      "epoch": 1.6304515975486404,
+      "grad_norm": 0.41820535006387766,
+      "learning_rate": 8.696853529198829e-07,
+      "loss": 0.2485,
+      "step": 6850
+    },
+    {
+      "epoch": 1.6306895936217054,
+      "grad_norm": 0.3883852616080865,
+      "learning_rate": 8.685993105399104e-07,
+      "loss": 0.324,
+      "step": 6851
+    },
+    {
+      "epoch": 1.6309275896947701,
+      "grad_norm": 0.3924806045692454,
+      "learning_rate": 8.675138821882856e-07,
+      "loss": 0.3843,
+      "step": 6852
+    },
+    {
+      "epoch": 1.6311655857678349,
+      "grad_norm": 0.3886779181476492,
+      "learning_rate": 8.664290680263321e-07,
+      "loss": 0.2817,
+      "step": 6853
+    },
+    {
+      "epoch": 1.6314035818408996,
+      "grad_norm": 0.41445337145235195,
+      "learning_rate": 8.653448682152782e-07,
+      "loss": 0.2835,
+      "step": 6854
+    },
+    {
+      "epoch": 1.6316415779139644,
+      "grad_norm": 0.35876208864526293,
+      "learning_rate": 8.642612829162639e-07,
+      "loss": 0.3438,
+      "step": 6855
+    },
+    {
+      "epoch": 1.631879573987029,
+      "grad_norm": 0.3470504035205238,
+      "learning_rate": 8.631783122903353e-07,
+      "loss": 0.3624,
+      "step": 6856
+    },
+    {
+      "epoch": 1.632117570060094,
+      "grad_norm": 0.373620998827096,
+      "learning_rate": 8.620959564984504e-07,
+      "loss": 0.2853,
+      "step": 6857
+    },
+    {
+      "epoch": 1.6323555661331588,
+      "grad_norm": 0.40447652300735565,
+      "learning_rate": 8.610142157014728e-07,
+      "loss": 0.3606,
+      "step": 6858
+    },
+    {
+      "epoch": 1.6325935622062238,
+      "grad_norm": 0.39593981996363464,
+      "learning_rate": 8.599330900601766e-07,
+      "loss": 0.3431,
+      "step": 6859
+    },
+    {
+      "epoch": 1.6328315582792885,
+      "grad_norm": 0.3592935150194725,
+      "learning_rate": 8.588525797352432e-07,
+      "loss": 0.3056,
+      "step": 6860
+    },
+    {
+      "epoch": 1.6330695543523532,
+      "grad_norm": 0.37888268748077286,
+      "learning_rate": 8.577726848872636e-07,
+      "loss": 0.2669,
+      "step": 6861
+    },
+    {
+      "epoch": 1.633307550425418,
+      "grad_norm": 0.3640350475107053,
+      "learning_rate": 8.566934056767368e-07,
+      "loss": 0.3408,
+      "step": 6862
+    },
+    {
+      "epoch": 1.6335455464984827,
+      "grad_norm": 0.385536304300356,
+      "learning_rate": 8.556147422640704e-07,
+      "loss": 0.3648,
+      "step": 6863
+    },
+    {
+      "epoch": 1.6337835425715475,
+      "grad_norm": 0.3924071742555982,
+      "learning_rate": 8.545366948095802e-07,
+      "loss": 0.2619,
+      "step": 6864
+    },
+    {
+      "epoch": 1.6340215386446124,
+      "grad_norm": 0.40696056451988827,
+      "learning_rate": 8.5345926347349e-07,
+      "loss": 0.323,
+      "step": 6865
+    },
+    {
+      "epoch": 1.6342595347176772,
+      "grad_norm": 0.42407201911219133,
+      "learning_rate": 8.523824484159348e-07,
+      "loss": 0.3849,
+      "step": 6866
+    },
+    {
+      "epoch": 1.6344975307907421,
+      "grad_norm": 0.36284614007552857,
+      "learning_rate": 8.513062497969554e-07,
+      "loss": 0.337,
+      "step": 6867
+    },
+    {
+      "epoch": 1.6347355268638069,
+      "grad_norm": 0.3887082396171208,
+      "learning_rate": 8.502306677765004e-07,
+      "loss": 0.2736,
+      "step": 6868
+    },
+    {
+      "epoch": 1.6349735229368716,
+      "grad_norm": 0.38345494649062606,
+      "learning_rate": 8.491557025144276e-07,
+      "loss": 0.3312,
+      "step": 6869
+    },
+    {
+      "epoch": 1.6352115190099363,
+      "grad_norm": 0.38028798193743946,
+      "learning_rate": 8.480813541705057e-07,
+      "loss": 0.375,
+      "step": 6870
+    },
+    {
+      "epoch": 1.635449515083001,
+      "grad_norm": 0.38125998205133177,
+      "learning_rate": 8.470076229044077e-07,
+      "loss": 0.3121,
+      "step": 6871
+    },
+    {
+      "epoch": 1.6356875111560658,
+      "grad_norm": 0.3801307871037892,
+      "learning_rate": 8.45934508875717e-07,
+      "loss": 0.3276,
+      "step": 6872
+    },
+    {
+      "epoch": 1.6359255072291308,
+      "grad_norm": 0.4262493108275734,
+      "learning_rate": 8.448620122439255e-07,
+      "loss": 0.384,
+      "step": 6873
+    },
+    {
+      "epoch": 1.6361635033021955,
+      "grad_norm": 0.3533409555026496,
+      "learning_rate": 8.437901331684317e-07,
+      "loss": 0.3331,
+      "step": 6874
+    },
+    {
+      "epoch": 1.6364014993752605,
+      "grad_norm": 0.3740778241308746,
+      "learning_rate": 8.427188718085438e-07,
+      "loss": 0.2702,
+      "step": 6875
+    },
+    {
+      "epoch": 1.6366394954483252,
+      "grad_norm": 0.3899836491292678,
+      "learning_rate": 8.416482283234778e-07,
+      "loss": 0.3263,
+      "step": 6876
+    },
+    {
+      "epoch": 1.63687749152139,
+      "grad_norm": 0.37954634736543486,
+      "learning_rate": 8.405782028723575e-07,
+      "loss": 0.38,
+      "step": 6877
+    },
+    {
+      "epoch": 1.6371154875944547,
+      "grad_norm": 0.3537922473440547,
+      "learning_rate": 8.395087956142156e-07,
+      "loss": 0.2528,
+      "step": 6878
+    },
+    {
+      "epoch": 1.6373534836675194,
+      "grad_norm": 0.3919326904956134,
+      "learning_rate": 8.384400067079923e-07,
+      "loss": 0.2933,
+      "step": 6879
+    },
+    {
+      "epoch": 1.6375914797405842,
+      "grad_norm": 0.3740391889328203,
+      "learning_rate": 8.37371836312535e-07,
+      "loss": 0.3497,
+      "step": 6880
+    },
+    {
+      "epoch": 1.6378294758136491,
+      "grad_norm": 0.37761989502751364,
+      "learning_rate": 8.363042845866021e-07,
+      "loss": 0.3382,
+      "step": 6881
+    },
+    {
+      "epoch": 1.6380674718867139,
+      "grad_norm": 0.38783064891752894,
+      "learning_rate": 8.352373516888573e-07,
+      "loss": 0.2931,
+      "step": 6882
+    },
+    {
+      "epoch": 1.6383054679597788,
+      "grad_norm": 0.37480525199164966,
+      "learning_rate": 8.341710377778739e-07,
+      "loss": 0.3041,
+      "step": 6883
+    },
+    {
+      "epoch": 1.6385434640328436,
+      "grad_norm": 0.37392017588457593,
+      "learning_rate": 8.331053430121317e-07,
+      "loss": 0.3653,
+      "step": 6884
+    },
+    {
+      "epoch": 1.6387814601059083,
+      "grad_norm": 0.37701145032799044,
+      "learning_rate": 8.320402675500195e-07,
+      "loss": 0.2937,
+      "step": 6885
+    },
+    {
+      "epoch": 1.639019456178973,
+      "grad_norm": 0.3811824330700152,
+      "learning_rate": 8.309758115498334e-07,
+      "loss": 0.2808,
+      "step": 6886
+    },
+    {
+      "epoch": 1.6392574522520378,
+      "grad_norm": 0.3643999529788084,
+      "learning_rate": 8.299119751697788e-07,
+      "loss": 0.3547,
+      "step": 6887
+    },
+    {
+      "epoch": 1.6394954483251025,
+      "grad_norm": 0.36502990685126363,
+      "learning_rate": 8.288487585679677e-07,
+      "loss": 0.3857,
+      "step": 6888
+    },
+    {
+      "epoch": 1.6397334443981675,
+      "grad_norm": 0.3658225172262794,
+      "learning_rate": 8.277861619024208e-07,
+      "loss": 0.2652,
+      "step": 6889
+    },
+    {
+      "epoch": 1.6399714404712322,
+      "grad_norm": 0.38230360243140593,
+      "learning_rate": 8.267241853310654e-07,
+      "loss": 0.3224,
+      "step": 6890
+    },
+    {
+      "epoch": 1.6402094365442972,
+      "grad_norm": 0.38637764499463456,
+      "learning_rate": 8.25662829011738e-07,
+      "loss": 0.3791,
+      "step": 6891
+    },
+    {
+      "epoch": 1.640447432617362,
+      "grad_norm": 0.38029656509707016,
+      "learning_rate": 8.246020931021808e-07,
+      "loss": 0.2924,
+      "step": 6892
+    },
+    {
+      "epoch": 1.6406854286904267,
+      "grad_norm": 0.3747338210302809,
+      "learning_rate": 8.235419777600484e-07,
+      "loss": 0.2697,
+      "step": 6893
+    },
+    {
+      "epoch": 1.6409234247634914,
+      "grad_norm": 0.3600089909702746,
+      "learning_rate": 8.224824831428991e-07,
+      "loss": 0.3374,
+      "step": 6894
+    },
+    {
+      "epoch": 1.6411614208365561,
+      "grad_norm": 0.38204493116592925,
+      "learning_rate": 8.21423609408199e-07,
+      "loss": 0.3695,
+      "step": 6895
+    },
+    {
+      "epoch": 1.6413994169096209,
+      "grad_norm": 0.3865945626227688,
+      "learning_rate": 8.203653567133224e-07,
+      "loss": 0.2677,
+      "step": 6896
+    },
+    {
+      "epoch": 1.6416374129826858,
+      "grad_norm": 0.37313722469830596,
+      "learning_rate": 8.193077252155545e-07,
+      "loss": 0.2893,
+      "step": 6897
+    },
+    {
+      "epoch": 1.6418754090557506,
+      "grad_norm": 0.3817816737768635,
+      "learning_rate": 8.182507150720837e-07,
+      "loss": 0.3498,
+      "step": 6898
+    },
+    {
+      "epoch": 1.6421134051288155,
+      "grad_norm": 0.36328532420628223,
+      "learning_rate": 8.171943264400084e-07,
+      "loss": 0.3091,
+      "step": 6899
+    },
+    {
+      "epoch": 1.6423514012018803,
+      "grad_norm": 0.40910572921600497,
+      "learning_rate": 8.161385594763338e-07,
+      "loss": 0.2732,
+      "step": 6900
+    },
+    {
+      "epoch": 1.642589397274945,
+      "grad_norm": 0.4173927289371764,
+      "learning_rate": 8.150834143379726e-07,
+      "loss": 0.3009,
+      "step": 6901
+    },
+    {
+      "epoch": 1.6428273933480098,
+      "grad_norm": 0.4031792044847936,
+      "learning_rate": 8.140288911817462e-07,
+      "loss": 0.3839,
+      "step": 6902
+    },
+    {
+      "epoch": 1.6430653894210745,
+      "grad_norm": 0.36972889214424126,
+      "learning_rate": 8.129749901643824e-07,
+      "loss": 0.314,
+      "step": 6903
+    },
+    {
+      "epoch": 1.6433033854941392,
+      "grad_norm": 0.3875768012510301,
+      "learning_rate": 8.119217114425171e-07,
+      "loss": 0.2766,
+      "step": 6904
+    },
+    {
+      "epoch": 1.6435413815672042,
+      "grad_norm": 0.36865187596941734,
+      "learning_rate": 8.108690551726938e-07,
+      "loss": 0.3303,
+      "step": 6905
+    },
+    {
+      "epoch": 1.643779377640269,
+      "grad_norm": 0.3707979238870622,
+      "learning_rate": 8.098170215113627e-07,
+      "loss": 0.3094,
+      "step": 6906
+    },
+    {
+      "epoch": 1.644017373713334,
+      "grad_norm": 0.3961111362306213,
+      "learning_rate": 8.087656106148811e-07,
+      "loss": 0.262,
+      "step": 6907
+    },
+    {
+      "epoch": 1.6442553697863986,
+      "grad_norm": 0.37905376446342726,
+      "learning_rate": 8.077148226395171e-07,
+      "loss": 0.2942,
+      "step": 6908
+    },
+    {
+      "epoch": 1.6444933658594634,
+      "grad_norm": 0.3891989136340424,
+      "learning_rate": 8.066646577414427e-07,
+      "loss": 0.3634,
+      "step": 6909
+    },
+    {
+      "epoch": 1.6447313619325281,
+      "grad_norm": 0.3635988958220919,
+      "learning_rate": 8.056151160767384e-07,
+      "loss": 0.3054,
+      "step": 6910
+    },
+    {
+      "epoch": 1.6449693580055929,
+      "grad_norm": 0.3976482528225413,
+      "learning_rate": 8.045661978013919e-07,
+      "loss": 0.2771,
+      "step": 6911
+    },
+    {
+      "epoch": 1.6452073540786576,
+      "grad_norm": 0.36402867904746283,
+      "learning_rate": 8.035179030712981e-07,
+      "loss": 0.343,
+      "step": 6912
+    },
+    {
+      "epoch": 1.6454453501517223,
+      "grad_norm": 0.41373321397240326,
+      "learning_rate": 8.0247023204226e-07,
+      "loss": 0.3902,
+      "step": 6913
+    },
+    {
+      "epoch": 1.6456833462247873,
+      "grad_norm": 0.36420015222611274,
+      "learning_rate": 8.014231848699877e-07,
+      "loss": 0.2814,
+      "step": 6914
+    },
+    {
+      "epoch": 1.645921342297852,
+      "grad_norm": 0.3831807168567994,
+      "learning_rate": 8.003767617100977e-07,
+      "loss": 0.2808,
+      "step": 6915
+    },
+    {
+      "epoch": 1.646159338370917,
+      "grad_norm": 0.5084315028646742,
+      "learning_rate": 7.993309627181145e-07,
+      "loss": 0.3751,
+      "step": 6916
+    },
+    {
+      "epoch": 1.6463973344439817,
+      "grad_norm": 0.37628750772345293,
+      "learning_rate": 7.982857880494699e-07,
+      "loss": 0.3378,
+      "step": 6917
+    },
+    {
+      "epoch": 1.6466353305170465,
+      "grad_norm": 0.47297885018084573,
+      "learning_rate": 7.972412378595024e-07,
+      "loss": 0.2825,
+      "step": 6918
+    },
+    {
+      "epoch": 1.6468733265901112,
+      "grad_norm": 0.40733630305637625,
+      "learning_rate": 7.961973123034572e-07,
+      "loss": 0.3292,
+      "step": 6919
+    },
+    {
+      "epoch": 1.647111322663176,
+      "grad_norm": 0.3636765632920138,
+      "learning_rate": 7.951540115364892e-07,
+      "loss": 0.3736,
+      "step": 6920
+    },
+    {
+      "epoch": 1.6473493187362407,
+      "grad_norm": 0.3599660837275332,
+      "learning_rate": 7.941113357136587e-07,
+      "loss": 0.2892,
+      "step": 6921
+    },
+    {
+      "epoch": 1.6475873148093056,
+      "grad_norm": 0.3781156441865842,
+      "learning_rate": 7.930692849899319e-07,
+      "loss": 0.2934,
+      "step": 6922
+    },
+    {
+      "epoch": 1.6478253108823704,
+      "grad_norm": 0.39701363259275885,
+      "learning_rate": 7.920278595201825e-07,
+      "loss": 0.369,
+      "step": 6923
+    },
+    {
+      "epoch": 1.6480633069554353,
+      "grad_norm": 0.37156701594123176,
+      "learning_rate": 7.909870594591951e-07,
+      "loss": 0.3413,
+      "step": 6924
+    },
+    {
+      "epoch": 1.6483013030285,
+      "grad_norm": 0.41917032707799795,
+      "learning_rate": 7.89946884961656e-07,
+      "loss": 0.2803,
+      "step": 6925
+    },
+    {
+      "epoch": 1.6485392991015648,
+      "grad_norm": 0.4065671522305405,
+      "learning_rate": 7.88907336182162e-07,
+      "loss": 0.3162,
+      "step": 6926
+    },
+    {
+      "epoch": 1.6487772951746296,
+      "grad_norm": 0.3790168785337112,
+      "learning_rate": 7.878684132752152e-07,
+      "loss": 0.3934,
+      "step": 6927
+    },
+    {
+      "epoch": 1.6490152912476943,
+      "grad_norm": 0.3708736912100001,
+      "learning_rate": 7.868301163952253e-07,
+      "loss": 0.2832,
+      "step": 6928
+    },
+    {
+      "epoch": 1.649253287320759,
+      "grad_norm": 0.3935121315020632,
+      "learning_rate": 7.857924456965083e-07,
+      "loss": 0.3096,
+      "step": 6929
+    },
+    {
+      "epoch": 1.649491283393824,
+      "grad_norm": 0.3818251147778019,
+      "learning_rate": 7.847554013332892e-07,
+      "loss": 0.3547,
+      "step": 6930
+    },
+    {
+      "epoch": 1.6497292794668887,
+      "grad_norm": 0.38974578717813124,
+      "learning_rate": 7.83718983459697e-07,
+      "loss": 0.3627,
+      "step": 6931
+    },
+    {
+      "epoch": 1.6499672755399537,
+      "grad_norm": 0.3757552764551271,
+      "learning_rate": 7.8268319222977e-07,
+      "loss": 0.3034,
+      "step": 6932
+    },
+    {
+      "epoch": 1.6502052716130184,
+      "grad_norm": 0.384816533453232,
+      "learning_rate": 7.81648027797452e-07,
+      "loss": 0.3372,
+      "step": 6933
+    },
+    {
+      "epoch": 1.6504432676860832,
+      "grad_norm": 0.3667467367151375,
+      "learning_rate": 7.806134903165935e-07,
+      "loss": 0.3632,
+      "step": 6934
+    },
+    {
+      "epoch": 1.650681263759148,
+      "grad_norm": 0.3586707968143236,
+      "learning_rate": 7.795795799409522e-07,
+      "loss": 0.301,
+      "step": 6935
+    },
+    {
+      "epoch": 1.6509192598322127,
+      "grad_norm": 0.40472475833277655,
+      "learning_rate": 7.785462968241947e-07,
+      "loss": 0.2845,
+      "step": 6936
+    },
+    {
+      "epoch": 1.6511572559052774,
+      "grad_norm": 0.47875293593979357,
+      "learning_rate": 7.775136411198914e-07,
+      "loss": 0.3537,
+      "step": 6937
+    },
+    {
+      "epoch": 1.6513952519783424,
+      "grad_norm": 0.3744577504265213,
+      "learning_rate": 7.764816129815201e-07,
+      "loss": 0.3607,
+      "step": 6938
+    },
+    {
+      "epoch": 1.651633248051407,
+      "grad_norm": 0.37901981094401166,
+      "learning_rate": 7.754502125624658e-07,
+      "loss": 0.2718,
+      "step": 6939
+    },
+    {
+      "epoch": 1.651871244124472,
+      "grad_norm": 0.373755700776998,
+      "learning_rate": 7.744194400160204e-07,
+      "loss": 0.2972,
+      "step": 6940
+    },
+    {
+      "epoch": 1.6521092401975368,
+      "grad_norm": 0.39240847812458995,
+      "learning_rate": 7.73389295495382e-07,
+      "loss": 0.3436,
+      "step": 6941
+    },
+    {
+      "epoch": 1.6523472362706015,
+      "grad_norm": 0.39288015487594496,
+      "learning_rate": 7.723597791536553e-07,
+      "loss": 0.3164,
+      "step": 6942
+    },
+    {
+      "epoch": 1.6525852323436663,
+      "grad_norm": 0.3862794779553621,
+      "learning_rate": 7.713308911438527e-07,
+      "loss": 0.3161,
+      "step": 6943
+    },
+    {
+      "epoch": 1.652823228416731,
+      "grad_norm": 0.4601597581604055,
+      "learning_rate": 7.703026316188916e-07,
+      "loss": 0.3288,
+      "step": 6944
+    },
+    {
+      "epoch": 1.6530612244897958,
+      "grad_norm": 0.38403595804884394,
+      "learning_rate": 7.692750007315969e-07,
+      "loss": 0.4199,
+      "step": 6945
+    },
+    {
+      "epoch": 1.6532992205628607,
+      "grad_norm": 0.3689442026310426,
+      "learning_rate": 7.682479986346996e-07,
+      "loss": 0.2934,
+      "step": 6946
+    },
+    {
+      "epoch": 1.6535372166359255,
+      "grad_norm": 0.364745434398555,
+      "learning_rate": 7.67221625480839e-07,
+      "loss": 0.2731,
+      "step": 6947
+    },
+    {
+      "epoch": 1.6537752127089904,
+      "grad_norm": 0.438406775655767,
+      "learning_rate": 7.66195881422559e-07,
+      "loss": 0.3691,
+      "step": 6948
+    },
+    {
+      "epoch": 1.6540132087820552,
+      "grad_norm": 0.3721049405974233,
+      "learning_rate": 7.651707666123098e-07,
+      "loss": 0.3326,
+      "step": 6949
+    },
+    {
+      "epoch": 1.65425120485512,
+      "grad_norm": 0.3894342319600942,
+      "learning_rate": 7.641462812024486e-07,
+      "loss": 0.2969,
+      "step": 6950
+    },
+    {
+      "epoch": 1.6544892009281846,
+      "grad_norm": 0.3805823643582205,
+      "learning_rate": 7.631224253452408e-07,
+      "loss": 0.3233,
+      "step": 6951
+    },
+    {
+      "epoch": 1.6547271970012494,
+      "grad_norm": 0.42081777893965966,
+      "learning_rate": 7.620991991928561e-07,
+      "loss": 0.3882,
+      "step": 6952
+    },
+    {
+      "epoch": 1.6549651930743141,
+      "grad_norm": 0.37868479763553864,
+      "learning_rate": 7.61076602897371e-07,
+      "loss": 0.3121,
+      "step": 6953
+    },
+    {
+      "epoch": 1.655203189147379,
+      "grad_norm": 0.4244582354427234,
+      "learning_rate": 7.600546366107686e-07,
+      "loss": 0.2845,
+      "step": 6954
+    },
+    {
+      "epoch": 1.6554411852204438,
+      "grad_norm": 0.47303463276050134,
+      "learning_rate": 7.590333004849387e-07,
+      "loss": 0.377,
+      "step": 6955
+    },
+    {
+      "epoch": 1.6556791812935088,
+      "grad_norm": 0.3665165569695061,
+      "learning_rate": 7.580125946716765e-07,
+      "loss": 0.3637,
+      "step": 6956
+    },
+    {
+      "epoch": 1.6559171773665735,
+      "grad_norm": 0.379471968924663,
+      "learning_rate": 7.569925193226846e-07,
+      "loss": 0.2889,
+      "step": 6957
+    },
+    {
+      "epoch": 1.6561551734396383,
+      "grad_norm": 0.38456329877923556,
+      "learning_rate": 7.55973074589571e-07,
+      "loss": 0.3245,
+      "step": 6958
+    },
+    {
+      "epoch": 1.656393169512703,
+      "grad_norm": 0.3852464205929202,
+      "learning_rate": 7.549542606238508e-07,
+      "loss": 0.388,
+      "step": 6959
+    },
+    {
+      "epoch": 1.6566311655857677,
+      "grad_norm": 0.35614680443106633,
+      "learning_rate": 7.53936077576945e-07,
+      "loss": 0.3175,
+      "step": 6960
+    },
+    {
+      "epoch": 1.6568691616588325,
+      "grad_norm": 0.3892713822547582,
+      "learning_rate": 7.529185256001803e-07,
+      "loss": 0.2797,
+      "step": 6961
+    },
+    {
+      "epoch": 1.6571071577318974,
+      "grad_norm": 0.3748547210585344,
+      "learning_rate": 7.519016048447908e-07,
+      "loss": 0.3167,
+      "step": 6962
+    },
+    {
+      "epoch": 1.6573451538049622,
+      "grad_norm": 0.37044782848474667,
+      "learning_rate": 7.508853154619145e-07,
+      "loss": 0.3462,
+      "step": 6963
+    },
+    {
+      "epoch": 1.6575831498780271,
+      "grad_norm": 0.3581984373631393,
+      "learning_rate": 7.498696576025993e-07,
+      "loss": 0.2675,
+      "step": 6964
+    },
+    {
+      "epoch": 1.6578211459510919,
+      "grad_norm": 0.3814135531783868,
+      "learning_rate": 7.488546314177964e-07,
+      "loss": 0.3111,
+      "step": 6965
+    },
+    {
+      "epoch": 1.6580591420241566,
+      "grad_norm": 0.3994529997922176,
+      "learning_rate": 7.478402370583631e-07,
+      "loss": 0.3391,
+      "step": 6966
+    },
+    {
+      "epoch": 1.6582971380972213,
+      "grad_norm": 0.35098177887130955,
+      "learning_rate": 7.468264746750642e-07,
+      "loss": 0.2783,
+      "step": 6967
+    },
+    {
+      "epoch": 1.658535134170286,
+      "grad_norm": 0.3932454998383334,
+      "learning_rate": 7.458133444185694e-07,
+      "loss": 0.2877,
+      "step": 6968
+    },
+    {
+      "epoch": 1.6587731302433508,
+      "grad_norm": 0.4124862836963177,
+      "learning_rate": 7.448008464394557e-07,
+      "loss": 0.3172,
+      "step": 6969
+    },
+    {
+      "epoch": 1.6590111263164158,
+      "grad_norm": 0.4156043963623505,
+      "learning_rate": 7.437889808882043e-07,
+      "loss": 0.3898,
+      "step": 6970
+    },
+    {
+      "epoch": 1.6592491223894805,
+      "grad_norm": 0.3664091808631948,
+      "learning_rate": 7.427777479152043e-07,
+      "loss": 0.2909,
+      "step": 6971
+    },
+    {
+      "epoch": 1.6594871184625455,
+      "grad_norm": 0.3897124758944131,
+      "learning_rate": 7.417671476707489e-07,
+      "loss": 0.3297,
+      "step": 6972
+    },
+    {
+      "epoch": 1.6597251145356102,
+      "grad_norm": 0.3893820163964789,
+      "learning_rate": 7.407571803050384e-07,
+      "loss": 0.3597,
+      "step": 6973
+    },
+    {
+      "epoch": 1.659963110608675,
+      "grad_norm": 0.3695387069491869,
+      "learning_rate": 7.397478459681806e-07,
+      "loss": 0.3694,
+      "step": 6974
+    },
+    {
+      "epoch": 1.6602011066817397,
+      "grad_norm": 0.39522818795728604,
+      "learning_rate": 7.387391448101861e-07,
+      "loss": 0.282,
+      "step": 6975
+    },
+    {
+      "epoch": 1.6604391027548044,
+      "grad_norm": 0.40266534305717183,
+      "learning_rate": 7.377310769809736e-07,
+      "loss": 0.3463,
+      "step": 6976
+    },
+    {
+      "epoch": 1.6606770988278692,
+      "grad_norm": 0.3563962402516382,
+      "learning_rate": 7.367236426303653e-07,
+      "loss": 0.3897,
+      "step": 6977
+    },
+    {
+      "epoch": 1.6609150949009341,
+      "grad_norm": 0.35469458517661445,
+      "learning_rate": 7.357168419080929e-07,
+      "loss": 0.2726,
+      "step": 6978
+    },
+    {
+      "epoch": 1.6611530909739989,
+      "grad_norm": 0.38451642874884295,
+      "learning_rate": 7.347106749637911e-07,
+      "loss": 0.3007,
+      "step": 6979
+    },
+    {
+      "epoch": 1.6613910870470638,
+      "grad_norm": 0.3867505955294661,
+      "learning_rate": 7.337051419470014e-07,
+      "loss": 0.3803,
+      "step": 6980
+    },
+    {
+      "epoch": 1.6616290831201286,
+      "grad_norm": 0.36921539064006076,
+      "learning_rate": 7.327002430071706e-07,
+      "loss": 0.3494,
+      "step": 6981
+    },
+    {
+      "epoch": 1.6618670791931933,
+      "grad_norm": 0.3865817459852862,
+      "learning_rate": 7.316959782936516e-07,
+      "loss": 0.2724,
+      "step": 6982
+    },
+    {
+      "epoch": 1.662105075266258,
+      "grad_norm": 0.38384403692006913,
+      "learning_rate": 7.306923479557032e-07,
+      "loss": 0.3179,
+      "step": 6983
+    },
+    {
+      "epoch": 1.6623430713393228,
+      "grad_norm": 0.3753237383050048,
+      "learning_rate": 7.296893521424891e-07,
+      "loss": 0.3753,
+      "step": 6984
+    },
+    {
+      "epoch": 1.6625810674123875,
+      "grad_norm": 0.37042483156179623,
+      "learning_rate": 7.286869910030797e-07,
+      "loss": 0.307,
+      "step": 6985
+    },
+    {
+      "epoch": 1.6628190634854525,
+      "grad_norm": 0.3789501748283688,
+      "learning_rate": 7.276852646864507e-07,
+      "loss": 0.3027,
+      "step": 6986
+    },
+    {
+      "epoch": 1.6630570595585172,
+      "grad_norm": 0.4135444826893321,
+      "learning_rate": 7.266841733414837e-07,
+      "loss": 0.3552,
+      "step": 6987
+    },
+    {
+      "epoch": 1.6632950556315822,
+      "grad_norm": 0.39788532592242914,
+      "learning_rate": 7.256837171169651e-07,
+      "loss": 0.3437,
+      "step": 6988
+    },
+    {
+      "epoch": 1.663533051704647,
+      "grad_norm": 0.3593638442974499,
+      "learning_rate": 7.246838961615877e-07,
+      "loss": 0.3051,
+      "step": 6989
+    },
+    {
+      "epoch": 1.6637710477777117,
+      "grad_norm": 0.4159858649381969,
+      "learning_rate": 7.23684710623948e-07,
+      "loss": 0.3217,
+      "step": 6990
+    },
+    {
+      "epoch": 1.6640090438507764,
+      "grad_norm": 0.3672974524927207,
+      "learning_rate": 7.226861606525526e-07,
+      "loss": 0.3805,
+      "step": 6991
+    },
+    {
+      "epoch": 1.6642470399238412,
+      "grad_norm": 0.3787693493770054,
+      "learning_rate": 7.216882463958091e-07,
+      "loss": 0.3179,
+      "step": 6992
+    },
+    {
+      "epoch": 1.664485035996906,
+      "grad_norm": 0.3796098325253126,
+      "learning_rate": 7.20690968002033e-07,
+      "loss": 0.2713,
+      "step": 6993
+    },
+    {
+      "epoch": 1.6647230320699709,
+      "grad_norm": 0.38051154387730174,
+      "learning_rate": 7.196943256194439e-07,
+      "loss": 0.2955,
+      "step": 6994
+    },
+    {
+      "epoch": 1.6649610281430356,
+      "grad_norm": 0.38260035859588687,
+      "learning_rate": 7.186983193961677e-07,
+      "loss": 0.3802,
+      "step": 6995
+    },
+    {
+      "epoch": 1.6651990242161006,
+      "grad_norm": 0.3560984048327247,
+      "learning_rate": 7.177029494802351e-07,
+      "loss": 0.3178,
+      "step": 6996
+    },
+    {
+      "epoch": 1.6654370202891653,
+      "grad_norm": 0.38924386944904715,
+      "learning_rate": 7.167082160195837e-07,
+      "loss": 0.2741,
+      "step": 6997
+    },
+    {
+      "epoch": 1.66567501636223,
+      "grad_norm": 0.4039947533316631,
+      "learning_rate": 7.157141191620548e-07,
+      "loss": 0.3479,
+      "step": 6998
+    },
+    {
+      "epoch": 1.6659130124352948,
+      "grad_norm": 0.38643709485362443,
+      "learning_rate": 7.147206590553956e-07,
+      "loss": 0.3369,
+      "step": 6999
+    },
+    {
+      "epoch": 1.6661510085083595,
+      "grad_norm": 0.40129045045890266,
+      "learning_rate": 7.137278358472583e-07,
+      "loss": 0.2818,
+      "step": 7000
+    },
+    {
+      "epoch": 1.6663890045814242,
+      "grad_norm": 0.4499031615543838,
+      "learning_rate": 7.127356496852029e-07,
+      "loss": 0.3035,
+      "step": 7001
+    },
+    {
+      "epoch": 1.6666270006544892,
+      "grad_norm": 0.3748939127556976,
+      "learning_rate": 7.117441007166919e-07,
+      "loss": 0.3819,
+      "step": 7002
+    },
+    {
+      "epoch": 1.666864996727554,
+      "grad_norm": 0.35881304656890556,
+      "learning_rate": 7.10753189089094e-07,
+      "loss": 0.2835,
+      "step": 7003
+    },
+    {
+      "epoch": 1.667102992800619,
+      "grad_norm": 0.41355655256875934,
+      "learning_rate": 7.097629149496815e-07,
+      "loss": 0.2988,
+      "step": 7004
+    },
+    {
+      "epoch": 1.6673409888736836,
+      "grad_norm": 0.6309979385551239,
+      "learning_rate": 7.087732784456369e-07,
+      "loss": 0.3538,
+      "step": 7005
+    },
+    {
+      "epoch": 1.6675789849467484,
+      "grad_norm": 0.37545921085712725,
+      "learning_rate": 7.077842797240426e-07,
+      "loss": 0.3416,
+      "step": 7006
+    },
+    {
+      "epoch": 1.6678169810198131,
+      "grad_norm": 0.365443297712444,
+      "learning_rate": 7.067959189318885e-07,
+      "loss": 0.2977,
+      "step": 7007
+    },
+    {
+      "epoch": 1.6680549770928779,
+      "grad_norm": 0.37039435287713296,
+      "learning_rate": 7.058081962160696e-07,
+      "loss": 0.3164,
+      "step": 7008
+    },
+    {
+      "epoch": 1.6682929731659426,
+      "grad_norm": 0.37427937679786044,
+      "learning_rate": 7.048211117233861e-07,
+      "loss": 0.3486,
+      "step": 7009
+    },
+    {
+      "epoch": 1.6685309692390076,
+      "grad_norm": 0.3729855514273312,
+      "learning_rate": 7.038346656005429e-07,
+      "loss": 0.2989,
+      "step": 7010
+    },
+    {
+      "epoch": 1.6687689653120723,
+      "grad_norm": 0.374090406829846,
+      "learning_rate": 7.028488579941506e-07,
+      "loss": 0.2792,
+      "step": 7011
+    },
+    {
+      "epoch": 1.6690069613851373,
+      "grad_norm": 0.39515663629452236,
+      "learning_rate": 7.018636890507241e-07,
+      "loss": 0.3659,
+      "step": 7012
+    },
+    {
+      "epoch": 1.669244957458202,
+      "grad_norm": 0.35372595983883504,
+      "learning_rate": 7.008791589166847e-07,
+      "loss": 0.3607,
+      "step": 7013
+    },
+    {
+      "epoch": 1.6694829535312667,
+      "grad_norm": 0.3533302020072786,
+      "learning_rate": 6.998952677383569e-07,
+      "loss": 0.276,
+      "step": 7014
+    },
+    {
+      "epoch": 1.6697209496043315,
+      "grad_norm": 0.42475735192235214,
+      "learning_rate": 6.989120156619717e-07,
+      "loss": 0.3186,
+      "step": 7015
+    },
+    {
+      "epoch": 1.6699589456773962,
+      "grad_norm": 0.38764230011040096,
+      "learning_rate": 6.979294028336652e-07,
+      "loss": 0.3573,
+      "step": 7016
+    },
+    {
+      "epoch": 1.670196941750461,
+      "grad_norm": 0.3611283372327025,
+      "learning_rate": 6.969474293994771e-07,
+      "loss": 0.3349,
+      "step": 7017
+    },
+    {
+      "epoch": 1.670434937823526,
+      "grad_norm": 0.36366689855170664,
+      "learning_rate": 6.959660955053527e-07,
+      "loss": 0.296,
+      "step": 7018
+    },
+    {
+      "epoch": 1.6706729338965907,
+      "grad_norm": 0.5422835131958719,
+      "learning_rate": 6.949854012971441e-07,
+      "loss": 0.3383,
+      "step": 7019
+    },
+    {
+      "epoch": 1.6709109299696556,
+      "grad_norm": 0.35747903910606177,
+      "learning_rate": 6.94005346920606e-07,
+      "loss": 0.4007,
+      "step": 7020
+    },
+    {
+      "epoch": 1.6711489260427204,
+      "grad_norm": 0.36742844091875815,
+      "learning_rate": 6.930259325213978e-07,
+      "loss": 0.2987,
+      "step": 7021
+    },
+    {
+      "epoch": 1.671386922115785,
+      "grad_norm": 0.40350636301013826,
+      "learning_rate": 6.920471582450861e-07,
+      "loss": 0.3304,
+      "step": 7022
+    },
+    {
+      "epoch": 1.6716249181888498,
+      "grad_norm": 0.40006906898219396,
+      "learning_rate": 6.910690242371404e-07,
+      "loss": 0.3446,
+      "step": 7023
+    },
+    {
+      "epoch": 1.6718629142619146,
+      "grad_norm": 0.35727853106570195,
+      "learning_rate": 6.900915306429351e-07,
+      "loss": 0.3153,
+      "step": 7024
+    },
+    {
+      "epoch": 1.6721009103349793,
+      "grad_norm": 0.382181007945197,
+      "learning_rate": 6.891146776077507e-07,
+      "loss": 0.2882,
+      "step": 7025
+    },
+    {
+      "epoch": 1.6723389064080443,
+      "grad_norm": 0.4060235764358048,
+      "learning_rate": 6.881384652767709e-07,
+      "loss": 0.3207,
+      "step": 7026
+    },
+    {
+      "epoch": 1.672576902481109,
+      "grad_norm": 0.37097326742176884,
+      "learning_rate": 6.871628937950848e-07,
+      "loss": 0.3922,
+      "step": 7027
+    },
+    {
+      "epoch": 1.672814898554174,
+      "grad_norm": 0.40285375164369536,
+      "learning_rate": 6.861879633076878e-07,
+      "loss": 0.2923,
+      "step": 7028
+    },
+    {
+      "epoch": 1.6730528946272387,
+      "grad_norm": 0.37801910915821857,
+      "learning_rate": 6.852136739594783e-07,
+      "loss": 0.2882,
+      "step": 7029
+    },
+    {
+      "epoch": 1.6732908907003035,
+      "grad_norm": 0.36728422194745036,
+      "learning_rate": 6.842400258952597e-07,
+      "loss": 0.3977,
+      "step": 7030
+    },
+    {
+      "epoch": 1.6735288867733682,
+      "grad_norm": 0.39561139457941275,
+      "learning_rate": 6.832670192597396e-07,
+      "loss": 0.3719,
+      "step": 7031
+    },
+    {
+      "epoch": 1.673766882846433,
+      "grad_norm": 0.37538581127020354,
+      "learning_rate": 6.822946541975306e-07,
+      "loss": 0.2774,
+      "step": 7032
+    },
+    {
+      "epoch": 1.6740048789194977,
+      "grad_norm": 0.38177066214722755,
+      "learning_rate": 6.813229308531516e-07,
+      "loss": 0.3272,
+      "step": 7033
+    },
+    {
+      "epoch": 1.6742428749925626,
+      "grad_norm": 0.38578694289159443,
+      "learning_rate": 6.803518493710243e-07,
+      "loss": 0.4126,
+      "step": 7034
+    },
+    {
+      "epoch": 1.6744808710656274,
+      "grad_norm": 0.36307309359597634,
+      "learning_rate": 6.793814098954749e-07,
+      "loss": 0.331,
+      "step": 7035
+    },
+    {
+      "epoch": 1.6747188671386923,
+      "grad_norm": 0.4000089226486024,
+      "learning_rate": 6.784116125707352e-07,
+      "loss": 0.267,
+      "step": 7036
+    },
+    {
+      "epoch": 1.674956863211757,
+      "grad_norm": 0.36487357951661303,
+      "learning_rate": 6.774424575409405e-07,
+      "loss": 0.3127,
+      "step": 7037
+    },
+    {
+      "epoch": 1.6751948592848218,
+      "grad_norm": 0.35391638388560076,
+      "learning_rate": 6.764739449501317e-07,
+      "loss": 0.3769,
+      "step": 7038
+    },
+    {
+      "epoch": 1.6754328553578866,
+      "grad_norm": 0.35186947926475576,
+      "learning_rate": 6.755060749422537e-07,
+      "loss": 0.2923,
+      "step": 7039
+    },
+    {
+      "epoch": 1.6756708514309513,
+      "grad_norm": 0.44135361184534583,
+      "learning_rate": 6.745388476611553e-07,
+      "loss": 0.2981,
+      "step": 7040
+    },
+    {
+      "epoch": 1.675908847504016,
+      "grad_norm": 0.45872802613072794,
+      "learning_rate": 6.735722632505915e-07,
+      "loss": 0.37,
+      "step": 7041
+    },
+    {
+      "epoch": 1.676146843577081,
+      "grad_norm": 0.3580957612187762,
+      "learning_rate": 6.726063218542195e-07,
+      "loss": 0.314,
+      "step": 7042
+    },
+    {
+      "epoch": 1.6763848396501457,
+      "grad_norm": 0.35669432760875613,
+      "learning_rate": 6.716410236156029e-07,
+      "loss": 0.3014,
+      "step": 7043
+    },
+    {
+      "epoch": 1.6766228357232107,
+      "grad_norm": 0.38931922786060297,
+      "learning_rate": 6.706763686782086e-07,
+      "loss": 0.3144,
+      "step": 7044
+    },
+    {
+      "epoch": 1.6768608317962754,
+      "grad_norm": 0.40124294635503877,
+      "learning_rate": 6.697123571854075e-07,
+      "loss": 0.3697,
+      "step": 7045
+    },
+    {
+      "epoch": 1.6770988278693402,
+      "grad_norm": 0.4969890421919046,
+      "learning_rate": 6.687489892804766e-07,
+      "loss": 0.3089,
+      "step": 7046
+    },
+    {
+      "epoch": 1.677336823942405,
+      "grad_norm": 0.3566446713995905,
+      "learning_rate": 6.677862651065964e-07,
+      "loss": 0.2827,
+      "step": 7047
+    },
+    {
+      "epoch": 1.6775748200154696,
+      "grad_norm": 0.38848733751656256,
+      "learning_rate": 6.668241848068507e-07,
+      "loss": 0.3592,
+      "step": 7048
+    },
+    {
+      "epoch": 1.6778128160885344,
+      "grad_norm": 0.5950856594781679,
+      "learning_rate": 6.658627485242291e-07,
+      "loss": 0.3444,
+      "step": 7049
+    },
+    {
+      "epoch": 1.6780508121615993,
+      "grad_norm": 0.36830146545513864,
+      "learning_rate": 6.649019564016246e-07,
+      "loss": 0.2709,
+      "step": 7050
+    },
+    {
+      "epoch": 1.678288808234664,
+      "grad_norm": 0.37103816586070554,
+      "learning_rate": 6.639418085818339e-07,
+      "loss": 0.3196,
+      "step": 7051
+    },
+    {
+      "epoch": 1.678526804307729,
+      "grad_norm": 0.38203373085536374,
+      "learning_rate": 6.629823052075602e-07,
+      "loss": 0.402,
+      "step": 7052
+    },
+    {
+      "epoch": 1.6787648003807938,
+      "grad_norm": 0.3476339385449799,
+      "learning_rate": 6.620234464214076e-07,
+      "loss": 0.3066,
+      "step": 7053
+    },
+    {
+      "epoch": 1.6790027964538585,
+      "grad_norm": 0.3792905804858758,
+      "learning_rate": 6.610652323658867e-07,
+      "loss": 0.2825,
+      "step": 7054
+    },
+    {
+      "epoch": 1.6792407925269233,
+      "grad_norm": 0.43917028015511345,
+      "learning_rate": 6.601076631834135e-07,
+      "loss": 0.3524,
+      "step": 7055
+    },
+    {
+      "epoch": 1.679478788599988,
+      "grad_norm": 0.35999526768311224,
+      "learning_rate": 6.591507390163049e-07,
+      "loss": 0.3362,
+      "step": 7056
+    },
+    {
+      "epoch": 1.6797167846730527,
+      "grad_norm": 0.3958931440700458,
+      "learning_rate": 6.581944600067847e-07,
+      "loss": 0.299,
+      "step": 7057
+    },
+    {
+      "epoch": 1.6799547807461177,
+      "grad_norm": 0.40723904451700604,
+      "learning_rate": 6.572388262969781e-07,
+      "loss": 0.3382,
+      "step": 7058
+    },
+    {
+      "epoch": 1.6801927768191824,
+      "grad_norm": 0.37213316814749897,
+      "learning_rate": 6.562838380289155e-07,
+      "loss": 0.3613,
+      "step": 7059
+    },
+    {
+      "epoch": 1.6804307728922474,
+      "grad_norm": 0.3775949694068757,
+      "learning_rate": 6.553294953445344e-07,
+      "loss": 0.3037,
+      "step": 7060
+    },
+    {
+      "epoch": 1.6806687689653121,
+      "grad_norm": 0.4302521600801045,
+      "learning_rate": 6.543757983856724e-07,
+      "loss": 0.2837,
+      "step": 7061
+    },
+    {
+      "epoch": 1.6809067650383769,
+      "grad_norm": 0.3744153721393427,
+      "learning_rate": 6.534227472940718e-07,
+      "loss": 0.3633,
+      "step": 7062
+    },
+    {
+      "epoch": 1.6811447611114416,
+      "grad_norm": 0.3601321543703086,
+      "learning_rate": 6.524703422113803e-07,
+      "loss": 0.3534,
+      "step": 7063
+    },
+    {
+      "epoch": 1.6813827571845064,
+      "grad_norm": 0.3978043541470634,
+      "learning_rate": 6.515185832791493e-07,
+      "loss": 0.2833,
+      "step": 7064
+    },
+    {
+      "epoch": 1.681620753257571,
+      "grad_norm": 0.3650335732859001,
+      "learning_rate": 6.50567470638832e-07,
+      "loss": 0.3149,
+      "step": 7065
+    },
+    {
+      "epoch": 1.681858749330636,
+      "grad_norm": 0.38338937802115375,
+      "learning_rate": 6.49617004431789e-07,
+      "loss": 0.3667,
+      "step": 7066
+    },
+    {
+      "epoch": 1.6820967454037008,
+      "grad_norm": 0.40126636400041343,
+      "learning_rate": 6.486671847992826e-07,
+      "loss": 0.3275,
+      "step": 7067
+    },
+    {
+      "epoch": 1.6823347414767658,
+      "grad_norm": 0.4648961323218715,
+      "learning_rate": 6.477180118824788e-07,
+      "loss": 0.2984,
+      "step": 7068
+    },
+    {
+      "epoch": 1.6825727375498305,
+      "grad_norm": 0.3776079096365654,
+      "learning_rate": 6.467694858224488e-07,
+      "loss": 0.3161,
+      "step": 7069
+    },
+    {
+      "epoch": 1.6828107336228952,
+      "grad_norm": 0.3654293611333949,
+      "learning_rate": 6.458216067601669e-07,
+      "loss": 0.3563,
+      "step": 7070
+    },
+    {
+      "epoch": 1.68304872969596,
+      "grad_norm": 0.3938912402346958,
+      "learning_rate": 6.448743748365116e-07,
+      "loss": 0.2843,
+      "step": 7071
+    },
+    {
+      "epoch": 1.6832867257690247,
+      "grad_norm": 0.40857518732879755,
+      "learning_rate": 6.439277901922647e-07,
+      "loss": 0.2857,
+      "step": 7072
+    },
+    {
+      "epoch": 1.6835247218420895,
+      "grad_norm": 0.37369837115535326,
+      "learning_rate": 6.429818529681115e-07,
+      "loss": 0.3408,
+      "step": 7073
+    },
+    {
+      "epoch": 1.6837627179151544,
+      "grad_norm": 0.566489154234049,
+      "learning_rate": 6.420365633046433e-07,
+      "loss": 0.3525,
+      "step": 7074
+    },
+    {
+      "epoch": 1.6840007139882192,
+      "grad_norm": 0.40096527952286826,
+      "learning_rate": 6.410919213423522e-07,
+      "loss": 0.2766,
+      "step": 7075
+    },
+    {
+      "epoch": 1.6842387100612841,
+      "grad_norm": 0.3835017576074419,
+      "learning_rate": 6.40147927221636e-07,
+      "loss": 0.3331,
+      "step": 7076
+    },
+    {
+      "epoch": 1.6844767061343489,
+      "grad_norm": 0.4193203253283927,
+      "learning_rate": 6.392045810827957e-07,
+      "loss": 0.3989,
+      "step": 7077
+    },
+    {
+      "epoch": 1.6847147022074136,
+      "grad_norm": 0.3714016805455296,
+      "learning_rate": 6.382618830660353e-07,
+      "loss": 0.2883,
+      "step": 7078
+    },
+    {
+      "epoch": 1.6849526982804783,
+      "grad_norm": 0.381171748269651,
+      "learning_rate": 6.373198333114633e-07,
+      "loss": 0.2767,
+      "step": 7079
+    },
+    {
+      "epoch": 1.685190694353543,
+      "grad_norm": 0.39675713102514104,
+      "learning_rate": 6.363784319590916e-07,
+      "loss": 0.3425,
+      "step": 7080
+    },
+    {
+      "epoch": 1.6854286904266078,
+      "grad_norm": 0.4002931733011814,
+      "learning_rate": 6.354376791488343e-07,
+      "loss": 0.365,
+      "step": 7081
+    },
+    {
+      "epoch": 1.6856666864996728,
+      "grad_norm": 0.37518227406606824,
+      "learning_rate": 6.344975750205129e-07,
+      "loss": 0.2542,
+      "step": 7082
+    },
+    {
+      "epoch": 1.6859046825727375,
+      "grad_norm": 0.3638772212144532,
+      "learning_rate": 6.335581197138496e-07,
+      "loss": 0.3343,
+      "step": 7083
+    },
+    {
+      "epoch": 1.6861426786458025,
+      "grad_norm": 0.3890298957493163,
+      "learning_rate": 6.326193133684705e-07,
+      "loss": 0.3964,
+      "step": 7084
+    },
+    {
+      "epoch": 1.6863806747188672,
+      "grad_norm": 0.36557376045651974,
+      "learning_rate": 6.31681156123905e-07,
+      "loss": 0.2866,
+      "step": 7085
+    },
+    {
+      "epoch": 1.686618670791932,
+      "grad_norm": 0.43312850934266955,
+      "learning_rate": 6.307436481195866e-07,
+      "loss": 0.2639,
+      "step": 7086
+    },
+    {
+      "epoch": 1.6868566668649967,
+      "grad_norm": 0.3528390333914725,
+      "learning_rate": 6.298067894948512e-07,
+      "loss": 0.3374,
+      "step": 7087
+    },
+    {
+      "epoch": 1.6870946629380614,
+      "grad_norm": 0.34344204575606535,
+      "learning_rate": 6.288705803889411e-07,
+      "loss": 0.3649,
+      "step": 7088
+    },
+    {
+      "epoch": 1.6873326590111262,
+      "grad_norm": 0.38507748165524075,
+      "learning_rate": 6.279350209409995e-07,
+      "loss": 0.2789,
+      "step": 7089
+    },
+    {
+      "epoch": 1.6875706550841911,
+      "grad_norm": 0.3803975185088582,
+      "learning_rate": 6.270001112900736e-07,
+      "loss": 0.3158,
+      "step": 7090
+    },
+    {
+      "epoch": 1.6878086511572559,
+      "grad_norm": 0.3872558923912277,
+      "learning_rate": 6.260658515751139e-07,
+      "loss": 0.3569,
+      "step": 7091
+    },
+    {
+      "epoch": 1.6880466472303208,
+      "grad_norm": 0.3794470357203257,
+      "learning_rate": 6.251322419349748e-07,
+      "loss": 0.2971,
+      "step": 7092
+    },
+    {
+      "epoch": 1.6882846433033856,
+      "grad_norm": 0.38546650413181716,
+      "learning_rate": 6.241992825084131e-07,
+      "loss": 0.2802,
+      "step": 7093
+    },
+    {
+      "epoch": 1.6885226393764503,
+      "grad_norm": 0.37651470887537414,
+      "learning_rate": 6.232669734340907e-07,
+      "loss": 0.2954,
+      "step": 7094
+    },
+    {
+      "epoch": 1.688760635449515,
+      "grad_norm": 0.3885988985334594,
+      "learning_rate": 6.223353148505706e-07,
+      "loss": 0.3914,
+      "step": 7095
+    },
+    {
+      "epoch": 1.6889986315225798,
+      "grad_norm": 0.3511153525219931,
+      "learning_rate": 6.21404306896321e-07,
+      "loss": 0.2711,
+      "step": 7096
+    },
+    {
+      "epoch": 1.6892366275956445,
+      "grad_norm": 0.37884092771461375,
+      "learning_rate": 6.204739497097129e-07,
+      "loss": 0.2851,
+      "step": 7097
+    },
+    {
+      "epoch": 1.6894746236687095,
+      "grad_norm": 0.387337614231269,
+      "learning_rate": 6.1954424342902e-07,
+      "loss": 0.3698,
+      "step": 7098
+    },
+    {
+      "epoch": 1.6897126197417742,
+      "grad_norm": 0.3546161227406718,
+      "learning_rate": 6.186151881924202e-07,
+      "loss": 0.3163,
+      "step": 7099
+    },
+    {
+      "epoch": 1.6899506158148392,
+      "grad_norm": 0.3730093557742399,
+      "learning_rate": 6.176867841379919e-07,
+      "loss": 0.2882,
+      "step": 7100
+    },
+    {
+      "epoch": 1.690188611887904,
+      "grad_norm": 0.38800933074459065,
+      "learning_rate": 6.16759031403722e-07,
+      "loss": 0.3274,
+      "step": 7101
+    },
+    {
+      "epoch": 1.6904266079609687,
+      "grad_norm": 0.39567528441392813,
+      "learning_rate": 6.158319301274962e-07,
+      "loss": 0.3702,
+      "step": 7102
+    },
+    {
+      "epoch": 1.6906646040340334,
+      "grad_norm": 0.3813083923210492,
+      "learning_rate": 6.14905480447105e-07,
+      "loss": 0.2895,
+      "step": 7103
+    },
+    {
+      "epoch": 1.6909026001070981,
+      "grad_norm": 0.383759332023671,
+      "learning_rate": 6.139796825002409e-07,
+      "loss": 0.2933,
+      "step": 7104
+    },
+    {
+      "epoch": 1.6911405961801629,
+      "grad_norm": 0.3886124154209392,
+      "learning_rate": 6.130545364245011e-07,
+      "loss": 0.3486,
+      "step": 7105
+    },
+    {
+      "epoch": 1.6913785922532278,
+      "grad_norm": 0.3865197625654805,
+      "learning_rate": 6.121300423573851e-07,
+      "loss": 0.3864,
+      "step": 7106
+    },
+    {
+      "epoch": 1.6916165883262926,
+      "grad_norm": 0.4342380259704541,
+      "learning_rate": 6.112062004362957e-07,
+      "loss": 0.2811,
+      "step": 7107
+    },
+    {
+      "epoch": 1.6918545843993575,
+      "grad_norm": 0.38137795050806866,
+      "learning_rate": 6.102830107985369e-07,
+      "loss": 0.3386,
+      "step": 7108
+    },
+    {
+      "epoch": 1.6920925804724223,
+      "grad_norm": 0.39290592099891664,
+      "learning_rate": 6.093604735813202e-07,
+      "loss": 0.3844,
+      "step": 7109
+    },
+    {
+      "epoch": 1.692330576545487,
+      "grad_norm": 0.36869581745026214,
+      "learning_rate": 6.084385889217565e-07,
+      "loss": 0.3033,
+      "step": 7110
+    },
+    {
+      "epoch": 1.6925685726185518,
+      "grad_norm": 0.3781454086182421,
+      "learning_rate": 6.075173569568605e-07,
+      "loss": 0.2898,
+      "step": 7111
+    },
+    {
+      "epoch": 1.6928065686916165,
+      "grad_norm": 0.404850762268356,
+      "learning_rate": 6.065967778235499e-07,
+      "loss": 0.3371,
+      "step": 7112
+    },
+    {
+      "epoch": 1.6930445647646812,
+      "grad_norm": 0.41163273024333985,
+      "learning_rate": 6.056768516586453e-07,
+      "loss": 0.3546,
+      "step": 7113
+    },
+    {
+      "epoch": 1.6932825608377462,
+      "grad_norm": 0.3498312393278164,
+      "learning_rate": 6.047575785988702e-07,
+      "loss": 0.3086,
+      "step": 7114
+    },
+    {
+      "epoch": 1.693520556910811,
+      "grad_norm": 0.39282228454424806,
+      "learning_rate": 6.038389587808535e-07,
+      "loss": 0.2948,
+      "step": 7115
+    },
+    {
+      "epoch": 1.693758552983876,
+      "grad_norm": 0.3988205195809829,
+      "learning_rate": 6.029209923411228e-07,
+      "loss": 0.3773,
+      "step": 7116
+    },
+    {
+      "epoch": 1.6939965490569406,
+      "grad_norm": 0.3666168567052365,
+      "learning_rate": 6.02003679416111e-07,
+      "loss": 0.3061,
+      "step": 7117
+    },
+    {
+      "epoch": 1.6942345451300054,
+      "grad_norm": 0.4069498736744767,
+      "learning_rate": 6.010870201421537e-07,
+      "loss": 0.2989,
+      "step": 7118
+    },
+    {
+      "epoch": 1.6944725412030701,
+      "grad_norm": 0.3967722726787557,
+      "learning_rate": 6.001710146554896e-07,
+      "loss": 0.3228,
+      "step": 7119
+    },
+    {
+      "epoch": 1.6947105372761349,
+      "grad_norm": 0.36380296589170585,
+      "learning_rate": 5.992556630922585e-07,
+      "loss": 0.3656,
+      "step": 7120
+    },
+    {
+      "epoch": 1.6949485333491996,
+      "grad_norm": 0.3690082479582151,
+      "learning_rate": 5.983409655885053e-07,
+      "loss": 0.3001,
+      "step": 7121
+    },
+    {
+      "epoch": 1.6951865294222646,
+      "grad_norm": 0.506841855343455,
+      "learning_rate": 5.974269222801765e-07,
+      "loss": 0.2926,
+      "step": 7122
+    },
+    {
+      "epoch": 1.6954245254953293,
+      "grad_norm": 0.4096814364673883,
+      "learning_rate": 5.965135333031213e-07,
+      "loss": 0.3593,
+      "step": 7123
+    },
+    {
+      "epoch": 1.6956625215683943,
+      "grad_norm": 0.38736545080244283,
+      "learning_rate": 5.956007987930923e-07,
+      "loss": 0.3661,
+      "step": 7124
+    },
+    {
+      "epoch": 1.695900517641459,
+      "grad_norm": 0.3611883206027303,
+      "learning_rate": 5.946887188857442e-07,
+      "loss": 0.2677,
+      "step": 7125
+    },
+    {
+      "epoch": 1.6961385137145237,
+      "grad_norm": 0.38872165541623244,
+      "learning_rate": 5.937772937166342e-07,
+      "loss": 0.3316,
+      "step": 7126
+    },
+    {
+      "epoch": 1.6963765097875885,
+      "grad_norm": 0.36656583624158623,
+      "learning_rate": 5.928665234212233e-07,
+      "loss": 0.4022,
+      "step": 7127
+    },
+    {
+      "epoch": 1.6966145058606532,
+      "grad_norm": 0.35861126589804326,
+      "learning_rate": 5.919564081348733e-07,
+      "loss": 0.2982,
+      "step": 7128
+    },
+    {
+      "epoch": 1.696852501933718,
+      "grad_norm": 0.36794026503347826,
+      "learning_rate": 5.910469479928521e-07,
+      "loss": 0.3044,
+      "step": 7129
+    },
+    {
+      "epoch": 1.697090498006783,
+      "grad_norm": 0.3868057111195459,
+      "learning_rate": 5.90138143130326e-07,
+      "loss": 0.3525,
+      "step": 7130
+    },
+    {
+      "epoch": 1.6973284940798476,
+      "grad_norm": 0.3694887390714355,
+      "learning_rate": 5.89229993682367e-07,
+      "loss": 0.3685,
+      "step": 7131
+    },
+    {
+      "epoch": 1.6975664901529126,
+      "grad_norm": 0.3863367650360045,
+      "learning_rate": 5.883224997839482e-07,
+      "loss": 0.2786,
+      "step": 7132
+    },
+    {
+      "epoch": 1.6978044862259773,
+      "grad_norm": 0.40288142462099663,
+      "learning_rate": 5.874156615699455e-07,
+      "loss": 0.3231,
+      "step": 7133
+    },
+    {
+      "epoch": 1.698042482299042,
+      "grad_norm": 0.39261984171695463,
+      "learning_rate": 5.865094791751375e-07,
+      "loss": 0.3858,
+      "step": 7134
+    },
+    {
+      "epoch": 1.6982804783721068,
+      "grad_norm": 0.38986076248425855,
+      "learning_rate": 5.856039527342044e-07,
+      "loss": 0.3153,
+      "step": 7135
+    },
+    {
+      "epoch": 1.6985184744451716,
+      "grad_norm": 0.3869365959701934,
+      "learning_rate": 5.846990823817316e-07,
+      "loss": 0.2779,
+      "step": 7136
+    },
+    {
+      "epoch": 1.6987564705182363,
+      "grad_norm": 0.38638500561242545,
+      "learning_rate": 5.837948682522048e-07,
+      "loss": 0.3351,
+      "step": 7137
+    },
+    {
+      "epoch": 1.6989944665913013,
+      "grad_norm": 0.36095089742651565,
+      "learning_rate": 5.828913104800121e-07,
+      "loss": 0.3707,
+      "step": 7138
+    },
+    {
+      "epoch": 1.699232462664366,
+      "grad_norm": 0.37138461340811796,
+      "learning_rate": 5.819884091994444e-07,
+      "loss": 0.2501,
+      "step": 7139
+    },
+    {
+      "epoch": 1.699470458737431,
+      "grad_norm": 0.3601206372592032,
+      "learning_rate": 5.810861645446958e-07,
+      "loss": 0.2903,
+      "step": 7140
+    },
+    {
+      "epoch": 1.6997084548104957,
+      "grad_norm": 0.3929879463329381,
+      "learning_rate": 5.801845766498615e-07,
+      "loss": 0.3893,
+      "step": 7141
+    },
+    {
+      "epoch": 1.6999464508835604,
+      "grad_norm": 0.36103944808905125,
+      "learning_rate": 5.792836456489392e-07,
+      "loss": 0.3077,
+      "step": 7142
+    },
+    {
+      "epoch": 1.7001844469566252,
+      "grad_norm": 0.41181030157128806,
+      "learning_rate": 5.783833716758314e-07,
+      "loss": 0.2777,
+      "step": 7143
+    },
+    {
+      "epoch": 1.70042244302969,
+      "grad_norm": 0.36292079451006326,
+      "learning_rate": 5.774837548643403e-07,
+      "loss": 0.3314,
+      "step": 7144
+    },
+    {
+      "epoch": 1.7006604391027547,
+      "grad_norm": 0.38043247398479435,
+      "learning_rate": 5.765847953481707e-07,
+      "loss": 0.3875,
+      "step": 7145
+    },
+    {
+      "epoch": 1.7008984351758196,
+      "grad_norm": 0.38907488954363156,
+      "learning_rate": 5.756864932609307e-07,
+      "loss": 0.2598,
+      "step": 7146
+    },
+    {
+      "epoch": 1.7011364312488844,
+      "grad_norm": 0.4014263611690131,
+      "learning_rate": 5.747888487361303e-07,
+      "loss": 0.2803,
+      "step": 7147
+    },
+    {
+      "epoch": 1.7013744273219493,
+      "grad_norm": 0.37583669721002044,
+      "learning_rate": 5.738918619071809e-07,
+      "loss": 0.3621,
+      "step": 7148
+    },
+    {
+      "epoch": 1.701612423395014,
+      "grad_norm": 0.3633761512343576,
+      "learning_rate": 5.729955329073978e-07,
+      "loss": 0.3107,
+      "step": 7149
+    },
+    {
+      "epoch": 1.7018504194680788,
+      "grad_norm": 0.3922899931026979,
+      "learning_rate": 5.720998618699974e-07,
+      "loss": 0.2716,
+      "step": 7150
+    },
+    {
+      "epoch": 1.7020884155411435,
+      "grad_norm": 0.3822926529873602,
+      "learning_rate": 5.712048489280981e-07,
+      "loss": 0.3296,
+      "step": 7151
+    },
+    {
+      "epoch": 1.7023264116142083,
+      "grad_norm": 0.3755985930742893,
+      "learning_rate": 5.703104942147214e-07,
+      "loss": 0.3824,
+      "step": 7152
+    },
+    {
+      "epoch": 1.702564407687273,
+      "grad_norm": 0.36980830385212704,
+      "learning_rate": 5.694167978627907e-07,
+      "loss": 0.312,
+      "step": 7153
+    },
+    {
+      "epoch": 1.702802403760338,
+      "grad_norm": 0.432008104893982,
+      "learning_rate": 5.685237600051314e-07,
+      "loss": 0.2991,
+      "step": 7154
+    },
+    {
+      "epoch": 1.7030403998334027,
+      "grad_norm": 0.3870322232137778,
+      "learning_rate": 5.676313807744705e-07,
+      "loss": 0.3511,
+      "step": 7155
+    },
+    {
+      "epoch": 1.7032783959064677,
+      "grad_norm": 0.3975439565591834,
+      "learning_rate": 5.667396603034369e-07,
+      "loss": 0.3807,
+      "step": 7156
+    },
+    {
+      "epoch": 1.7035163919795324,
+      "grad_norm": 0.3536210783757654,
+      "learning_rate": 5.658485987245648e-07,
+      "loss": 0.2829,
+      "step": 7157
+    },
+    {
+      "epoch": 1.7037543880525972,
+      "grad_norm": 0.38718378271877263,
+      "learning_rate": 5.64958196170286e-07,
+      "loss": 0.3196,
+      "step": 7158
+    },
+    {
+      "epoch": 1.703992384125662,
+      "grad_norm": 0.40914299947041166,
+      "learning_rate": 5.640684527729373e-07,
+      "loss": 0.3649,
+      "step": 7159
+    },
+    {
+      "epoch": 1.7042303801987266,
+      "grad_norm": 0.37188576462750905,
+      "learning_rate": 5.631793686647558e-07,
+      "loss": 0.316,
+      "step": 7160
+    },
+    {
+      "epoch": 1.7044683762717914,
+      "grad_norm": 0.37633208910101296,
+      "learning_rate": 5.622909439778817e-07,
+      "loss": 0.2676,
+      "step": 7161
+    },
+    {
+      "epoch": 1.7047063723448563,
+      "grad_norm": 0.38454873470389395,
+      "learning_rate": 5.614031788443563e-07,
+      "loss": 0.3507,
+      "step": 7162
+    },
+    {
+      "epoch": 1.704944368417921,
+      "grad_norm": 0.3977198853767724,
+      "learning_rate": 5.605160733961252e-07,
+      "loss": 0.3885,
+      "step": 7163
+    },
+    {
+      "epoch": 1.705182364490986,
+      "grad_norm": 0.3933331704312476,
+      "learning_rate": 5.596296277650332e-07,
+      "loss": 0.2696,
+      "step": 7164
+    },
+    {
+      "epoch": 1.7054203605640508,
+      "grad_norm": 0.3822921368220508,
+      "learning_rate": 5.587438420828273e-07,
+      "loss": 0.3277,
+      "step": 7165
+    },
+    {
+      "epoch": 1.7056583566371155,
+      "grad_norm": 0.3861803935557107,
+      "learning_rate": 5.578587164811583e-07,
+      "loss": 0.3735,
+      "step": 7166
+    },
+    {
+      "epoch": 1.7058963527101803,
+      "grad_norm": 0.34946179290548746,
+      "learning_rate": 5.569742510915776e-07,
+      "loss": 0.292,
+      "step": 7167
+    },
+    {
+      "epoch": 1.706134348783245,
+      "grad_norm": 0.3777091948818582,
+      "learning_rate": 5.56090446045538e-07,
+      "loss": 0.2919,
+      "step": 7168
+    },
+    {
+      "epoch": 1.7063723448563097,
+      "grad_norm": 0.39415615795103837,
+      "learning_rate": 5.552073014743942e-07,
+      "loss": 0.3222,
+      "step": 7169
+    },
+    {
+      "epoch": 1.7066103409293747,
+      "grad_norm": 0.38807660041300635,
+      "learning_rate": 5.543248175094051e-07,
+      "loss": 0.3784,
+      "step": 7170
+    },
+    {
+      "epoch": 1.7068483370024394,
+      "grad_norm": 0.36815447326497414,
+      "learning_rate": 5.534429942817293e-07,
+      "loss": 0.2682,
+      "step": 7171
+    },
+    {
+      "epoch": 1.7070863330755044,
+      "grad_norm": 0.3782786257896689,
+      "learning_rate": 5.525618319224269e-07,
+      "loss": 0.3099,
+      "step": 7172
+    },
+    {
+      "epoch": 1.7073243291485691,
+      "grad_norm": 0.3774779346032418,
+      "learning_rate": 5.516813305624602e-07,
+      "loss": 0.3569,
+      "step": 7173
+    },
+    {
+      "epoch": 1.7075623252216339,
+      "grad_norm": 0.3825672141737441,
+      "learning_rate": 5.508014903326941e-07,
+      "loss": 0.339,
+      "step": 7174
+    },
+    {
+      "epoch": 1.7078003212946986,
+      "grad_norm": 0.38678799012825976,
+      "learning_rate": 5.499223113638946e-07,
+      "loss": 0.2895,
+      "step": 7175
+    },
+    {
+      "epoch": 1.7080383173677633,
+      "grad_norm": 0.45675241641075215,
+      "learning_rate": 5.490437937867287e-07,
+      "loss": 0.3156,
+      "step": 7176
+    },
+    {
+      "epoch": 1.708276313440828,
+      "grad_norm": 0.3990012412168306,
+      "learning_rate": 5.481659377317672e-07,
+      "loss": 0.4195,
+      "step": 7177
+    },
+    {
+      "epoch": 1.708514309513893,
+      "grad_norm": 0.361980501450529,
+      "learning_rate": 5.472887433294799e-07,
+      "loss": 0.2905,
+      "step": 7178
+    },
+    {
+      "epoch": 1.7087523055869578,
+      "grad_norm": 0.3934064270709845,
+      "learning_rate": 5.464122107102399e-07,
+      "loss": 0.2835,
+      "step": 7179
+    },
+    {
+      "epoch": 1.7089903016600227,
+      "grad_norm": 0.39868212939298053,
+      "learning_rate": 5.455363400043223e-07,
+      "loss": 0.3682,
+      "step": 7180
+    },
+    {
+      "epoch": 1.7092282977330875,
+      "grad_norm": 0.37266869092332056,
+      "learning_rate": 5.446611313419026e-07,
+      "loss": 0.3462,
+      "step": 7181
+    },
+    {
+      "epoch": 1.7094662938061522,
+      "grad_norm": 0.3622657497062413,
+      "learning_rate": 5.437865848530588e-07,
+      "loss": 0.2699,
+      "step": 7182
+    },
+    {
+      "epoch": 1.709704289879217,
+      "grad_norm": 0.37818014208202183,
+      "learning_rate": 5.429127006677681e-07,
+      "loss": 0.3102,
+      "step": 7183
+    },
+    {
+      "epoch": 1.7099422859522817,
+      "grad_norm": 0.4190339066632442,
+      "learning_rate": 5.420394789159151e-07,
+      "loss": 0.3571,
+      "step": 7184
+    },
+    {
+      "epoch": 1.7101802820253464,
+      "grad_norm": 0.4294561415916188,
+      "learning_rate": 5.411669197272795e-07,
+      "loss": 0.2779,
+      "step": 7185
+    },
+    {
+      "epoch": 1.7104182780984114,
+      "grad_norm": 0.3737799198142464,
+      "learning_rate": 5.402950232315457e-07,
+      "loss": 0.2662,
+      "step": 7186
+    },
+    {
+      "epoch": 1.7106562741714761,
+      "grad_norm": 0.3802806165041149,
+      "learning_rate": 5.394237895582999e-07,
+      "loss": 0.359,
+      "step": 7187
+    },
+    {
+      "epoch": 1.710894270244541,
+      "grad_norm": 0.3824886968859595,
+      "learning_rate": 5.385532188370279e-07,
+      "loss": 0.401,
+      "step": 7188
+    },
+    {
+      "epoch": 1.7111322663176058,
+      "grad_norm": 0.3682750615164672,
+      "learning_rate": 5.376833111971175e-07,
+      "loss": 0.2592,
+      "step": 7189
+    },
+    {
+      "epoch": 1.7113702623906706,
+      "grad_norm": 0.4080980675820481,
+      "learning_rate": 5.368140667678607e-07,
+      "loss": 0.3052,
+      "step": 7190
+    },
+    {
+      "epoch": 1.7116082584637353,
+      "grad_norm": 0.4315797792499158,
+      "learning_rate": 5.359454856784469e-07,
+      "loss": 0.3828,
+      "step": 7191
+    },
+    {
+      "epoch": 1.7118462545368,
+      "grad_norm": 0.3873206456391449,
+      "learning_rate": 5.350775680579695e-07,
+      "loss": 0.2887,
+      "step": 7192
+    },
+    {
+      "epoch": 1.7120842506098648,
+      "grad_norm": 0.39557005651691873,
+      "learning_rate": 5.342103140354226e-07,
+      "loss": 0.2943,
+      "step": 7193
+    },
+    {
+      "epoch": 1.7123222466829298,
+      "grad_norm": 0.3875924022674923,
+      "learning_rate": 5.333437237397015e-07,
+      "loss": 0.3258,
+      "step": 7194
+    },
+    {
+      "epoch": 1.7125602427559945,
+      "grad_norm": 0.39590844421878507,
+      "learning_rate": 5.324777972996026e-07,
+      "loss": 0.3854,
+      "step": 7195
+    },
+    {
+      "epoch": 1.7127982388290595,
+      "grad_norm": 0.37400174892709304,
+      "learning_rate": 5.316125348438239e-07,
+      "loss": 0.2619,
+      "step": 7196
+    },
+    {
+      "epoch": 1.7130362349021242,
+      "grad_norm": 0.40507111453694483,
+      "learning_rate": 5.307479365009644e-07,
+      "loss": 0.2943,
+      "step": 7197
+    },
+    {
+      "epoch": 1.713274230975189,
+      "grad_norm": 0.41809861423219297,
+      "learning_rate": 5.298840023995267e-07,
+      "loss": 0.3762,
+      "step": 7198
+    },
+    {
+      "epoch": 1.7135122270482537,
+      "grad_norm": 0.3901582679717628,
+      "learning_rate": 5.290207326679109e-07,
+      "loss": 0.3525,
+      "step": 7199
+    },
+    {
+      "epoch": 1.7137502231213184,
+      "grad_norm": 0.37330950872565466,
+      "learning_rate": 5.281581274344216e-07,
+      "loss": 0.2854,
+      "step": 7200
+    },
+    {
+      "epoch": 1.7139882191943832,
+      "grad_norm": 0.3781668431340482,
+      "learning_rate": 5.272961868272625e-07,
+      "loss": 0.3256,
+      "step": 7201
+    },
+    {
+      "epoch": 1.7142262152674481,
+      "grad_norm": 0.38276965275384967,
+      "learning_rate": 5.264349109745392e-07,
+      "loss": 0.3804,
+      "step": 7202
+    },
+    {
+      "epoch": 1.7144642113405129,
+      "grad_norm": 0.37802592125250717,
+      "learning_rate": 5.25574300004259e-07,
+      "loss": 0.2817,
+      "step": 7203
+    },
+    {
+      "epoch": 1.7147022074135778,
+      "grad_norm": 0.39422912395070164,
+      "learning_rate": 5.247143540443295e-07,
+      "loss": 0.3247,
+      "step": 7204
+    },
+    {
+      "epoch": 1.7149402034866426,
+      "grad_norm": 0.35708805870299815,
+      "learning_rate": 5.2385507322256e-07,
+      "loss": 0.3755,
+      "step": 7205
+    },
+    {
+      "epoch": 1.7151781995597073,
+      "grad_norm": 0.3733848792164133,
+      "learning_rate": 5.229964576666618e-07,
+      "loss": 0.3567,
+      "step": 7206
+    },
+    {
+      "epoch": 1.715416195632772,
+      "grad_norm": 0.4043891988399861,
+      "learning_rate": 5.221385075042451e-07,
+      "loss": 0.2927,
+      "step": 7207
+    },
+    {
+      "epoch": 1.7156541917058368,
+      "grad_norm": 0.36868782474294576,
+      "learning_rate": 5.212812228628234e-07,
+      "loss": 0.3259,
+      "step": 7208
+    },
+    {
+      "epoch": 1.7158921877789015,
+      "grad_norm": 0.38185612816752407,
+      "learning_rate": 5.204246038698102e-07,
+      "loss": 0.39,
+      "step": 7209
+    },
+    {
+      "epoch": 1.7161301838519665,
+      "grad_norm": 0.3691193320372495,
+      "learning_rate": 5.195686506525205e-07,
+      "loss": 0.306,
+      "step": 7210
+    },
+    {
+      "epoch": 1.7163681799250312,
+      "grad_norm": 0.4010679948783104,
+      "learning_rate": 5.187133633381686e-07,
+      "loss": 0.2615,
+      "step": 7211
+    },
+    {
+      "epoch": 1.7166061759980962,
+      "grad_norm": 0.3691780944853636,
+      "learning_rate": 5.178587420538733e-07,
+      "loss": 0.3175,
+      "step": 7212
+    },
+    {
+      "epoch": 1.716844172071161,
+      "grad_norm": 0.3869003239326718,
+      "learning_rate": 5.17004786926652e-07,
+      "loss": 0.3637,
+      "step": 7213
+    },
+    {
+      "epoch": 1.7170821681442257,
+      "grad_norm": 0.39395271592154035,
+      "learning_rate": 5.161514980834232e-07,
+      "loss": 0.2891,
+      "step": 7214
+    },
+    {
+      "epoch": 1.7173201642172904,
+      "grad_norm": 0.3683495140171302,
+      "learning_rate": 5.152988756510063e-07,
+      "loss": 0.3021,
+      "step": 7215
+    },
+    {
+      "epoch": 1.7175581602903551,
+      "grad_norm": 0.37064038864613685,
+      "learning_rate": 5.144469197561231e-07,
+      "loss": 0.3637,
+      "step": 7216
+    },
+    {
+      "epoch": 1.7177961563634199,
+      "grad_norm": 0.387245884205873,
+      "learning_rate": 5.135956305253953e-07,
+      "loss": 0.3318,
+      "step": 7217
+    },
+    {
+      "epoch": 1.7180341524364848,
+      "grad_norm": 0.3648749289239642,
+      "learning_rate": 5.127450080853447e-07,
+      "loss": 0.2687,
+      "step": 7218
+    },
+    {
+      "epoch": 1.7182721485095496,
+      "grad_norm": 0.3709832312749167,
+      "learning_rate": 5.118950525623955e-07,
+      "loss": 0.315,
+      "step": 7219
+    },
+    {
+      "epoch": 1.7185101445826145,
+      "grad_norm": 0.3771433733813478,
+      "learning_rate": 5.110457640828714e-07,
+      "loss": 0.3693,
+      "step": 7220
+    },
+    {
+      "epoch": 1.7187481406556793,
+      "grad_norm": 0.37194166595478395,
+      "learning_rate": 5.101971427729985e-07,
+      "loss": 0.306,
+      "step": 7221
+    },
+    {
+      "epoch": 1.718986136728744,
+      "grad_norm": 0.4198291358677723,
+      "learning_rate": 5.09349188758903e-07,
+      "loss": 0.2858,
+      "step": 7222
+    },
+    {
+      "epoch": 1.7192241328018087,
+      "grad_norm": 0.4062451136719214,
+      "learning_rate": 5.085019021666104e-07,
+      "loss": 0.3829,
+      "step": 7223
+    },
+    {
+      "epoch": 1.7194621288748735,
+      "grad_norm": 0.38423840200623954,
+      "learning_rate": 5.076552831220505e-07,
+      "loss": 0.3337,
+      "step": 7224
+    },
+    {
+      "epoch": 1.7197001249479382,
+      "grad_norm": 0.3799743269319547,
+      "learning_rate": 5.068093317510492e-07,
+      "loss": 0.2786,
+      "step": 7225
+    },
+    {
+      "epoch": 1.7199381210210032,
+      "grad_norm": 0.378826385850787,
+      "learning_rate": 5.059640481793382e-07,
+      "loss": 0.3434,
+      "step": 7226
+    },
+    {
+      "epoch": 1.720176117094068,
+      "grad_norm": 0.38264535191194776,
+      "learning_rate": 5.05119432532547e-07,
+      "loss": 0.3617,
+      "step": 7227
+    },
+    {
+      "epoch": 1.7204141131671329,
+      "grad_norm": 0.4302946941716753,
+      "learning_rate": 5.042754849362063e-07,
+      "loss": 0.2898,
+      "step": 7228
+    },
+    {
+      "epoch": 1.7206521092401976,
+      "grad_norm": 0.35178529312525003,
+      "learning_rate": 5.03432205515747e-07,
+      "loss": 0.2742,
+      "step": 7229
+    },
+    {
+      "epoch": 1.7208901053132624,
+      "grad_norm": 0.3644446480635659,
+      "learning_rate": 5.025895943965021e-07,
+      "loss": 0.3651,
+      "step": 7230
+    },
+    {
+      "epoch": 1.721128101386327,
+      "grad_norm": 0.35818553869400055,
+      "learning_rate": 5.01747651703704e-07,
+      "loss": 0.3395,
+      "step": 7231
+    },
+    {
+      "epoch": 1.7213660974593918,
+      "grad_norm": 0.3840129788680463,
+      "learning_rate": 5.009063775624857e-07,
+      "loss": 0.2691,
+      "step": 7232
+    },
+    {
+      "epoch": 1.7216040935324566,
+      "grad_norm": 0.3982821339866859,
+      "learning_rate": 5.000657720978824e-07,
+      "loss": 0.3457,
+      "step": 7233
+    },
+    {
+      "epoch": 1.7218420896055215,
+      "grad_norm": 0.40131118490637396,
+      "learning_rate": 4.992258354348284e-07,
+      "loss": 0.3598,
+      "step": 7234
+    },
+    {
+      "epoch": 1.7220800856785863,
+      "grad_norm": 0.38055567769870746,
+      "learning_rate": 4.983865676981586e-07,
+      "loss": 0.2975,
+      "step": 7235
+    },
+    {
+      "epoch": 1.7223180817516512,
+      "grad_norm": 0.41080399456136174,
+      "learning_rate": 4.9754796901261e-07,
+      "loss": 0.2739,
+      "step": 7236
+    },
+    {
+      "epoch": 1.722556077824716,
+      "grad_norm": 0.3662126709880048,
+      "learning_rate": 4.96710039502818e-07,
+      "loss": 0.3467,
+      "step": 7237
+    },
+    {
+      "epoch": 1.7227940738977807,
+      "grad_norm": 0.3938043911249227,
+      "learning_rate": 4.958727792933194e-07,
+      "loss": 0.3637,
+      "step": 7238
+    },
+    {
+      "epoch": 1.7230320699708455,
+      "grad_norm": 0.4031184138478153,
+      "learning_rate": 4.950361885085536e-07,
+      "loss": 0.2873,
+      "step": 7239
+    },
+    {
+      "epoch": 1.7232700660439102,
+      "grad_norm": 0.4025403942805333,
+      "learning_rate": 4.942002672728575e-07,
+      "loss": 0.2969,
+      "step": 7240
+    },
+    {
+      "epoch": 1.723508062116975,
+      "grad_norm": 0.3959523479792395,
+      "learning_rate": 4.933650157104697e-07,
+      "loss": 0.3719,
+      "step": 7241
+    },
+    {
+      "epoch": 1.72374605819004,
+      "grad_norm": 0.36786887087342524,
+      "learning_rate": 4.925304339455289e-07,
+      "loss": 0.3423,
+      "step": 7242
+    },
+    {
+      "epoch": 1.7239840542631046,
+      "grad_norm": 0.3874665309151205,
+      "learning_rate": 4.916965221020753e-07,
+      "loss": 0.2973,
+      "step": 7243
+    },
+    {
+      "epoch": 1.7242220503361696,
+      "grad_norm": 0.3880295711150305,
+      "learning_rate": 4.908632803040492e-07,
+      "loss": 0.3129,
+      "step": 7244
+    },
+    {
+      "epoch": 1.7244600464092343,
+      "grad_norm": 0.3748397703051552,
+      "learning_rate": 4.900307086752898e-07,
+      "loss": 0.3819,
+      "step": 7245
+    },
+    {
+      "epoch": 1.724698042482299,
+      "grad_norm": 0.359732975677239,
+      "learning_rate": 4.891988073395382e-07,
+      "loss": 0.2676,
+      "step": 7246
+    },
+    {
+      "epoch": 1.7249360385553638,
+      "grad_norm": 0.35344522976400145,
+      "learning_rate": 4.88367576420436e-07,
+      "loss": 0.286,
+      "step": 7247
+    },
+    {
+      "epoch": 1.7251740346284286,
+      "grad_norm": 0.3568660899520547,
+      "learning_rate": 4.875370160415243e-07,
+      "loss": 0.3399,
+      "step": 7248
+    },
+    {
+      "epoch": 1.7254120307014933,
+      "grad_norm": 0.371131074798834,
+      "learning_rate": 4.867071263262452e-07,
+      "loss": 0.3362,
+      "step": 7249
+    },
+    {
+      "epoch": 1.7256500267745583,
+      "grad_norm": 0.3910514202327229,
+      "learning_rate": 4.858779073979408e-07,
+      "loss": 0.2673,
+      "step": 7250
+    },
+    {
+      "epoch": 1.725888022847623,
+      "grad_norm": 0.3545398248925455,
+      "learning_rate": 4.850493593798528e-07,
+      "loss": 0.3272,
+      "step": 7251
+    },
+    {
+      "epoch": 1.726126018920688,
+      "grad_norm": 0.3979049249797813,
+      "learning_rate": 4.842214823951236e-07,
+      "loss": 0.359,
+      "step": 7252
+    },
+    {
+      "epoch": 1.7263640149937527,
+      "grad_norm": 0.3517488882321978,
+      "learning_rate": 4.833942765667981e-07,
+      "loss": 0.2883,
+      "step": 7253
+    },
+    {
+      "epoch": 1.7266020110668174,
+      "grad_norm": 0.37339923080107096,
+      "learning_rate": 4.825677420178187e-07,
+      "loss": 0.2837,
+      "step": 7254
+    },
+    {
+      "epoch": 1.7268400071398822,
+      "grad_norm": 0.3888091649358466,
+      "learning_rate": 4.817418788710287e-07,
+      "loss": 0.3544,
+      "step": 7255
+    },
+    {
+      "epoch": 1.727078003212947,
+      "grad_norm": 0.37811647136667986,
+      "learning_rate": 4.809166872491716e-07,
+      "loss": 0.3118,
+      "step": 7256
+    },
+    {
+      "epoch": 1.7273159992860116,
+      "grad_norm": 0.3737143324650708,
+      "learning_rate": 4.800921672748921e-07,
+      "loss": 0.267,
+      "step": 7257
+    },
+    {
+      "epoch": 1.7275539953590766,
+      "grad_norm": 0.48902821392635987,
+      "learning_rate": 4.792683190707331e-07,
+      "loss": 0.3132,
+      "step": 7258
+    },
+    {
+      "epoch": 1.7277919914321413,
+      "grad_norm": 0.3593715757632593,
+      "learning_rate": 4.784451427591396e-07,
+      "loss": 0.3794,
+      "step": 7259
+    },
+    {
+      "epoch": 1.7280299875052063,
+      "grad_norm": 0.37488004076926507,
+      "learning_rate": 4.776226384624555e-07,
+      "loss": 0.2982,
+      "step": 7260
+    },
+    {
+      "epoch": 1.728267983578271,
+      "grad_norm": 0.4022945491362904,
+      "learning_rate": 4.7680080630292613e-07,
+      "loss": 0.3157,
+      "step": 7261
+    },
+    {
+      "epoch": 1.7285059796513358,
+      "grad_norm": 0.39397197065419187,
+      "learning_rate": 4.75979646402695e-07,
+      "loss": 0.3123,
+      "step": 7262
+    },
+    {
+      "epoch": 1.7287439757244005,
+      "grad_norm": 0.40175779262681727,
+      "learning_rate": 4.7515915888380724e-07,
+      "loss": 0.3971,
+      "step": 7263
+    },
+    {
+      "epoch": 1.7289819717974653,
+      "grad_norm": 0.4001343316717571,
+      "learning_rate": 4.7433934386820813e-07,
+      "loss": 0.2903,
+      "step": 7264
+    },
+    {
+      "epoch": 1.72921996787053,
+      "grad_norm": 0.39178481360532014,
+      "learning_rate": 4.7352020147774067e-07,
+      "loss": 0.3215,
+      "step": 7265
+    },
+    {
+      "epoch": 1.729457963943595,
+      "grad_norm": 0.3668638438326148,
+      "learning_rate": 4.7270173183415203e-07,
+      "loss": 0.3626,
+      "step": 7266
+    },
+    {
+      "epoch": 1.7296959600166597,
+      "grad_norm": 0.381515302995729,
+      "learning_rate": 4.7188393505908594e-07,
+      "loss": 0.3434,
+      "step": 7267
+    },
+    {
+      "epoch": 1.7299339560897247,
+      "grad_norm": 0.38224529914825683,
+      "learning_rate": 4.710668112740874e-07,
+      "loss": 0.2686,
+      "step": 7268
+    },
+    {
+      "epoch": 1.7301719521627894,
+      "grad_norm": 0.39868968626428386,
+      "learning_rate": 4.7025036060059983e-07,
+      "loss": 0.3514,
+      "step": 7269
+    },
+    {
+      "epoch": 1.7304099482358541,
+      "grad_norm": 0.3877292058096328,
+      "learning_rate": 4.694345831599706e-07,
+      "loss": 0.373,
+      "step": 7270
+    },
+    {
+      "epoch": 1.7306479443089189,
+      "grad_norm": 0.364088046457619,
+      "learning_rate": 4.686194790734427e-07,
+      "loss": 0.2926,
+      "step": 7271
+    },
+    {
+      "epoch": 1.7308859403819836,
+      "grad_norm": 0.3655121541245187,
+      "learning_rate": 4.6780504846216155e-07,
+      "loss": 0.3324,
+      "step": 7272
+    },
+    {
+      "epoch": 1.7311239364550484,
+      "grad_norm": 0.37380101505935226,
+      "learning_rate": 4.6699129144717135e-07,
+      "loss": 0.3662,
+      "step": 7273
+    },
+    {
+      "epoch": 1.7313619325281133,
+      "grad_norm": 0.35389840940426953,
+      "learning_rate": 4.6617820814941594e-07,
+      "loss": 0.3308,
+      "step": 7274
+    },
+    {
+      "epoch": 1.731599928601178,
+      "grad_norm": 0.37982947460360994,
+      "learning_rate": 4.6536579868974083e-07,
+      "loss": 0.2662,
+      "step": 7275
+    },
+    {
+      "epoch": 1.731837924674243,
+      "grad_norm": 0.38660456006967564,
+      "learning_rate": 4.6455406318888896e-07,
+      "loss": 0.3559,
+      "step": 7276
+    },
+    {
+      "epoch": 1.7320759207473078,
+      "grad_norm": 0.42338649966136893,
+      "learning_rate": 4.6374300176750484e-07,
+      "loss": 0.3795,
+      "step": 7277
+    },
+    {
+      "epoch": 1.7323139168203725,
+      "grad_norm": 0.364628505385949,
+      "learning_rate": 4.629326145461327e-07,
+      "loss": 0.2946,
+      "step": 7278
+    },
+    {
+      "epoch": 1.7325519128934372,
+      "grad_norm": 0.36740380393685235,
+      "learning_rate": 4.6212290164521554e-07,
+      "loss": 0.3445,
+      "step": 7279
+    },
+    {
+      "epoch": 1.732789908966502,
+      "grad_norm": 0.40395669425071407,
+      "learning_rate": 4.613138631850955e-07,
+      "loss": 0.3394,
+      "step": 7280
+    },
+    {
+      "epoch": 1.7330279050395667,
+      "grad_norm": 0.3954298208790007,
+      "learning_rate": 4.6050549928601864e-07,
+      "loss": 0.3488,
+      "step": 7281
+    },
+    {
+      "epoch": 1.7332659011126317,
+      "grad_norm": 0.3667594695117182,
+      "learning_rate": 4.59697810068126e-07,
+      "loss": 0.3099,
+      "step": 7282
+    },
+    {
+      "epoch": 1.7335038971856964,
+      "grad_norm": 0.37536423714685246,
+      "learning_rate": 4.588907956514599e-07,
+      "loss": 0.3318,
+      "step": 7283
+    },
+    {
+      "epoch": 1.7337418932587614,
+      "grad_norm": 0.4090601125824329,
+      "learning_rate": 4.5808445615596386e-07,
+      "loss": 0.3491,
+      "step": 7284
+    },
+    {
+      "epoch": 1.7339798893318261,
+      "grad_norm": 0.40724673701790437,
+      "learning_rate": 4.5727879170147927e-07,
+      "loss": 0.2832,
+      "step": 7285
+    },
+    {
+      "epoch": 1.7342178854048909,
+      "grad_norm": 0.4192280684140144,
+      "learning_rate": 4.564738024077475e-07,
+      "loss": 0.2829,
+      "step": 7286
+    },
+    {
+      "epoch": 1.7344558814779556,
+      "grad_norm": 0.3683379060762051,
+      "learning_rate": 4.5566948839441014e-07,
+      "loss": 0.3196,
+      "step": 7287
+    },
+    {
+      "epoch": 1.7346938775510203,
+      "grad_norm": 0.37268180835647635,
+      "learning_rate": 4.5486584978100766e-07,
+      "loss": 0.3712,
+      "step": 7288
+    },
+    {
+      "epoch": 1.734931873624085,
+      "grad_norm": 0.3896434055889146,
+      "learning_rate": 4.5406288668698175e-07,
+      "loss": 0.2722,
+      "step": 7289
+    },
+    {
+      "epoch": 1.73516986969715,
+      "grad_norm": 0.37283077254852875,
+      "learning_rate": 4.5326059923167185e-07,
+      "loss": 0.2979,
+      "step": 7290
+    },
+    {
+      "epoch": 1.7354078657702148,
+      "grad_norm": 0.40147117472261845,
+      "learning_rate": 4.524589875343177e-07,
+      "loss": 0.3637,
+      "step": 7291
+    },
+    {
+      "epoch": 1.7356458618432797,
+      "grad_norm": 0.3635087671699396,
+      "learning_rate": 4.5165805171405786e-07,
+      "loss": 0.3019,
+      "step": 7292
+    },
+    {
+      "epoch": 1.7358838579163445,
+      "grad_norm": 0.359877509083195,
+      "learning_rate": 4.508577918899326e-07,
+      "loss": 0.2844,
+      "step": 7293
+    },
+    {
+      "epoch": 1.7361218539894092,
+      "grad_norm": 0.3989712129889001,
+      "learning_rate": 4.500582081808802e-07,
+      "loss": 0.3003,
+      "step": 7294
+    },
+    {
+      "epoch": 1.736359850062474,
+      "grad_norm": 0.43670169269521797,
+      "learning_rate": 4.492593007057383e-07,
+      "loss": 0.3806,
+      "step": 7295
+    },
+    {
+      "epoch": 1.7365978461355387,
+      "grad_norm": 0.4151227878379709,
+      "learning_rate": 4.484610695832431e-07,
+      "loss": 0.2829,
+      "step": 7296
+    },
+    {
+      "epoch": 1.7368358422086034,
+      "grad_norm": 0.39427740496318525,
+      "learning_rate": 4.476635149320341e-07,
+      "loss": 0.2821,
+      "step": 7297
+    },
+    {
+      "epoch": 1.7370738382816684,
+      "grad_norm": 0.39483412534077605,
+      "learning_rate": 4.4686663687064537e-07,
+      "loss": 0.3707,
+      "step": 7298
+    },
+    {
+      "epoch": 1.7373118343547331,
+      "grad_norm": 0.3811903403162823,
+      "learning_rate": 4.4607043551751385e-07,
+      "loss": 0.3131,
+      "step": 7299
+    },
+    {
+      "epoch": 1.737549830427798,
+      "grad_norm": 0.3787340344910432,
+      "learning_rate": 4.452749109909743e-07,
+      "loss": 0.2882,
+      "step": 7300
+    },
+    {
+      "epoch": 1.7377878265008628,
+      "grad_norm": 0.3977977827225253,
+      "learning_rate": 4.4448006340926163e-07,
+      "loss": 0.2914,
+      "step": 7301
+    },
+    {
+      "epoch": 1.7380258225739276,
+      "grad_norm": 0.3471686985062694,
+      "learning_rate": 4.4368589289050966e-07,
+      "loss": 0.3736,
+      "step": 7302
+    },
+    {
+      "epoch": 1.7382638186469923,
+      "grad_norm": 0.3567461740090027,
+      "learning_rate": 4.428923995527512e-07,
+      "loss": 0.2677,
+      "step": 7303
+    },
+    {
+      "epoch": 1.738501814720057,
+      "grad_norm": 0.3895718708014465,
+      "learning_rate": 4.4209958351392024e-07,
+      "loss": 0.2761,
+      "step": 7304
+    },
+    {
+      "epoch": 1.7387398107931218,
+      "grad_norm": 0.4146289851718209,
+      "learning_rate": 4.4130744489184805e-07,
+      "loss": 0.336,
+      "step": 7305
+    },
+    {
+      "epoch": 1.7389778068661867,
+      "grad_norm": 0.36115931059450657,
+      "learning_rate": 4.4051598380426606e-07,
+      "loss": 0.3318,
+      "step": 7306
+    },
+    {
+      "epoch": 1.7392158029392515,
+      "grad_norm": 0.3901592985931496,
+      "learning_rate": 4.3972520036880406e-07,
+      "loss": 0.2939,
+      "step": 7307
+    },
+    {
+      "epoch": 1.7394537990123164,
+      "grad_norm": 0.3861549670050002,
+      "learning_rate": 4.389350947029941e-07,
+      "loss": 0.3264,
+      "step": 7308
+    },
+    {
+      "epoch": 1.7396917950853812,
+      "grad_norm": 0.39366268242787633,
+      "learning_rate": 4.381456669242645e-07,
+      "loss": 0.3827,
+      "step": 7309
+    },
+    {
+      "epoch": 1.739929791158446,
+      "grad_norm": 0.4147331435398097,
+      "learning_rate": 4.3735691714994366e-07,
+      "loss": 0.3196,
+      "step": 7310
+    },
+    {
+      "epoch": 1.7401677872315107,
+      "grad_norm": 0.4033560437583594,
+      "learning_rate": 4.365688454972589e-07,
+      "loss": 0.284,
+      "step": 7311
+    },
+    {
+      "epoch": 1.7404057833045754,
+      "grad_norm": 0.36568041838421067,
+      "learning_rate": 4.357814520833381e-07,
+      "loss": 0.3219,
+      "step": 7312
+    },
+    {
+      "epoch": 1.7406437793776401,
+      "grad_norm": 0.33395093697829437,
+      "learning_rate": 4.349947370252067e-07,
+      "loss": 0.3806,
+      "step": 7313
+    },
+    {
+      "epoch": 1.740881775450705,
+      "grad_norm": 0.37826064768858925,
+      "learning_rate": 4.342087004397899e-07,
+      "loss": 0.3458,
+      "step": 7314
+    },
+    {
+      "epoch": 1.7411197715237698,
+      "grad_norm": 0.36042290741386074,
+      "learning_rate": 4.3342334244391215e-07,
+      "loss": 0.2922,
+      "step": 7315
+    },
+    {
+      "epoch": 1.7413577675968348,
+      "grad_norm": 0.3742173084173869,
+      "learning_rate": 4.326386631542978e-07,
+      "loss": 0.3492,
+      "step": 7316
+    },
+    {
+      "epoch": 1.7415957636698995,
+      "grad_norm": 0.3611783561237634,
+      "learning_rate": 4.3185466268756916e-07,
+      "loss": 0.3284,
+      "step": 7317
+    },
+    {
+      "epoch": 1.7418337597429643,
+      "grad_norm": 0.36444937618317436,
+      "learning_rate": 4.310713411602485e-07,
+      "loss": 0.284,
+      "step": 7318
+    },
+    {
+      "epoch": 1.742071755816029,
+      "grad_norm": 0.3619558848348502,
+      "learning_rate": 4.30288698688755e-07,
+      "loss": 0.3108,
+      "step": 7319
+    },
+    {
+      "epoch": 1.7423097518890938,
+      "grad_norm": 0.40390658897210363,
+      "learning_rate": 4.295067353894111e-07,
+      "loss": 0.3621,
+      "step": 7320
+    },
+    {
+      "epoch": 1.7425477479621585,
+      "grad_norm": 0.40363738867672744,
+      "learning_rate": 4.28725451378435e-07,
+      "loss": 0.2972,
+      "step": 7321
+    },
+    {
+      "epoch": 1.7427857440352235,
+      "grad_norm": 0.3606197822662685,
+      "learning_rate": 4.279448467719444e-07,
+      "loss": 0.2996,
+      "step": 7322
+    },
+    {
+      "epoch": 1.7430237401082882,
+      "grad_norm": 0.37968170265939494,
+      "learning_rate": 4.271649216859558e-07,
+      "loss": 0.3488,
+      "step": 7323
+    },
+    {
+      "epoch": 1.7432617361813532,
+      "grad_norm": 0.3581004782699803,
+      "learning_rate": 4.263856762363877e-07,
+      "loss": 0.3275,
+      "step": 7324
+    },
+    {
+      "epoch": 1.743499732254418,
+      "grad_norm": 0.36888664733014503,
+      "learning_rate": 4.256071105390536e-07,
+      "loss": 0.2782,
+      "step": 7325
+    },
+    {
+      "epoch": 1.7437377283274826,
+      "grad_norm": 0.3736404487957657,
+      "learning_rate": 4.2482922470966804e-07,
+      "loss": 0.3517,
+      "step": 7326
+    },
+    {
+      "epoch": 1.7439757244005474,
+      "grad_norm": 0.3838130180426928,
+      "learning_rate": 4.2405201886384364e-07,
+      "loss": 0.3888,
+      "step": 7327
+    },
+    {
+      "epoch": 1.7442137204736121,
+      "grad_norm": 0.39228239050902514,
+      "learning_rate": 4.232754931170929e-07,
+      "loss": 0.2859,
+      "step": 7328
+    },
+    {
+      "epoch": 1.7444517165466769,
+      "grad_norm": 0.38813077534276835,
+      "learning_rate": 4.22499647584827e-07,
+      "loss": 0.2981,
+      "step": 7329
+    },
+    {
+      "epoch": 1.7446897126197418,
+      "grad_norm": 0.37696497911373844,
+      "learning_rate": 4.2172448238235464e-07,
+      "loss": 0.3654,
+      "step": 7330
+    },
+    {
+      "epoch": 1.7449277086928066,
+      "grad_norm": 0.39042415238351036,
+      "learning_rate": 4.2094999762488597e-07,
+      "loss": 0.3358,
+      "step": 7331
+    },
+    {
+      "epoch": 1.7451657047658715,
+      "grad_norm": 0.3811058522780075,
+      "learning_rate": 4.2017619342752723e-07,
+      "loss": 0.2515,
+      "step": 7332
+    },
+    {
+      "epoch": 1.7454037008389363,
+      "grad_norm": 0.3830369971256063,
+      "learning_rate": 4.194030699052859e-07,
+      "loss": 0.3463,
+      "step": 7333
+    },
+    {
+      "epoch": 1.745641696912001,
+      "grad_norm": 0.3863649827385546,
+      "learning_rate": 4.1863062717306724e-07,
+      "loss": 0.3765,
+      "step": 7334
+    },
+    {
+      "epoch": 1.7458796929850657,
+      "grad_norm": 0.3740836674286429,
+      "learning_rate": 4.178588653456733e-07,
+      "loss": 0.2836,
+      "step": 7335
+    },
+    {
+      "epoch": 1.7461176890581305,
+      "grad_norm": 0.40533187846971,
+      "learning_rate": 4.1708778453781017e-07,
+      "loss": 0.285,
+      "step": 7336
+    },
+    {
+      "epoch": 1.7463556851311952,
+      "grad_norm": 0.37887619317785987,
+      "learning_rate": 4.163173848640778e-07,
+      "loss": 0.356,
+      "step": 7337
+    },
+    {
+      "epoch": 1.7465936812042602,
+      "grad_norm": 0.3617723264544318,
+      "learning_rate": 4.155476664389768e-07,
+      "loss": 0.3547,
+      "step": 7338
+    },
+    {
+      "epoch": 1.746831677277325,
+      "grad_norm": 0.37992103116699155,
+      "learning_rate": 4.147786293769068e-07,
+      "loss": 0.3065,
+      "step": 7339
+    },
+    {
+      "epoch": 1.7470696733503899,
+      "grad_norm": 0.3848821730711763,
+      "learning_rate": 4.140102737921653e-07,
+      "loss": 0.2986,
+      "step": 7340
+    },
+    {
+      "epoch": 1.7473076694234546,
+      "grad_norm": 0.37504711339234986,
+      "learning_rate": 4.1324259979894865e-07,
+      "loss": 0.3887,
+      "step": 7341
+    },
+    {
+      "epoch": 1.7475456654965194,
+      "grad_norm": 0.38275904514715237,
+      "learning_rate": 4.1247560751135283e-07,
+      "loss": 0.3301,
+      "step": 7342
+    },
+    {
+      "epoch": 1.747783661569584,
+      "grad_norm": 0.38728504743609043,
+      "learning_rate": 4.117092970433717e-07,
+      "loss": 0.2825,
+      "step": 7343
+    },
+    {
+      "epoch": 1.7480216576426488,
+      "grad_norm": 0.3869386086927362,
+      "learning_rate": 4.10943668508898e-07,
+      "loss": 0.3368,
+      "step": 7344
+    },
+    {
+      "epoch": 1.7482596537157136,
+      "grad_norm": 0.41661324253240994,
+      "learning_rate": 4.101787220217229e-07,
+      "loss": 0.3815,
+      "step": 7345
+    },
+    {
+      "epoch": 1.7484976497887785,
+      "grad_norm": 0.34535425484790955,
+      "learning_rate": 4.09414457695535e-07,
+      "loss": 0.2993,
+      "step": 7346
+    },
+    {
+      "epoch": 1.7487356458618433,
+      "grad_norm": 0.3806804219160961,
+      "learning_rate": 4.0865087564392556e-07,
+      "loss": 0.3206,
+      "step": 7347
+    },
+    {
+      "epoch": 1.7489736419349082,
+      "grad_norm": 0.39462354482109935,
+      "learning_rate": 4.0788797598038054e-07,
+      "loss": 0.3489,
+      "step": 7348
+    },
+    {
+      "epoch": 1.749211638007973,
+      "grad_norm": 0.3527473981630956,
+      "learning_rate": 4.0712575881828585e-07,
+      "loss": 0.3032,
+      "step": 7349
+    },
+    {
+      "epoch": 1.7494496340810377,
+      "grad_norm": 0.37882523971462584,
+      "learning_rate": 4.0636422427092483e-07,
+      "loss": 0.2507,
+      "step": 7350
+    },
+    {
+      "epoch": 1.7496876301541024,
+      "grad_norm": 0.4076311890215,
+      "learning_rate": 4.056033724514813e-07,
+      "loss": 0.3176,
+      "step": 7351
+    },
+    {
+      "epoch": 1.7499256262271672,
+      "grad_norm": 0.38737649623817405,
+      "learning_rate": 4.048432034730371e-07,
+      "loss": 0.3973,
+      "step": 7352
+    },
+    {
+      "epoch": 1.750163622300232,
+      "grad_norm": 0.36019762556095697,
+      "learning_rate": 4.040837174485718e-07,
+      "loss": 0.2926,
+      "step": 7353
+    },
+    {
+      "epoch": 1.7504016183732969,
+      "grad_norm": 0.3757932232128984,
+      "learning_rate": 4.033249144909629e-07,
+      "loss": 0.288,
+      "step": 7354
+    },
+    {
+      "epoch": 1.7506396144463616,
+      "grad_norm": 0.3784418578541022,
+      "learning_rate": 4.0256679471298856e-07,
+      "loss": 0.3387,
+      "step": 7355
+    },
+    {
+      "epoch": 1.7508776105194266,
+      "grad_norm": 0.36293048107545833,
+      "learning_rate": 4.01809358227323e-07,
+      "loss": 0.3649,
+      "step": 7356
+    },
+    {
+      "epoch": 1.7511156065924913,
+      "grad_norm": 0.37609039329825666,
+      "learning_rate": 4.0105260514654076e-07,
+      "loss": 0.3159,
+      "step": 7357
+    },
+    {
+      "epoch": 1.751353602665556,
+      "grad_norm": 0.4037114023056679,
+      "learning_rate": 4.00296535583114e-07,
+      "loss": 0.3318,
+      "step": 7358
+    },
+    {
+      "epoch": 1.7515915987386208,
+      "grad_norm": 0.42110395329000605,
+      "learning_rate": 3.9954114964941336e-07,
+      "loss": 0.3875,
+      "step": 7359
+    },
+    {
+      "epoch": 1.7518295948116855,
+      "grad_norm": 0.39591006197133266,
+      "learning_rate": 3.9878644745770745e-07,
+      "loss": 0.3029,
+      "step": 7360
+    },
+    {
+      "epoch": 1.7520675908847503,
+      "grad_norm": 0.395799256886029,
+      "learning_rate": 3.9803242912016427e-07,
+      "loss": 0.288,
+      "step": 7361
+    },
+    {
+      "epoch": 1.7523055869578152,
+      "grad_norm": 0.397500164340299,
+      "learning_rate": 3.972790947488481e-07,
+      "loss": 0.3269,
+      "step": 7362
+    },
+    {
+      "epoch": 1.75254358303088,
+      "grad_norm": 0.40196952358276516,
+      "learning_rate": 3.965264444557254e-07,
+      "loss": 0.3584,
+      "step": 7363
+    },
+    {
+      "epoch": 1.752781579103945,
+      "grad_norm": 0.37722120669417086,
+      "learning_rate": 3.9577447835265734e-07,
+      "loss": 0.2634,
+      "step": 7364
+    },
+    {
+      "epoch": 1.7530195751770097,
+      "grad_norm": 0.40416706874239217,
+      "learning_rate": 3.950231965514051e-07,
+      "loss": 0.3041,
+      "step": 7365
+    },
+    {
+      "epoch": 1.7532575712500744,
+      "grad_norm": 0.3524829672281888,
+      "learning_rate": 3.94272599163627e-07,
+      "loss": 0.3895,
+      "step": 7366
+    },
+    {
+      "epoch": 1.7534955673231392,
+      "grad_norm": 0.40869397302371735,
+      "learning_rate": 3.935226863008812e-07,
+      "loss": 0.2801,
+      "step": 7367
+    },
+    {
+      "epoch": 1.753733563396204,
+      "grad_norm": 0.3844052458895688,
+      "learning_rate": 3.9277345807462285e-07,
+      "loss": 0.2741,
+      "step": 7368
+    },
+    {
+      "epoch": 1.7539715594692686,
+      "grad_norm": 0.3940308898609167,
+      "learning_rate": 3.920249145962063e-07,
+      "loss": 0.3044,
+      "step": 7369
+    },
+    {
+      "epoch": 1.7542095555423336,
+      "grad_norm": 0.37516037241099764,
+      "learning_rate": 3.9127705597688305e-07,
+      "loss": 0.3628,
+      "step": 7370
+    },
+    {
+      "epoch": 1.7544475516153983,
+      "grad_norm": 0.36404405464445827,
+      "learning_rate": 3.9052988232780364e-07,
+      "loss": 0.2845,
+      "step": 7371
+    },
+    {
+      "epoch": 1.7546855476884633,
+      "grad_norm": 0.37799405975146255,
+      "learning_rate": 3.8978339376001594e-07,
+      "loss": 0.2724,
+      "step": 7372
+    },
+    {
+      "epoch": 1.754923543761528,
+      "grad_norm": 0.3851431947005174,
+      "learning_rate": 3.8903759038446676e-07,
+      "loss": 0.3726,
+      "step": 7373
+    },
+    {
+      "epoch": 1.7551615398345928,
+      "grad_norm": 0.35648046802246963,
+      "learning_rate": 3.882924723120024e-07,
+      "loss": 0.3407,
+      "step": 7374
+    },
+    {
+      "epoch": 1.7553995359076575,
+      "grad_norm": 0.3805853248880385,
+      "learning_rate": 3.875480396533648e-07,
+      "loss": 0.276,
+      "step": 7375
+    },
+    {
+      "epoch": 1.7556375319807223,
+      "grad_norm": 0.43047194201246963,
+      "learning_rate": 3.8680429251919504e-07,
+      "loss": 0.3141,
+      "step": 7376
+    },
+    {
+      "epoch": 1.755875528053787,
+      "grad_norm": 0.3672797811079228,
+      "learning_rate": 3.8606123102003124e-07,
+      "loss": 0.3954,
+      "step": 7377
+    },
+    {
+      "epoch": 1.756113524126852,
+      "grad_norm": 0.3655093853884352,
+      "learning_rate": 3.8531885526631286e-07,
+      "loss": 0.266,
+      "step": 7378
+    },
+    {
+      "epoch": 1.7563515201999167,
+      "grad_norm": 0.4008961738434252,
+      "learning_rate": 3.845771653683744e-07,
+      "loss": 0.2754,
+      "step": 7379
+    },
+    {
+      "epoch": 1.7565895162729817,
+      "grad_norm": 0.3754989479985082,
+      "learning_rate": 3.8383616143644884e-07,
+      "loss": 0.3759,
+      "step": 7380
+    },
+    {
+      "epoch": 1.7568275123460464,
+      "grad_norm": 0.3998697477070844,
+      "learning_rate": 3.8309584358066866e-07,
+      "loss": 0.3457,
+      "step": 7381
+    },
+    {
+      "epoch": 1.7570655084191111,
+      "grad_norm": 0.3805072843335442,
+      "learning_rate": 3.823562119110624e-07,
+      "loss": 0.305,
+      "step": 7382
+    },
+    {
+      "epoch": 1.7573035044921759,
+      "grad_norm": 0.38537738584226183,
+      "learning_rate": 3.816172665375584e-07,
+      "loss": 0.3079,
+      "step": 7383
+    },
+    {
+      "epoch": 1.7575415005652406,
+      "grad_norm": 0.37422390675327777,
+      "learning_rate": 3.808790075699814e-07,
+      "loss": 0.3951,
+      "step": 7384
+    },
+    {
+      "epoch": 1.7577794966383053,
+      "grad_norm": 0.3857626430504062,
+      "learning_rate": 3.80141435118056e-07,
+      "loss": 0.2866,
+      "step": 7385
+    },
+    {
+      "epoch": 1.7580174927113703,
+      "grad_norm": 0.3848693921935248,
+      "learning_rate": 3.794045492914028e-07,
+      "loss": 0.2799,
+      "step": 7386
+    },
+    {
+      "epoch": 1.758255488784435,
+      "grad_norm": 0.37773875751263675,
+      "learning_rate": 3.786683501995414e-07,
+      "loss": 0.3419,
+      "step": 7387
+    },
+    {
+      "epoch": 1.7584934848575,
+      "grad_norm": 0.39284310713498916,
+      "learning_rate": 3.779328379518898e-07,
+      "loss": 0.3779,
+      "step": 7388
+    },
+    {
+      "epoch": 1.7587314809305648,
+      "grad_norm": 0.38708082552128015,
+      "learning_rate": 3.771980126577629e-07,
+      "loss": 0.2952,
+      "step": 7389
+    },
+    {
+      "epoch": 1.7589694770036295,
+      "grad_norm": 0.3840167669470753,
+      "learning_rate": 3.7646387442637266e-07,
+      "loss": 0.3068,
+      "step": 7390
+    },
+    {
+      "epoch": 1.7592074730766942,
+      "grad_norm": 0.37577064742826666,
+      "learning_rate": 3.757304233668324e-07,
+      "loss": 0.3682,
+      "step": 7391
+    },
+    {
+      "epoch": 1.759445469149759,
+      "grad_norm": 0.37383506429697316,
+      "learning_rate": 3.749976595881505e-07,
+      "loss": 0.2946,
+      "step": 7392
+    },
+    {
+      "epoch": 1.7596834652228237,
+      "grad_norm": 0.3938798316973006,
+      "learning_rate": 3.742655831992331e-07,
+      "loss": 0.2602,
+      "step": 7393
+    },
+    {
+      "epoch": 1.7599214612958887,
+      "grad_norm": 0.382277877373375,
+      "learning_rate": 3.735341943088855e-07,
+      "loss": 0.3399,
+      "step": 7394
+    },
+    {
+      "epoch": 1.7601594573689534,
+      "grad_norm": 0.4438405373662082,
+      "learning_rate": 3.7280349302580954e-07,
+      "loss": 0.4037,
+      "step": 7395
+    },
+    {
+      "epoch": 1.7603974534420184,
+      "grad_norm": 0.38449221017426694,
+      "learning_rate": 3.720734794586062e-07,
+      "loss": 0.303,
+      "step": 7396
+    },
+    {
+      "epoch": 1.760635449515083,
+      "grad_norm": 0.37868139806924034,
+      "learning_rate": 3.7134415371577303e-07,
+      "loss": 0.2986,
+      "step": 7397
+    },
+    {
+      "epoch": 1.7608734455881478,
+      "grad_norm": 0.3943383369047064,
+      "learning_rate": 3.7061551590570565e-07,
+      "loss": 0.3536,
+      "step": 7398
+    },
+    {
+      "epoch": 1.7611114416612126,
+      "grad_norm": 0.36959772575231714,
+      "learning_rate": 3.698875661366985e-07,
+      "loss": 0.2997,
+      "step": 7399
+    },
+    {
+      "epoch": 1.7613494377342773,
+      "grad_norm": 0.40020141349107846,
+      "learning_rate": 3.691603045169417e-07,
+      "loss": 0.2694,
+      "step": 7400
+    },
+    {
+      "epoch": 1.761587433807342,
+      "grad_norm": 0.37634863654754686,
+      "learning_rate": 3.684337311545261e-07,
+      "loss": 0.3371,
+      "step": 7401
+    },
+    {
+      "epoch": 1.761825429880407,
+      "grad_norm": 0.38956164399358506,
+      "learning_rate": 3.677078461574368e-07,
+      "loss": 0.3591,
+      "step": 7402
+    },
+    {
+      "epoch": 1.7620634259534718,
+      "grad_norm": 0.3602522289158944,
+      "learning_rate": 3.6698264963355936e-07,
+      "loss": 0.2936,
+      "step": 7403
+    },
+    {
+      "epoch": 1.7623014220265367,
+      "grad_norm": 0.3761002128092343,
+      "learning_rate": 3.662581416906746e-07,
+      "loss": 0.3009,
+      "step": 7404
+    },
+    {
+      "epoch": 1.7625394180996015,
+      "grad_norm": 0.4007846365643332,
+      "learning_rate": 3.6553432243646435e-07,
+      "loss": 0.3523,
+      "step": 7405
+    },
+    {
+      "epoch": 1.7627774141726662,
+      "grad_norm": 0.35494662132278193,
+      "learning_rate": 3.6481119197850466e-07,
+      "loss": 0.3274,
+      "step": 7406
+    },
+    {
+      "epoch": 1.763015410245731,
+      "grad_norm": 0.3790695232145245,
+      "learning_rate": 3.640887504242707e-07,
+      "loss": 0.2875,
+      "step": 7407
+    },
+    {
+      "epoch": 1.7632534063187957,
+      "grad_norm": 0.4084614475813453,
+      "learning_rate": 3.6336699788113605e-07,
+      "loss": 0.3147,
+      "step": 7408
+    },
+    {
+      "epoch": 1.7634914023918604,
+      "grad_norm": 0.37225641815520183,
+      "learning_rate": 3.626459344563693e-07,
+      "loss": 0.384,
+      "step": 7409
+    },
+    {
+      "epoch": 1.7637293984649254,
+      "grad_norm": 0.36150847703005184,
+      "learning_rate": 3.619255602571403e-07,
+      "loss": 0.3092,
+      "step": 7410
+    },
+    {
+      "epoch": 1.7639673945379901,
+      "grad_norm": 0.3522846334053475,
+      "learning_rate": 3.612058753905129e-07,
+      "loss": 0.2549,
+      "step": 7411
+    },
+    {
+      "epoch": 1.764205390611055,
+      "grad_norm": 0.36618028447488987,
+      "learning_rate": 3.604868799634509e-07,
+      "loss": 0.3265,
+      "step": 7412
+    },
+    {
+      "epoch": 1.7644433866841198,
+      "grad_norm": 0.3566376234452962,
+      "learning_rate": 3.597685740828144e-07,
+      "loss": 0.3793,
+      "step": 7413
+    },
+    {
+      "epoch": 1.7646813827571846,
+      "grad_norm": 0.36692847158051417,
+      "learning_rate": 3.5905095785536135e-07,
+      "loss": 0.2872,
+      "step": 7414
+    },
+    {
+      "epoch": 1.7649193788302493,
+      "grad_norm": 0.37257656439382925,
+      "learning_rate": 3.5833403138774756e-07,
+      "loss": 0.3225,
+      "step": 7415
+    },
+    {
+      "epoch": 1.765157374903314,
+      "grad_norm": 0.3749686078408963,
+      "learning_rate": 3.5761779478652614e-07,
+      "loss": 0.3883,
+      "step": 7416
+    },
+    {
+      "epoch": 1.7653953709763788,
+      "grad_norm": 0.3583625843405949,
+      "learning_rate": 3.56902248158148e-07,
+      "loss": 0.2977,
+      "step": 7417
+    },
+    {
+      "epoch": 1.7656333670494437,
+      "grad_norm": 0.38524990025382316,
+      "learning_rate": 3.5618739160895864e-07,
+      "loss": 0.2877,
+      "step": 7418
+    },
+    {
+      "epoch": 1.7658713631225085,
+      "grad_norm": 0.37604720661475144,
+      "learning_rate": 3.55473225245207e-07,
+      "loss": 0.3219,
+      "step": 7419
+    },
+    {
+      "epoch": 1.7661093591955734,
+      "grad_norm": 0.3846604314430845,
+      "learning_rate": 3.5475974917303366e-07,
+      "loss": 0.3566,
+      "step": 7420
+    },
+    {
+      "epoch": 1.7663473552686382,
+      "grad_norm": 0.3680527964727019,
+      "learning_rate": 3.5404696349847944e-07,
+      "loss": 0.2627,
+      "step": 7421
+    },
+    {
+      "epoch": 1.766585351341703,
+      "grad_norm": 0.3991689202195992,
+      "learning_rate": 3.5333486832748176e-07,
+      "loss": 0.2867,
+      "step": 7422
+    },
+    {
+      "epoch": 1.7668233474147677,
+      "grad_norm": 0.3999787206661557,
+      "learning_rate": 3.5262346376587544e-07,
+      "loss": 0.3645,
+      "step": 7423
+    },
+    {
+      "epoch": 1.7670613434878324,
+      "grad_norm": 0.3757038581644989,
+      "learning_rate": 3.5191274991939306e-07,
+      "loss": 0.3368,
+      "step": 7424
+    },
+    {
+      "epoch": 1.7672993395608971,
+      "grad_norm": 0.382050405949238,
+      "learning_rate": 3.512027268936641e-07,
+      "loss": 0.2718,
+      "step": 7425
+    },
+    {
+      "epoch": 1.767537335633962,
+      "grad_norm": 0.38279994990044025,
+      "learning_rate": 3.504933947942157e-07,
+      "loss": 0.3359,
+      "step": 7426
+    },
+    {
+      "epoch": 1.7677753317070268,
+      "grad_norm": 0.38419625146613884,
+      "learning_rate": 3.4978475372647145e-07,
+      "loss": 0.3792,
+      "step": 7427
+    },
+    {
+      "epoch": 1.7680133277800918,
+      "grad_norm": 0.3984627542168908,
+      "learning_rate": 3.4907680379575426e-07,
+      "loss": 0.2835,
+      "step": 7428
+    },
+    {
+      "epoch": 1.7682513238531565,
+      "grad_norm": 0.37364849406584333,
+      "learning_rate": 3.4836954510728215e-07,
+      "loss": 0.2711,
+      "step": 7429
+    },
+    {
+      "epoch": 1.7684893199262213,
+      "grad_norm": 0.42900751703465184,
+      "learning_rate": 3.476629777661716e-07,
+      "loss": 0.3473,
+      "step": 7430
+    },
+    {
+      "epoch": 1.768727315999286,
+      "grad_norm": 0.3800418075868142,
+      "learning_rate": 3.469571018774348e-07,
+      "loss": 0.3456,
+      "step": 7431
+    },
+    {
+      "epoch": 1.7689653120723507,
+      "grad_norm": 0.43203134753145855,
+      "learning_rate": 3.462519175459844e-07,
+      "loss": 0.2851,
+      "step": 7432
+    },
+    {
+      "epoch": 1.7692033081454155,
+      "grad_norm": 0.40985553269034813,
+      "learning_rate": 3.4554742487662716e-07,
+      "loss": 0.3091,
+      "step": 7433
+    },
+    {
+      "epoch": 1.7694413042184804,
+      "grad_norm": 0.38887010959356444,
+      "learning_rate": 3.448436239740682e-07,
+      "loss": 0.3986,
+      "step": 7434
+    },
+    {
+      "epoch": 1.7696793002915452,
+      "grad_norm": 0.36915654756950994,
+      "learning_rate": 3.4414051494291e-07,
+      "loss": 0.2855,
+      "step": 7435
+    },
+    {
+      "epoch": 1.7699172963646101,
+      "grad_norm": 0.386681678539916,
+      "learning_rate": 3.434380978876517e-07,
+      "loss": 0.3074,
+      "step": 7436
+    },
+    {
+      "epoch": 1.7701552924376749,
+      "grad_norm": 0.37941576584294967,
+      "learning_rate": 3.4273637291268926e-07,
+      "loss": 0.3312,
+      "step": 7437
+    },
+    {
+      "epoch": 1.7703932885107396,
+      "grad_norm": 0.3603145085548946,
+      "learning_rate": 3.4203534012231753e-07,
+      "loss": 0.3694,
+      "step": 7438
+    },
+    {
+      "epoch": 1.7706312845838044,
+      "grad_norm": 0.38237676060980935,
+      "learning_rate": 3.413349996207266e-07,
+      "loss": 0.2842,
+      "step": 7439
+    },
+    {
+      "epoch": 1.770869280656869,
+      "grad_norm": 0.39329602306550704,
+      "learning_rate": 3.4063535151200424e-07,
+      "loss": 0.3089,
+      "step": 7440
+    },
+    {
+      "epoch": 1.7711072767299338,
+      "grad_norm": 0.3520088960926671,
+      "learning_rate": 3.3993639590013615e-07,
+      "loss": 0.367,
+      "step": 7441
+    },
+    {
+      "epoch": 1.7713452728029988,
+      "grad_norm": 0.3549709470925406,
+      "learning_rate": 3.3923813288900376e-07,
+      "loss": 0.303,
+      "step": 7442
+    },
+    {
+      "epoch": 1.7715832688760635,
+      "grad_norm": 0.3698693474185596,
+      "learning_rate": 3.3854056258238675e-07,
+      "loss": 0.2703,
+      "step": 7443
+    },
+    {
+      "epoch": 1.7718212649491285,
+      "grad_norm": 0.3744847290334765,
+      "learning_rate": 3.378436850839612e-07,
+      "loss": 0.3261,
+      "step": 7444
+    },
+    {
+      "epoch": 1.7720592610221932,
+      "grad_norm": 0.3841273561075149,
+      "learning_rate": 3.3714750049729903e-07,
+      "loss": 0.3532,
+      "step": 7445
+    },
+    {
+      "epoch": 1.772297257095258,
+      "grad_norm": 0.3673423324794477,
+      "learning_rate": 3.364520089258727e-07,
+      "loss": 0.2856,
+      "step": 7446
+    },
+    {
+      "epoch": 1.7725352531683227,
+      "grad_norm": 0.39543112396129604,
+      "learning_rate": 3.3575721047304887e-07,
+      "loss": 0.3264,
+      "step": 7447
+    },
+    {
+      "epoch": 1.7727732492413875,
+      "grad_norm": 0.38069356140983673,
+      "learning_rate": 3.350631052420911e-07,
+      "loss": 0.3384,
+      "step": 7448
+    },
+    {
+      "epoch": 1.7730112453144522,
+      "grad_norm": 0.39721194025902407,
+      "learning_rate": 3.3436969333616064e-07,
+      "loss": 0.3504,
+      "step": 7449
+    },
+    {
+      "epoch": 1.7732492413875172,
+      "grad_norm": 0.41294146060598996,
+      "learning_rate": 3.3367697485831573e-07,
+      "loss": 0.309,
+      "step": 7450
+    },
+    {
+      "epoch": 1.773487237460582,
+      "grad_norm": 0.4004419253097933,
+      "learning_rate": 3.3298494991151234e-07,
+      "loss": 0.3409,
+      "step": 7451
+    },
+    {
+      "epoch": 1.7737252335336469,
+      "grad_norm": 0.3787521305790956,
+      "learning_rate": 3.322936185986009e-07,
+      "loss": 0.4034,
+      "step": 7452
+    },
+    {
+      "epoch": 1.7739632296067116,
+      "grad_norm": 0.36357233108658643,
+      "learning_rate": 3.31602981022332e-07,
+      "loss": 0.2957,
+      "step": 7453
+    },
+    {
+      "epoch": 1.7742012256797763,
+      "grad_norm": 0.3702141009628778,
+      "learning_rate": 3.309130372853492e-07,
+      "loss": 0.2788,
+      "step": 7454
+    },
+    {
+      "epoch": 1.774439221752841,
+      "grad_norm": 0.41114542834457846,
+      "learning_rate": 3.302237874901981e-07,
+      "loss": 0.3846,
+      "step": 7455
+    },
+    {
+      "epoch": 1.7746772178259058,
+      "grad_norm": 0.35691967017327136,
+      "learning_rate": 3.2953523173931633e-07,
+      "loss": 0.353,
+      "step": 7456
+    },
+    {
+      "epoch": 1.7749152138989706,
+      "grad_norm": 0.3949364534978649,
+      "learning_rate": 3.2884737013504143e-07,
+      "loss": 0.2854,
+      "step": 7457
+    },
+    {
+      "epoch": 1.7751532099720353,
+      "grad_norm": 0.38350872066977443,
+      "learning_rate": 3.2816020277960604e-07,
+      "loss": 0.3232,
+      "step": 7458
+    },
+    {
+      "epoch": 1.7753912060451003,
+      "grad_norm": 0.4021154125147116,
+      "learning_rate": 3.2747372977513905e-07,
+      "loss": 0.3592,
+      "step": 7459
+    },
+    {
+      "epoch": 1.775629202118165,
+      "grad_norm": 0.3879804363704007,
+      "learning_rate": 3.2678795122366933e-07,
+      "loss": 0.2945,
+      "step": 7460
+    },
+    {
+      "epoch": 1.77586719819123,
+      "grad_norm": 0.4237627394310612,
+      "learning_rate": 3.2610286722711993e-07,
+      "loss": 0.2732,
+      "step": 7461
+    },
+    {
+      "epoch": 1.7761051942642947,
+      "grad_norm": 0.3924917993001975,
+      "learning_rate": 3.2541847788731153e-07,
+      "loss": 0.3461,
+      "step": 7462
+    },
+    {
+      "epoch": 1.7763431903373594,
+      "grad_norm": 0.3627340154994653,
+      "learning_rate": 3.2473478330595996e-07,
+      "loss": 0.3414,
+      "step": 7463
+    },
+    {
+      "epoch": 1.7765811864104242,
+      "grad_norm": 0.3879803415205633,
+      "learning_rate": 3.240517835846807e-07,
+      "loss": 0.2865,
+      "step": 7464
+    },
+    {
+      "epoch": 1.776819182483489,
+      "grad_norm": 0.36721296027414413,
+      "learning_rate": 3.233694788249836e-07,
+      "loss": 0.3411,
+      "step": 7465
+    },
+    {
+      "epoch": 1.7770571785565537,
+      "grad_norm": 0.37560707117027603,
+      "learning_rate": 3.2268786912827645e-07,
+      "loss": 0.3818,
+      "step": 7466
+    },
+    {
+      "epoch": 1.7772951746296186,
+      "grad_norm": 0.37942928089496036,
+      "learning_rate": 3.220069545958632e-07,
+      "loss": 0.3262,
+      "step": 7467
+    },
+    {
+      "epoch": 1.7775331707026834,
+      "grad_norm": 0.35866524083415724,
+      "learning_rate": 3.2132673532894397e-07,
+      "loss": 0.2738,
+      "step": 7468
+    },
+    {
+      "epoch": 1.7777711667757483,
+      "grad_norm": 0.4579563149174641,
+      "learning_rate": 3.206472114286169e-07,
+      "loss": 0.3463,
+      "step": 7469
+    },
+    {
+      "epoch": 1.778009162848813,
+      "grad_norm": 0.374846770314416,
+      "learning_rate": 3.1996838299587604e-07,
+      "loss": 0.369,
+      "step": 7470
+    },
+    {
+      "epoch": 1.7782471589218778,
+      "grad_norm": 0.3575595417175333,
+      "learning_rate": 3.192902501316114e-07,
+      "loss": 0.322,
+      "step": 7471
+    },
+    {
+      "epoch": 1.7784851549949425,
+      "grad_norm": 0.3793283868853329,
+      "learning_rate": 3.186128129366112e-07,
+      "loss": 0.2936,
+      "step": 7472
+    },
+    {
+      "epoch": 1.7787231510680073,
+      "grad_norm": 0.4086026355069854,
+      "learning_rate": 3.179360715115576e-07,
+      "loss": 0.355,
+      "step": 7473
+    },
+    {
+      "epoch": 1.778961147141072,
+      "grad_norm": 0.35603960063048445,
+      "learning_rate": 3.172600259570335e-07,
+      "loss": 0.3446,
+      "step": 7474
+    },
+    {
+      "epoch": 1.779199143214137,
+      "grad_norm": 0.3594124787406389,
+      "learning_rate": 3.165846763735153e-07,
+      "loss": 0.2763,
+      "step": 7475
+    },
+    {
+      "epoch": 1.7794371392872017,
+      "grad_norm": 0.35661873853503745,
+      "learning_rate": 3.1591002286137597e-07,
+      "loss": 0.3205,
+      "step": 7476
+    },
+    {
+      "epoch": 1.7796751353602667,
+      "grad_norm": 0.37799283767267144,
+      "learning_rate": 3.152360655208864e-07,
+      "loss": 0.3899,
+      "step": 7477
+    },
+    {
+      "epoch": 1.7799131314333314,
+      "grad_norm": 0.3943463609437691,
+      "learning_rate": 3.1456280445221256e-07,
+      "loss": 0.3142,
+      "step": 7478
+    },
+    {
+      "epoch": 1.7801511275063961,
+      "grad_norm": 0.3930479973881174,
+      "learning_rate": 3.138902397554183e-07,
+      "loss": 0.2794,
+      "step": 7479
+    },
+    {
+      "epoch": 1.7803891235794609,
+      "grad_norm": 0.37084479307111856,
+      "learning_rate": 3.132183715304632e-07,
+      "loss": 0.3584,
+      "step": 7480
+    },
+    {
+      "epoch": 1.7806271196525256,
+      "grad_norm": 0.3640689537300512,
+      "learning_rate": 3.125471998772023e-07,
+      "loss": 0.3562,
+      "step": 7481
+    },
+    {
+      "epoch": 1.7808651157255904,
+      "grad_norm": 0.3798227659828128,
+      "learning_rate": 3.118767248953908e-07,
+      "loss": 0.2747,
+      "step": 7482
+    },
+    {
+      "epoch": 1.7811031117986553,
+      "grad_norm": 0.39461973353159785,
+      "learning_rate": 3.112069466846762e-07,
+      "loss": 0.3147,
+      "step": 7483
+    },
+    {
+      "epoch": 1.78134110787172,
+      "grad_norm": 0.39876904022229337,
+      "learning_rate": 3.105378653446045e-07,
+      "loss": 0.4053,
+      "step": 7484
+    },
+    {
+      "epoch": 1.781579103944785,
+      "grad_norm": 0.363220255107949,
+      "learning_rate": 3.098694809746183e-07,
+      "loss": 0.3208,
+      "step": 7485
+    },
+    {
+      "epoch": 1.7818171000178498,
+      "grad_norm": 0.4052186140806434,
+      "learning_rate": 3.092017936740549e-07,
+      "loss": 0.2928,
+      "step": 7486
+    },
+    {
+      "epoch": 1.7820550960909145,
+      "grad_norm": 0.3622874186549825,
+      "learning_rate": 3.085348035421487e-07,
+      "loss": 0.3231,
+      "step": 7487
+    },
+    {
+      "epoch": 1.7822930921639792,
+      "grad_norm": 0.39892992338775146,
+      "learning_rate": 3.0786851067803326e-07,
+      "loss": 0.3581,
+      "step": 7488
+    },
+    {
+      "epoch": 1.782531088237044,
+      "grad_norm": 0.4289428994215822,
+      "learning_rate": 3.0720291518073485e-07,
+      "loss": 0.2747,
+      "step": 7489
+    },
+    {
+      "epoch": 1.7827690843101087,
+      "grad_norm": 0.4117645952035017,
+      "learning_rate": 3.065380171491772e-07,
+      "loss": 0.3095,
+      "step": 7490
+    },
+    {
+      "epoch": 1.7830070803831737,
+      "grad_norm": 0.3635531180905779,
+      "learning_rate": 3.0587381668218117e-07,
+      "loss": 0.3608,
+      "step": 7491
+    },
+    {
+      "epoch": 1.7832450764562384,
+      "grad_norm": 0.3678683816926744,
+      "learning_rate": 3.052103138784629e-07,
+      "loss": 0.3101,
+      "step": 7492
+    },
+    {
+      "epoch": 1.7834830725293034,
+      "grad_norm": 0.3893055958544627,
+      "learning_rate": 3.045475088366351e-07,
+      "loss": 0.2619,
+      "step": 7493
+    },
+    {
+      "epoch": 1.7837210686023681,
+      "grad_norm": 0.3745763428587718,
+      "learning_rate": 3.038854016552079e-07,
+      "loss": 0.3315,
+      "step": 7494
+    },
+    {
+      "epoch": 1.7839590646754329,
+      "grad_norm": 0.3797581457610855,
+      "learning_rate": 3.0322399243258583e-07,
+      "loss": 0.3711,
+      "step": 7495
+    },
+    {
+      "epoch": 1.7841970607484976,
+      "grad_norm": 0.39193763190558745,
+      "learning_rate": 3.0256328126707147e-07,
+      "loss": 0.3174,
+      "step": 7496
+    },
+    {
+      "epoch": 1.7844350568215623,
+      "grad_norm": 0.3855942012320072,
+      "learning_rate": 3.0190326825686234e-07,
+      "loss": 0.3247,
+      "step": 7497
+    },
+    {
+      "epoch": 1.784673052894627,
+      "grad_norm": 0.3719708597803907,
+      "learning_rate": 3.012439535000533e-07,
+      "loss": 0.3487,
+      "step": 7498
+    },
+    {
+      "epoch": 1.784911048967692,
+      "grad_norm": 0.3389285568386847,
+      "learning_rate": 3.005853370946338e-07,
+      "loss": 0.3018,
+      "step": 7499
+    },
+    {
+      "epoch": 1.7851490450407568,
+      "grad_norm": 0.3686403079627146,
+      "learning_rate": 2.9992741913849044e-07,
+      "loss": 0.2694,
+      "step": 7500
+    },
+    {
+      "epoch": 1.7853870411138217,
+      "grad_norm": 0.3739646758479323,
+      "learning_rate": 2.9927019972940785e-07,
+      "loss": 0.3262,
+      "step": 7501
+    },
+    {
+      "epoch": 1.7856250371868865,
+      "grad_norm": 0.3733801667577494,
+      "learning_rate": 2.9861367896506397e-07,
+      "loss": 0.3706,
+      "step": 7502
+    },
+    {
+      "epoch": 1.7858630332599512,
+      "grad_norm": 0.3896825661654586,
+      "learning_rate": 2.9795785694303413e-07,
+      "loss": 0.3129,
+      "step": 7503
+    },
+    {
+      "epoch": 1.786101029333016,
+      "grad_norm": 0.41296329329174347,
+      "learning_rate": 2.9730273376078923e-07,
+      "loss": 0.2796,
+      "step": 7504
+    },
+    {
+      "epoch": 1.7863390254060807,
+      "grad_norm": 0.3719130039198455,
+      "learning_rate": 2.9664830951569743e-07,
+      "loss": 0.3465,
+      "step": 7505
+    },
+    {
+      "epoch": 1.7865770214791454,
+      "grad_norm": 0.3562312311470605,
+      "learning_rate": 2.959945843050227e-07,
+      "loss": 0.3219,
+      "step": 7506
+    },
+    {
+      "epoch": 1.7868150175522104,
+      "grad_norm": 0.38540030460210417,
+      "learning_rate": 2.9534155822592336e-07,
+      "loss": 0.2877,
+      "step": 7507
+    },
+    {
+      "epoch": 1.7870530136252751,
+      "grad_norm": 0.378036461217732,
+      "learning_rate": 2.9468923137545626e-07,
+      "loss": 0.3406,
+      "step": 7508
+    },
+    {
+      "epoch": 1.78729100969834,
+      "grad_norm": 0.402519853904251,
+      "learning_rate": 2.940376038505732e-07,
+      "loss": 0.3822,
+      "step": 7509
+    },
+    {
+      "epoch": 1.7875290057714048,
+      "grad_norm": 0.37522413264398013,
+      "learning_rate": 2.933866757481224e-07,
+      "loss": 0.3143,
+      "step": 7510
+    },
+    {
+      "epoch": 1.7877670018444696,
+      "grad_norm": 0.3791804024356112,
+      "learning_rate": 2.9273644716484753e-07,
+      "loss": 0.2764,
+      "step": 7511
+    },
+    {
+      "epoch": 1.7880049979175343,
+      "grad_norm": 0.40095271597833637,
+      "learning_rate": 2.9208691819738844e-07,
+      "loss": 0.3297,
+      "step": 7512
+    },
+    {
+      "epoch": 1.788242993990599,
+      "grad_norm": 0.39865047915199964,
+      "learning_rate": 2.914380889422819e-07,
+      "loss": 0.3583,
+      "step": 7513
+    },
+    {
+      "epoch": 1.7884809900636638,
+      "grad_norm": 0.35173957915577636,
+      "learning_rate": 2.9078995949595847e-07,
+      "loss": 0.289,
+      "step": 7514
+    },
+    {
+      "epoch": 1.7887189861367288,
+      "grad_norm": 0.3817357829174564,
+      "learning_rate": 2.901425299547483e-07,
+      "loss": 0.3104,
+      "step": 7515
+    },
+    {
+      "epoch": 1.7889569822097935,
+      "grad_norm": 0.37478155473085156,
+      "learning_rate": 2.8949580041487457e-07,
+      "loss": 0.391,
+      "step": 7516
+    },
+    {
+      "epoch": 1.7891949782828585,
+      "grad_norm": 0.3725958003240469,
+      "learning_rate": 2.8884977097245694e-07,
+      "loss": 0.3156,
+      "step": 7517
+    },
+    {
+      "epoch": 1.7894329743559232,
+      "grad_norm": 0.3801605189515548,
+      "learning_rate": 2.8820444172351137e-07,
+      "loss": 0.281,
+      "step": 7518
+    },
+    {
+      "epoch": 1.789670970428988,
+      "grad_norm": 0.3623288254644941,
+      "learning_rate": 2.8755981276395005e-07,
+      "loss": 0.3224,
+      "step": 7519
+    },
+    {
+      "epoch": 1.7899089665020527,
+      "grad_norm": 0.3776619119078162,
+      "learning_rate": 2.869158841895808e-07,
+      "loss": 0.3689,
+      "step": 7520
+    },
+    {
+      "epoch": 1.7901469625751174,
+      "grad_norm": 0.35832837886741215,
+      "learning_rate": 2.862726560961071e-07,
+      "loss": 0.2744,
+      "step": 7521
+    },
+    {
+      "epoch": 1.7903849586481821,
+      "grad_norm": 0.38760652888202674,
+      "learning_rate": 2.8563012857912906e-07,
+      "loss": 0.303,
+      "step": 7522
+    },
+    {
+      "epoch": 1.790622954721247,
+      "grad_norm": 0.3754292098719095,
+      "learning_rate": 2.849883017341415e-07,
+      "loss": 0.3525,
+      "step": 7523
+    },
+    {
+      "epoch": 1.7908609507943118,
+      "grad_norm": 0.3492865594689972,
+      "learning_rate": 2.8434717565653635e-07,
+      "loss": 0.337,
+      "step": 7524
+    },
+    {
+      "epoch": 1.7910989468673768,
+      "grad_norm": 0.3666136541523488,
+      "learning_rate": 2.837067504416002e-07,
+      "loss": 0.3009,
+      "step": 7525
+    },
+    {
+      "epoch": 1.7913369429404415,
+      "grad_norm": 0.37743641568773106,
+      "learning_rate": 2.830670261845164e-07,
+      "loss": 0.3058,
+      "step": 7526
+    },
+    {
+      "epoch": 1.7915749390135063,
+      "grad_norm": 0.4041150603093485,
+      "learning_rate": 2.8242800298036443e-07,
+      "loss": 0.3588,
+      "step": 7527
+    },
+    {
+      "epoch": 1.791812935086571,
+      "grad_norm": 0.3678697803590038,
+      "learning_rate": 2.8178968092411717e-07,
+      "loss": 0.2898,
+      "step": 7528
+    },
+    {
+      "epoch": 1.7920509311596358,
+      "grad_norm": 0.40176844938042455,
+      "learning_rate": 2.8115206011064655e-07,
+      "loss": 0.2558,
+      "step": 7529
+    },
+    {
+      "epoch": 1.7922889272327005,
+      "grad_norm": 0.3917439000088701,
+      "learning_rate": 2.80515140634719e-07,
+      "loss": 0.3533,
+      "step": 7530
+    },
+    {
+      "epoch": 1.7925269233057655,
+      "grad_norm": 0.36028134434707115,
+      "learning_rate": 2.798789225909959e-07,
+      "loss": 0.3531,
+      "step": 7531
+    },
+    {
+      "epoch": 1.7927649193788302,
+      "grad_norm": 0.37200867528731646,
+      "learning_rate": 2.792434060740351e-07,
+      "loss": 0.2834,
+      "step": 7532
+    },
+    {
+      "epoch": 1.7930029154518952,
+      "grad_norm": 0.40052191062969406,
+      "learning_rate": 2.7860859117828985e-07,
+      "loss": 0.3288,
+      "step": 7533
+    },
+    {
+      "epoch": 1.79324091152496,
+      "grad_norm": 0.36692664516075085,
+      "learning_rate": 2.779744779981097e-07,
+      "loss": 0.4008,
+      "step": 7534
+    },
+    {
+      "epoch": 1.7934789075980246,
+      "grad_norm": 0.37862040939536806,
+      "learning_rate": 2.773410666277382e-07,
+      "loss": 0.3141,
+      "step": 7535
+    },
+    {
+      "epoch": 1.7937169036710894,
+      "grad_norm": 0.4291885543813882,
+      "learning_rate": 2.767083571613183e-07,
+      "loss": 0.2753,
+      "step": 7536
+    },
+    {
+      "epoch": 1.7939548997441541,
+      "grad_norm": 0.3603867639177607,
+      "learning_rate": 2.7607634969288535e-07,
+      "loss": 0.3465,
+      "step": 7537
+    },
+    {
+      "epoch": 1.7941928958172189,
+      "grad_norm": 0.3909700151084424,
+      "learning_rate": 2.7544504431637085e-07,
+      "loss": 0.4081,
+      "step": 7538
+    },
+    {
+      "epoch": 1.7944308918902838,
+      "grad_norm": 0.37728790668800677,
+      "learning_rate": 2.748144411256026e-07,
+      "loss": 0.2785,
+      "step": 7539
+    },
+    {
+      "epoch": 1.7946688879633486,
+      "grad_norm": 0.36294912722499145,
+      "learning_rate": 2.74184540214304e-07,
+      "loss": 0.3051,
+      "step": 7540
+    },
+    {
+      "epoch": 1.7949068840364135,
+      "grad_norm": 0.35721883529172505,
+      "learning_rate": 2.7355534167609334e-07,
+      "loss": 0.3684,
+      "step": 7541
+    },
+    {
+      "epoch": 1.7951448801094783,
+      "grad_norm": 0.3554032641789872,
+      "learning_rate": 2.7292684560448537e-07,
+      "loss": 0.2988,
+      "step": 7542
+    },
+    {
+      "epoch": 1.795382876182543,
+      "grad_norm": 0.43060025594932566,
+      "learning_rate": 2.722990520928903e-07,
+      "loss": 0.3254,
+      "step": 7543
+    },
+    {
+      "epoch": 1.7956208722556077,
+      "grad_norm": 0.380455936842252,
+      "learning_rate": 2.716719612346147e-07,
+      "loss": 0.3326,
+      "step": 7544
+    },
+    {
+      "epoch": 1.7958588683286725,
+      "grad_norm": 0.4177701326166782,
+      "learning_rate": 2.7104557312285786e-07,
+      "loss": 0.3889,
+      "step": 7545
+    },
+    {
+      "epoch": 1.7960968644017372,
+      "grad_norm": 0.3649041520875534,
+      "learning_rate": 2.7041988785071804e-07,
+      "loss": 0.2986,
+      "step": 7546
+    },
+    {
+      "epoch": 1.7963348604748022,
+      "grad_norm": 0.39864708497247475,
+      "learning_rate": 2.697949055111876e-07,
+      "loss": 0.2909,
+      "step": 7547
+    },
+    {
+      "epoch": 1.796572856547867,
+      "grad_norm": 0.380155366498838,
+      "learning_rate": 2.691706261971533e-07,
+      "loss": 0.363,
+      "step": 7548
+    },
+    {
+      "epoch": 1.7968108526209319,
+      "grad_norm": 0.3973304332381195,
+      "learning_rate": 2.685470500013987e-07,
+      "loss": 0.3397,
+      "step": 7549
+    },
+    {
+      "epoch": 1.7970488486939966,
+      "grad_norm": 0.3718159670234423,
+      "learning_rate": 2.679241770166036e-07,
+      "loss": 0.2915,
+      "step": 7550
+    },
+    {
+      "epoch": 1.7972868447670614,
+      "grad_norm": 0.43325048967411206,
+      "learning_rate": 2.673020073353411e-07,
+      "loss": 0.3344,
+      "step": 7551
+    },
+    {
+      "epoch": 1.797524840840126,
+      "grad_norm": 0.4179971224778607,
+      "learning_rate": 2.666805410500822e-07,
+      "loss": 0.4318,
+      "step": 7552
+    },
+    {
+      "epoch": 1.7977628369131908,
+      "grad_norm": 0.37436449309549835,
+      "learning_rate": 2.6605977825319094e-07,
+      "loss": 0.301,
+      "step": 7553
+    },
+    {
+      "epoch": 1.7980008329862556,
+      "grad_norm": 0.407183373159236,
+      "learning_rate": 2.6543971903692954e-07,
+      "loss": 0.2842,
+      "step": 7554
+    },
+    {
+      "epoch": 1.7982388290593205,
+      "grad_norm": 0.47155428570797986,
+      "learning_rate": 2.6482036349345265e-07,
+      "loss": 0.3289,
+      "step": 7555
+    },
+    {
+      "epoch": 1.7984768251323853,
+      "grad_norm": 0.3675704624500037,
+      "learning_rate": 2.642017117148116e-07,
+      "loss": 0.3144,
+      "step": 7556
+    },
+    {
+      "epoch": 1.7987148212054502,
+      "grad_norm": 0.3654773621312083,
+      "learning_rate": 2.63583763792955e-07,
+      "loss": 0.2829,
+      "step": 7557
+    },
+    {
+      "epoch": 1.798952817278515,
+      "grad_norm": 0.38467306352137903,
+      "learning_rate": 2.629665198197251e-07,
+      "loss": 0.3096,
+      "step": 7558
+    },
+    {
+      "epoch": 1.7991908133515797,
+      "grad_norm": 0.3771099539638736,
+      "learning_rate": 2.623499798868584e-07,
+      "loss": 0.3867,
+      "step": 7559
+    },
+    {
+      "epoch": 1.7994288094246444,
+      "grad_norm": 0.38421855757957385,
+      "learning_rate": 2.617341440859883e-07,
+      "loss": 0.2757,
+      "step": 7560
+    },
+    {
+      "epoch": 1.7996668054977092,
+      "grad_norm": 0.36935340044670584,
+      "learning_rate": 2.6111901250864325e-07,
+      "loss": 0.267,
+      "step": 7561
+    },
+    {
+      "epoch": 1.799904801570774,
+      "grad_norm": 0.3520631923142841,
+      "learning_rate": 2.6050458524624735e-07,
+      "loss": 0.3649,
+      "step": 7562
+    },
+    {
+      "epoch": 1.8001427976438389,
+      "grad_norm": 0.3642419618021161,
+      "learning_rate": 2.5989086239011975e-07,
+      "loss": 0.3463,
+      "step": 7563
+    },
+    {
+      "epoch": 1.8003807937169036,
+      "grad_norm": 0.3716992913649046,
+      "learning_rate": 2.5927784403147473e-07,
+      "loss": 0.2595,
+      "step": 7564
+    },
+    {
+      "epoch": 1.8006187897899686,
+      "grad_norm": 0.37707110768373664,
+      "learning_rate": 2.586655302614216e-07,
+      "loss": 0.3036,
+      "step": 7565
+    },
+    {
+      "epoch": 1.8008567858630333,
+      "grad_norm": 0.3722945299659591,
+      "learning_rate": 2.5805392117096597e-07,
+      "loss": 0.3856,
+      "step": 7566
+    },
+    {
+      "epoch": 1.801094781936098,
+      "grad_norm": 0.37881870157045106,
+      "learning_rate": 2.5744301685100727e-07,
+      "loss": 0.3017,
+      "step": 7567
+    },
+    {
+      "epoch": 1.8013327780091628,
+      "grad_norm": 0.3840903049457169,
+      "learning_rate": 2.5683281739234233e-07,
+      "loss": 0.2878,
+      "step": 7568
+    },
+    {
+      "epoch": 1.8015707740822275,
+      "grad_norm": 0.3737271689190538,
+      "learning_rate": 2.5622332288565975e-07,
+      "loss": 0.3099,
+      "step": 7569
+    },
+    {
+      "epoch": 1.8018087701552923,
+      "grad_norm": 0.410673409572434,
+      "learning_rate": 2.5561453342154763e-07,
+      "loss": 0.3553,
+      "step": 7570
+    },
+    {
+      "epoch": 1.8020467662283572,
+      "grad_norm": 0.3897577800417089,
+      "learning_rate": 2.5500644909048577e-07,
+      "loss": 0.2892,
+      "step": 7571
+    },
+    {
+      "epoch": 1.802284762301422,
+      "grad_norm": 0.3795777307796096,
+      "learning_rate": 2.543990699828519e-07,
+      "loss": 0.304,
+      "step": 7572
+    },
+    {
+      "epoch": 1.802522758374487,
+      "grad_norm": 0.3926970531258347,
+      "learning_rate": 2.5379239618891604e-07,
+      "loss": 0.3838,
+      "step": 7573
+    },
+    {
+      "epoch": 1.8027607544475517,
+      "grad_norm": 0.35843032279557285,
+      "learning_rate": 2.5318642779884605e-07,
+      "loss": 0.3366,
+      "step": 7574
+    },
+    {
+      "epoch": 1.8029987505206164,
+      "grad_norm": 0.3687701099994784,
+      "learning_rate": 2.525811649027032e-07,
+      "loss": 0.3038,
+      "step": 7575
+    },
+    {
+      "epoch": 1.8032367465936812,
+      "grad_norm": 0.3906979385647688,
+      "learning_rate": 2.5197660759044505e-07,
+      "loss": 0.3117,
+      "step": 7576
+    },
+    {
+      "epoch": 1.803474742666746,
+      "grad_norm": 0.434483976268019,
+      "learning_rate": 2.51372755951923e-07,
+      "loss": 0.382,
+      "step": 7577
+    },
+    {
+      "epoch": 1.8037127387398106,
+      "grad_norm": 0.3842828684371924,
+      "learning_rate": 2.5076961007688526e-07,
+      "loss": 0.2864,
+      "step": 7578
+    },
+    {
+      "epoch": 1.8039507348128756,
+      "grad_norm": 0.42535367755859443,
+      "learning_rate": 2.5016717005497347e-07,
+      "loss": 0.2755,
+      "step": 7579
+    },
+    {
+      "epoch": 1.8041887308859403,
+      "grad_norm": 0.376588372683319,
+      "learning_rate": 2.4956543597572546e-07,
+      "loss": 0.3643,
+      "step": 7580
+    },
+    {
+      "epoch": 1.8044267269590053,
+      "grad_norm": 0.3587861059882926,
+      "learning_rate": 2.4896440792857355e-07,
+      "loss": 0.3672,
+      "step": 7581
+    },
+    {
+      "epoch": 1.80466472303207,
+      "grad_norm": 0.40501079901398007,
+      "learning_rate": 2.483640860028458e-07,
+      "loss": 0.2751,
+      "step": 7582
+    },
+    {
+      "epoch": 1.8049027191051348,
+      "grad_norm": 0.3908238593281349,
+      "learning_rate": 2.4776447028776404e-07,
+      "loss": 0.3438,
+      "step": 7583
+    },
+    {
+      "epoch": 1.8051407151781995,
+      "grad_norm": 0.4008384905708165,
+      "learning_rate": 2.4716556087244716e-07,
+      "loss": 0.3841,
+      "step": 7584
+    },
+    {
+      "epoch": 1.8053787112512643,
+      "grad_norm": 0.363703944202256,
+      "learning_rate": 2.465673578459077e-07,
+      "loss": 0.2752,
+      "step": 7585
+    },
+    {
+      "epoch": 1.805616707324329,
+      "grad_norm": 0.40343185473148424,
+      "learning_rate": 2.459698612970529e-07,
+      "loss": 0.2837,
+      "step": 7586
+    },
+    {
+      "epoch": 1.805854703397394,
+      "grad_norm": 0.35789641416139295,
+      "learning_rate": 2.4537307131468566e-07,
+      "loss": 0.3523,
+      "step": 7587
+    },
+    {
+      "epoch": 1.8060926994704587,
+      "grad_norm": 0.3926977742362301,
+      "learning_rate": 2.447769879875039e-07,
+      "loss": 0.4072,
+      "step": 7588
+    },
+    {
+      "epoch": 1.8063306955435237,
+      "grad_norm": 0.3757795347751587,
+      "learning_rate": 2.441816114040996e-07,
+      "loss": 0.2978,
+      "step": 7589
+    },
+    {
+      "epoch": 1.8065686916165884,
+      "grad_norm": 0.3829503186037233,
+      "learning_rate": 2.435869416529618e-07,
+      "loss": 0.3131,
+      "step": 7590
+    },
+    {
+      "epoch": 1.8068066876896531,
+      "grad_norm": 0.41211846502557,
+      "learning_rate": 2.429929788224722e-07,
+      "loss": 0.3494,
+      "step": 7591
+    },
+    {
+      "epoch": 1.8070446837627179,
+      "grad_norm": 0.37268683629691657,
+      "learning_rate": 2.4239972300090897e-07,
+      "loss": 0.2897,
+      "step": 7592
+    },
+    {
+      "epoch": 1.8072826798357826,
+      "grad_norm": 0.3841386899279831,
+      "learning_rate": 2.418071742764444e-07,
+      "loss": 0.2877,
+      "step": 7593
+    },
+    {
+      "epoch": 1.8075206759088474,
+      "grad_norm": 0.36351385879987314,
+      "learning_rate": 2.4121533273714524e-07,
+      "loss": 0.3227,
+      "step": 7594
+    },
+    {
+      "epoch": 1.8077586719819123,
+      "grad_norm": 0.37194110953788245,
+      "learning_rate": 2.4062419847097507e-07,
+      "loss": 0.3593,
+      "step": 7595
+    },
+    {
+      "epoch": 1.807996668054977,
+      "grad_norm": 0.3905909596183342,
+      "learning_rate": 2.4003377156578967e-07,
+      "loss": 0.294,
+      "step": 7596
+    },
+    {
+      "epoch": 1.808234664128042,
+      "grad_norm": 0.37292298295629817,
+      "learning_rate": 2.3944405210934106e-07,
+      "loss": 0.3196,
+      "step": 7597
+    },
+    {
+      "epoch": 1.8084726602011068,
+      "grad_norm": 0.3933254447982632,
+      "learning_rate": 2.388550401892775e-07,
+      "loss": 0.3323,
+      "step": 7598
+    },
+    {
+      "epoch": 1.8087106562741715,
+      "grad_norm": 0.3710230013969392,
+      "learning_rate": 2.382667358931401e-07,
+      "loss": 0.3092,
+      "step": 7599
+    },
+    {
+      "epoch": 1.8089486523472362,
+      "grad_norm": 0.3914008246491944,
+      "learning_rate": 2.3767913930836552e-07,
+      "loss": 0.2787,
+      "step": 7600
+    },
+    {
+      "epoch": 1.809186648420301,
+      "grad_norm": 0.402510376468686,
+      "learning_rate": 2.37092250522285e-07,
+      "loss": 0.3047,
+      "step": 7601
+    },
+    {
+      "epoch": 1.8094246444933657,
+      "grad_norm": 0.3923964917222654,
+      "learning_rate": 2.3650606962212442e-07,
+      "loss": 0.3777,
+      "step": 7602
+    },
+    {
+      "epoch": 1.8096626405664307,
+      "grad_norm": 0.390152017290336,
+      "learning_rate": 2.3592059669500512e-07,
+      "loss": 0.2928,
+      "step": 7603
+    },
+    {
+      "epoch": 1.8099006366394954,
+      "grad_norm": 0.39619044275128396,
+      "learning_rate": 2.353358318279425e-07,
+      "loss": 0.2758,
+      "step": 7604
+    },
+    {
+      "epoch": 1.8101386327125604,
+      "grad_norm": 0.3778971109266139,
+      "learning_rate": 2.347517751078482e-07,
+      "loss": 0.373,
+      "step": 7605
+    },
+    {
+      "epoch": 1.810376628785625,
+      "grad_norm": 0.41970787322500813,
+      "learning_rate": 2.3416842662152606e-07,
+      "loss": 0.3298,
+      "step": 7606
+    },
+    {
+      "epoch": 1.8106146248586898,
+      "grad_norm": 0.37056563319434105,
+      "learning_rate": 2.3358578645567676e-07,
+      "loss": 0.276,
+      "step": 7607
+    },
+    {
+      "epoch": 1.8108526209317546,
+      "grad_norm": 0.4059835978022746,
+      "learning_rate": 2.3300385469689491e-07,
+      "loss": 0.32,
+      "step": 7608
+    },
+    {
+      "epoch": 1.8110906170048193,
+      "grad_norm": 0.3951261910417104,
+      "learning_rate": 2.324226314316702e-07,
+      "loss": 0.3878,
+      "step": 7609
+    },
+    {
+      "epoch": 1.811328613077884,
+      "grad_norm": 0.3525934295874246,
+      "learning_rate": 2.318421167463869e-07,
+      "loss": 0.2804,
+      "step": 7610
+    },
+    {
+      "epoch": 1.811566609150949,
+      "grad_norm": 0.4037839992260597,
+      "learning_rate": 2.3126231072732264e-07,
+      "loss": 0.3023,
+      "step": 7611
+    },
+    {
+      "epoch": 1.8118046052240138,
+      "grad_norm": 0.3608118803223138,
+      "learning_rate": 2.3068321346065236e-07,
+      "loss": 0.3508,
+      "step": 7612
+    },
+    {
+      "epoch": 1.8120426012970787,
+      "grad_norm": 0.3707655208185674,
+      "learning_rate": 2.3010482503244447e-07,
+      "loss": 0.3564,
+      "step": 7613
+    },
+    {
+      "epoch": 1.8122805973701435,
+      "grad_norm": 0.3741317976188311,
+      "learning_rate": 2.2952714552866017e-07,
+      "loss": 0.2794,
+      "step": 7614
+    },
+    {
+      "epoch": 1.8125185934432082,
+      "grad_norm": 0.39603813001981003,
+      "learning_rate": 2.2895017503515859e-07,
+      "loss": 0.2905,
+      "step": 7615
+    },
+    {
+      "epoch": 1.812756589516273,
+      "grad_norm": 0.3998027621001329,
+      "learning_rate": 2.2837391363769e-07,
+      "loss": 0.3547,
+      "step": 7616
+    },
+    {
+      "epoch": 1.8129945855893377,
+      "grad_norm": 0.37531093259849246,
+      "learning_rate": 2.2779836142190314e-07,
+      "loss": 0.3318,
+      "step": 7617
+    },
+    {
+      "epoch": 1.8132325816624024,
+      "grad_norm": 0.3743872139458294,
+      "learning_rate": 2.2722351847333844e-07,
+      "loss": 0.2864,
+      "step": 7618
+    },
+    {
+      "epoch": 1.8134705777354674,
+      "grad_norm": 0.3736656227074288,
+      "learning_rate": 2.266493848774315e-07,
+      "loss": 0.3149,
+      "step": 7619
+    },
+    {
+      "epoch": 1.8137085738085321,
+      "grad_norm": 0.3917461855932985,
+      "learning_rate": 2.2607596071951288e-07,
+      "loss": 0.3824,
+      "step": 7620
+    },
+    {
+      "epoch": 1.813946569881597,
+      "grad_norm": 0.38346753454473853,
+      "learning_rate": 2.255032460848078e-07,
+      "loss": 0.2818,
+      "step": 7621
+    },
+    {
+      "epoch": 1.8141845659546618,
+      "grad_norm": 0.38249166251681005,
+      "learning_rate": 2.2493124105843534e-07,
+      "loss": 0.2997,
+      "step": 7622
+    },
+    {
+      "epoch": 1.8144225620277266,
+      "grad_norm": 0.39871403476731015,
+      "learning_rate": 2.243599457254103e-07,
+      "loss": 0.3529,
+      "step": 7623
+    },
+    {
+      "epoch": 1.8146605581007913,
+      "grad_norm": 0.38237212611132454,
+      "learning_rate": 2.2378936017064035e-07,
+      "loss": 0.3541,
+      "step": 7624
+    },
+    {
+      "epoch": 1.814898554173856,
+      "grad_norm": 0.4176523388972532,
+      "learning_rate": 2.2321948447892984e-07,
+      "loss": 0.2793,
+      "step": 7625
+    },
+    {
+      "epoch": 1.8151365502469208,
+      "grad_norm": 0.3765532620043632,
+      "learning_rate": 2.22650318734976e-07,
+      "loss": 0.3217,
+      "step": 7626
+    },
+    {
+      "epoch": 1.8153745463199857,
+      "grad_norm": 0.5502986454158951,
+      "learning_rate": 2.2208186302337064e-07,
+      "loss": 0.363,
+      "step": 7627
+    },
+    {
+      "epoch": 1.8156125423930505,
+      "grad_norm": 0.3691080359982276,
+      "learning_rate": 2.2151411742860008e-07,
+      "loss": 0.3043,
+      "step": 7628
+    },
+    {
+      "epoch": 1.8158505384661154,
+      "grad_norm": 0.4185986270949634,
+      "learning_rate": 2.2094708203504623e-07,
+      "loss": 0.3095,
+      "step": 7629
+    },
+    {
+      "epoch": 1.8160885345391802,
+      "grad_norm": 0.39746628135216966,
+      "learning_rate": 2.2038075692698392e-07,
+      "loss": 0.3555,
+      "step": 7630
+    },
+    {
+      "epoch": 1.816326530612245,
+      "grad_norm": 0.38678877865572725,
+      "learning_rate": 2.1981514218858302e-07,
+      "loss": 0.3517,
+      "step": 7631
+    },
+    {
+      "epoch": 1.8165645266853097,
+      "grad_norm": 0.3839615523300972,
+      "learning_rate": 2.1925023790390797e-07,
+      "loss": 0.2845,
+      "step": 7632
+    },
+    {
+      "epoch": 1.8168025227583744,
+      "grad_norm": 0.3969919578339475,
+      "learning_rate": 2.1868604415691775e-07,
+      "loss": 0.3163,
+      "step": 7633
+    },
+    {
+      "epoch": 1.8170405188314391,
+      "grad_norm": 0.40042063644372505,
+      "learning_rate": 2.1812256103146523e-07,
+      "loss": 0.3821,
+      "step": 7634
+    },
+    {
+      "epoch": 1.817278514904504,
+      "grad_norm": 0.3595786143418067,
+      "learning_rate": 2.1755978861129846e-07,
+      "loss": 0.2926,
+      "step": 7635
+    },
+    {
+      "epoch": 1.8175165109775688,
+      "grad_norm": 0.4000391772571078,
+      "learning_rate": 2.1699772698005884e-07,
+      "loss": 0.2917,
+      "step": 7636
+    },
+    {
+      "epoch": 1.8177545070506338,
+      "grad_norm": 0.39209491326042045,
+      "learning_rate": 2.164363762212829e-07,
+      "loss": 0.3798,
+      "step": 7637
+    },
+    {
+      "epoch": 1.8179925031236985,
+      "grad_norm": 0.36009581505557847,
+      "learning_rate": 2.1587573641839999e-07,
+      "loss": 0.3549,
+      "step": 7638
+    },
+    {
+      "epoch": 1.8182304991967633,
+      "grad_norm": 0.3891544552915791,
+      "learning_rate": 2.1531580765473737e-07,
+      "loss": 0.2979,
+      "step": 7639
+    },
+    {
+      "epoch": 1.818468495269828,
+      "grad_norm": 0.34833762751724157,
+      "learning_rate": 2.147565900135129e-07,
+      "loss": 0.3132,
+      "step": 7640
+    },
+    {
+      "epoch": 1.8187064913428928,
+      "grad_norm": 0.4102701100568015,
+      "learning_rate": 2.141980835778401e-07,
+      "loss": 0.3732,
+      "step": 7641
+    },
+    {
+      "epoch": 1.8189444874159575,
+      "grad_norm": 0.374891996884048,
+      "learning_rate": 2.136402884307276e-07,
+      "loss": 0.3313,
+      "step": 7642
+    },
+    {
+      "epoch": 1.8191824834890225,
+      "grad_norm": 0.36912237819842353,
+      "learning_rate": 2.130832046550757e-07,
+      "loss": 0.3006,
+      "step": 7643
+    },
+    {
+      "epoch": 1.8194204795620872,
+      "grad_norm": 0.4056734928284316,
+      "learning_rate": 2.1252683233368377e-07,
+      "loss": 0.3093,
+      "step": 7644
+    },
+    {
+      "epoch": 1.8196584756351522,
+      "grad_norm": 0.37038662624715685,
+      "learning_rate": 2.1197117154924006e-07,
+      "loss": 0.3657,
+      "step": 7645
+    },
+    {
+      "epoch": 1.819896471708217,
+      "grad_norm": 0.3710751679952538,
+      "learning_rate": 2.114162223843308e-07,
+      "loss": 0.2794,
+      "step": 7646
+    },
+    {
+      "epoch": 1.8201344677812816,
+      "grad_norm": 0.3778232887871572,
+      "learning_rate": 2.1086198492143494e-07,
+      "loss": 0.2793,
+      "step": 7647
+    },
+    {
+      "epoch": 1.8203724638543464,
+      "grad_norm": 0.41311741726892776,
+      "learning_rate": 2.1030845924292553e-07,
+      "loss": 0.3549,
+      "step": 7648
+    },
+    {
+      "epoch": 1.820610459927411,
+      "grad_norm": 0.3539492859309911,
+      "learning_rate": 2.0975564543107007e-07,
+      "loss": 0.3133,
+      "step": 7649
+    },
+    {
+      "epoch": 1.8208484560004758,
+      "grad_norm": 0.3918727984124256,
+      "learning_rate": 2.0920354356803118e-07,
+      "loss": 0.282,
+      "step": 7650
+    },
+    {
+      "epoch": 1.8210864520735408,
+      "grad_norm": 0.3829591506819489,
+      "learning_rate": 2.0865215373586377e-07,
+      "loss": 0.3235,
+      "step": 7651
+    },
+    {
+      "epoch": 1.8213244481466055,
+      "grad_norm": 0.3848423040694204,
+      "learning_rate": 2.081014760165184e-07,
+      "loss": 0.3783,
+      "step": 7652
+    },
+    {
+      "epoch": 1.8215624442196705,
+      "grad_norm": 0.3682491320220431,
+      "learning_rate": 2.0755151049183963e-07,
+      "loss": 0.2923,
+      "step": 7653
+    },
+    {
+      "epoch": 1.8218004402927352,
+      "grad_norm": 0.3710419857253826,
+      "learning_rate": 2.070022572435665e-07,
+      "loss": 0.2838,
+      "step": 7654
+    },
+    {
+      "epoch": 1.8220384363658,
+      "grad_norm": 0.3940817637494536,
+      "learning_rate": 2.0645371635333032e-07,
+      "loss": 0.3577,
+      "step": 7655
+    },
+    {
+      "epoch": 1.8222764324388647,
+      "grad_norm": 0.34803178778167626,
+      "learning_rate": 2.0590588790265874e-07,
+      "loss": 0.3054,
+      "step": 7656
+    },
+    {
+      "epoch": 1.8225144285119295,
+      "grad_norm": 0.3559915393103781,
+      "learning_rate": 2.0535877197297271e-07,
+      "loss": 0.285,
+      "step": 7657
+    },
+    {
+      "epoch": 1.8227524245849942,
+      "grad_norm": 0.3952700916510172,
+      "learning_rate": 2.0481236864558663e-07,
+      "loss": 0.3013,
+      "step": 7658
+    },
+    {
+      "epoch": 1.8229904206580592,
+      "grad_norm": 0.37451269015209115,
+      "learning_rate": 2.0426667800170996e-07,
+      "loss": 0.3538,
+      "step": 7659
+    },
+    {
+      "epoch": 1.823228416731124,
+      "grad_norm": 0.37289422275429907,
+      "learning_rate": 2.0372170012244563e-07,
+      "loss": 0.3015,
+      "step": 7660
+    },
+    {
+      "epoch": 1.8234664128041889,
+      "grad_norm": 0.3849495025851358,
+      "learning_rate": 2.031774350887905e-07,
+      "loss": 0.2954,
+      "step": 7661
+    },
+    {
+      "epoch": 1.8237044088772536,
+      "grad_norm": 0.3772895053852847,
+      "learning_rate": 2.0263388298163655e-07,
+      "loss": 0.3223,
+      "step": 7662
+    },
+    {
+      "epoch": 1.8239424049503183,
+      "grad_norm": 0.35964348698217047,
+      "learning_rate": 2.0209104388176858e-07,
+      "loss": 0.3607,
+      "step": 7663
+    },
+    {
+      "epoch": 1.824180401023383,
+      "grad_norm": 0.37541695568154915,
+      "learning_rate": 2.0154891786986595e-07,
+      "loss": 0.2967,
+      "step": 7664
+    },
+    {
+      "epoch": 1.8244183970964478,
+      "grad_norm": 0.3958066722757771,
+      "learning_rate": 2.0100750502650258e-07,
+      "loss": 0.3289,
+      "step": 7665
+    },
+    {
+      "epoch": 1.8246563931695126,
+      "grad_norm": 0.4131768602497947,
+      "learning_rate": 2.0046680543214403e-07,
+      "loss": 0.3372,
+      "step": 7666
+    },
+    {
+      "epoch": 1.8248943892425775,
+      "grad_norm": 0.42405294941130334,
+      "learning_rate": 1.9992681916715385e-07,
+      "loss": 0.3229,
+      "step": 7667
+    },
+    {
+      "epoch": 1.8251323853156423,
+      "grad_norm": 0.40593441902537913,
+      "learning_rate": 1.993875463117867e-07,
+      "loss": 0.2836,
+      "step": 7668
+    },
+    {
+      "epoch": 1.8253703813887072,
+      "grad_norm": 0.37653545873556477,
+      "learning_rate": 1.9884898694619127e-07,
+      "loss": 0.3173,
+      "step": 7669
+    },
+    {
+      "epoch": 1.825608377461772,
+      "grad_norm": 0.41304477293098335,
+      "learning_rate": 1.9831114115041017e-07,
+      "loss": 0.3672,
+      "step": 7670
+    },
+    {
+      "epoch": 1.8258463735348367,
+      "grad_norm": 0.33869784993849245,
+      "learning_rate": 1.9777400900438283e-07,
+      "loss": 0.2859,
+      "step": 7671
+    },
+    {
+      "epoch": 1.8260843696079014,
+      "grad_norm": 0.3714603948136559,
+      "learning_rate": 1.9723759058793868e-07,
+      "loss": 0.2998,
+      "step": 7672
+    },
+    {
+      "epoch": 1.8263223656809662,
+      "grad_norm": 0.40178937677479326,
+      "learning_rate": 1.9670188598080342e-07,
+      "loss": 0.3435,
+      "step": 7673
+    },
+    {
+      "epoch": 1.826560361754031,
+      "grad_norm": 0.3755897088267994,
+      "learning_rate": 1.9616689526259557e-07,
+      "loss": 0.3184,
+      "step": 7674
+    },
+    {
+      "epoch": 1.8267983578270959,
+      "grad_norm": 0.3412859979077663,
+      "learning_rate": 1.9563261851282822e-07,
+      "loss": 0.2572,
+      "step": 7675
+    },
+    {
+      "epoch": 1.8270363539001606,
+      "grad_norm": 0.369227042248507,
+      "learning_rate": 1.9509905581090837e-07,
+      "loss": 0.3358,
+      "step": 7676
+    },
+    {
+      "epoch": 1.8272743499732256,
+      "grad_norm": 0.37802690397166483,
+      "learning_rate": 1.9456620723613596e-07,
+      "loss": 0.4107,
+      "step": 7677
+    },
+    {
+      "epoch": 1.8275123460462903,
+      "grad_norm": 1.2815424892667338,
+      "learning_rate": 1.9403407286770592e-07,
+      "loss": 0.2987,
+      "step": 7678
+    },
+    {
+      "epoch": 1.827750342119355,
+      "grad_norm": 0.37575903236434194,
+      "learning_rate": 1.935026527847067e-07,
+      "loss": 0.2836,
+      "step": 7679
+    },
+    {
+      "epoch": 1.8279883381924198,
+      "grad_norm": 0.37225786292995694,
+      "learning_rate": 1.9297194706612012e-07,
+      "loss": 0.3522,
+      "step": 7680
+    },
+    {
+      "epoch": 1.8282263342654845,
+      "grad_norm": 0.4069520217778423,
+      "learning_rate": 1.9244195579082193e-07,
+      "loss": 0.3215,
+      "step": 7681
+    },
+    {
+      "epoch": 1.8284643303385493,
+      "grad_norm": 0.3777178114060692,
+      "learning_rate": 1.9191267903758304e-07,
+      "loss": 0.2767,
+      "step": 7682
+    },
+    {
+      "epoch": 1.8287023264116142,
+      "grad_norm": 0.3802098683973281,
+      "learning_rate": 1.913841168850661e-07,
+      "loss": 0.308,
+      "step": 7683
+    },
+    {
+      "epoch": 1.828940322484679,
+      "grad_norm": 0.49271704172627867,
+      "learning_rate": 1.9085626941182932e-07,
+      "loss": 0.3787,
+      "step": 7684
+    },
+    {
+      "epoch": 1.829178318557744,
+      "grad_norm": 0.3772677332419929,
+      "learning_rate": 1.903291366963228e-07,
+      "loss": 0.2788,
+      "step": 7685
+    },
+    {
+      "epoch": 1.8294163146308087,
+      "grad_norm": 0.36065380388280177,
+      "learning_rate": 1.8980271881689216e-07,
+      "loss": 0.2706,
+      "step": 7686
+    },
+    {
+      "epoch": 1.8296543107038734,
+      "grad_norm": 0.377496168662484,
+      "learning_rate": 1.892770158517765e-07,
+      "loss": 0.3517,
+      "step": 7687
+    },
+    {
+      "epoch": 1.8298923067769381,
+      "grad_norm": 0.35698841434630135,
+      "learning_rate": 1.8875202787910774e-07,
+      "loss": 0.3576,
+      "step": 7688
+    },
+    {
+      "epoch": 1.8301303028500029,
+      "grad_norm": 0.366279133501771,
+      "learning_rate": 1.882277549769118e-07,
+      "loss": 0.2936,
+      "step": 7689
+    },
+    {
+      "epoch": 1.8303682989230676,
+      "grad_norm": 0.3754152778664585,
+      "learning_rate": 1.8770419722310916e-07,
+      "loss": 0.2969,
+      "step": 7690
+    },
+    {
+      "epoch": 1.8306062949961326,
+      "grad_norm": 0.3887220576270878,
+      "learning_rate": 1.8718135469551313e-07,
+      "loss": 0.3747,
+      "step": 7691
+    },
+    {
+      "epoch": 1.8308442910691973,
+      "grad_norm": 0.38990633714055073,
+      "learning_rate": 1.866592274718315e-07,
+      "loss": 0.3552,
+      "step": 7692
+    },
+    {
+      "epoch": 1.8310822871422623,
+      "grad_norm": 0.37677144527387463,
+      "learning_rate": 1.8613781562966392e-07,
+      "loss": 0.2833,
+      "step": 7693
+    },
+    {
+      "epoch": 1.831320283215327,
+      "grad_norm": 0.425287347061427,
+      "learning_rate": 1.8561711924650728e-07,
+      "loss": 0.3178,
+      "step": 7694
+    },
+    {
+      "epoch": 1.8315582792883918,
+      "grad_norm": 0.38942860989951505,
+      "learning_rate": 1.8509713839974852e-07,
+      "loss": 0.3685,
+      "step": 7695
+    },
+    {
+      "epoch": 1.8317962753614565,
+      "grad_norm": 0.3639559057337997,
+      "learning_rate": 1.8457787316667032e-07,
+      "loss": 0.2576,
+      "step": 7696
+    },
+    {
+      "epoch": 1.8320342714345212,
+      "grad_norm": 0.3541221788742815,
+      "learning_rate": 1.84059323624447e-07,
+      "loss": 0.2868,
+      "step": 7697
+    },
+    {
+      "epoch": 1.832272267507586,
+      "grad_norm": 0.3872637366190305,
+      "learning_rate": 1.835414898501492e-07,
+      "loss": 0.3573,
+      "step": 7698
+    },
+    {
+      "epoch": 1.832510263580651,
+      "grad_norm": 0.3678827746800215,
+      "learning_rate": 1.8302437192073975e-07,
+      "loss": 0.3303,
+      "step": 7699
+    },
+    {
+      "epoch": 1.8327482596537157,
+      "grad_norm": 0.36454295345870347,
+      "learning_rate": 1.8250796991307494e-07,
+      "loss": 0.296,
+      "step": 7700
+    },
+    {
+      "epoch": 1.8329862557267806,
+      "grad_norm": 0.37266982455556436,
+      "learning_rate": 1.8199228390390457e-07,
+      "loss": 0.3295,
+      "step": 7701
+    },
+    {
+      "epoch": 1.8332242517998454,
+      "grad_norm": 0.4376692678565518,
+      "learning_rate": 1.814773139698728e-07,
+      "loss": 0.3667,
+      "step": 7702
+    },
+    {
+      "epoch": 1.8334622478729101,
+      "grad_norm": 0.3683739228990514,
+      "learning_rate": 1.8096306018751675e-07,
+      "loss": 0.2748,
+      "step": 7703
+    },
+    {
+      "epoch": 1.8337002439459749,
+      "grad_norm": 0.4073240247250818,
+      "learning_rate": 1.80449522633267e-07,
+      "loss": 0.2995,
+      "step": 7704
+    },
+    {
+      "epoch": 1.8339382400190396,
+      "grad_norm": 0.39393736737366614,
+      "learning_rate": 1.7993670138344798e-07,
+      "loss": 0.3601,
+      "step": 7705
+    },
+    {
+      "epoch": 1.8341762360921043,
+      "grad_norm": 0.3767167773666181,
+      "learning_rate": 1.7942459651427825e-07,
+      "loss": 0.3498,
+      "step": 7706
+    },
+    {
+      "epoch": 1.8344142321651693,
+      "grad_norm": 0.4089511356922392,
+      "learning_rate": 1.789132081018674e-07,
+      "loss": 0.3199,
+      "step": 7707
+    },
+    {
+      "epoch": 1.834652228238234,
+      "grad_norm": 0.3812109106624504,
+      "learning_rate": 1.7840253622222303e-07,
+      "loss": 0.3192,
+      "step": 7708
+    },
+    {
+      "epoch": 1.834890224311299,
+      "grad_norm": 0.36263457008589506,
+      "learning_rate": 1.7789258095124217e-07,
+      "loss": 0.3782,
+      "step": 7709
+    },
+    {
+      "epoch": 1.8351282203843637,
+      "grad_norm": 0.3724803988209711,
+      "learning_rate": 1.77383342364717e-07,
+      "loss": 0.2963,
+      "step": 7710
+    },
+    {
+      "epoch": 1.8353662164574285,
+      "grad_norm": 0.3765829889635161,
+      "learning_rate": 1.7687482053833304e-07,
+      "loss": 0.2866,
+      "step": 7711
+    },
+    {
+      "epoch": 1.8356042125304932,
+      "grad_norm": 0.3644840472067164,
+      "learning_rate": 1.7636701554766877e-07,
+      "loss": 0.3348,
+      "step": 7712
+    },
+    {
+      "epoch": 1.835842208603558,
+      "grad_norm": 0.4078076416765129,
+      "learning_rate": 1.7585992746819713e-07,
+      "loss": 0.3769,
+      "step": 7713
+    },
+    {
+      "epoch": 1.8360802046766227,
+      "grad_norm": 0.37270746494698037,
+      "learning_rate": 1.753535563752834e-07,
+      "loss": 0.2779,
+      "step": 7714
+    },
+    {
+      "epoch": 1.8363182007496877,
+      "grad_norm": 0.38980061433134444,
+      "learning_rate": 1.7484790234418791e-07,
+      "loss": 0.3094,
+      "step": 7715
+    },
+    {
+      "epoch": 1.8365561968227524,
+      "grad_norm": 0.3952094967807275,
+      "learning_rate": 1.7434296545006224e-07,
+      "loss": 0.3964,
+      "step": 7716
+    },
+    {
+      "epoch": 1.8367941928958174,
+      "grad_norm": 0.3729540471305891,
+      "learning_rate": 1.73838745767953e-07,
+      "loss": 0.3146,
+      "step": 7717
+    },
+    {
+      "epoch": 1.837032188968882,
+      "grad_norm": 0.40875312184567286,
+      "learning_rate": 1.7333524337279918e-07,
+      "loss": 0.3322,
+      "step": 7718
+    },
+    {
+      "epoch": 1.8372701850419468,
+      "grad_norm": 0.40437790963447295,
+      "learning_rate": 1.7283245833943473e-07,
+      "loss": 0.327,
+      "step": 7719
+    },
+    {
+      "epoch": 1.8375081811150116,
+      "grad_norm": 0.37395892343895687,
+      "learning_rate": 1.723303907425844e-07,
+      "loss": 0.3752,
+      "step": 7720
+    },
+    {
+      "epoch": 1.8377461771880763,
+      "grad_norm": 0.35037991077375436,
+      "learning_rate": 1.7182904065686956e-07,
+      "loss": 0.289,
+      "step": 7721
+    },
+    {
+      "epoch": 1.837984173261141,
+      "grad_norm": 0.39297451358371965,
+      "learning_rate": 1.713284081568023e-07,
+      "loss": 0.3139,
+      "step": 7722
+    },
+    {
+      "epoch": 1.838222169334206,
+      "grad_norm": 0.4465414484001565,
+      "learning_rate": 1.708284933167892e-07,
+      "loss": 0.3373,
+      "step": 7723
+    },
+    {
+      "epoch": 1.8384601654072708,
+      "grad_norm": 0.3641160770057071,
+      "learning_rate": 1.703292962111297e-07,
+      "loss": 0.3339,
+      "step": 7724
+    },
+    {
+      "epoch": 1.8386981614803357,
+      "grad_norm": 0.37173673768425497,
+      "learning_rate": 1.6983081691401727e-07,
+      "loss": 0.2737,
+      "step": 7725
+    },
+    {
+      "epoch": 1.8389361575534005,
+      "grad_norm": 0.37865859937279667,
+      "learning_rate": 1.6933305549953817e-07,
+      "loss": 0.3234,
+      "step": 7726
+    },
+    {
+      "epoch": 1.8391741536264652,
+      "grad_norm": 0.3858905080470261,
+      "learning_rate": 1.688360120416721e-07,
+      "loss": 0.3567,
+      "step": 7727
+    },
+    {
+      "epoch": 1.83941214969953,
+      "grad_norm": 0.36393548918786084,
+      "learning_rate": 1.6833968661429168e-07,
+      "loss": 0.2794,
+      "step": 7728
+    },
+    {
+      "epoch": 1.8396501457725947,
+      "grad_norm": 0.38271700713219603,
+      "learning_rate": 1.6784407929116342e-07,
+      "loss": 0.2761,
+      "step": 7729
+    },
+    {
+      "epoch": 1.8398881418456594,
+      "grad_norm": 0.4049677071792201,
+      "learning_rate": 1.6734919014594674e-07,
+      "loss": 0.3441,
+      "step": 7730
+    },
+    {
+      "epoch": 1.8401261379187244,
+      "grad_norm": 0.36928102997215106,
+      "learning_rate": 1.668550192521945e-07,
+      "loss": 0.3377,
+      "step": 7731
+    },
+    {
+      "epoch": 1.840364133991789,
+      "grad_norm": 0.37686053081429005,
+      "learning_rate": 1.6636156668335236e-07,
+      "loss": 0.2798,
+      "step": 7732
+    },
+    {
+      "epoch": 1.840602130064854,
+      "grad_norm": 0.37607937056638047,
+      "learning_rate": 1.6586883251275998e-07,
+      "loss": 0.3279,
+      "step": 7733
+    },
+    {
+      "epoch": 1.8408401261379188,
+      "grad_norm": 0.4077890202353418,
+      "learning_rate": 1.6537681681364993e-07,
+      "loss": 0.3957,
+      "step": 7734
+    },
+    {
+      "epoch": 1.8410781222109835,
+      "grad_norm": 0.3665125487760819,
+      "learning_rate": 1.64885519659147e-07,
+      "loss": 0.2984,
+      "step": 7735
+    },
+    {
+      "epoch": 1.8413161182840483,
+      "grad_norm": 0.3856906973870745,
+      "learning_rate": 1.6439494112227173e-07,
+      "loss": 0.2753,
+      "step": 7736
+    },
+    {
+      "epoch": 1.841554114357113,
+      "grad_norm": 0.370985815282774,
+      "learning_rate": 1.6390508127593463e-07,
+      "loss": 0.3377,
+      "step": 7737
+    },
+    {
+      "epoch": 1.8417921104301778,
+      "grad_norm": 0.3622514504955658,
+      "learning_rate": 1.634159401929425e-07,
+      "loss": 0.3836,
+      "step": 7738
+    },
+    {
+      "epoch": 1.8420301065032427,
+      "grad_norm": 0.3860158792125223,
+      "learning_rate": 1.6292751794599216e-07,
+      "loss": 0.2713,
+      "step": 7739
+    },
+    {
+      "epoch": 1.8422681025763075,
+      "grad_norm": 0.3973498950561496,
+      "learning_rate": 1.6243981460767666e-07,
+      "loss": 0.3111,
+      "step": 7740
+    },
+    {
+      "epoch": 1.8425060986493724,
+      "grad_norm": 0.3638871826673044,
+      "learning_rate": 1.619528302504797e-07,
+      "loss": 0.3788,
+      "step": 7741
+    },
+    {
+      "epoch": 1.8427440947224372,
+      "grad_norm": 0.3705555624225468,
+      "learning_rate": 1.6146656494678003e-07,
+      "loss": 0.3012,
+      "step": 7742
+    },
+    {
+      "epoch": 1.842982090795502,
+      "grad_norm": 0.3584047014227033,
+      "learning_rate": 1.609810187688482e-07,
+      "loss": 0.2983,
+      "step": 7743
+    },
+    {
+      "epoch": 1.8432200868685666,
+      "grad_norm": 0.3877701919251873,
+      "learning_rate": 1.6049619178884868e-07,
+      "loss": 0.3207,
+      "step": 7744
+    },
+    {
+      "epoch": 1.8434580829416314,
+      "grad_norm": 0.3813895710119214,
+      "learning_rate": 1.6001208407883884e-07,
+      "loss": 0.3819,
+      "step": 7745
+    },
+    {
+      "epoch": 1.8436960790146961,
+      "grad_norm": 0.376270806834782,
+      "learning_rate": 1.5952869571076835e-07,
+      "loss": 0.2711,
+      "step": 7746
+    },
+    {
+      "epoch": 1.843934075087761,
+      "grad_norm": 0.3843012907616639,
+      "learning_rate": 1.5904602675648083e-07,
+      "loss": 0.2944,
+      "step": 7747
+    },
+    {
+      "epoch": 1.8441720711608258,
+      "grad_norm": 0.3918616144624181,
+      "learning_rate": 1.5856407728771394e-07,
+      "loss": 0.3551,
+      "step": 7748
+    },
+    {
+      "epoch": 1.8444100672338908,
+      "grad_norm": 0.3541860195426503,
+      "learning_rate": 1.5808284737609592e-07,
+      "loss": 0.3476,
+      "step": 7749
+    },
+    {
+      "epoch": 1.8446480633069555,
+      "grad_norm": 0.4021959688745201,
+      "learning_rate": 1.5760233709315064e-07,
+      "loss": 0.2617,
+      "step": 7750
+    },
+    {
+      "epoch": 1.8448860593800203,
+      "grad_norm": 0.4117631142432679,
+      "learning_rate": 1.571225465102927e-07,
+      "loss": 0.3208,
+      "step": 7751
+    },
+    {
+      "epoch": 1.845124055453085,
+      "grad_norm": 0.41626948663412294,
+      "learning_rate": 1.566434756988311e-07,
+      "loss": 0.3668,
+      "step": 7752
+    },
+    {
+      "epoch": 1.8453620515261497,
+      "grad_norm": 0.3503906065492793,
+      "learning_rate": 1.561651247299689e-07,
+      "loss": 0.2722,
+      "step": 7753
+    },
+    {
+      "epoch": 1.8456000475992145,
+      "grad_norm": 0.3878284628086832,
+      "learning_rate": 1.556874936747993e-07,
+      "loss": 0.3053,
+      "step": 7754
+    },
+    {
+      "epoch": 1.8458380436722794,
+      "grad_norm": 0.3664751216986663,
+      "learning_rate": 1.5521058260431043e-07,
+      "loss": 0.3567,
+      "step": 7755
+    },
+    {
+      "epoch": 1.8460760397453442,
+      "grad_norm": 0.38503978655873494,
+      "learning_rate": 1.5473439158938398e-07,
+      "loss": 0.3633,
+      "step": 7756
+    },
+    {
+      "epoch": 1.8463140358184091,
+      "grad_norm": 0.37539732052773167,
+      "learning_rate": 1.5425892070079274e-07,
+      "loss": 0.2755,
+      "step": 7757
+    },
+    {
+      "epoch": 1.8465520318914739,
+      "grad_norm": 0.3905485468486571,
+      "learning_rate": 1.5378417000920355e-07,
+      "loss": 0.3062,
+      "step": 7758
+    },
+    {
+      "epoch": 1.8467900279645386,
+      "grad_norm": 0.3717300860222802,
+      "learning_rate": 1.5331013958517604e-07,
+      "loss": 0.3826,
+      "step": 7759
+    },
+    {
+      "epoch": 1.8470280240376034,
+      "grad_norm": 0.3825543677637209,
+      "learning_rate": 1.528368294991639e-07,
+      "loss": 0.2877,
+      "step": 7760
+    },
+    {
+      "epoch": 1.847266020110668,
+      "grad_norm": 0.38491103167235174,
+      "learning_rate": 1.523642398215114e-07,
+      "loss": 0.2753,
+      "step": 7761
+    },
+    {
+      "epoch": 1.8475040161837328,
+      "grad_norm": 0.385893795694505,
+      "learning_rate": 1.5189237062245732e-07,
+      "loss": 0.3507,
+      "step": 7762
+    },
+    {
+      "epoch": 1.8477420122567978,
+      "grad_norm": 0.36863061120979945,
+      "learning_rate": 1.5142122197213338e-07,
+      "loss": 0.3815,
+      "step": 7763
+    },
+    {
+      "epoch": 1.8479800083298625,
+      "grad_norm": 0.371943151815035,
+      "learning_rate": 1.5095079394056466e-07,
+      "loss": 0.2779,
+      "step": 7764
+    },
+    {
+      "epoch": 1.8482180044029275,
+      "grad_norm": 0.41031142975223894,
+      "learning_rate": 1.5048108659766693e-07,
+      "loss": 0.2951,
+      "step": 7765
+    },
+    {
+      "epoch": 1.8484560004759922,
+      "grad_norm": 0.36678805969636596,
+      "learning_rate": 1.500121000132515e-07,
+      "loss": 0.3523,
+      "step": 7766
+    },
+    {
+      "epoch": 1.848693996549057,
+      "grad_norm": 0.3656310119733287,
+      "learning_rate": 1.4954383425702102e-07,
+      "loss": 0.3059,
+      "step": 7767
+    },
+    {
+      "epoch": 1.8489319926221217,
+      "grad_norm": 0.4742087761434599,
+      "learning_rate": 1.4907628939857087e-07,
+      "loss": 0.3005,
+      "step": 7768
+    },
+    {
+      "epoch": 1.8491699886951865,
+      "grad_norm": 0.3832956933946555,
+      "learning_rate": 1.4860946550739052e-07,
+      "loss": 0.3152,
+      "step": 7769
+    },
+    {
+      "epoch": 1.8494079847682512,
+      "grad_norm": 0.3707364198788508,
+      "learning_rate": 1.4814336265286112e-07,
+      "loss": 0.3801,
+      "step": 7770
+    },
+    {
+      "epoch": 1.8496459808413162,
+      "grad_norm": 0.3794533583364384,
+      "learning_rate": 1.476779809042572e-07,
+      "loss": 0.2977,
+      "step": 7771
+    },
+    {
+      "epoch": 1.849883976914381,
+      "grad_norm": 0.4142293643990621,
+      "learning_rate": 1.4721332033074575e-07,
+      "loss": 0.31,
+      "step": 7772
+    },
+    {
+      "epoch": 1.8501219729874459,
+      "grad_norm": 0.364668075185516,
+      "learning_rate": 1.467493810013876e-07,
+      "loss": 0.3325,
+      "step": 7773
+    },
+    {
+      "epoch": 1.8503599690605106,
+      "grad_norm": 0.36286815622733415,
+      "learning_rate": 1.462861629851342e-07,
+      "loss": 0.3554,
+      "step": 7774
+    },
+    {
+      "epoch": 1.8505979651335753,
+      "grad_norm": 0.3942139877708106,
+      "learning_rate": 1.4582366635083223e-07,
+      "loss": 0.2767,
+      "step": 7775
+    },
+    {
+      "epoch": 1.85083596120664,
+      "grad_norm": 0.37049438808023333,
+      "learning_rate": 1.4536189116722056e-07,
+      "loss": 0.2968,
+      "step": 7776
+    },
+    {
+      "epoch": 1.8510739572797048,
+      "grad_norm": 0.42757648044300495,
+      "learning_rate": 1.4490083750292984e-07,
+      "loss": 0.3868,
+      "step": 7777
+    },
+    {
+      "epoch": 1.8513119533527695,
+      "grad_norm": 0.3666344543082101,
+      "learning_rate": 1.4444050542648302e-07,
+      "loss": 0.289,
+      "step": 7778
+    },
+    {
+      "epoch": 1.8515499494258345,
+      "grad_norm": 0.4136576161910318,
+      "learning_rate": 1.4398089500629874e-07,
+      "loss": 0.2883,
+      "step": 7779
+    },
+    {
+      "epoch": 1.8517879454988992,
+      "grad_norm": 0.38875371128162894,
+      "learning_rate": 1.4352200631068515e-07,
+      "loss": 0.3883,
+      "step": 7780
+    },
+    {
+      "epoch": 1.8520259415719642,
+      "grad_norm": 0.37524678870829536,
+      "learning_rate": 1.430638394078454e-07,
+      "loss": 0.3115,
+      "step": 7781
+    },
+    {
+      "epoch": 1.852263937645029,
+      "grad_norm": 0.3915337777058191,
+      "learning_rate": 1.4260639436587398e-07,
+      "loss": 0.2911,
+      "step": 7782
+    },
+    {
+      "epoch": 1.8525019337180937,
+      "grad_norm": 0.4065274951070914,
+      "learning_rate": 1.4214967125275814e-07,
+      "loss": 0.3267,
+      "step": 7783
+    },
+    {
+      "epoch": 1.8527399297911584,
+      "grad_norm": 0.3753625153872455,
+      "learning_rate": 1.4169367013637857e-07,
+      "loss": 0.374,
+      "step": 7784
+    },
+    {
+      "epoch": 1.8529779258642232,
+      "grad_norm": 0.39450510369704206,
+      "learning_rate": 1.4123839108450832e-07,
+      "loss": 0.3065,
+      "step": 7785
+    },
+    {
+      "epoch": 1.853215921937288,
+      "grad_norm": 0.38706116936482216,
+      "learning_rate": 1.4078383416481321e-07,
+      "loss": 0.273,
+      "step": 7786
+    },
+    {
+      "epoch": 1.8534539180103529,
+      "grad_norm": 0.3993104090511267,
+      "learning_rate": 1.4032999944485203e-07,
+      "loss": 0.3124,
+      "step": 7787
+    },
+    {
+      "epoch": 1.8536919140834176,
+      "grad_norm": 0.389522809964699,
+      "learning_rate": 1.398768869920747e-07,
+      "loss": 0.3647,
+      "step": 7788
+    },
+    {
+      "epoch": 1.8539299101564826,
+      "grad_norm": 0.4039304741367637,
+      "learning_rate": 1.3942449687382565e-07,
+      "loss": 0.2761,
+      "step": 7789
+    },
+    {
+      "epoch": 1.8541679062295473,
+      "grad_norm": 0.42136351566968644,
+      "learning_rate": 1.389728291573411e-07,
+      "loss": 0.328,
+      "step": 7790
+    },
+    {
+      "epoch": 1.854405902302612,
+      "grad_norm": 0.4212557672996947,
+      "learning_rate": 1.3852188390975073e-07,
+      "loss": 0.3918,
+      "step": 7791
+    },
+    {
+      "epoch": 1.8546438983756768,
+      "grad_norm": 0.3603764523291382,
+      "learning_rate": 1.380716611980748e-07,
+      "loss": 0.3351,
+      "step": 7792
+    },
+    {
+      "epoch": 1.8548818944487415,
+      "grad_norm": 0.35966797992735267,
+      "learning_rate": 1.3762216108922922e-07,
+      "loss": 0.2732,
+      "step": 7793
+    },
+    {
+      "epoch": 1.8551198905218063,
+      "grad_norm": 0.36333670692392717,
+      "learning_rate": 1.3717338365001943e-07,
+      "loss": 0.329,
+      "step": 7794
+    },
+    {
+      "epoch": 1.8553578865948712,
+      "grad_norm": 0.38329227054003784,
+      "learning_rate": 1.367253289471454e-07,
+      "loss": 0.3802,
+      "step": 7795
+    },
+    {
+      "epoch": 1.855595882667936,
+      "grad_norm": 0.3514869841098985,
+      "learning_rate": 1.3627799704719947e-07,
+      "loss": 0.2712,
+      "step": 7796
+    },
+    {
+      "epoch": 1.855833878741001,
+      "grad_norm": 0.381197916996613,
+      "learning_rate": 1.358313880166656e-07,
+      "loss": 0.302,
+      "step": 7797
+    },
+    {
+      "epoch": 1.8560718748140657,
+      "grad_norm": 0.402462868821984,
+      "learning_rate": 1.3538550192192078e-07,
+      "loss": 0.3535,
+      "step": 7798
+    },
+    {
+      "epoch": 1.8563098708871304,
+      "grad_norm": 0.3514030830100683,
+      "learning_rate": 1.3494033882923586e-07,
+      "loss": 0.3455,
+      "step": 7799
+    },
+    {
+      "epoch": 1.8565478669601951,
+      "grad_norm": 0.3781598067445243,
+      "learning_rate": 1.3449589880477176e-07,
+      "loss": 0.2882,
+      "step": 7800
+    },
+    {
+      "epoch": 1.8567858630332599,
+      "grad_norm": 0.38189268313397884,
+      "learning_rate": 1.3405218191458402e-07,
+      "loss": 0.3177,
+      "step": 7801
+    },
+    {
+      "epoch": 1.8570238591063246,
+      "grad_norm": 0.3890034707887386,
+      "learning_rate": 1.3360918822461989e-07,
+      "loss": 0.3846,
+      "step": 7802
+    },
+    {
+      "epoch": 1.8572618551793896,
+      "grad_norm": 0.354555606174368,
+      "learning_rate": 1.331669178007189e-07,
+      "loss": 0.2828,
+      "step": 7803
+    },
+    {
+      "epoch": 1.8574998512524543,
+      "grad_norm": 0.39468523966196106,
+      "learning_rate": 1.32725370708614e-07,
+      "loss": 0.2641,
+      "step": 7804
+    },
+    {
+      "epoch": 1.8577378473255193,
+      "grad_norm": 0.4158745671593587,
+      "learning_rate": 1.322845470139289e-07,
+      "loss": 0.3666,
+      "step": 7805
+    },
+    {
+      "epoch": 1.857975843398584,
+      "grad_norm": 0.38138886762707447,
+      "learning_rate": 1.3184444678218223e-07,
+      "loss": 0.3276,
+      "step": 7806
+    },
+    {
+      "epoch": 1.8582138394716488,
+      "grad_norm": 0.35114076586398596,
+      "learning_rate": 1.314050700787828e-07,
+      "loss": 0.2949,
+      "step": 7807
+    },
+    {
+      "epoch": 1.8584518355447135,
+      "grad_norm": 0.36724157814202074,
+      "learning_rate": 1.3096641696903334e-07,
+      "loss": 0.3347,
+      "step": 7808
+    },
+    {
+      "epoch": 1.8586898316177782,
+      "grad_norm": 0.38853816925181384,
+      "learning_rate": 1.3052848751812842e-07,
+      "loss": 0.3861,
+      "step": 7809
+    },
+    {
+      "epoch": 1.858927827690843,
+      "grad_norm": 0.3903142761173342,
+      "learning_rate": 1.3009128179115539e-07,
+      "loss": 0.3111,
+      "step": 7810
+    },
+    {
+      "epoch": 1.859165823763908,
+      "grad_norm": 0.38876795886287885,
+      "learning_rate": 1.2965479985309338e-07,
+      "loss": 0.2738,
+      "step": 7811
+    },
+    {
+      "epoch": 1.8594038198369727,
+      "grad_norm": 0.4101960252962491,
+      "learning_rate": 1.2921904176881494e-07,
+      "loss": 0.3681,
+      "step": 7812
+    },
+    {
+      "epoch": 1.8596418159100376,
+      "grad_norm": 0.36520915728705955,
+      "learning_rate": 1.2878400760308385e-07,
+      "loss": 0.3571,
+      "step": 7813
+    },
+    {
+      "epoch": 1.8598798119831024,
+      "grad_norm": 0.40612725661487625,
+      "learning_rate": 1.2834969742055725e-07,
+      "loss": 0.2713,
+      "step": 7814
+    },
+    {
+      "epoch": 1.860117808056167,
+      "grad_norm": 0.6769806659333333,
+      "learning_rate": 1.2791611128578463e-07,
+      "loss": 0.3343,
+      "step": 7815
+    },
+    {
+      "epoch": 1.8603558041292318,
+      "grad_norm": 0.40138777376249124,
+      "learning_rate": 1.2748324926320777e-07,
+      "loss": 0.3592,
+      "step": 7816
+    },
+    {
+      "epoch": 1.8605938002022966,
+      "grad_norm": 0.37763740906507537,
+      "learning_rate": 1.270511114171591e-07,
+      "loss": 0.3313,
+      "step": 7817
+    },
+    {
+      "epoch": 1.8608317962753613,
+      "grad_norm": 0.4238056700691326,
+      "learning_rate": 1.2661969781186723e-07,
+      "loss": 0.2498,
+      "step": 7818
+    },
+    {
+      "epoch": 1.8610697923484263,
+      "grad_norm": 0.37963559080128195,
+      "learning_rate": 1.2618900851144976e-07,
+      "loss": 0.3321,
+      "step": 7819
+    },
+    {
+      "epoch": 1.861307788421491,
+      "grad_norm": 0.3885643564174667,
+      "learning_rate": 1.2575904357991775e-07,
+      "loss": 0.3975,
+      "step": 7820
+    },
+    {
+      "epoch": 1.861545784494556,
+      "grad_norm": 0.34011787013626105,
+      "learning_rate": 1.2532980308117503e-07,
+      "loss": 0.272,
+      "step": 7821
+    },
+    {
+      "epoch": 1.8617837805676207,
+      "grad_norm": 0.3891036676630237,
+      "learning_rate": 1.2490128707901727e-07,
+      "loss": 0.293,
+      "step": 7822
+    },
+    {
+      "epoch": 1.8620217766406855,
+      "grad_norm": 0.38193328581420083,
+      "learning_rate": 1.2447349563713186e-07,
+      "loss": 0.3707,
+      "step": 7823
+    },
+    {
+      "epoch": 1.8622597727137502,
+      "grad_norm": 0.3658925256499234,
+      "learning_rate": 1.2404642881910012e-07,
+      "loss": 0.3051,
+      "step": 7824
+    },
+    {
+      "epoch": 1.862497768786815,
+      "grad_norm": 0.3801598075071461,
+      "learning_rate": 1.236200866883941e-07,
+      "loss": 0.2614,
+      "step": 7825
+    },
+    {
+      "epoch": 1.8627357648598797,
+      "grad_norm": 0.36720137768282524,
+      "learning_rate": 1.231944693083792e-07,
+      "loss": 0.3171,
+      "step": 7826
+    },
+    {
+      "epoch": 1.8629737609329446,
+      "grad_norm": 0.4198297803663176,
+      "learning_rate": 1.2276957674231204e-07,
+      "loss": 0.3851,
+      "step": 7827
+    },
+    {
+      "epoch": 1.8632117570060094,
+      "grad_norm": 0.35023106406242704,
+      "learning_rate": 1.223454090533427e-07,
+      "loss": 0.2795,
+      "step": 7828
+    },
+    {
+      "epoch": 1.8634497530790743,
+      "grad_norm": 0.34830108173573165,
+      "learning_rate": 1.219219663045129e-07,
+      "loss": 0.2655,
+      "step": 7829
+    },
+    {
+      "epoch": 1.863687749152139,
+      "grad_norm": 0.40557375679941676,
+      "learning_rate": 1.2149924855875737e-07,
+      "loss": 0.386,
+      "step": 7830
+    },
+    {
+      "epoch": 1.8639257452252038,
+      "grad_norm": 0.38135226051859966,
+      "learning_rate": 1.210772558789014e-07,
+      "loss": 0.3438,
+      "step": 7831
+    },
+    {
+      "epoch": 1.8641637412982686,
+      "grad_norm": 0.3573950365210543,
+      "learning_rate": 1.2065598832766369e-07,
+      "loss": 0.2979,
+      "step": 7832
+    },
+    {
+      "epoch": 1.8644017373713333,
+      "grad_norm": 0.35921086327394625,
+      "learning_rate": 1.202354459676558e-07,
+      "loss": 0.2962,
+      "step": 7833
+    },
+    {
+      "epoch": 1.864639733444398,
+      "grad_norm": 0.3826967666594862,
+      "learning_rate": 1.1981562886137998e-07,
+      "loss": 0.375,
+      "step": 7834
+    },
+    {
+      "epoch": 1.864877729517463,
+      "grad_norm": 0.3639494907850754,
+      "learning_rate": 1.1939653707123132e-07,
+      "loss": 0.281,
+      "step": 7835
+    },
+    {
+      "epoch": 1.8651157255905277,
+      "grad_norm": 0.37370185714825943,
+      "learning_rate": 1.1897817065949835e-07,
+      "loss": 0.2973,
+      "step": 7836
+    },
+    {
+      "epoch": 1.8653537216635927,
+      "grad_norm": 0.5048736432794805,
+      "learning_rate": 1.1856052968835907e-07,
+      "loss": 0.3218,
+      "step": 7837
+    },
+    {
+      "epoch": 1.8655917177366574,
+      "grad_norm": 0.36966116531403587,
+      "learning_rate": 1.1814361421988662e-07,
+      "loss": 0.3956,
+      "step": 7838
+    },
+    {
+      "epoch": 1.8658297138097222,
+      "grad_norm": 0.3560493828010198,
+      "learning_rate": 1.1772742431604423e-07,
+      "loss": 0.2868,
+      "step": 7839
+    },
+    {
+      "epoch": 1.866067709882787,
+      "grad_norm": 0.5370977594652094,
+      "learning_rate": 1.1731196003868794e-07,
+      "loss": 0.3322,
+      "step": 7840
+    },
+    {
+      "epoch": 1.8663057059558517,
+      "grad_norm": 0.3922980128293181,
+      "learning_rate": 1.1689722144956672e-07,
+      "loss": 0.3587,
+      "step": 7841
+    },
+    {
+      "epoch": 1.8665437020289164,
+      "grad_norm": 0.32486417759838265,
+      "learning_rate": 1.164832086103207e-07,
+      "loss": 0.3013,
+      "step": 7842
+    },
+    {
+      "epoch": 1.8667816981019814,
+      "grad_norm": 0.38729519267969026,
+      "learning_rate": 1.1606992158248177e-07,
+      "loss": 0.2793,
+      "step": 7843
+    },
+    {
+      "epoch": 1.867019694175046,
+      "grad_norm": 0.37956208767813593,
+      "learning_rate": 1.1565736042747522e-07,
+      "loss": 0.3162,
+      "step": 7844
+    },
+    {
+      "epoch": 1.867257690248111,
+      "grad_norm": 0.4116795086325041,
+      "learning_rate": 1.1524552520661702e-07,
+      "loss": 0.3733,
+      "step": 7845
+    },
+    {
+      "epoch": 1.8674956863211758,
+      "grad_norm": 0.39437952445062263,
+      "learning_rate": 1.1483441598111766e-07,
+      "loss": 0.3036,
+      "step": 7846
+    },
+    {
+      "epoch": 1.8677336823942405,
+      "grad_norm": 0.39085643388306573,
+      "learning_rate": 1.1442403281207714e-07,
+      "loss": 0.3244,
+      "step": 7847
+    },
+    {
+      "epoch": 1.8679716784673053,
+      "grad_norm": 0.4014816625744995,
+      "learning_rate": 1.140143757604889e-07,
+      "loss": 0.3738,
+      "step": 7848
+    },
+    {
+      "epoch": 1.86820967454037,
+      "grad_norm": 0.3769060375261056,
+      "learning_rate": 1.1360544488723756e-07,
+      "loss": 0.3044,
+      "step": 7849
+    },
+    {
+      "epoch": 1.8684476706134348,
+      "grad_norm": 0.3904744507763302,
+      "learning_rate": 1.1319724025310063e-07,
+      "loss": 0.2969,
+      "step": 7850
+    },
+    {
+      "epoch": 1.8686856666864997,
+      "grad_norm": 0.3773232634572751,
+      "learning_rate": 1.1278976191874735e-07,
+      "loss": 0.3294,
+      "step": 7851
+    },
+    {
+      "epoch": 1.8689236627595645,
+      "grad_norm": 0.3775912202758717,
+      "learning_rate": 1.1238300994473983e-07,
+      "loss": 0.3817,
+      "step": 7852
+    },
+    {
+      "epoch": 1.8691616588326294,
+      "grad_norm": 0.36177147793921166,
+      "learning_rate": 1.1197698439153027e-07,
+      "loss": 0.2878,
+      "step": 7853
+    },
+    {
+      "epoch": 1.8693996549056942,
+      "grad_norm": 0.3942909323087775,
+      "learning_rate": 1.1157168531946483e-07,
+      "loss": 0.2825,
+      "step": 7854
+    },
+    {
+      "epoch": 1.869637650978759,
+      "grad_norm": 0.38479407162289014,
+      "learning_rate": 1.1116711278878034e-07,
+      "loss": 0.3565,
+      "step": 7855
+    },
+    {
+      "epoch": 1.8698756470518236,
+      "grad_norm": 0.37536080195437777,
+      "learning_rate": 1.1076326685960758e-07,
+      "loss": 0.3585,
+      "step": 7856
+    },
+    {
+      "epoch": 1.8701136431248884,
+      "grad_norm": 0.4072733768621761,
+      "learning_rate": 1.1036014759196689e-07,
+      "loss": 0.2605,
+      "step": 7857
+    },
+    {
+      "epoch": 1.870351639197953,
+      "grad_norm": 0.3794776719624951,
+      "learning_rate": 1.09957755045772e-07,
+      "loss": 0.3195,
+      "step": 7858
+    },
+    {
+      "epoch": 1.870589635271018,
+      "grad_norm": 0.3781656516512505,
+      "learning_rate": 1.095560892808284e-07,
+      "loss": 0.3695,
+      "step": 7859
+    },
+    {
+      "epoch": 1.8708276313440828,
+      "grad_norm": 0.3477470340626564,
+      "learning_rate": 1.0915515035683444e-07,
+      "loss": 0.2976,
+      "step": 7860
+    },
+    {
+      "epoch": 1.8710656274171478,
+      "grad_norm": 0.39868562095433935,
+      "learning_rate": 1.0875493833337803e-07,
+      "loss": 0.2701,
+      "step": 7861
+    },
+    {
+      "epoch": 1.8713036234902125,
+      "grad_norm": 0.4054869894626742,
+      "learning_rate": 1.0835545326994213e-07,
+      "loss": 0.3525,
+      "step": 7862
+    },
+    {
+      "epoch": 1.8715416195632772,
+      "grad_norm": 0.36521275528859204,
+      "learning_rate": 1.079566952258987e-07,
+      "loss": 0.3625,
+      "step": 7863
+    },
+    {
+      "epoch": 1.871779615636342,
+      "grad_norm": 0.3635342656993625,
+      "learning_rate": 1.075586642605142e-07,
+      "loss": 0.268,
+      "step": 7864
+    },
+    {
+      "epoch": 1.8720176117094067,
+      "grad_norm": 0.37356788241263744,
+      "learning_rate": 1.0716136043294468e-07,
+      "loss": 0.2874,
+      "step": 7865
+    },
+    {
+      "epoch": 1.8722556077824715,
+      "grad_norm": 0.3961313481479438,
+      "learning_rate": 1.0676478380224065e-07,
+      "loss": 0.3772,
+      "step": 7866
+    },
+    {
+      "epoch": 1.8724936038555364,
+      "grad_norm": 0.37248128983442935,
+      "learning_rate": 1.063689344273422e-07,
+      "loss": 0.291,
+      "step": 7867
+    },
+    {
+      "epoch": 1.8727315999286012,
+      "grad_norm": 0.42564199958608884,
+      "learning_rate": 1.0597381236708282e-07,
+      "loss": 0.2925,
+      "step": 7868
+    },
+    {
+      "epoch": 1.8729695960016661,
+      "grad_norm": 0.3945675089261431,
+      "learning_rate": 1.055794176801872e-07,
+      "loss": 0.3273,
+      "step": 7869
+    },
+    {
+      "epoch": 1.8732075920747309,
+      "grad_norm": 0.36858121380582176,
+      "learning_rate": 1.0518575042527235e-07,
+      "loss": 0.425,
+      "step": 7870
+    },
+    {
+      "epoch": 1.8734455881477956,
+      "grad_norm": 0.4008228433529883,
+      "learning_rate": 1.0479281066084701e-07,
+      "loss": 0.2753,
+      "step": 7871
+    },
+    {
+      "epoch": 1.8736835842208603,
+      "grad_norm": 0.3970329160462698,
+      "learning_rate": 1.044005984453117e-07,
+      "loss": 0.3065,
+      "step": 7872
+    },
+    {
+      "epoch": 1.873921580293925,
+      "grad_norm": 0.3752973477487869,
+      "learning_rate": 1.0400911383695756e-07,
+      "loss": 0.3393,
+      "step": 7873
+    },
+    {
+      "epoch": 1.8741595763669898,
+      "grad_norm": 0.34507839687636127,
+      "learning_rate": 1.0361835689397137e-07,
+      "loss": 0.3171,
+      "step": 7874
+    },
+    {
+      "epoch": 1.8743975724400548,
+      "grad_norm": 0.3956732220335401,
+      "learning_rate": 1.0322832767442726e-07,
+      "loss": 0.2712,
+      "step": 7875
+    },
+    {
+      "epoch": 1.8746355685131195,
+      "grad_norm": 0.3478429638128932,
+      "learning_rate": 1.0283902623629439e-07,
+      "loss": 0.3163,
+      "step": 7876
+    },
+    {
+      "epoch": 1.8748735645861845,
+      "grad_norm": 0.43937488881046505,
+      "learning_rate": 1.0245045263743203e-07,
+      "loss": 0.3866,
+      "step": 7877
+    },
+    {
+      "epoch": 1.8751115606592492,
+      "grad_norm": 0.3712312999960514,
+      "learning_rate": 1.0206260693559234e-07,
+      "loss": 0.2936,
+      "step": 7878
+    },
+    {
+      "epoch": 1.875349556732314,
+      "grad_norm": 0.3667219720522898,
+      "learning_rate": 1.0167548918841752e-07,
+      "loss": 0.2841,
+      "step": 7879
+    },
+    {
+      "epoch": 1.8755875528053787,
+      "grad_norm": 0.3758275879637338,
+      "learning_rate": 1.0128909945344433e-07,
+      "loss": 0.3117,
+      "step": 7880
+    },
+    {
+      "epoch": 1.8758255488784434,
+      "grad_norm": 0.3693311113464016,
+      "learning_rate": 1.0090343778809908e-07,
+      "loss": 0.3607,
+      "step": 7881
+    },
+    {
+      "epoch": 1.8760635449515082,
+      "grad_norm": 0.3689248776098616,
+      "learning_rate": 1.0051850424970034e-07,
+      "loss": 0.287,
+      "step": 7882
+    },
+    {
+      "epoch": 1.8763015410245731,
+      "grad_norm": 0.3877904699869412,
+      "learning_rate": 1.0013429889546011e-07,
+      "loss": 0.3327,
+      "step": 7883
+    },
+    {
+      "epoch": 1.8765395370976379,
+      "grad_norm": 0.3813592994481828,
+      "learning_rate": 9.975082178247942e-08,
+      "loss": 0.3817,
+      "step": 7884
+    },
+    {
+      "epoch": 1.8767775331707028,
+      "grad_norm": 0.39447485339729055,
+      "learning_rate": 9.936807296775264e-08,
+      "loss": 0.3179,
+      "step": 7885
+    },
+    {
+      "epoch": 1.8770155292437676,
+      "grad_norm": 0.36838965887067854,
+      "learning_rate": 9.898605250816596e-08,
+      "loss": 0.2767,
+      "step": 7886
+    },
+    {
+      "epoch": 1.8772535253168323,
+      "grad_norm": 0.40695179381212354,
+      "learning_rate": 9.860476046049783e-08,
+      "loss": 0.3116,
+      "step": 7887
+    },
+    {
+      "epoch": 1.877491521389897,
+      "grad_norm": 0.37922684546428304,
+      "learning_rate": 9.822419688141627e-08,
+      "loss": 0.371,
+      "step": 7888
+    },
+    {
+      "epoch": 1.8777295174629618,
+      "grad_norm": 0.3605426474272245,
+      "learning_rate": 9.784436182748381e-08,
+      "loss": 0.3157,
+      "step": 7889
+    },
+    {
+      "epoch": 1.8779675135360265,
+      "grad_norm": 0.3855508493344283,
+      "learning_rate": 9.74652553551525e-08,
+      "loss": 0.3126,
+      "step": 7890
+    },
+    {
+      "epoch": 1.8782055096090915,
+      "grad_norm": 0.36654027571515835,
+      "learning_rate": 9.708687752076673e-08,
+      "loss": 0.3796,
+      "step": 7891
+    },
+    {
+      "epoch": 1.8784435056821562,
+      "grad_norm": 0.37798898644072126,
+      "learning_rate": 9.670922838056374e-08,
+      "loss": 0.3268,
+      "step": 7892
+    },
+    {
+      "epoch": 1.8786815017552212,
+      "grad_norm": 0.37273754822520255,
+      "learning_rate": 9.633230799067084e-08,
+      "loss": 0.2805,
+      "step": 7893
+    },
+    {
+      "epoch": 1.878919497828286,
+      "grad_norm": 0.38458823706391126,
+      "learning_rate": 9.595611640710767e-08,
+      "loss": 0.3494,
+      "step": 7894
+    },
+    {
+      "epoch": 1.8791574939013507,
+      "grad_norm": 0.39167494539281444,
+      "learning_rate": 9.558065368578561e-08,
+      "loss": 0.3506,
+      "step": 7895
+    },
+    {
+      "epoch": 1.8793954899744154,
+      "grad_norm": 0.3574825980828279,
+      "learning_rate": 9.520591988250838e-08,
+      "loss": 0.2689,
+      "step": 7896
+    },
+    {
+      "epoch": 1.8796334860474802,
+      "grad_norm": 0.37303769358161776,
+      "learning_rate": 9.483191505296974e-08,
+      "loss": 0.3235,
+      "step": 7897
+    },
+    {
+      "epoch": 1.879871482120545,
+      "grad_norm": 0.3651354155087522,
+      "learning_rate": 9.44586392527569e-08,
+      "loss": 0.3755,
+      "step": 7898
+    },
+    {
+      "epoch": 1.8801094781936099,
+      "grad_norm": 0.39194478737773597,
+      "learning_rate": 9.408609253734713e-08,
+      "loss": 0.3255,
+      "step": 7899
+    },
+    {
+      "epoch": 1.8803474742666746,
+      "grad_norm": 0.3713897973414568,
+      "learning_rate": 9.371427496211061e-08,
+      "loss": 0.2869,
+      "step": 7900
+    },
+    {
+      "epoch": 1.8805854703397396,
+      "grad_norm": 0.36090648920272766,
+      "learning_rate": 9.334318658230867e-08,
+      "loss": 0.3191,
+      "step": 7901
+    },
+    {
+      "epoch": 1.8808234664128043,
+      "grad_norm": 0.41318377703708387,
+      "learning_rate": 9.297282745309389e-08,
+      "loss": 0.3773,
+      "step": 7902
+    },
+    {
+      "epoch": 1.881061462485869,
+      "grad_norm": 0.3713769438671181,
+      "learning_rate": 9.260319762951109e-08,
+      "loss": 0.3186,
+      "step": 7903
+    },
+    {
+      "epoch": 1.8812994585589338,
+      "grad_norm": 0.3858895133643765,
+      "learning_rate": 9.223429716649634e-08,
+      "loss": 0.2755,
+      "step": 7904
+    },
+    {
+      "epoch": 1.8815374546319985,
+      "grad_norm": 0.422907620593951,
+      "learning_rate": 9.186612611887691e-08,
+      "loss": 0.3659,
+      "step": 7905
+    },
+    {
+      "epoch": 1.8817754507050632,
+      "grad_norm": 0.4024148982753146,
+      "learning_rate": 9.14986845413729e-08,
+      "loss": 0.3693,
+      "step": 7906
+    },
+    {
+      "epoch": 1.8820134467781282,
+      "grad_norm": 0.3803493722755298,
+      "learning_rate": 9.113197248859451e-08,
+      "loss": 0.2724,
+      "step": 7907
+    },
+    {
+      "epoch": 1.882251442851193,
+      "grad_norm": 0.4118593592302728,
+      "learning_rate": 9.076599001504482e-08,
+      "loss": 0.3282,
+      "step": 7908
+    },
+    {
+      "epoch": 1.882489438924258,
+      "grad_norm": 0.3953308094670974,
+      "learning_rate": 9.0400737175117e-08,
+      "loss": 0.3764,
+      "step": 7909
+    },
+    {
+      "epoch": 1.8827274349973226,
+      "grad_norm": 0.3696749681639606,
+      "learning_rate": 9.003621402309815e-08,
+      "loss": 0.277,
+      "step": 7910
+    },
+    {
+      "epoch": 1.8829654310703874,
+      "grad_norm": 0.3615369062856125,
+      "learning_rate": 8.96724206131644e-08,
+      "loss": 0.2998,
+      "step": 7911
+    },
+    {
+      "epoch": 1.8832034271434521,
+      "grad_norm": 0.41428190444047164,
+      "learning_rate": 8.930935699938415e-08,
+      "loss": 0.3354,
+      "step": 7912
+    },
+    {
+      "epoch": 1.8834414232165169,
+      "grad_norm": 0.3929646777015546,
+      "learning_rate": 8.89470232357187e-08,
+      "loss": 0.3684,
+      "step": 7913
+    },
+    {
+      "epoch": 1.8836794192895816,
+      "grad_norm": 0.3589833395799497,
+      "learning_rate": 8.858541937601827e-08,
+      "loss": 0.2739,
+      "step": 7914
+    },
+    {
+      "epoch": 1.8839174153626466,
+      "grad_norm": 0.3889496001692624,
+      "learning_rate": 8.82245454740277e-08,
+      "loss": 0.2876,
+      "step": 7915
+    },
+    {
+      "epoch": 1.8841554114357113,
+      "grad_norm": 0.42247196753368266,
+      "learning_rate": 8.786440158338072e-08,
+      "loss": 0.3592,
+      "step": 7916
+    },
+    {
+      "epoch": 1.8843934075087763,
+      "grad_norm": 0.36693957355643053,
+      "learning_rate": 8.750498775760453e-08,
+      "loss": 0.3229,
+      "step": 7917
+    },
+    {
+      "epoch": 1.884631403581841,
+      "grad_norm": 0.37575006564385804,
+      "learning_rate": 8.714630405011637e-08,
+      "loss": 0.2745,
+      "step": 7918
+    },
+    {
+      "epoch": 1.8848693996549057,
+      "grad_norm": 0.3699668444902597,
+      "learning_rate": 8.678835051422585e-08,
+      "loss": 0.3479,
+      "step": 7919
+    },
+    {
+      "epoch": 1.8851073957279705,
+      "grad_norm": 0.37951849531963455,
+      "learning_rate": 8.643112720313262e-08,
+      "loss": 0.3748,
+      "step": 7920
+    },
+    {
+      "epoch": 1.8853453918010352,
+      "grad_norm": 0.401706429927378,
+      "learning_rate": 8.607463416993034e-08,
+      "loss": 0.2929,
+      "step": 7921
+    },
+    {
+      "epoch": 1.8855833878741,
+      "grad_norm": 0.40506962687618053,
+      "learning_rate": 8.571887146760217e-08,
+      "loss": 0.3326,
+      "step": 7922
+    },
+    {
+      "epoch": 1.885821383947165,
+      "grad_norm": 0.38167411063561124,
+      "learning_rate": 8.536383914902301e-08,
+      "loss": 0.3743,
+      "step": 7923
+    },
+    {
+      "epoch": 1.8860593800202297,
+      "grad_norm": 0.37951337247623673,
+      "learning_rate": 8.500953726695959e-08,
+      "loss": 0.3368,
+      "step": 7924
+    },
+    {
+      "epoch": 1.8862973760932946,
+      "grad_norm": 0.3807047055623607,
+      "learning_rate": 8.465596587406977e-08,
+      "loss": 0.2633,
+      "step": 7925
+    },
+    {
+      "epoch": 1.8865353721663594,
+      "grad_norm": 0.37255448111092837,
+      "learning_rate": 8.430312502290316e-08,
+      "loss": 0.3475,
+      "step": 7926
+    },
+    {
+      "epoch": 1.886773368239424,
+      "grad_norm": 0.3904240340365388,
+      "learning_rate": 8.395101476590062e-08,
+      "loss": 0.3999,
+      "step": 7927
+    },
+    {
+      "epoch": 1.8870113643124888,
+      "grad_norm": 0.3590248194841884,
+      "learning_rate": 8.359963515539416e-08,
+      "loss": 0.2924,
+      "step": 7928
+    },
+    {
+      "epoch": 1.8872493603855536,
+      "grad_norm": 0.38702468210446556,
+      "learning_rate": 8.324898624360867e-08,
+      "loss": 0.2733,
+      "step": 7929
+    },
+    {
+      "epoch": 1.8874873564586183,
+      "grad_norm": 0.3814942319217911,
+      "learning_rate": 8.289906808265746e-08,
+      "loss": 0.3281,
+      "step": 7930
+    },
+    {
+      "epoch": 1.8877253525316833,
+      "grad_norm": 0.40004094569529125,
+      "learning_rate": 8.254988072454895e-08,
+      "loss": 0.3284,
+      "step": 7931
+    },
+    {
+      "epoch": 1.887963348604748,
+      "grad_norm": 0.3723689654618086,
+      "learning_rate": 8.220142422117939e-08,
+      "loss": 0.2681,
+      "step": 7932
+    },
+    {
+      "epoch": 1.888201344677813,
+      "grad_norm": 0.44726403333459647,
+      "learning_rate": 8.185369862433845e-08,
+      "loss": 0.3056,
+      "step": 7933
+    },
+    {
+      "epoch": 1.8884393407508777,
+      "grad_norm": 0.3806982139299245,
+      "learning_rate": 8.150670398570759e-08,
+      "loss": 0.3877,
+      "step": 7934
+    },
+    {
+      "epoch": 1.8886773368239425,
+      "grad_norm": 0.35216108572000016,
+      "learning_rate": 8.116044035685777e-08,
+      "loss": 0.2884,
+      "step": 7935
+    },
+    {
+      "epoch": 1.8889153328970072,
+      "grad_norm": 0.39643635386909837,
+      "learning_rate": 8.081490778925283e-08,
+      "loss": 0.2902,
+      "step": 7936
+    },
+    {
+      "epoch": 1.889153328970072,
+      "grad_norm": 0.4923610296056875,
+      "learning_rate": 8.047010633424723e-08,
+      "loss": 0.3359,
+      "step": 7937
+    },
+    {
+      "epoch": 1.8893913250431367,
+      "grad_norm": 0.379519284347565,
+      "learning_rate": 8.012603604308721e-08,
+      "loss": 0.3535,
+      "step": 7938
+    },
+    {
+      "epoch": 1.8896293211162016,
+      "grad_norm": 0.3635457197510201,
+      "learning_rate": 7.978269696691021e-08,
+      "loss": 0.2947,
+      "step": 7939
+    },
+    {
+      "epoch": 1.8898673171892664,
+      "grad_norm": 0.4033859742678164,
+      "learning_rate": 7.944008915674484e-08,
+      "loss": 0.311,
+      "step": 7940
+    },
+    {
+      "epoch": 1.8901053132623313,
+      "grad_norm": 0.3982200170533466,
+      "learning_rate": 7.909821266351092e-08,
+      "loss": 0.3703,
+      "step": 7941
+    },
+    {
+      "epoch": 1.890343309335396,
+      "grad_norm": 0.35373099489430676,
+      "learning_rate": 7.875706753801949e-08,
+      "loss": 0.3103,
+      "step": 7942
+    },
+    {
+      "epoch": 1.8905813054084608,
+      "grad_norm": 0.36747336775173683,
+      "learning_rate": 7.841665383097386e-08,
+      "loss": 0.2825,
+      "step": 7943
+    },
+    {
+      "epoch": 1.8908193014815255,
+      "grad_norm": 0.3650587611745797,
+      "learning_rate": 7.807697159296746e-08,
+      "loss": 0.3499,
+      "step": 7944
+    },
+    {
+      "epoch": 1.8910572975545903,
+      "grad_norm": 0.40187256746234257,
+      "learning_rate": 7.773802087448545e-08,
+      "loss": 0.356,
+      "step": 7945
+    },
+    {
+      "epoch": 1.891295293627655,
+      "grad_norm": 0.37435669326276105,
+      "learning_rate": 7.739980172590477e-08,
+      "loss": 0.275,
+      "step": 7946
+    },
+    {
+      "epoch": 1.89153328970072,
+      "grad_norm": 0.36212200367835184,
+      "learning_rate": 7.706231419749243e-08,
+      "loss": 0.3142,
+      "step": 7947
+    },
+    {
+      "epoch": 1.8917712857737847,
+      "grad_norm": 0.39951418029554553,
+      "learning_rate": 7.672555833940832e-08,
+      "loss": 0.3621,
+      "step": 7948
+    },
+    {
+      "epoch": 1.8920092818468497,
+      "grad_norm": 0.3786672956664201,
+      "learning_rate": 7.638953420170181e-08,
+      "loss": 0.3187,
+      "step": 7949
+    },
+    {
+      "epoch": 1.8922472779199144,
+      "grad_norm": 0.3923921053093728,
+      "learning_rate": 7.605424183431464e-08,
+      "loss": 0.2786,
+      "step": 7950
+    },
+    {
+      "epoch": 1.8924852739929792,
+      "grad_norm": 0.38868628053542476,
+      "learning_rate": 7.57196812870803e-08,
+      "loss": 0.3226,
+      "step": 7951
+    },
+    {
+      "epoch": 1.892723270066044,
+      "grad_norm": 0.3985157429933502,
+      "learning_rate": 7.538585260972175e-08,
+      "loss": 0.365,
+      "step": 7952
+    },
+    {
+      "epoch": 1.8929612661391086,
+      "grad_norm": 0.41970802109407324,
+      "learning_rate": 7.505275585185434e-08,
+      "loss": 0.2964,
+      "step": 7953
+    },
+    {
+      "epoch": 1.8931992622121734,
+      "grad_norm": 0.38733261918545264,
+      "learning_rate": 7.472039106298512e-08,
+      "loss": 0.3075,
+      "step": 7954
+    },
+    {
+      "epoch": 1.8934372582852383,
+      "grad_norm": 0.4141672520453948,
+      "learning_rate": 7.438875829251069e-08,
+      "loss": 0.3563,
+      "step": 7955
+    },
+    {
+      "epoch": 1.893675254358303,
+      "grad_norm": 0.3763459702365932,
+      "learning_rate": 7.405785758972106e-08,
+      "loss": 0.3256,
+      "step": 7956
+    },
+    {
+      "epoch": 1.893913250431368,
+      "grad_norm": 0.41340503131702716,
+      "learning_rate": 7.372768900379579e-08,
+      "loss": 0.276,
+      "step": 7957
+    },
+    {
+      "epoch": 1.8941512465044328,
+      "grad_norm": 0.3690483636794343,
+      "learning_rate": 7.339825258380618e-08,
+      "loss": 0.3253,
+      "step": 7958
+    },
+    {
+      "epoch": 1.8943892425774975,
+      "grad_norm": 0.3890667108974889,
+      "learning_rate": 7.306954837871415e-08,
+      "loss": 0.3858,
+      "step": 7959
+    },
+    {
+      "epoch": 1.8946272386505623,
+      "grad_norm": 0.3686373090623398,
+      "learning_rate": 7.2741576437374e-08,
+      "loss": 0.3156,
+      "step": 7960
+    },
+    {
+      "epoch": 1.894865234723627,
+      "grad_norm": 0.3662856876936937,
+      "learning_rate": 7.241433680852949e-08,
+      "loss": 0.2893,
+      "step": 7961
+    },
+    {
+      "epoch": 1.8951032307966917,
+      "grad_norm": 0.4337210767045377,
+      "learning_rate": 7.208782954081784e-08,
+      "loss": 0.347,
+      "step": 7962
+    },
+    {
+      "epoch": 1.8953412268697567,
+      "grad_norm": 0.37458324393332965,
+      "learning_rate": 7.176205468276465e-08,
+      "loss": 0.395,
+      "step": 7963
+    },
+    {
+      "epoch": 1.8955792229428214,
+      "grad_norm": 0.36173109464150077,
+      "learning_rate": 7.143701228278899e-08,
+      "loss": 0.2887,
+      "step": 7964
+    },
+    {
+      "epoch": 1.8958172190158864,
+      "grad_norm": 0.38262759512926825,
+      "learning_rate": 7.111270238920054e-08,
+      "loss": 0.2915,
+      "step": 7965
+    },
+    {
+      "epoch": 1.8960552150889511,
+      "grad_norm": 0.3827036517273416,
+      "learning_rate": 7.078912505019908e-08,
+      "loss": 0.3438,
+      "step": 7966
+    },
+    {
+      "epoch": 1.8962932111620159,
+      "grad_norm": 0.3848814530905264,
+      "learning_rate": 7.046628031387615e-08,
+      "loss": 0.301,
+      "step": 7967
+    },
+    {
+      "epoch": 1.8965312072350806,
+      "grad_norm": 0.4067525298074377,
+      "learning_rate": 7.014416822821557e-08,
+      "loss": 0.2432,
+      "step": 7968
+    },
+    {
+      "epoch": 1.8967692033081454,
+      "grad_norm": 0.3812327537956296,
+      "learning_rate": 6.982278884108907e-08,
+      "loss": 0.3234,
+      "step": 7969
+    },
+    {
+      "epoch": 1.89700719938121,
+      "grad_norm": 0.3605852849420692,
+      "learning_rate": 6.950214220026397e-08,
+      "loss": 0.4051,
+      "step": 7970
+    },
+    {
+      "epoch": 1.897245195454275,
+      "grad_norm": 0.3958665551698103,
+      "learning_rate": 6.918222835339438e-08,
+      "loss": 0.2686,
+      "step": 7971
+    },
+    {
+      "epoch": 1.8974831915273398,
+      "grad_norm": 0.38258672492708956,
+      "learning_rate": 6.886304734802896e-08,
+      "loss": 0.3141,
+      "step": 7972
+    },
+    {
+      "epoch": 1.8977211876004048,
+      "grad_norm": 0.3981035167445995,
+      "learning_rate": 6.854459923160472e-08,
+      "loss": 0.3582,
+      "step": 7973
+    },
+    {
+      "epoch": 1.8979591836734695,
+      "grad_norm": 0.36696157556372344,
+      "learning_rate": 6.822688405145161e-08,
+      "loss": 0.3243,
+      "step": 7974
+    },
+    {
+      "epoch": 1.8981971797465342,
+      "grad_norm": 0.3713373387211765,
+      "learning_rate": 6.790990185478963e-08,
+      "loss": 0.2546,
+      "step": 7975
+    },
+    {
+      "epoch": 1.898435175819599,
+      "grad_norm": 0.37672668107775037,
+      "learning_rate": 6.759365268872997e-08,
+      "loss": 0.3332,
+      "step": 7976
+    },
+    {
+      "epoch": 1.8986731718926637,
+      "grad_norm": 0.359979308125644,
+      "learning_rate": 6.727813660027616e-08,
+      "loss": 0.3621,
+      "step": 7977
+    },
+    {
+      "epoch": 1.8989111679657285,
+      "grad_norm": 0.3608027535934777,
+      "learning_rate": 6.696335363632012e-08,
+      "loss": 0.2963,
+      "step": 7978
+    },
+    {
+      "epoch": 1.8991491640387934,
+      "grad_norm": 0.37376170686319327,
+      "learning_rate": 6.66493038436472e-08,
+      "loss": 0.2775,
+      "step": 7979
+    },
+    {
+      "epoch": 1.8993871601118582,
+      "grad_norm": 0.38424326860289765,
+      "learning_rate": 6.633598726893342e-08,
+      "loss": 0.3589,
+      "step": 7980
+    },
+    {
+      "epoch": 1.8996251561849231,
+      "grad_norm": 0.394059524418179,
+      "learning_rate": 6.602340395874484e-08,
+      "loss": 0.3771,
+      "step": 7981
+    },
+    {
+      "epoch": 1.8998631522579879,
+      "grad_norm": 0.37382562755513166,
+      "learning_rate": 6.571155395953877e-08,
+      "loss": 0.2874,
+      "step": 7982
+    },
+    {
+      "epoch": 1.9001011483310526,
+      "grad_norm": 0.3874100204834066,
+      "learning_rate": 6.540043731766421e-08,
+      "loss": 0.3195,
+      "step": 7983
+    },
+    {
+      "epoch": 1.9003391444041173,
+      "grad_norm": 0.48921624228327715,
+      "learning_rate": 6.509005407936087e-08,
+      "loss": 0.3603,
+      "step": 7984
+    },
+    {
+      "epoch": 1.900577140477182,
+      "grad_norm": 0.35858308962539986,
+      "learning_rate": 6.478040429075961e-08,
+      "loss": 0.2859,
+      "step": 7985
+    },
+    {
+      "epoch": 1.9008151365502468,
+      "grad_norm": 0.3932336834908693,
+      "learning_rate": 6.447148799788139e-08,
+      "loss": 0.2843,
+      "step": 7986
+    },
+    {
+      "epoch": 1.9010531326233118,
+      "grad_norm": 0.36226032388387175,
+      "learning_rate": 6.416330524663895e-08,
+      "loss": 0.3366,
+      "step": 7987
+    },
+    {
+      "epoch": 1.9012911286963765,
+      "grad_norm": 0.41269859656532076,
+      "learning_rate": 6.385585608283673e-08,
+      "loss": 0.3668,
+      "step": 7988
+    },
+    {
+      "epoch": 1.9015291247694415,
+      "grad_norm": 0.3845523305990824,
+      "learning_rate": 6.35491405521682e-08,
+      "loss": 0.2758,
+      "step": 7989
+    },
+    {
+      "epoch": 1.9017671208425062,
+      "grad_norm": 0.40301323636395514,
+      "learning_rate": 6.324315870021858e-08,
+      "loss": 0.3034,
+      "step": 7990
+    },
+    {
+      "epoch": 1.902005116915571,
+      "grad_norm": 0.3907642073015502,
+      "learning_rate": 6.29379105724659e-08,
+      "loss": 0.3739,
+      "step": 7991
+    },
+    {
+      "epoch": 1.9022431129886357,
+      "grad_norm": 0.3606625473176197,
+      "learning_rate": 6.263339621427666e-08,
+      "loss": 0.3433,
+      "step": 7992
+    },
+    {
+      "epoch": 1.9024811090617004,
+      "grad_norm": 0.4090491853767471,
+      "learning_rate": 6.232961567090912e-08,
+      "loss": 0.2904,
+      "step": 7993
+    },
+    {
+      "epoch": 1.9027191051347652,
+      "grad_norm": 0.3812316669315538,
+      "learning_rate": 6.202656898751324e-08,
+      "loss": 0.3236,
+      "step": 7994
+    },
+    {
+      "epoch": 1.9029571012078301,
+      "grad_norm": 0.37564697095435645,
+      "learning_rate": 6.172425620912859e-08,
+      "loss": 0.3895,
+      "step": 7995
+    },
+    {
+      "epoch": 1.9031950972808949,
+      "grad_norm": 0.38971270053161045,
+      "learning_rate": 6.142267738068641e-08,
+      "loss": 0.2818,
+      "step": 7996
+    },
+    {
+      "epoch": 1.9034330933539596,
+      "grad_norm": 0.4139644761266531,
+      "learning_rate": 6.112183254700866e-08,
+      "loss": 0.3051,
+      "step": 7997
+    },
+    {
+      "epoch": 1.9036710894270246,
+      "grad_norm": 0.4000612669015111,
+      "learning_rate": 6.082172175280843e-08,
+      "loss": 0.3858,
+      "step": 7998
+    },
+    {
+      "epoch": 1.9039090855000893,
+      "grad_norm": 0.39138000482768825,
+      "learning_rate": 6.052234504269006e-08,
+      "loss": 0.2829,
+      "step": 7999
+    },
+    {
+      "epoch": 1.904147081573154,
+      "grad_norm": 0.4111965153470907,
+      "learning_rate": 6.022370246114795e-08,
+      "loss": 0.2599,
+      "step": 8000
+    },
+    {
+      "epoch": 1.9043850776462188,
+      "grad_norm": 0.38739615879574596,
+      "learning_rate": 5.992579405256826e-08,
+      "loss": 0.3336,
+      "step": 8001
+    },
+    {
+      "epoch": 1.9046230737192835,
+      "grad_norm": 0.35399850972575364,
+      "learning_rate": 5.96286198612267e-08,
+      "loss": 0.4036,
+      "step": 8002
+    },
+    {
+      "epoch": 1.9048610697923483,
+      "grad_norm": 0.3501272012525231,
+      "learning_rate": 5.933217993129126e-08,
+      "loss": 0.2928,
+      "step": 8003
+    },
+    {
+      "epoch": 1.9050990658654132,
+      "grad_norm": 0.3940432999472092,
+      "learning_rate": 5.903647430682002e-08,
+      "loss": 0.2813,
+      "step": 8004
+    },
+    {
+      "epoch": 1.905337061938478,
+      "grad_norm": 0.37153990539467646,
+      "learning_rate": 5.8741503031762294e-08,
+      "loss": 0.3495,
+      "step": 8005
+    },
+    {
+      "epoch": 1.905575058011543,
+      "grad_norm": 0.3650091291465742,
+      "learning_rate": 5.844726614995799e-08,
+      "loss": 0.3493,
+      "step": 8006
+    },
+    {
+      "epoch": 1.9058130540846077,
+      "grad_norm": 0.35578280548382063,
+      "learning_rate": 5.815376370513825e-08,
+      "loss": 0.2911,
+      "step": 8007
+    },
+    {
+      "epoch": 1.9060510501576724,
+      "grad_norm": 0.4008983674611723,
+      "learning_rate": 5.7860995740924296e-08,
+      "loss": 0.3143,
+      "step": 8008
+    },
+    {
+      "epoch": 1.9062890462307371,
+      "grad_norm": 0.38432833244657477,
+      "learning_rate": 5.756896230082909e-08,
+      "loss": 0.3776,
+      "step": 8009
+    },
+    {
+      "epoch": 1.9065270423038019,
+      "grad_norm": 0.3632804695720169,
+      "learning_rate": 5.7277663428256245e-08,
+      "loss": 0.2985,
+      "step": 8010
+    },
+    {
+      "epoch": 1.9067650383768666,
+      "grad_norm": 0.3906672244339681,
+      "learning_rate": 5.698709916649892e-08,
+      "loss": 0.2724,
+      "step": 8011
+    },
+    {
+      "epoch": 1.9070030344499316,
+      "grad_norm": 0.38051344490028755,
+      "learning_rate": 5.669726955874366e-08,
+      "loss": 0.3162,
+      "step": 8012
+    },
+    {
+      "epoch": 1.9072410305229963,
+      "grad_norm": 0.35282134896325107,
+      "learning_rate": 5.640817464806547e-08,
+      "loss": 0.3467,
+      "step": 8013
+    },
+    {
+      "epoch": 1.9074790265960613,
+      "grad_norm": 0.38263778679767824,
+      "learning_rate": 5.611981447743109e-08,
+      "loss": 0.2883,
+      "step": 8014
+    },
+    {
+      "epoch": 1.907717022669126,
+      "grad_norm": 0.37534374522335623,
+      "learning_rate": 5.5832189089697895e-08,
+      "loss": 0.3081,
+      "step": 8015
+    },
+    {
+      "epoch": 1.9079550187421908,
+      "grad_norm": 0.3824203414300469,
+      "learning_rate": 5.554529852761337e-08,
+      "loss": 0.3706,
+      "step": 8016
+    },
+    {
+      "epoch": 1.9081930148152555,
+      "grad_norm": 0.37799082292639585,
+      "learning_rate": 5.525914283381839e-08,
+      "loss": 0.3255,
+      "step": 8017
+    },
+    {
+      "epoch": 1.9084310108883202,
+      "grad_norm": 0.3920668628878149,
+      "learning_rate": 5.497372205084173e-08,
+      "loss": 0.2966,
+      "step": 8018
+    },
+    {
+      "epoch": 1.908669006961385,
+      "grad_norm": 0.34537104511664757,
+      "learning_rate": 5.468903622110389e-08,
+      "loss": 0.3038,
+      "step": 8019
+    },
+    {
+      "epoch": 1.90890700303445,
+      "grad_norm": 0.3841678485218043,
+      "learning_rate": 5.440508538691658e-08,
+      "loss": 0.3513,
+      "step": 8020
+    },
+    {
+      "epoch": 1.9091449991075147,
+      "grad_norm": 0.34656903960442637,
+      "learning_rate": 5.412186959048105e-08,
+      "loss": 0.2848,
+      "step": 8021
+    },
+    {
+      "epoch": 1.9093829951805796,
+      "grad_norm": 0.3940028400776739,
+      "learning_rate": 5.38393888738914e-08,
+      "loss": 0.2805,
+      "step": 8022
+    },
+    {
+      "epoch": 1.9096209912536444,
+      "grad_norm": 0.3819034767089527,
+      "learning_rate": 5.355764327913071e-08,
+      "loss": 0.3605,
+      "step": 8023
+    },
+    {
+      "epoch": 1.9098589873267091,
+      "grad_norm": 0.4050020295830134,
+      "learning_rate": 5.3276632848072716e-08,
+      "loss": 0.3406,
+      "step": 8024
+    },
+    {
+      "epoch": 1.9100969833997739,
+      "grad_norm": 0.3711188459584683,
+      "learning_rate": 5.299635762248345e-08,
+      "loss": 0.2649,
+      "step": 8025
+    },
+    {
+      "epoch": 1.9103349794728386,
+      "grad_norm": 0.3825043059653578,
+      "learning_rate": 5.271681764401848e-08,
+      "loss": 0.297,
+      "step": 8026
+    },
+    {
+      "epoch": 1.9105729755459033,
+      "grad_norm": 0.3872078261635392,
+      "learning_rate": 5.243801295422457e-08,
+      "loss": 0.3961,
+      "step": 8027
+    },
+    {
+      "epoch": 1.9108109716189683,
+      "grad_norm": 0.3675581463780866,
+      "learning_rate": 5.215994359453858e-08,
+      "loss": 0.2728,
+      "step": 8028
+    },
+    {
+      "epoch": 1.911048967692033,
+      "grad_norm": 0.39794626943566946,
+      "learning_rate": 5.188260960628855e-08,
+      "loss": 0.2939,
+      "step": 8029
+    },
+    {
+      "epoch": 1.911286963765098,
+      "grad_norm": 0.3881930755761182,
+      "learning_rate": 5.1606011030693184e-08,
+      "loss": 0.3508,
+      "step": 8030
+    },
+    {
+      "epoch": 1.9115249598381627,
+      "grad_norm": 0.34966274001131864,
+      "learning_rate": 5.1330147908861814e-08,
+      "loss": 0.3254,
+      "step": 8031
+    },
+    {
+      "epoch": 1.9117629559112275,
+      "grad_norm": 0.3810722565822783,
+      "learning_rate": 5.1055020281794987e-08,
+      "loss": 0.2676,
+      "step": 8032
+    },
+    {
+      "epoch": 1.9120009519842922,
+      "grad_norm": 0.3913423806435706,
+      "learning_rate": 5.078062819038332e-08,
+      "loss": 0.3274,
+      "step": 8033
+    },
+    {
+      "epoch": 1.912238948057357,
+      "grad_norm": 0.3957900932077373,
+      "learning_rate": 5.0506971675407526e-08,
+      "loss": 0.3859,
+      "step": 8034
+    },
+    {
+      "epoch": 1.9124769441304217,
+      "grad_norm": 0.41296040287870084,
+      "learning_rate": 5.0234050777540625e-08,
+      "loss": 0.2761,
+      "step": 8035
+    },
+    {
+      "epoch": 1.9127149402034866,
+      "grad_norm": 0.3617049666384471,
+      "learning_rate": 4.996186553734517e-08,
+      "loss": 0.258,
+      "step": 8036
+    },
+    {
+      "epoch": 1.9129529362765514,
+      "grad_norm": 0.4056471829864505,
+      "learning_rate": 4.9690415995274354e-08,
+      "loss": 0.3416,
+      "step": 8037
+    },
+    {
+      "epoch": 1.9131909323496163,
+      "grad_norm": 0.3575725862341246,
+      "learning_rate": 4.941970219167203e-08,
+      "loss": 0.3405,
+      "step": 8038
+    },
+    {
+      "epoch": 1.913428928422681,
+      "grad_norm": 0.3654418605929413,
+      "learning_rate": 4.914972416677433e-08,
+      "loss": 0.2714,
+      "step": 8039
+    },
+    {
+      "epoch": 1.9136669244957458,
+      "grad_norm": 0.3928718133203763,
+      "learning_rate": 4.8880481960705274e-08,
+      "loss": 0.3084,
+      "step": 8040
+    },
+    {
+      "epoch": 1.9139049205688106,
+      "grad_norm": 0.3860998147396003,
+      "learning_rate": 4.861197561348119e-08,
+      "loss": 0.3449,
+      "step": 8041
+    },
+    {
+      "epoch": 1.9141429166418753,
+      "grad_norm": 0.3828325589358172,
+      "learning_rate": 4.83442051650096e-08,
+      "loss": 0.3055,
+      "step": 8042
+    },
+    {
+      "epoch": 1.91438091271494,
+      "grad_norm": 0.39352418664155536,
+      "learning_rate": 4.8077170655086436e-08,
+      "loss": 0.2848,
+      "step": 8043
+    },
+    {
+      "epoch": 1.914618908788005,
+      "grad_norm": 0.4101375769165912,
+      "learning_rate": 4.781087212340052e-08,
+      "loss": 0.3486,
+      "step": 8044
+    },
+    {
+      "epoch": 1.9148569048610697,
+      "grad_norm": 0.39533684005769876,
+      "learning_rate": 4.754530960953074e-08,
+      "loss": 0.3536,
+      "step": 8045
+    },
+    {
+      "epoch": 1.9150949009341347,
+      "grad_norm": 0.37662010337344887,
+      "learning_rate": 4.728048315294553e-08,
+      "loss": 0.315,
+      "step": 8046
+    },
+    {
+      "epoch": 1.9153328970071994,
+      "grad_norm": 0.3934959783113809,
+      "learning_rate": 4.701639279300507e-08,
+      "loss": 0.3085,
+      "step": 8047
+    },
+    {
+      "epoch": 1.9155708930802642,
+      "grad_norm": 0.37377434092719386,
+      "learning_rate": 4.675303856895907e-08,
+      "loss": 0.381,
+      "step": 8048
+    },
+    {
+      "epoch": 1.915808889153329,
+      "grad_norm": 0.37694394724451996,
+      "learning_rate": 4.649042051994956e-08,
+      "loss": 0.3324,
+      "step": 8049
+    },
+    {
+      "epoch": 1.9160468852263937,
+      "grad_norm": 0.37595294353982567,
+      "learning_rate": 4.622853868500699e-08,
+      "loss": 0.2523,
+      "step": 8050
+    },
+    {
+      "epoch": 1.9162848812994584,
+      "grad_norm": 0.40725905564623294,
+      "learning_rate": 4.596739310305409e-08,
+      "loss": 0.3108,
+      "step": 8051
+    },
+    {
+      "epoch": 1.9165228773725234,
+      "grad_norm": 0.38543823318432485,
+      "learning_rate": 4.570698381290317e-08,
+      "loss": 0.3867,
+      "step": 8052
+    },
+    {
+      "epoch": 1.916760873445588,
+      "grad_norm": 0.3693394614353527,
+      "learning_rate": 4.5447310853258265e-08,
+      "loss": 0.306,
+      "step": 8053
+    },
+    {
+      "epoch": 1.916998869518653,
+      "grad_norm": 0.374918757902991,
+      "learning_rate": 4.5188374262712385e-08,
+      "loss": 0.2909,
+      "step": 8054
+    },
+    {
+      "epoch": 1.9172368655917178,
+      "grad_norm": 0.4268585385107588,
+      "learning_rate": 4.493017407975087e-08,
+      "loss": 0.3706,
+      "step": 8055
+    },
+    {
+      "epoch": 1.9174748616647825,
+      "grad_norm": 0.3550622241647807,
+      "learning_rate": 4.467271034274745e-08,
+      "loss": 0.3147,
+      "step": 8056
+    },
+    {
+      "epoch": 1.9177128577378473,
+      "grad_norm": 0.3673246840410985,
+      "learning_rate": 4.44159830899682e-08,
+      "loss": 0.2714,
+      "step": 8057
+    },
+    {
+      "epoch": 1.917950853810912,
+      "grad_norm": 0.38166747800870954,
+      "learning_rate": 4.415999235956925e-08,
+      "loss": 0.3132,
+      "step": 8058
+    },
+    {
+      "epoch": 1.9181888498839768,
+      "grad_norm": 0.37082445385628915,
+      "learning_rate": 4.390473818959684e-08,
+      "loss": 0.3427,
+      "step": 8059
+    },
+    {
+      "epoch": 1.9184268459570417,
+      "grad_norm": 0.3502851478207926,
+      "learning_rate": 4.3650220617988404e-08,
+      "loss": 0.3055,
+      "step": 8060
+    },
+    {
+      "epoch": 1.9186648420301065,
+      "grad_norm": 0.38376398396770767,
+      "learning_rate": 4.3396439682570904e-08,
+      "loss": 0.2819,
+      "step": 8061
+    },
+    {
+      "epoch": 1.9189028381031714,
+      "grad_norm": 0.37104295492217804,
+      "learning_rate": 4.3143395421063607e-08,
+      "loss": 0.3136,
+      "step": 8062
+    },
+    {
+      "epoch": 1.9191408341762362,
+      "grad_norm": 0.39064486463386716,
+      "learning_rate": 4.2891087871073656e-08,
+      "loss": 0.3645,
+      "step": 8063
+    },
+    {
+      "epoch": 1.919378830249301,
+      "grad_norm": 0.3959614629937633,
+      "learning_rate": 4.263951707010161e-08,
+      "loss": 0.2889,
+      "step": 8064
+    },
+    {
+      "epoch": 1.9196168263223656,
+      "grad_norm": 0.3976404851904098,
+      "learning_rate": 4.238868305553645e-08,
+      "loss": 0.2977,
+      "step": 8065
+    },
+    {
+      "epoch": 1.9198548223954304,
+      "grad_norm": 0.4187234518989236,
+      "learning_rate": 4.2138585864658354e-08,
+      "loss": 0.3897,
+      "step": 8066
+    },
+    {
+      "epoch": 1.920092818468495,
+      "grad_norm": 0.37755488054016684,
+      "learning_rate": 4.188922553463759e-08,
+      "loss": 0.3252,
+      "step": 8067
+    },
+    {
+      "epoch": 1.92033081454156,
+      "grad_norm": 0.3915281355437284,
+      "learning_rate": 4.164060210253618e-08,
+      "loss": 0.2707,
+      "step": 8068
+    },
+    {
+      "epoch": 1.9205688106146248,
+      "grad_norm": 0.39725191077466404,
+      "learning_rate": 4.1392715605305114e-08,
+      "loss": 0.346,
+      "step": 8069
+    },
+    {
+      "epoch": 1.9208068066876898,
+      "grad_norm": 0.3714723524123798,
+      "learning_rate": 4.1145566079786034e-08,
+      "loss": 0.3478,
+      "step": 8070
+    },
+    {
+      "epoch": 1.9210448027607545,
+      "grad_norm": 0.3707101045708407,
+      "learning_rate": 4.089915356271234e-08,
+      "loss": 0.3095,
+      "step": 8071
+    },
+    {
+      "epoch": 1.9212827988338192,
+      "grad_norm": 0.4055723864611074,
+      "learning_rate": 4.0653478090706965e-08,
+      "loss": 0.2812,
+      "step": 8072
+    },
+    {
+      "epoch": 1.921520794906884,
+      "grad_norm": 0.3801871133061815,
+      "learning_rate": 4.040853970028291e-08,
+      "loss": 0.3632,
+      "step": 8073
+    },
+    {
+      "epoch": 1.9217587909799487,
+      "grad_norm": 0.3539168169356299,
+      "learning_rate": 4.01643384278444e-08,
+      "loss": 0.3052,
+      "step": 8074
+    },
+    {
+      "epoch": 1.9219967870530135,
+      "grad_norm": 0.3611686215376351,
+      "learning_rate": 3.992087430968516e-08,
+      "loss": 0.2806,
+      "step": 8075
+    },
+    {
+      "epoch": 1.9222347831260784,
+      "grad_norm": 0.3913596751920842,
+      "learning_rate": 3.967814738199072e-08,
+      "loss": 0.3105,
+      "step": 8076
+    },
+    {
+      "epoch": 1.9224727791991432,
+      "grad_norm": 0.39705193175327785,
+      "learning_rate": 3.943615768083609e-08,
+      "loss": 0.3657,
+      "step": 8077
+    },
+    {
+      "epoch": 1.9227107752722081,
+      "grad_norm": 0.35680102901673433,
+      "learning_rate": 3.919490524218694e-08,
+      "loss": 0.2939,
+      "step": 8078
+    },
+    {
+      "epoch": 1.9229487713452729,
+      "grad_norm": 0.3936950953927909,
+      "learning_rate": 3.89543901018985e-08,
+      "loss": 0.2862,
+      "step": 8079
+    },
+    {
+      "epoch": 1.9231867674183376,
+      "grad_norm": 0.41880114246300687,
+      "learning_rate": 3.8714612295718824e-08,
+      "loss": 0.3537,
+      "step": 8080
+    },
+    {
+      "epoch": 1.9234247634914023,
+      "grad_norm": 0.3639233905209205,
+      "learning_rate": 3.8475571859283855e-08,
+      "loss": 0.3672,
+      "step": 8081
+    },
+    {
+      "epoch": 1.923662759564467,
+      "grad_norm": 0.3903343153212714,
+      "learning_rate": 3.823726882812129e-08,
+      "loss": 0.2758,
+      "step": 8082
+    },
+    {
+      "epoch": 1.9239007556375318,
+      "grad_norm": 0.3900369810611726,
+      "learning_rate": 3.7999703237648346e-08,
+      "loss": 0.3241,
+      "step": 8083
+    },
+    {
+      "epoch": 1.9241387517105968,
+      "grad_norm": 0.39117092554001687,
+      "learning_rate": 3.776287512317345e-08,
+      "loss": 0.3595,
+      "step": 8084
+    },
+    {
+      "epoch": 1.9243767477836615,
+      "grad_norm": 0.36654968786594255,
+      "learning_rate": 3.752678451989567e-08,
+      "loss": 0.2833,
+      "step": 8085
+    },
+    {
+      "epoch": 1.9246147438567265,
+      "grad_norm": 0.3809066672279184,
+      "learning_rate": 3.729143146290304e-08,
+      "loss": 0.2862,
+      "step": 8086
+    },
+    {
+      "epoch": 1.9248527399297912,
+      "grad_norm": 0.3490062069699892,
+      "learning_rate": 3.705681598717481e-08,
+      "loss": 0.3525,
+      "step": 8087
+    },
+    {
+      "epoch": 1.925090736002856,
+      "grad_norm": 0.379879949063031,
+      "learning_rate": 3.682293812758142e-08,
+      "loss": 0.3721,
+      "step": 8088
+    },
+    {
+      "epoch": 1.9253287320759207,
+      "grad_norm": 0.3778359839016277,
+      "learning_rate": 3.658979791888284e-08,
+      "loss": 0.3119,
+      "step": 8089
+    },
+    {
+      "epoch": 1.9255667281489854,
+      "grad_norm": 0.36956910970266094,
+      "learning_rate": 3.635739539572858e-08,
+      "loss": 0.3345,
+      "step": 8090
+    },
+    {
+      "epoch": 1.9258047242220502,
+      "grad_norm": 0.41589648566449683,
+      "learning_rate": 3.6125730592660445e-08,
+      "loss": 0.4026,
+      "step": 8091
+    },
+    {
+      "epoch": 1.9260427202951151,
+      "grad_norm": 0.37675858564217196,
+      "learning_rate": 3.589480354410868e-08,
+      "loss": 0.3182,
+      "step": 8092
+    },
+    {
+      "epoch": 1.9262807163681799,
+      "grad_norm": 0.3885940611238435,
+      "learning_rate": 3.5664614284395274e-08,
+      "loss": 0.2901,
+      "step": 8093
+    },
+    {
+      "epoch": 1.9265187124412448,
+      "grad_norm": 0.4267648886909655,
+      "learning_rate": 3.543516284773174e-08,
+      "loss": 0.3237,
+      "step": 8094
+    },
+    {
+      "epoch": 1.9267567085143096,
+      "grad_norm": 0.37790981095161896,
+      "learning_rate": 3.520644926822081e-08,
+      "loss": 0.4189,
+      "step": 8095
+    },
+    {
+      "epoch": 1.9269947045873743,
+      "grad_norm": 0.35608236467137666,
+      "learning_rate": 3.497847357985418e-08,
+      "loss": 0.3035,
+      "step": 8096
+    },
+    {
+      "epoch": 1.927232700660439,
+      "grad_norm": 0.35125924042990425,
+      "learning_rate": 3.475123581651529e-08,
+      "loss": 0.2894,
+      "step": 8097
+    },
+    {
+      "epoch": 1.9274706967335038,
+      "grad_norm": 0.3867965627570178,
+      "learning_rate": 3.452473601197659e-08,
+      "loss": 0.3634,
+      "step": 8098
+    },
+    {
+      "epoch": 1.9277086928065685,
+      "grad_norm": 0.35092993273506357,
+      "learning_rate": 3.42989741999028e-08,
+      "loss": 0.3621,
+      "step": 8099
+    },
+    {
+      "epoch": 1.9279466888796335,
+      "grad_norm": 0.37737964092644094,
+      "learning_rate": 3.4073950413846536e-08,
+      "loss": 0.2531,
+      "step": 8100
+    },
+    {
+      "epoch": 1.9281846849526982,
+      "grad_norm": 0.37016176061370865,
+      "learning_rate": 3.38496646872527e-08,
+      "loss": 0.3185,
+      "step": 8101
+    },
+    {
+      "epoch": 1.9284226810257632,
+      "grad_norm": 0.41856577365765557,
+      "learning_rate": 3.36261170534552e-08,
+      "loss": 0.3914,
+      "step": 8102
+    },
+    {
+      "epoch": 1.928660677098828,
+      "grad_norm": 0.350550972272308,
+      "learning_rate": 3.340330754567911e-08,
+      "loss": 0.295,
+      "step": 8103
+    },
+    {
+      "epoch": 1.9288986731718927,
+      "grad_norm": 0.37139534698154514,
+      "learning_rate": 3.3181236197038505e-08,
+      "loss": 0.295,
+      "step": 8104
+    },
+    {
+      "epoch": 1.9291366692449574,
+      "grad_norm": 0.4026861513587683,
+      "learning_rate": 3.295990304054031e-08,
+      "loss": 0.374,
+      "step": 8105
+    },
+    {
+      "epoch": 1.9293746653180222,
+      "grad_norm": 0.36621879141961106,
+      "learning_rate": 3.2739308109078215e-08,
+      "loss": 0.3791,
+      "step": 8106
+    },
+    {
+      "epoch": 1.929612661391087,
+      "grad_norm": 0.37842728624086314,
+      "learning_rate": 3.251945143543933e-08,
+      "loss": 0.2516,
+      "step": 8107
+    },
+    {
+      "epoch": 1.9298506574641519,
+      "grad_norm": 0.4177544855172531,
+      "learning_rate": 3.230033305229974e-08,
+      "loss": 0.3046,
+      "step": 8108
+    },
+    {
+      "epoch": 1.9300886535372166,
+      "grad_norm": 0.37909632728985976,
+      "learning_rate": 3.2081952992225605e-08,
+      "loss": 0.374,
+      "step": 8109
+    },
+    {
+      "epoch": 1.9303266496102816,
+      "grad_norm": 0.4361384086133121,
+      "learning_rate": 3.186431128767375e-08,
+      "loss": 0.3021,
+      "step": 8110
+    },
+    {
+      "epoch": 1.9305646456833463,
+      "grad_norm": 0.370480970702569,
+      "learning_rate": 3.164740797099053e-08,
+      "loss": 0.2905,
+      "step": 8111
+    },
+    {
+      "epoch": 1.930802641756411,
+      "grad_norm": 0.41765762330060385,
+      "learning_rate": 3.143124307441403e-08,
+      "loss": 0.3399,
+      "step": 8112
+    },
+    {
+      "epoch": 1.9310406378294758,
+      "grad_norm": 0.36824830434988537,
+      "learning_rate": 3.121581663007134e-08,
+      "loss": 0.3557,
+      "step": 8113
+    },
+    {
+      "epoch": 1.9312786339025405,
+      "grad_norm": 0.3842531168711303,
+      "learning_rate": 3.100112866997962e-08,
+      "loss": 0.2953,
+      "step": 8114
+    },
+    {
+      "epoch": 1.9315166299756052,
+      "grad_norm": 0.3817089022291876,
+      "learning_rate": 3.078717922604779e-08,
+      "loss": 0.3135,
+      "step": 8115
+    },
+    {
+      "epoch": 1.9317546260486702,
+      "grad_norm": 0.3710256940967274,
+      "learning_rate": 3.05739683300732e-08,
+      "loss": 0.3639,
+      "step": 8116
+    },
+    {
+      "epoch": 1.931992622121735,
+      "grad_norm": 0.39056691671635796,
+      "learning_rate": 3.036149601374494e-08,
+      "loss": 0.3277,
+      "step": 8117
+    },
+    {
+      "epoch": 1.9322306181948,
+      "grad_norm": 0.48173619690845865,
+      "learning_rate": 3.0149762308641083e-08,
+      "loss": 0.2814,
+      "step": 8118
+    },
+    {
+      "epoch": 1.9324686142678646,
+      "grad_norm": 0.3992688487992293,
+      "learning_rate": 2.993876724623035e-08,
+      "loss": 0.3308,
+      "step": 8119
+    },
+    {
+      "epoch": 1.9327066103409294,
+      "grad_norm": 0.3846765879945684,
+      "learning_rate": 2.972851085787265e-08,
+      "loss": 0.3655,
+      "step": 8120
+    },
+    {
+      "epoch": 1.9329446064139941,
+      "grad_norm": 0.335030815815972,
+      "learning_rate": 2.9518993174816323e-08,
+      "loss": 0.2838,
+      "step": 8121
+    },
+    {
+      "epoch": 1.9331826024870589,
+      "grad_norm": 0.39220164580944505,
+      "learning_rate": 2.9310214228202016e-08,
+      "loss": 0.2987,
+      "step": 8122
+    },
+    {
+      "epoch": 1.9334205985601236,
+      "grad_norm": 0.38126990911469083,
+      "learning_rate": 2.9102174049058796e-08,
+      "loss": 0.3555,
+      "step": 8123
+    },
+    {
+      "epoch": 1.9336585946331886,
+      "grad_norm": 0.37534376126487273,
+      "learning_rate": 2.8894872668305816e-08,
+      "loss": 0.3167,
+      "step": 8124
+    },
+    {
+      "epoch": 1.9338965907062533,
+      "grad_norm": 0.3782285028819716,
+      "learning_rate": 2.8688310116754546e-08,
+      "loss": 0.2798,
+      "step": 8125
+    },
+    {
+      "epoch": 1.9341345867793183,
+      "grad_norm": 0.4017596879571187,
+      "learning_rate": 2.848248642510487e-08,
+      "loss": 0.3332,
+      "step": 8126
+    },
+    {
+      "epoch": 1.934372582852383,
+      "grad_norm": 0.385095717269012,
+      "learning_rate": 2.8277401623946764e-08,
+      "loss": 0.3786,
+      "step": 8127
+    },
+    {
+      "epoch": 1.9346105789254477,
+      "grad_norm": 0.3681786253505828,
+      "learning_rate": 2.8073055743761956e-08,
+      "loss": 0.2888,
+      "step": 8128
+    },
+    {
+      "epoch": 1.9348485749985125,
+      "grad_norm": 0.40494290707801794,
+      "learning_rate": 2.7869448814920042e-08,
+      "loss": 0.2698,
+      "step": 8129
+    },
+    {
+      "epoch": 1.9350865710715772,
+      "grad_norm": 0.39678215471443756,
+      "learning_rate": 2.766658086768237e-08,
+      "loss": 0.3654,
+      "step": 8130
+    },
+    {
+      "epoch": 1.935324567144642,
+      "grad_norm": 0.38096391433914467,
+      "learning_rate": 2.746445193220093e-08,
+      "loss": 0.3331,
+      "step": 8131
+    },
+    {
+      "epoch": 1.935562563217707,
+      "grad_norm": 0.4207033973400047,
+      "learning_rate": 2.726306203851614e-08,
+      "loss": 0.291,
+      "step": 8132
+    },
+    {
+      "epoch": 1.9358005592907717,
+      "grad_norm": 0.35674390767561825,
+      "learning_rate": 2.706241121656017e-08,
+      "loss": 0.2914,
+      "step": 8133
+    },
+    {
+      "epoch": 1.9360385553638366,
+      "grad_norm": 0.37966258077198645,
+      "learning_rate": 2.6862499496154713e-08,
+      "loss": 0.4079,
+      "step": 8134
+    },
+    {
+      "epoch": 1.9362765514369014,
+      "grad_norm": 0.366878622762437,
+      "learning_rate": 2.6663326907010457e-08,
+      "loss": 0.2792,
+      "step": 8135
+    },
+    {
+      "epoch": 1.936514547509966,
+      "grad_norm": 0.3694716880277891,
+      "learning_rate": 2.6464893478730936e-08,
+      "loss": 0.2941,
+      "step": 8136
+    },
+    {
+      "epoch": 1.9367525435830308,
+      "grad_norm": 0.36316133821050267,
+      "learning_rate": 2.6267199240807563e-08,
+      "loss": 0.3586,
+      "step": 8137
+    },
+    {
+      "epoch": 1.9369905396560956,
+      "grad_norm": 0.37714358252792873,
+      "learning_rate": 2.6070244222622387e-08,
+      "loss": 0.3521,
+      "step": 8138
+    },
+    {
+      "epoch": 1.9372285357291603,
+      "grad_norm": 0.3846436958910601,
+      "learning_rate": 2.5874028453448106e-08,
+      "loss": 0.2677,
+      "step": 8139
+    },
+    {
+      "epoch": 1.9374665318022253,
+      "grad_norm": 0.3714862068184088,
+      "learning_rate": 2.567855196244695e-08,
+      "loss": 0.2874,
+      "step": 8140
+    },
+    {
+      "epoch": 1.93770452787529,
+      "grad_norm": 0.3857626420570937,
+      "learning_rate": 2.548381477867179e-08,
+      "loss": 0.3467,
+      "step": 8141
+    },
+    {
+      "epoch": 1.937942523948355,
+      "grad_norm": 0.3698968880808057,
+      "learning_rate": 2.528981693106558e-08,
+      "loss": 0.3375,
+      "step": 8142
+    },
+    {
+      "epoch": 1.9381805200214197,
+      "grad_norm": 0.3576794108662344,
+      "learning_rate": 2.5096558448460817e-08,
+      "loss": 0.2723,
+      "step": 8143
+    },
+    {
+      "epoch": 1.9384185160944845,
+      "grad_norm": 0.3902217274757619,
+      "learning_rate": 2.4904039359580635e-08,
+      "loss": 0.3334,
+      "step": 8144
+    },
+    {
+      "epoch": 1.9386565121675492,
+      "grad_norm": 0.3916377148051716,
+      "learning_rate": 2.4712259693038254e-08,
+      "loss": 0.3588,
+      "step": 8145
+    },
+    {
+      "epoch": 1.938894508240614,
+      "grad_norm": 0.3658288720995766,
+      "learning_rate": 2.452121947733699e-08,
+      "loss": 0.2968,
+      "step": 8146
+    },
+    {
+      "epoch": 1.9391325043136787,
+      "grad_norm": 0.4357430764134552,
+      "learning_rate": 2.4330918740869125e-08,
+      "loss": 0.314,
+      "step": 8147
+    },
+    {
+      "epoch": 1.9393705003867436,
+      "grad_norm": 0.40149441715579426,
+      "learning_rate": 2.414135751191926e-08,
+      "loss": 0.374,
+      "step": 8148
+    },
+    {
+      "epoch": 1.9396084964598084,
+      "grad_norm": 0.3611645193048641,
+      "learning_rate": 2.395253581866097e-08,
+      "loss": 0.3328,
+      "step": 8149
+    },
+    {
+      "epoch": 1.9398464925328733,
+      "grad_norm": 0.39467006084967055,
+      "learning_rate": 2.3764453689156808e-08,
+      "loss": 0.2535,
+      "step": 8150
+    },
+    {
+      "epoch": 1.940084488605938,
+      "grad_norm": 0.4124421719142737,
+      "learning_rate": 2.3577111151361078e-08,
+      "loss": 0.3229,
+      "step": 8151
+    },
+    {
+      "epoch": 1.9403224846790028,
+      "grad_norm": 0.3578248332226124,
+      "learning_rate": 2.3390508233117615e-08,
+      "loss": 0.4014,
+      "step": 8152
+    },
+    {
+      "epoch": 1.9405604807520676,
+      "grad_norm": 0.35974756105539724,
+      "learning_rate": 2.3204644962159793e-08,
+      "loss": 0.2963,
+      "step": 8153
+    },
+    {
+      "epoch": 1.9407984768251323,
+      "grad_norm": 0.3930501887956103,
+      "learning_rate": 2.301952136611163e-08,
+      "loss": 0.255,
+      "step": 8154
+    },
+    {
+      "epoch": 1.941036472898197,
+      "grad_norm": 0.39726170086743057,
+      "learning_rate": 2.2835137472487223e-08,
+      "loss": 0.3411,
+      "step": 8155
+    },
+    {
+      "epoch": 1.941274468971262,
+      "grad_norm": 0.36986727713553336,
+      "learning_rate": 2.2651493308690765e-08,
+      "loss": 0.3355,
+      "step": 8156
+    },
+    {
+      "epoch": 1.9415124650443267,
+      "grad_norm": 0.35051444026980566,
+      "learning_rate": 2.2468588902015975e-08,
+      "loss": 0.2846,
+      "step": 8157
+    },
+    {
+      "epoch": 1.9417504611173917,
+      "grad_norm": 0.37221414208027037,
+      "learning_rate": 2.2286424279646668e-08,
+      "loss": 0.3196,
+      "step": 8158
+    },
+    {
+      "epoch": 1.9419884571904564,
+      "grad_norm": 0.38381167717462056,
+      "learning_rate": 2.2104999468657852e-08,
+      "loss": 0.3715,
+      "step": 8159
+    },
+    {
+      "epoch": 1.9422264532635212,
+      "grad_norm": 0.3787505872883276,
+      "learning_rate": 2.192431449601351e-08,
+      "loss": 0.301,
+      "step": 8160
+    },
+    {
+      "epoch": 1.942464449336586,
+      "grad_norm": 0.36125888810575163,
+      "learning_rate": 2.1744369388567167e-08,
+      "loss": 0.2484,
+      "step": 8161
+    },
+    {
+      "epoch": 1.9427024454096506,
+      "grad_norm": 0.4252944882802151,
+      "learning_rate": 2.1565164173063536e-08,
+      "loss": 0.343,
+      "step": 8162
+    },
+    {
+      "epoch": 1.9429404414827154,
+      "grad_norm": 0.38454550235674245,
+      "learning_rate": 2.1386698876137424e-08,
+      "loss": 0.3642,
+      "step": 8163
+    },
+    {
+      "epoch": 1.9431784375557803,
+      "grad_norm": 0.3722701639410283,
+      "learning_rate": 2.1208973524312616e-08,
+      "loss": 0.3339,
+      "step": 8164
+    },
+    {
+      "epoch": 1.943416433628845,
+      "grad_norm": 0.39979583818697223,
+      "learning_rate": 2.103198814400409e-08,
+      "loss": 0.2935,
+      "step": 8165
+    },
+    {
+      "epoch": 1.94365442970191,
+      "grad_norm": 0.3753339389528479,
+      "learning_rate": 2.0855742761515808e-08,
+      "loss": 0.3629,
+      "step": 8166
+    },
+    {
+      "epoch": 1.9438924257749748,
+      "grad_norm": 0.36576216964926017,
+      "learning_rate": 2.0680237403041815e-08,
+      "loss": 0.3144,
+      "step": 8167
+    },
+    {
+      "epoch": 1.9441304218480395,
+      "grad_norm": 0.40044108633786357,
+      "learning_rate": 2.0505472094667356e-08,
+      "loss": 0.3043,
+      "step": 8168
+    },
+    {
+      "epoch": 1.9443684179211043,
+      "grad_norm": 0.39121889961558187,
+      "learning_rate": 2.0331446862366098e-08,
+      "loss": 0.3215,
+      "step": 8169
+    },
+    {
+      "epoch": 1.944606413994169,
+      "grad_norm": 0.3966140516297892,
+      "learning_rate": 2.0158161732003467e-08,
+      "loss": 0.3878,
+      "step": 8170
+    },
+    {
+      "epoch": 1.9448444100672337,
+      "grad_norm": 0.36014975036534674,
+      "learning_rate": 1.9985616729332747e-08,
+      "loss": 0.2826,
+      "step": 8171
+    },
+    {
+      "epoch": 1.9450824061402987,
+      "grad_norm": 0.3712453532736723,
+      "learning_rate": 1.9813811879999533e-08,
+      "loss": 0.2923,
+      "step": 8172
+    },
+    {
+      "epoch": 1.9453204022133634,
+      "grad_norm": 0.4017756063096361,
+      "learning_rate": 1.9642747209537295e-08,
+      "loss": 0.338,
+      "step": 8173
+    },
+    {
+      "epoch": 1.9455583982864284,
+      "grad_norm": 0.37378197959060855,
+      "learning_rate": 1.9472422743371245e-08,
+      "loss": 0.3247,
+      "step": 8174
+    },
+    {
+      "epoch": 1.9457963943594931,
+      "grad_norm": 0.4124391596209017,
+      "learning_rate": 1.9302838506815026e-08,
+      "loss": 0.2884,
+      "step": 8175
+    },
+    {
+      "epoch": 1.9460343904325579,
+      "grad_norm": 0.4033106524212199,
+      "learning_rate": 1.913399452507403e-08,
+      "loss": 0.3301,
+      "step": 8176
+    },
+    {
+      "epoch": 1.9462723865056226,
+      "grad_norm": 0.3966965084126736,
+      "learning_rate": 1.8965890823242072e-08,
+      "loss": 0.3688,
+      "step": 8177
+    },
+    {
+      "epoch": 1.9465103825786874,
+      "grad_norm": 0.33755930518081106,
+      "learning_rate": 1.879852742630306e-08,
+      "loss": 0.2721,
+      "step": 8178
+    },
+    {
+      "epoch": 1.946748378651752,
+      "grad_norm": 0.36588329404854725,
+      "learning_rate": 1.8631904359132646e-08,
+      "loss": 0.2917,
+      "step": 8179
+    },
+    {
+      "epoch": 1.946986374724817,
+      "grad_norm": 0.3587494911753443,
+      "learning_rate": 1.8466021646493802e-08,
+      "loss": 0.343,
+      "step": 8180
+    },
+    {
+      "epoch": 1.9472243707978818,
+      "grad_norm": 0.38139976856694574,
+      "learning_rate": 1.830087931304181e-08,
+      "loss": 0.3703,
+      "step": 8181
+    },
+    {
+      "epoch": 1.9474623668709468,
+      "grad_norm": 0.4031204694527784,
+      "learning_rate": 1.8136477383319805e-08,
+      "loss": 0.2817,
+      "step": 8182
+    },
+    {
+      "epoch": 1.9477003629440115,
+      "grad_norm": 0.3888962919598794,
+      "learning_rate": 1.797281588176325e-08,
+      "loss": 0.299,
+      "step": 8183
+    },
+    {
+      "epoch": 1.9479383590170762,
+      "grad_norm": 0.4082574436849942,
+      "learning_rate": 1.7809894832695463e-08,
+      "loss": 0.3874,
+      "step": 8184
+    },
+    {
+      "epoch": 1.948176355090141,
+      "grad_norm": 0.3623742069248171,
+      "learning_rate": 1.7647714260330407e-08,
+      "loss": 0.3092,
+      "step": 8185
+    },
+    {
+      "epoch": 1.9484143511632057,
+      "grad_norm": 0.39366833031118265,
+      "learning_rate": 1.748627418877269e-08,
+      "loss": 0.2897,
+      "step": 8186
+    },
+    {
+      "epoch": 1.9486523472362705,
+      "grad_norm": 0.3874663652919054,
+      "learning_rate": 1.7325574642016453e-08,
+      "loss": 0.3632,
+      "step": 8187
+    },
+    {
+      "epoch": 1.9488903433093354,
+      "grad_norm": 0.4022386962255987,
+      "learning_rate": 1.716561564394481e-08,
+      "loss": 0.3729,
+      "step": 8188
+    },
+    {
+      "epoch": 1.9491283393824002,
+      "grad_norm": 0.3760810663300212,
+      "learning_rate": 1.7006397218332084e-08,
+      "loss": 0.2598,
+      "step": 8189
+    },
+    {
+      "epoch": 1.9493663354554651,
+      "grad_norm": 0.3838950169229048,
+      "learning_rate": 1.6847919388842115e-08,
+      "loss": 0.3023,
+      "step": 8190
+    },
+    {
+      "epoch": 1.9496043315285299,
+      "grad_norm": 0.3762323973720483,
+      "learning_rate": 1.6690182179028845e-08,
+      "loss": 0.358,
+      "step": 8191
+    },
+    {
+      "epoch": 1.9498423276015946,
+      "grad_norm": 0.40170151586759195,
+      "learning_rate": 1.6533185612335188e-08,
+      "loss": 0.2977,
+      "step": 8192
+    },
+    {
+      "epoch": 1.9500803236746593,
+      "grad_norm": 0.38021345018959474,
+      "learning_rate": 1.6376929712095813e-08,
+      "loss": 0.2929,
+      "step": 8193
+    },
+    {
+      "epoch": 1.950318319747724,
+      "grad_norm": 0.3951132465369512,
+      "learning_rate": 1.6221414501532694e-08,
+      "loss": 0.3475,
+      "step": 8194
+    },
+    {
+      "epoch": 1.9505563158207888,
+      "grad_norm": 0.38568162349813606,
+      "learning_rate": 1.606664000376068e-08,
+      "loss": 0.3766,
+      "step": 8195
+    },
+    {
+      "epoch": 1.9507943118938538,
+      "grad_norm": 0.3780659119974086,
+      "learning_rate": 1.591260624178248e-08,
+      "loss": 0.2771,
+      "step": 8196
+    },
+    {
+      "epoch": 1.9510323079669185,
+      "grad_norm": 0.3958482547520693,
+      "learning_rate": 1.5759313238491447e-08,
+      "loss": 0.2964,
+      "step": 8197
+    },
+    {
+      "epoch": 1.9512703040399835,
+      "grad_norm": 0.39144688735676086,
+      "learning_rate": 1.5606761016670467e-08,
+      "loss": 0.3604,
+      "step": 8198
+    },
+    {
+      "epoch": 1.9515083001130482,
+      "grad_norm": 0.4028520449978215,
+      "learning_rate": 1.5454949598993075e-08,
+      "loss": 0.3177,
+      "step": 8199
+    },
+    {
+      "epoch": 1.951746296186113,
+      "grad_norm": 0.3749353165874801,
+      "learning_rate": 1.5303879008021773e-08,
+      "loss": 0.2832,
+      "step": 8200
+    },
+    {
+      "epoch": 1.9519842922591777,
+      "grad_norm": 0.39758890288042986,
+      "learning_rate": 1.5153549266209154e-08,
+      "loss": 0.3128,
+      "step": 8201
+    },
+    {
+      "epoch": 1.9522222883322424,
+      "grad_norm": 0.3508156781517721,
+      "learning_rate": 1.5003960395898465e-08,
+      "loss": 0.4,
+      "step": 8202
+    },
+    {
+      "epoch": 1.9524602844053072,
+      "grad_norm": 0.35738164902463865,
+      "learning_rate": 1.4855112419321916e-08,
+      "loss": 0.3172,
+      "step": 8203
+    },
+    {
+      "epoch": 1.9526982804783721,
+      "grad_norm": 0.3828648741919708,
+      "learning_rate": 1.4707005358602367e-08,
+      "loss": 0.2917,
+      "step": 8204
+    },
+    {
+      "epoch": 1.9529362765514369,
+      "grad_norm": 0.4386268966748521,
+      "learning_rate": 1.4559639235751654e-08,
+      "loss": 0.3149,
+      "step": 8205
+    },
+    {
+      "epoch": 1.9531742726245018,
+      "grad_norm": 0.3676937839979994,
+      "learning_rate": 1.4413014072672816e-08,
+      "loss": 0.3793,
+      "step": 8206
+    },
+    {
+      "epoch": 1.9534122686975666,
+      "grad_norm": 0.3692252990914482,
+      "learning_rate": 1.426712989115786e-08,
+      "loss": 0.2811,
+      "step": 8207
+    },
+    {
+      "epoch": 1.9536502647706313,
+      "grad_norm": 0.37336901880476925,
+      "learning_rate": 1.4121986712888336e-08,
+      "loss": 0.316,
+      "step": 8208
+    },
+    {
+      "epoch": 1.953888260843696,
+      "grad_norm": 0.40798571701364533,
+      "learning_rate": 1.3977584559435874e-08,
+      "loss": 0.374,
+      "step": 8209
+    },
+    {
+      "epoch": 1.9541262569167608,
+      "grad_norm": 0.36868651161847604,
+      "learning_rate": 1.3833923452262754e-08,
+      "loss": 0.3055,
+      "step": 8210
+    },
+    {
+      "epoch": 1.9543642529898255,
+      "grad_norm": 0.38007255786831645,
+      "learning_rate": 1.3691003412720783e-08,
+      "loss": 0.2946,
+      "step": 8211
+    },
+    {
+      "epoch": 1.9546022490628905,
+      "grad_norm": 0.35949632759077443,
+      "learning_rate": 1.3548824462050747e-08,
+      "loss": 0.3456,
+      "step": 8212
+    },
+    {
+      "epoch": 1.9548402451359552,
+      "grad_norm": 0.37479143142903726,
+      "learning_rate": 1.3407386621384078e-08,
+      "loss": 0.4028,
+      "step": 8213
+    },
+    {
+      "epoch": 1.9550782412090202,
+      "grad_norm": 0.3666312123014584,
+      "learning_rate": 1.3266689911742291e-08,
+      "loss": 0.2829,
+      "step": 8214
+    },
+    {
+      "epoch": 1.955316237282085,
+      "grad_norm": 0.4020518839752495,
+      "learning_rate": 1.312673435403644e-08,
+      "loss": 0.3137,
+      "step": 8215
+    },
+    {
+      "epoch": 1.9555542333551497,
+      "grad_norm": 0.4198314847703424,
+      "learning_rate": 1.2987519969067109e-08,
+      "loss": 0.3647,
+      "step": 8216
+    },
+    {
+      "epoch": 1.9557922294282144,
+      "grad_norm": 0.37449752005583575,
+      "learning_rate": 1.284904677752441e-08,
+      "loss": 0.3045,
+      "step": 8217
+    },
+    {
+      "epoch": 1.9560302255012791,
+      "grad_norm": 0.3788052759823918,
+      "learning_rate": 1.2711314799990216e-08,
+      "loss": 0.2877,
+      "step": 8218
+    },
+    {
+      "epoch": 1.9562682215743439,
+      "grad_norm": 0.3865156770622441,
+      "learning_rate": 1.2574324056934262e-08,
+      "loss": 0.3561,
+      "step": 8219
+    },
+    {
+      "epoch": 1.9565062176474088,
+      "grad_norm": 0.3781434327446371,
+      "learning_rate": 1.2438074568716374e-08,
+      "loss": 0.3798,
+      "step": 8220
+    },
+    {
+      "epoch": 1.9567442137204736,
+      "grad_norm": 0.36790890617104577,
+      "learning_rate": 1.230256635558702e-08,
+      "loss": 0.2838,
+      "step": 8221
+    },
+    {
+      "epoch": 1.9569822097935385,
+      "grad_norm": 0.39260234101874714,
+      "learning_rate": 1.21677994376862e-08,
+      "loss": 0.3047,
+      "step": 8222
+    },
+    {
+      "epoch": 1.9572202058666033,
+      "grad_norm": 0.3574524348159013,
+      "learning_rate": 1.2033773835042894e-08,
+      "loss": 0.3746,
+      "step": 8223
+    },
+    {
+      "epoch": 1.957458201939668,
+      "grad_norm": 0.3612325904669084,
+      "learning_rate": 1.1900489567577277e-08,
+      "loss": 0.3464,
+      "step": 8224
+    },
+    {
+      "epoch": 1.9576961980127328,
+      "grad_norm": 0.38077468030121714,
+      "learning_rate": 1.1767946655099061e-08,
+      "loss": 0.2666,
+      "step": 8225
+    },
+    {
+      "epoch": 1.9579341940857975,
+      "grad_norm": 0.368563482866289,
+      "learning_rate": 1.1636145117306374e-08,
+      "loss": 0.3406,
+      "step": 8226
+    },
+    {
+      "epoch": 1.9581721901588622,
+      "grad_norm": 0.42477667717805123,
+      "learning_rate": 1.1505084973789105e-08,
+      "loss": 0.3935,
+      "step": 8227
+    },
+    {
+      "epoch": 1.9584101862319272,
+      "grad_norm": 0.3438577540547306,
+      "learning_rate": 1.1374766244025003e-08,
+      "loss": 0.2929,
+      "step": 8228
+    },
+    {
+      "epoch": 1.958648182304992,
+      "grad_norm": 0.3782776548245008,
+      "learning_rate": 1.1245188947384133e-08,
+      "loss": 0.2818,
+      "step": 8229
+    },
+    {
+      "epoch": 1.958886178378057,
+      "grad_norm": 0.36863077536612854,
+      "learning_rate": 1.1116353103123312e-08,
+      "loss": 0.3458,
+      "step": 8230
+    },
+    {
+      "epoch": 1.9591241744511216,
+      "grad_norm": 0.3659543529085472,
+      "learning_rate": 1.0988258730391665e-08,
+      "loss": 0.3483,
+      "step": 8231
+    },
+    {
+      "epoch": 1.9593621705241864,
+      "grad_norm": 0.38260543409474423,
+      "learning_rate": 1.0860905848227298e-08,
+      "loss": 0.2871,
+      "step": 8232
+    },
+    {
+      "epoch": 1.9596001665972511,
+      "grad_norm": 0.41483183809961477,
+      "learning_rate": 1.0734294475557294e-08,
+      "loss": 0.3166,
+      "step": 8233
+    },
+    {
+      "epoch": 1.9598381626703159,
+      "grad_norm": 0.39222204343126627,
+      "learning_rate": 1.0608424631199376e-08,
+      "loss": 0.3949,
+      "step": 8234
+    },
+    {
+      "epoch": 1.9600761587433806,
+      "grad_norm": 0.3711141699280681,
+      "learning_rate": 1.0483296333861914e-08,
+      "loss": 0.3006,
+      "step": 8235
+    },
+    {
+      "epoch": 1.9603141548164456,
+      "grad_norm": 0.3763503542360473,
+      "learning_rate": 1.0358909602140588e-08,
+      "loss": 0.3047,
+      "step": 8236
+    },
+    {
+      "epoch": 1.9605521508895103,
+      "grad_norm": 0.3969121772442897,
+      "learning_rate": 1.0235264454523385e-08,
+      "loss": 0.3416,
+      "step": 8237
+    },
+    {
+      "epoch": 1.9607901469625753,
+      "grad_norm": 0.3816122834946716,
+      "learning_rate": 1.011236090938672e-08,
+      "loss": 0.3718,
+      "step": 8238
+    },
+    {
+      "epoch": 1.96102814303564,
+      "grad_norm": 0.38396544116333275,
+      "learning_rate": 9.99019898499709e-09,
+      "loss": 0.2717,
+      "step": 8239
+    },
+    {
+      "epoch": 1.9612661391087047,
+      "grad_norm": 0.44666178093194486,
+      "learning_rate": 9.868778699511083e-09,
+      "loss": 0.3074,
+      "step": 8240
+    },
+    {
+      "epoch": 1.9615041351817695,
+      "grad_norm": 0.4143851166339734,
+      "learning_rate": 9.748100070974265e-09,
+      "loss": 0.3759,
+      "step": 8241
+    },
+    {
+      "epoch": 1.9617421312548342,
+      "grad_norm": 0.3921730498280631,
+      "learning_rate": 9.628163117322286e-09,
+      "loss": 0.3331,
+      "step": 8242
+    },
+    {
+      "epoch": 1.961980127327899,
+      "grad_norm": 0.3726163829999817,
+      "learning_rate": 9.508967856381445e-09,
+      "loss": 0.2988,
+      "step": 8243
+    },
+    {
+      "epoch": 1.962218123400964,
+      "grad_norm": 0.39354313502014604,
+      "learning_rate": 9.390514305867015e-09,
+      "loss": 0.3275,
+      "step": 8244
+    },
+    {
+      "epoch": 1.9624561194740286,
+      "grad_norm": 0.37253432454149743,
+      "learning_rate": 9.272802483383248e-09,
+      "loss": 0.3857,
+      "step": 8245
+    },
+    {
+      "epoch": 1.9626941155470936,
+      "grad_norm": 0.3981252148153215,
+      "learning_rate": 9.155832406426147e-09,
+      "loss": 0.265,
+      "step": 8246
+    },
+    {
+      "epoch": 1.9629321116201583,
+      "grad_norm": 0.39768413263890057,
+      "learning_rate": 9.039604092379583e-09,
+      "loss": 0.2868,
+      "step": 8247
+    },
+    {
+      "epoch": 1.963170107693223,
+      "grad_norm": 0.4075926684645318,
+      "learning_rate": 8.92411755851863e-09,
+      "loss": 0.3498,
+      "step": 8248
+    },
+    {
+      "epoch": 1.9634081037662878,
+      "grad_norm": 0.3715207251676553,
+      "learning_rate": 8.809372822006779e-09,
+      "loss": 0.3479,
+      "step": 8249
+    },
+    {
+      "epoch": 1.9636460998393526,
+      "grad_norm": 0.3869938419383704,
+      "learning_rate": 8.69536989989872e-09,
+      "loss": 0.2767,
+      "step": 8250
+    },
+    {
+      "epoch": 1.9638840959124173,
+      "grad_norm": 0.37358902937459904,
+      "learning_rate": 8.582108809137013e-09,
+      "loss": 0.3084,
+      "step": 8251
+    },
+    {
+      "epoch": 1.9641220919854823,
+      "grad_norm": 0.4092803779218533,
+      "learning_rate": 8.469589566555968e-09,
+      "loss": 0.3679,
+      "step": 8252
+    },
+    {
+      "epoch": 1.964360088058547,
+      "grad_norm": 0.36047460363307426,
+      "learning_rate": 8.357812188878323e-09,
+      "loss": 0.2806,
+      "step": 8253
+    },
+    {
+      "epoch": 1.964598084131612,
+      "grad_norm": 0.3678747427954382,
+      "learning_rate": 8.246776692716896e-09,
+      "loss": 0.2568,
+      "step": 8254
+    },
+    {
+      "epoch": 1.9648360802046767,
+      "grad_norm": 0.40832629901247686,
+      "learning_rate": 8.1364830945746e-09,
+      "loss": 0.333,
+      "step": 8255
+    },
+    {
+      "epoch": 1.9650740762777414,
+      "grad_norm": 0.37115745740832373,
+      "learning_rate": 8.026931410843874e-09,
+      "loss": 0.3326,
+      "step": 8256
+    },
+    {
+      "epoch": 1.9653120723508062,
+      "grad_norm": 0.37423048974217404,
+      "learning_rate": 7.918121657806699e-09,
+      "loss": 0.2807,
+      "step": 8257
+    },
+    {
+      "epoch": 1.965550068423871,
+      "grad_norm": 0.3864398760686447,
+      "learning_rate": 7.81005385163458e-09,
+      "loss": 0.3276,
+      "step": 8258
+    },
+    {
+      "epoch": 1.9657880644969357,
+      "grad_norm": 0.388715422162384,
+      "learning_rate": 7.702728008389116e-09,
+      "loss": 0.3942,
+      "step": 8259
+    },
+    {
+      "epoch": 1.9660260605700006,
+      "grad_norm": 0.3429967987431223,
+      "learning_rate": 7.596144144021988e-09,
+      "loss": 0.2696,
+      "step": 8260
+    },
+    {
+      "epoch": 1.9662640566430654,
+      "grad_norm": 0.3884832865704859,
+      "learning_rate": 7.490302274373862e-09,
+      "loss": 0.2719,
+      "step": 8261
+    },
+    {
+      "epoch": 1.9665020527161303,
+      "grad_norm": 0.3804734222913664,
+      "learning_rate": 7.385202415175485e-09,
+      "loss": 0.3317,
+      "step": 8262
+    },
+    {
+      "epoch": 1.966740048789195,
+      "grad_norm": 0.36816873312551346,
+      "learning_rate": 7.280844582047142e-09,
+      "loss": 0.3705,
+      "step": 8263
+    },
+    {
+      "epoch": 1.9669780448622598,
+      "grad_norm": 0.36983723483013164,
+      "learning_rate": 7.1772287904997575e-09,
+      "loss": 0.3021,
+      "step": 8264
+    },
+    {
+      "epoch": 1.9672160409353245,
+      "grad_norm": 0.38603048071401885,
+      "learning_rate": 7.07435505593268e-09,
+      "loss": 0.2876,
+      "step": 8265
+    },
+    {
+      "epoch": 1.9674540370083893,
+      "grad_norm": 0.4003003944120439,
+      "learning_rate": 6.972223393634792e-09,
+      "loss": 0.3849,
+      "step": 8266
+    },
+    {
+      "epoch": 1.967692033081454,
+      "grad_norm": 0.3602611770065517,
+      "learning_rate": 6.870833818786727e-09,
+      "loss": 0.2768,
+      "step": 8267
+    },
+    {
+      "epoch": 1.967930029154519,
+      "grad_norm": 0.3884374384546158,
+      "learning_rate": 6.770186346456431e-09,
+      "loss": 0.2914,
+      "step": 8268
+    },
+    {
+      "epoch": 1.9681680252275837,
+      "grad_norm": 0.3744232480095091,
+      "learning_rate": 6.670280991603606e-09,
+      "loss": 0.3281,
+      "step": 8269
+    },
+    {
+      "epoch": 1.9684060213006487,
+      "grad_norm": 0.42375613579156163,
+      "learning_rate": 6.571117769075264e-09,
+      "loss": 0.3981,
+      "step": 8270
+    },
+    {
+      "epoch": 1.9686440173737134,
+      "grad_norm": 0.35315666764990467,
+      "learning_rate": 6.472696693610725e-09,
+      "loss": 0.2874,
+      "step": 8271
+    },
+    {
+      "epoch": 1.9688820134467782,
+      "grad_norm": 0.3452396601565405,
+      "learning_rate": 6.375017779837178e-09,
+      "loss": 0.2947,
+      "step": 8272
+    },
+    {
+      "epoch": 1.969120009519843,
+      "grad_norm": 0.3557979924222308,
+      "learning_rate": 6.278081042272455e-09,
+      "loss": 0.3548,
+      "step": 8273
+    },
+    {
+      "epoch": 1.9693580055929076,
+      "grad_norm": 0.3864839080975044,
+      "learning_rate": 6.181886495323364e-09,
+      "loss": 0.3309,
+      "step": 8274
+    },
+    {
+      "epoch": 1.9695960016659724,
+      "grad_norm": 0.4527567996253969,
+      "learning_rate": 6.086434153287357e-09,
+      "loss": 0.2887,
+      "step": 8275
+    },
+    {
+      "epoch": 1.9698339977390373,
+      "grad_norm": 0.4701596918796658,
+      "learning_rate": 5.991724030350865e-09,
+      "loss": 0.2977,
+      "step": 8276
+    },
+    {
+      "epoch": 1.970071993812102,
+      "grad_norm": 0.40077106587480205,
+      "learning_rate": 5.8977561405898496e-09,
+      "loss": 0.37,
+      "step": 8277
+    },
+    {
+      "epoch": 1.970309989885167,
+      "grad_norm": 0.36798667561511195,
+      "learning_rate": 5.804530497970362e-09,
+      "loss": 0.3101,
+      "step": 8278
+    },
+    {
+      "epoch": 1.9705479859582318,
+      "grad_norm": 0.4143172864878241,
+      "learning_rate": 5.712047116347985e-09,
+      "loss": 0.2745,
+      "step": 8279
+    },
+    {
+      "epoch": 1.9707859820312965,
+      "grad_norm": 0.3852482915444085,
+      "learning_rate": 5.620306009467835e-09,
+      "loss": 0.3629,
+      "step": 8280
+    },
+    {
+      "epoch": 1.9710239781043613,
+      "grad_norm": 0.37496502608000093,
+      "learning_rate": 5.529307190965671e-09,
+      "loss": 0.344,
+      "step": 8281
+    },
+    {
+      "epoch": 1.971261974177426,
+      "grad_norm": 0.3832082730157154,
+      "learning_rate": 5.439050674365676e-09,
+      "loss": 0.2708,
+      "step": 8282
+    },
+    {
+      "epoch": 1.9714999702504907,
+      "grad_norm": 0.40313323861297806,
+      "learning_rate": 5.349536473082118e-09,
+      "loss": 0.3132,
+      "step": 8283
+    },
+    {
+      "epoch": 1.9717379663235557,
+      "grad_norm": 0.3707010760106356,
+      "learning_rate": 5.260764600419354e-09,
+      "loss": 0.4018,
+      "step": 8284
+    },
+    {
+      "epoch": 1.9719759623966204,
+      "grad_norm": 0.43019076768807374,
+      "learning_rate": 5.172735069570722e-09,
+      "loss": 0.3143,
+      "step": 8285
+    },
+    {
+      "epoch": 1.9722139584696854,
+      "grad_norm": 0.46022861405119286,
+      "learning_rate": 5.0854478936190884e-09,
+      "loss": 0.2626,
+      "step": 8286
+    },
+    {
+      "epoch": 1.9724519545427501,
+      "grad_norm": 0.3912662063723855,
+      "learning_rate": 4.998903085539075e-09,
+      "loss": 0.319,
+      "step": 8287
+    },
+    {
+      "epoch": 1.9726899506158149,
+      "grad_norm": 0.3793180333217797,
+      "learning_rate": 4.913100658192061e-09,
+      "loss": 0.3523,
+      "step": 8288
+    },
+    {
+      "epoch": 1.9729279466888796,
+      "grad_norm": 0.3865661630275148,
+      "learning_rate": 4.828040624330621e-09,
+      "loss": 0.2683,
+      "step": 8289
+    },
+    {
+      "epoch": 1.9731659427619443,
+      "grad_norm": 0.37603160286322296,
+      "learning_rate": 4.743722996597422e-09,
+      "loss": 0.3024,
+      "step": 8290
+    },
+    {
+      "epoch": 1.973403938835009,
+      "grad_norm": 0.3751677087026661,
+      "learning_rate": 4.6601477875235505e-09,
+      "loss": 0.34,
+      "step": 8291
+    },
+    {
+      "epoch": 1.973641934908074,
+      "grad_norm": 0.376225703087134,
+      "learning_rate": 4.577315009530181e-09,
+      "loss": 0.3369,
+      "step": 8292
+    },
+    {
+      "epoch": 1.9738799309811388,
+      "grad_norm": 0.3927228549545841,
+      "learning_rate": 4.495224674928578e-09,
+      "loss": 0.2813,
+      "step": 8293
+    },
+    {
+      "epoch": 1.9741179270542037,
+      "grad_norm": 0.3827330662097907,
+      "learning_rate": 4.413876795919536e-09,
+      "loss": 0.3303,
+      "step": 8294
+    },
+    {
+      "epoch": 1.9743559231272685,
+      "grad_norm": 0.3862347082915683,
+      "learning_rate": 4.333271384593385e-09,
+      "loss": 0.3676,
+      "step": 8295
+    },
+    {
+      "epoch": 1.9745939192003332,
+      "grad_norm": 0.40932286583844174,
+      "learning_rate": 4.253408452929986e-09,
+      "loss": 0.27,
+      "step": 8296
+    },
+    {
+      "epoch": 1.974831915273398,
+      "grad_norm": 0.4000770232069227,
+      "learning_rate": 4.174288012798733e-09,
+      "loss": 0.3247,
+      "step": 8297
+    },
+    {
+      "epoch": 1.9750699113464627,
+      "grad_norm": 0.36945290953037635,
+      "learning_rate": 4.095910075959108e-09,
+      "loss": 0.3289,
+      "step": 8298
+    },
+    {
+      "epoch": 1.9753079074195274,
+      "grad_norm": 0.4060885406421163,
+      "learning_rate": 4.018274654059573e-09,
+      "loss": 0.3047,
+      "step": 8299
+    },
+    {
+      "epoch": 1.9755459034925924,
+      "grad_norm": 0.4119247754609785,
+      "learning_rate": 3.941381758639784e-09,
+      "loss": 0.2876,
+      "step": 8300
+    },
+    {
+      "epoch": 1.9757838995656571,
+      "grad_norm": 0.39855318475174173,
+      "learning_rate": 3.865231401126712e-09,
+      "loss": 0.3215,
+      "step": 8301
+    },
+    {
+      "epoch": 1.976021895638722,
+      "grad_norm": 0.3834493602365561,
+      "learning_rate": 3.789823592838526e-09,
+      "loss": 0.381,
+      "step": 8302
+    },
+    {
+      "epoch": 1.9762598917117868,
+      "grad_norm": 0.39370603464488824,
+      "learning_rate": 3.7151583449834826e-09,
+      "loss": 0.2854,
+      "step": 8303
+    },
+    {
+      "epoch": 1.9764978877848516,
+      "grad_norm": 0.38092804030560223,
+      "learning_rate": 3.6412356686577056e-09,
+      "loss": 0.2866,
+      "step": 8304
+    },
+    {
+      "epoch": 1.9767358838579163,
+      "grad_norm": 0.359154440466524,
+      "learning_rate": 3.5680555748479617e-09,
+      "loss": 0.3367,
+      "step": 8305
+    },
+    {
+      "epoch": 1.976973879930981,
+      "grad_norm": 0.3866822121685151,
+      "learning_rate": 3.4956180744311063e-09,
+      "loss": 0.3345,
+      "step": 8306
+    },
+    {
+      "epoch": 1.9772118760040458,
+      "grad_norm": 0.3877651597061357,
+      "learning_rate": 3.423923178172972e-09,
+      "loss": 0.2589,
+      "step": 8307
+    },
+    {
+      "epoch": 1.9774498720771108,
+      "grad_norm": 0.3888359989652597,
+      "learning_rate": 3.3529708967294794e-09,
+      "loss": 0.3375,
+      "step": 8308
+    },
+    {
+      "epoch": 1.9776878681501755,
+      "grad_norm": 0.38304911568033706,
+      "learning_rate": 3.282761240645527e-09,
+      "loss": 0.3554,
+      "step": 8309
+    },
+    {
+      "epoch": 1.9779258642232405,
+      "grad_norm": 0.3470736673254614,
+      "learning_rate": 3.213294220355545e-09,
+      "loss": 0.3175,
+      "step": 8310
+    },
+    {
+      "epoch": 1.9781638602963052,
+      "grad_norm": 0.377926105048321,
+      "learning_rate": 3.1445698461851638e-09,
+      "loss": 0.2487,
+      "step": 8311
+    },
+    {
+      "epoch": 1.97840185636937,
+      "grad_norm": 0.3806616330829527,
+      "learning_rate": 3.0765881283478794e-09,
+      "loss": 0.3329,
+      "step": 8312
+    },
+    {
+      "epoch": 1.9786398524424347,
+      "grad_norm": 0.3620860977402568,
+      "learning_rate": 3.0093490769472765e-09,
+      "loss": 0.3399,
+      "step": 8313
+    },
+    {
+      "epoch": 1.9788778485154994,
+      "grad_norm": 0.38013912104937325,
+      "learning_rate": 2.942852701977028e-09,
+      "loss": 0.2705,
+      "step": 8314
+    },
+    {
+      "epoch": 1.9791158445885642,
+      "grad_norm": 0.3818113771924823,
+      "learning_rate": 2.8770990133203392e-09,
+      "loss": 0.3248,
+      "step": 8315
+    },
+    {
+      "epoch": 1.9793538406616291,
+      "grad_norm": 0.4113225030445951,
+      "learning_rate": 2.8120880207493928e-09,
+      "loss": 0.3487,
+      "step": 8316
+    },
+    {
+      "epoch": 1.9795918367346939,
+      "grad_norm": 0.3703805736030292,
+      "learning_rate": 2.747819733927015e-09,
+      "loss": 0.3121,
+      "step": 8317
+    },
+    {
+      "epoch": 1.9798298328077588,
+      "grad_norm": 0.368702788372816,
+      "learning_rate": 2.6842941624044548e-09,
+      "loss": 0.2654,
+      "step": 8318
+    },
+    {
+      "epoch": 1.9800678288808236,
+      "grad_norm": 0.39480635521831586,
+      "learning_rate": 2.6215113156230487e-09,
+      "loss": 0.3279,
+      "step": 8319
+    },
+    {
+      "epoch": 1.9803058249538883,
+      "grad_norm": 0.35852264603642603,
+      "learning_rate": 2.559471202914776e-09,
+      "loss": 0.393,
+      "step": 8320
+    },
+    {
+      "epoch": 1.980543821026953,
+      "grad_norm": 0.3594864841251863,
+      "learning_rate": 2.498173833499484e-09,
+      "loss": 0.2659,
+      "step": 8321
+    },
+    {
+      "epoch": 1.9807818171000178,
+      "grad_norm": 0.40546718269932125,
+      "learning_rate": 2.4376192164882183e-09,
+      "loss": 0.3409,
+      "step": 8322
+    },
+    {
+      "epoch": 1.9810198131730825,
+      "grad_norm": 0.5532056344379126,
+      "learning_rate": 2.3778073608798914e-09,
+      "loss": 0.358,
+      "step": 8323
+    },
+    {
+      "epoch": 1.9812578092461475,
+      "grad_norm": 0.3546557183922418,
+      "learning_rate": 2.3187382755651687e-09,
+      "loss": 0.3298,
+      "step": 8324
+    },
+    {
+      "epoch": 1.9814958053192122,
+      "grad_norm": 0.37203732480489365,
+      "learning_rate": 2.2604119693220295e-09,
+      "loss": 0.2895,
+      "step": 8325
+    },
+    {
+      "epoch": 1.9817338013922772,
+      "grad_norm": 0.4172958208934213,
+      "learning_rate": 2.202828450820205e-09,
+      "loss": 0.3164,
+      "step": 8326
+    },
+    {
+      "epoch": 1.981971797465342,
+      "grad_norm": 0.3625690030083054,
+      "learning_rate": 2.1459877286172935e-09,
+      "loss": 0.3785,
+      "step": 8327
+    },
+    {
+      "epoch": 1.9822097935384067,
+      "grad_norm": 0.3603665358371541,
+      "learning_rate": 2.0898898111620935e-09,
+      "loss": 0.2816,
+      "step": 8328
+    },
+    {
+      "epoch": 1.9824477896114714,
+      "grad_norm": 0.3780300709578647,
+      "learning_rate": 2.034534706791269e-09,
+      "loss": 0.2825,
+      "step": 8329
+    },
+    {
+      "epoch": 1.9826857856845361,
+      "grad_norm": 0.37028301830691185,
+      "learning_rate": 1.979922423732128e-09,
+      "loss": 0.3403,
+      "step": 8330
+    },
+    {
+      "epoch": 1.9829237817576009,
+      "grad_norm": 0.3638771091784381,
+      "learning_rate": 1.9260529701015105e-09,
+      "loss": 0.3491,
+      "step": 8331
+    },
+    {
+      "epoch": 1.9831617778306658,
+      "grad_norm": 0.41702450685086956,
+      "learning_rate": 1.8729263539063457e-09,
+      "loss": 0.2864,
+      "step": 8332
+    },
+    {
+      "epoch": 1.9833997739037306,
+      "grad_norm": 0.37868852344723125,
+      "learning_rate": 1.820542583041429e-09,
+      "loss": 0.3317,
+      "step": 8333
+    },
+    {
+      "epoch": 1.9836377699767955,
+      "grad_norm": 0.3666826086769097,
+      "learning_rate": 1.76890166529331e-09,
+      "loss": 0.3906,
+      "step": 8334
+    },
+    {
+      "epoch": 1.9838757660498603,
+      "grad_norm": 0.35122611004045284,
+      "learning_rate": 1.718003608336405e-09,
+      "loss": 0.2839,
+      "step": 8335
+    },
+    {
+      "epoch": 1.984113762122925,
+      "grad_norm": 0.4202455413793847,
+      "learning_rate": 1.6678484197357737e-09,
+      "loss": 0.262,
+      "step": 8336
+    },
+    {
+      "epoch": 1.9843517581959897,
+      "grad_norm": 0.37863495269066555,
+      "learning_rate": 1.6184361069460085e-09,
+      "loss": 0.3527,
+      "step": 8337
+    },
+    {
+      "epoch": 1.9845897542690545,
+      "grad_norm": 0.35319699077317435,
+      "learning_rate": 1.569766677310125e-09,
+      "loss": 0.3451,
+      "step": 8338
+    },
+    {
+      "epoch": 1.9848277503421192,
+      "grad_norm": 0.34742580202122714,
+      "learning_rate": 1.521840138062336e-09,
+      "loss": 0.2781,
+      "step": 8339
+    },
+    {
+      "epoch": 1.9850657464151842,
+      "grad_norm": 0.38670204572295525,
+      "learning_rate": 1.4746564963258325e-09,
+      "loss": 0.2925,
+      "step": 8340
+    },
+    {
+      "epoch": 1.985303742488249,
+      "grad_norm": 0.3666529795353334,
+      "learning_rate": 1.4282157591122282e-09,
+      "loss": 0.3708,
+      "step": 8341
+    },
+    {
+      "epoch": 1.9855417385613139,
+      "grad_norm": 0.3696854659046601,
+      "learning_rate": 1.3825179333248895e-09,
+      "loss": 0.2828,
+      "step": 8342
+    },
+    {
+      "epoch": 1.9857797346343786,
+      "grad_norm": 0.406404518542554,
+      "learning_rate": 1.3375630257550509e-09,
+      "loss": 0.3009,
+      "step": 8343
+    },
+    {
+      "epoch": 1.9860177307074434,
+      "grad_norm": 0.385912160739807,
+      "learning_rate": 1.2933510430845898e-09,
+      "loss": 0.3693,
+      "step": 8344
+    },
+    {
+      "epoch": 1.986255726780508,
+      "grad_norm": 0.3925971908766997,
+      "learning_rate": 1.2498819918843609e-09,
+      "loss": 0.4039,
+      "step": 8345
+    },
+    {
+      "epoch": 1.9864937228535728,
+      "grad_norm": 0.3652691608712696,
+      "learning_rate": 1.2071558786141969e-09,
+      "loss": 0.273,
+      "step": 8346
+    },
+    {
+      "epoch": 1.9867317189266376,
+      "grad_norm": 0.38597466303071754,
+      "learning_rate": 1.1651727096251287e-09,
+      "loss": 0.3049,
+      "step": 8347
+    },
+    {
+      "epoch": 1.9869697149997025,
+      "grad_norm": 0.3810587229670749,
+      "learning_rate": 1.1239324911566096e-09,
+      "loss": 0.3677,
+      "step": 8348
+    },
+    {
+      "epoch": 1.9872077110727673,
+      "grad_norm": 0.40007821179401476,
+      "learning_rate": 1.083435229338181e-09,
+      "loss": 0.3497,
+      "step": 8349
+    },
+    {
+      "epoch": 1.9874457071458322,
+      "grad_norm": 0.40580202241849406,
+      "learning_rate": 1.043680930187807e-09,
+      "loss": 0.2742,
+      "step": 8350
+    },
+    {
+      "epoch": 1.987683703218897,
+      "grad_norm": 0.37923487971766323,
+      "learning_rate": 1.0046695996152046e-09,
+      "loss": 0.3465,
+      "step": 8351
+    },
+    {
+      "epoch": 1.9879216992919617,
+      "grad_norm": 0.37215554302402476,
+      "learning_rate": 9.66401243417958e-10,
+      "loss": 0.3798,
+      "step": 8352
+    },
+    {
+      "epoch": 1.9881596953650265,
+      "grad_norm": 0.39321133558205074,
+      "learning_rate": 9.288758672837406e-10,
+      "loss": 0.3067,
+      "step": 8353
+    },
+    {
+      "epoch": 1.9883976914380912,
+      "grad_norm": 0.3748563987454188,
+      "learning_rate": 8.92093476789202e-10,
+      "loss": 0.2742,
+      "step": 8354
+    },
+    {
+      "epoch": 1.988635687511156,
+      "grad_norm": 0.3891255633326672,
+      "learning_rate": 8.560540774016357e-10,
+      "loss": 0.3438,
+      "step": 8355
+    },
+    {
+      "epoch": 1.988873683584221,
+      "grad_norm": 0.3646078625631843,
+      "learning_rate": 8.207576744773127e-10,
+      "loss": 0.3308,
+      "step": 8356
+    },
+    {
+      "epoch": 1.9891116796572856,
+      "grad_norm": 0.36793355963820623,
+      "learning_rate": 7.862042732620367e-10,
+      "loss": 0.2491,
+      "step": 8357
+    },
+    {
+      "epoch": 1.9893496757303506,
+      "grad_norm": 0.3672753281223139,
+      "learning_rate": 7.523938788916996e-10,
+      "loss": 0.3369,
+      "step": 8358
+    },
+    {
+      "epoch": 1.9895876718034153,
+      "grad_norm": 0.39245638391380494,
+      "learning_rate": 7.193264963911706e-10,
+      "loss": 0.4073,
+      "step": 8359
+    },
+    {
+      "epoch": 1.98982566787648,
+      "grad_norm": 0.3865165837782479,
+      "learning_rate": 6.870021306742968e-10,
+      "loss": 0.2962,
+      "step": 8360
+    },
+    {
+      "epoch": 1.9900636639495448,
+      "grad_norm": 0.41219133327793983,
+      "learning_rate": 6.554207865466788e-10,
+      "loss": 0.302,
+      "step": 8361
+    },
+    {
+      "epoch": 1.9903016600226096,
+      "grad_norm": 0.4223513362715057,
+      "learning_rate": 6.245824687006741e-10,
+      "loss": 0.3742,
+      "step": 8362
+    },
+    {
+      "epoch": 1.9905396560956743,
+      "grad_norm": 0.3839303730562907,
+      "learning_rate": 5.944871817209486e-10,
+      "loss": 0.383,
+      "step": 8363
+    },
+    {
+      "epoch": 1.9907776521687393,
+      "grad_norm": 0.3926286842555442,
+      "learning_rate": 5.651349300794806e-10,
+      "loss": 0.292,
+      "step": 8364
+    },
+    {
+      "epoch": 1.991015648241804,
+      "grad_norm": 0.397668403394273,
+      "learning_rate": 5.365257181388917e-10,
+      "loss": 0.2986,
+      "step": 8365
+    },
+    {
+      "epoch": 1.991253644314869,
+      "grad_norm": 0.391019416533082,
+      "learning_rate": 5.086595501513358e-10,
+      "loss": 0.3498,
+      "step": 8366
+    },
+    {
+      "epoch": 1.9914916403879337,
+      "grad_norm": 0.42914431554738636,
+      "learning_rate": 4.815364302590553e-10,
+      "loss": 0.339,
+      "step": 8367
+    },
+    {
+      "epoch": 1.9917296364609984,
+      "grad_norm": 0.37889014824362105,
+      "learning_rate": 4.5515636249160446e-10,
+      "loss": 0.2723,
+      "step": 8368
+    },
+    {
+      "epoch": 1.9919676325340632,
+      "grad_norm": 0.38046794389103084,
+      "learning_rate": 4.2951935077140127e-10,
+      "loss": 0.3101,
+      "step": 8369
+    },
+    {
+      "epoch": 1.992205628607128,
+      "grad_norm": 0.40863630016035024,
+      "learning_rate": 4.04625398907621e-10,
+      "loss": 0.3595,
+      "step": 8370
+    },
+    {
+      "epoch": 1.9924436246801926,
+      "grad_norm": 0.34145231287271066,
+      "learning_rate": 3.80474510601192e-10,
+      "loss": 0.3063,
+      "step": 8371
+    },
+    {
+      "epoch": 1.9926816207532576,
+      "grad_norm": 0.37706309665477467,
+      "learning_rate": 3.5706668944035517e-10,
+      "loss": 0.3212,
+      "step": 8372
+    },
+    {
+      "epoch": 1.9929196168263223,
+      "grad_norm": 0.4166841421991216,
+      "learning_rate": 3.344019389045494e-10,
+      "loss": 0.3757,
+      "step": 8373
+    },
+    {
+      "epoch": 1.9931576128993873,
+      "grad_norm": 0.36453047466675503,
+      "learning_rate": 3.124802623627465e-10,
+      "loss": 0.318,
+      "step": 8374
+    },
+    {
+      "epoch": 1.993395608972452,
+      "grad_norm": 0.40191372694668726,
+      "learning_rate": 2.913016630723409e-10,
+      "loss": 0.2785,
+      "step": 8375
+    },
+    {
+      "epoch": 1.9936336050455168,
+      "grad_norm": 0.38151201580835575,
+      "learning_rate": 2.708661441813698e-10,
+      "loss": 0.3213,
+      "step": 8376
+    },
+    {
+      "epoch": 1.9938716011185815,
+      "grad_norm": 0.38965306740860267,
+      "learning_rate": 2.5117370872684843e-10,
+      "loss": 0.3845,
+      "step": 8377
+    },
+    {
+      "epoch": 1.9941095971916463,
+      "grad_norm": 0.3625652747058628,
+      "learning_rate": 2.3222435963643485e-10,
+      "loss": 0.298,
+      "step": 8378
+    },
+    {
+      "epoch": 1.994347593264711,
+      "grad_norm": 0.3795900323777344,
+      "learning_rate": 2.1401809972509957e-10,
+      "loss": 0.2865,
+      "step": 8379
+    },
+    {
+      "epoch": 1.994585589337776,
+      "grad_norm": 0.3930027956116569,
+      "learning_rate": 1.965549316995663e-10,
+      "loss": 0.3304,
+      "step": 8380
+    },
+    {
+      "epoch": 1.9948235854108407,
+      "grad_norm": 0.37631234555304693,
+      "learning_rate": 1.7983485815553646e-10,
+      "loss": 0.3261,
+      "step": 8381
+    },
+    {
+      "epoch": 1.9950615814839057,
+      "grad_norm": 0.4000269616966013,
+      "learning_rate": 1.6385788157713413e-10,
+      "loss": 0.2903,
+      "step": 8382
+    },
+    {
+      "epoch": 1.9952995775569704,
+      "grad_norm": 0.3941543276970109,
+      "learning_rate": 1.486240043396814e-10,
+      "loss": 0.3109,
+      "step": 8383
+    },
+    {
+      "epoch": 1.9955375736300351,
+      "grad_norm": 0.36557779503672044,
+      "learning_rate": 1.3413322870692304e-10,
+      "loss": 0.3941,
+      "step": 8384
+    },
+    {
+      "epoch": 1.9957755697030999,
+      "grad_norm": 0.3717543720800664,
+      "learning_rate": 1.203855568326917e-10,
+      "loss": 0.3063,
+      "step": 8385
+    },
+    {
+      "epoch": 1.9960135657761646,
+      "grad_norm": 0.37253382680661573,
+      "learning_rate": 1.0738099076035291e-10,
+      "loss": 0.3121,
+      "step": 8386
+    },
+    {
+      "epoch": 1.9962515618492294,
+      "grad_norm": 0.3692513186862969,
+      "learning_rate": 9.511953242280492e-11,
+      "loss": 0.3249,
+      "step": 8387
+    },
+    {
+      "epoch": 1.9964895579222943,
+      "grad_norm": 0.42548108491286396,
+      "learning_rate": 8.360118364192371e-11,
+      "loss": 0.3672,
+      "step": 8388
+    },
+    {
+      "epoch": 1.996727553995359,
+      "grad_norm": 0.381706175323407,
+      "learning_rate": 7.28259461296732e-11,
+      "loss": 0.2803,
+      "step": 8389
+    },
+    {
+      "epoch": 1.996965550068424,
+      "grad_norm": 0.3901616736475803,
+      "learning_rate": 6.27938214881052e-11,
+      "loss": 0.3152,
+      "step": 8390
+    },
+    {
+      "epoch": 1.9972035461414888,
+      "grad_norm": 0.36504854093795047,
+      "learning_rate": 5.350481120769413e-11,
+      "loss": 0.3544,
+      "step": 8391
+    },
+    {
+      "epoch": 1.9974415422145535,
+      "grad_norm": 0.37140042079834173,
+      "learning_rate": 4.4958916669002315e-11,
+      "loss": 0.3046,
+      "step": 8392
+    },
+    {
+      "epoch": 1.9976795382876182,
+      "grad_norm": 0.42614737328351837,
+      "learning_rate": 3.7156139142680014e-11,
+      "loss": 0.2803,
+      "step": 8393
+    },
+    {
+      "epoch": 1.997917534360683,
+      "grad_norm": 0.4103735541875691,
+      "learning_rate": 3.009647978780006e-11,
+      "loss": 0.3432,
+      "step": 8394
+    },
+    {
+      "epoch": 1.9981555304337477,
+      "grad_norm": 0.4593503351998582,
+      "learning_rate": 2.377993965407832e-11,
+      "loss": 0.3484,
+      "step": 8395
+    },
+    {
+      "epoch": 1.9983935265068127,
+      "grad_norm": 0.38126825431046657,
+      "learning_rate": 1.8206519680208368e-11,
+      "loss": 0.2837,
+      "step": 8396
+    },
+    {
+      "epoch": 1.9986315225798774,
+      "grad_norm": 0.3896583084295616,
+      "learning_rate": 1.3376220694416575e-11,
+      "loss": 0.2981,
+      "step": 8397
+    },
+    {
+      "epoch": 1.9988695186529424,
+      "grad_norm": 0.3774261304077494,
+      "learning_rate": 9.289043414462128e-12,
+      "loss": 0.3539,
+      "step": 8398
+    },
+    {
+      "epoch": 1.9991075147260071,
+      "grad_norm": 0.3547871077621748,
+      "learning_rate": 5.944988447637024e-12,
+      "loss": 0.3137,
+      "step": 8399
+    },
+    {
+      "epoch": 1.9993455107990719,
+      "grad_norm": 0.37445511560704847,
+      "learning_rate": 3.3440562918762855e-12,
+      "loss": 0.2957,
+      "step": 8400
+    },
+    {
+      "epoch": 1.9995835068721366,
+      "grad_norm": 0.37305187911281695,
+      "learning_rate": 1.4862473329824156e-12,
+      "loss": 0.3171,
+      "step": 8401
+    },
+    {
+      "epoch": 1.9998215029452013,
+      "grad_norm": 0.3725046051286145,
+      "learning_rate": 3.7156184684583597e-13,
+      "loss": 0.3945,
+      "step": 8402
+    },
+    {
+      "epoch": 1.9998215029452013,
+      "step": 8402,
+      "total_flos": 1.2470400040881357e+17,
+      "train_loss": 0.3506650565916145,
+      "train_runtime": 79311.4461,
+      "train_samples_per_second": 108.495,
+      "train_steps_per_second": 0.106
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 8402,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.2470400040881357e+17,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}