diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,36428 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 5198,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00019238168526356292,
+      "grad_norm": 4.92498779296875,
+      "learning_rate": 1.282051282051282e-06,
+      "loss": 1.5772,
+      "step": 1
+    },
+    {
+      "epoch": 0.00038476337052712584,
+      "grad_norm": 12.497864723205566,
+      "learning_rate": 2.564102564102564e-06,
+      "loss": 1.717,
+      "step": 2
+    },
+    {
+      "epoch": 0.0005771450557906887,
+      "grad_norm": 12.55647087097168,
+      "learning_rate": 3.846153846153847e-06,
+      "loss": 1.6658,
+      "step": 3
+    },
+    {
+      "epoch": 0.0007695267410542517,
+      "grad_norm": 12.172236442565918,
+      "learning_rate": 5.128205128205128e-06,
+      "loss": 1.59,
+      "step": 4
+    },
+    {
+      "epoch": 0.0009619084263178145,
+      "grad_norm": 8.63586711883545,
+      "learning_rate": 6.41025641025641e-06,
+      "loss": 1.659,
+      "step": 5
+    },
+    {
+      "epoch": 0.0011542901115813775,
+      "grad_norm": 10.547835350036621,
+      "learning_rate": 7.692307692307694e-06,
+      "loss": 1.5772,
+      "step": 6
+    },
+    {
+      "epoch": 0.0013466717968449403,
+      "grad_norm": 9.318448066711426,
+      "learning_rate": 8.974358974358976e-06,
+      "loss": 1.5338,
+      "step": 7
+    },
+    {
+      "epoch": 0.0015390534821085034,
+      "grad_norm": 5.935449123382568,
+      "learning_rate": 1.0256410256410256e-05,
+      "loss": 1.4657,
+      "step": 8
+    },
+    {
+      "epoch": 0.0017314351673720662,
+      "grad_norm": 3.443028450012207,
+      "learning_rate": 1.153846153846154e-05,
+      "loss": 1.4751,
+      "step": 9
+    },
+    {
+      "epoch": 0.001923816852635629,
+      "grad_norm": 4.789201736450195,
+      "learning_rate": 1.282051282051282e-05,
+      "loss": 1.3456,
+      "step": 10
+    },
+    {
+      "epoch": 0.002116198537899192,
+      "grad_norm": 2.8618321418762207,
+      "learning_rate": 1.4102564102564104e-05,
+      "loss": 1.338,
+      "step": 11
+    },
+    {
+      "epoch": 0.002308580223162755,
+      "grad_norm": 3.6797053813934326,
+      "learning_rate": 1.5384615384615387e-05,
+      "loss": 1.4073,
+      "step": 12
+    },
+    {
+      "epoch": 0.002500961908426318,
+      "grad_norm": 3.941969633102417,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 1.4233,
+      "step": 13
+    },
+    {
+      "epoch": 0.0026933435936898806,
+      "grad_norm": 3.190112590789795,
+      "learning_rate": 1.794871794871795e-05,
+      "loss": 1.3141,
+      "step": 14
+    },
+    {
+      "epoch": 0.0028857252789534437,
+      "grad_norm": 4.279757499694824,
+      "learning_rate": 1.923076923076923e-05,
+      "loss": 1.3558,
+      "step": 15
+    },
+    {
+      "epoch": 0.0030781069642170067,
+      "grad_norm": 3.963376045227051,
+      "learning_rate": 2.0512820512820512e-05,
+      "loss": 1.3764,
+      "step": 16
+    },
+    {
+      "epoch": 0.0032704886494805694,
+      "grad_norm": 1.3308634757995605,
+      "learning_rate": 2.1794871794871795e-05,
+      "loss": 0.7492,
+      "step": 17
+    },
+    {
+      "epoch": 0.0034628703347441324,
+      "grad_norm": 3.4888436794281006,
+      "learning_rate": 2.307692307692308e-05,
+      "loss": 1.2644,
+      "step": 18
+    },
+    {
+      "epoch": 0.0036552520200076955,
+      "grad_norm": 2.3414809703826904,
+      "learning_rate": 2.435897435897436e-05,
+      "loss": 1.2878,
+      "step": 19
+    },
+    {
+      "epoch": 0.003847633705271258,
+      "grad_norm": 2.5538511276245117,
+      "learning_rate": 2.564102564102564e-05,
+      "loss": 1.2858,
+      "step": 20
+    },
+    {
+      "epoch": 0.004040015390534821,
+      "grad_norm": 2.734639883041382,
+      "learning_rate": 2.6923076923076923e-05,
+      "loss": 1.2464,
+      "step": 21
+    },
+    {
+      "epoch": 0.004232397075798384,
+      "grad_norm": 2.0652518272399902,
+      "learning_rate": 2.8205128205128207e-05,
+      "loss": 1.2913,
+      "step": 22
+    },
+    {
+      "epoch": 0.004424778761061947,
+      "grad_norm": 2.4652435779571533,
+      "learning_rate": 2.948717948717949e-05,
+      "loss": 1.1348,
+      "step": 23
+    },
+    {
+      "epoch": 0.00461716044632551,
+      "grad_norm": 1.8661668300628662,
+      "learning_rate": 3.0769230769230774e-05,
+      "loss": 1.2506,
+      "step": 24
+    },
+    {
+      "epoch": 0.004809542131589073,
+      "grad_norm": 1.7143030166625977,
+      "learning_rate": 3.205128205128206e-05,
+      "loss": 1.2665,
+      "step": 25
+    },
+    {
+      "epoch": 0.005001923816852636,
+      "grad_norm": 1.8765703439712524,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 1.2066,
+      "step": 26
+    },
+    {
+      "epoch": 0.005194305502116198,
+      "grad_norm": 1.9898818731307983,
+      "learning_rate": 3.461538461538462e-05,
+      "loss": 1.2298,
+      "step": 27
+    },
+    {
+      "epoch": 0.005386687187379761,
+      "grad_norm": 2.0084235668182373,
+      "learning_rate": 3.58974358974359e-05,
+      "loss": 1.2147,
+      "step": 28
+    },
+    {
+      "epoch": 0.005579068872643324,
+      "grad_norm": 1.2268122434616089,
+      "learning_rate": 3.717948717948718e-05,
+      "loss": 1.1637,
+      "step": 29
+    },
+    {
+      "epoch": 0.005771450557906887,
+      "grad_norm": 2.4675722122192383,
+      "learning_rate": 3.846153846153846e-05,
+      "loss": 1.2264,
+      "step": 30
+    },
+    {
+      "epoch": 0.00596383224317045,
+      "grad_norm": 1.9280927181243896,
+      "learning_rate": 3.974358974358974e-05,
+      "loss": 1.1645,
+      "step": 31
+    },
+    {
+      "epoch": 0.0061562139284340135,
+      "grad_norm": 2.3071112632751465,
+      "learning_rate": 4.1025641025641023e-05,
+      "loss": 1.2016,
+      "step": 32
+    },
+    {
+      "epoch": 0.006348595613697576,
+      "grad_norm": 2.146149158477783,
+      "learning_rate": 4.230769230769231e-05,
+      "loss": 1.172,
+      "step": 33
+    },
+    {
+      "epoch": 0.006540977298961139,
+      "grad_norm": 1.942165493965149,
+      "learning_rate": 4.358974358974359e-05,
+      "loss": 1.13,
+      "step": 34
+    },
+    {
+      "epoch": 0.006733358984224702,
+      "grad_norm": 2.0738024711608887,
+      "learning_rate": 4.4871794871794874e-05,
+      "loss": 1.128,
+      "step": 35
+    },
+    {
+      "epoch": 0.006925740669488265,
+      "grad_norm": 2.229220390319824,
+      "learning_rate": 4.615384615384616e-05,
+      "loss": 1.1389,
+      "step": 36
+    },
+    {
+      "epoch": 0.007118122354751828,
+      "grad_norm": 1.8627092838287354,
+      "learning_rate": 4.7435897435897435e-05,
+      "loss": 1.2044,
+      "step": 37
+    },
+    {
+      "epoch": 0.007310504040015391,
+      "grad_norm": 1.918159008026123,
+      "learning_rate": 4.871794871794872e-05,
+      "loss": 1.1606,
+      "step": 38
+    },
+    {
+      "epoch": 0.007502885725278953,
+      "grad_norm": 1.8836833238601685,
+      "learning_rate": 5e-05,
+      "loss": 1.16,
+      "step": 39
+    },
+    {
+      "epoch": 0.007695267410542516,
+      "grad_norm": 1.9975242614746094,
+      "learning_rate": 5.128205128205128e-05,
+      "loss": 1.1144,
+      "step": 40
+    },
+    {
+      "epoch": 0.007887649095806078,
+      "grad_norm": 1.812211275100708,
+      "learning_rate": 5.256410256410257e-05,
+      "loss": 1.1127,
+      "step": 41
+    },
+    {
+      "epoch": 0.008080030781069641,
+      "grad_norm": 1.3491607904434204,
+      "learning_rate": 5.384615384615385e-05,
+      "loss": 1.1111,
+      "step": 42
+    },
+    {
+      "epoch": 0.008272412466333205,
+      "grad_norm": 1.6214349269866943,
+      "learning_rate": 5.512820512820514e-05,
+      "loss": 1.1722,
+      "step": 43
+    },
+    {
+      "epoch": 0.008464794151596768,
+      "grad_norm": 1.7671188116073608,
+      "learning_rate": 5.6410256410256414e-05,
+      "loss": 1.1099,
+      "step": 44
+    },
+    {
+      "epoch": 0.00865717583686033,
+      "grad_norm": 2.1512198448181152,
+      "learning_rate": 5.769230769230769e-05,
+      "loss": 1.1119,
+      "step": 45
+    },
+    {
+      "epoch": 0.008849557522123894,
+      "grad_norm": 1.9436262845993042,
+      "learning_rate": 5.897435897435898e-05,
+      "loss": 1.1383,
+      "step": 46
+    },
+    {
+      "epoch": 0.009041939207387457,
+      "grad_norm": 2.0559234619140625,
+      "learning_rate": 6.025641025641026e-05,
+      "loss": 1.2193,
+      "step": 47
+    },
+    {
+      "epoch": 0.00923432089265102,
+      "grad_norm": 1.9375413656234741,
+      "learning_rate": 6.153846153846155e-05,
+      "loss": 1.0862,
+      "step": 48
+    },
+    {
+      "epoch": 0.009426702577914583,
+      "grad_norm": 1.699538230895996,
+      "learning_rate": 6.282051282051282e-05,
+      "loss": 1.178,
+      "step": 49
+    },
+    {
+      "epoch": 0.009619084263178146,
+      "grad_norm": 1.714860200881958,
+      "learning_rate": 6.410256410256412e-05,
+      "loss": 1.1476,
+      "step": 50
+    },
+    {
+      "epoch": 0.009811465948441709,
+      "grad_norm": 1.8182289600372314,
+      "learning_rate": 6.538461538461539e-05,
+      "loss": 1.1427,
+      "step": 51
+    },
+    {
+      "epoch": 0.010003847633705272,
+      "grad_norm": 1.8188971281051636,
+      "learning_rate": 6.666666666666667e-05,
+      "loss": 1.1269,
+      "step": 52
+    },
+    {
+      "epoch": 0.010196229318968833,
+      "grad_norm": 1.4701728820800781,
+      "learning_rate": 6.794871794871795e-05,
+      "loss": 1.0904,
+      "step": 53
+    },
+    {
+      "epoch": 0.010388611004232396,
+      "grad_norm": 1.8664014339447021,
+      "learning_rate": 6.923076923076924e-05,
+      "loss": 1.1275,
+      "step": 54
+    },
+    {
+      "epoch": 0.01058099268949596,
+      "grad_norm": 1.7978571653366089,
+      "learning_rate": 7.051282051282052e-05,
+      "loss": 1.2013,
+      "step": 55
+    },
+    {
+      "epoch": 0.010773374374759523,
+      "grad_norm": 1.404748797416687,
+      "learning_rate": 7.17948717948718e-05,
+      "loss": 1.1761,
+      "step": 56
+    },
+    {
+      "epoch": 0.010965756060023086,
+      "grad_norm": 1.8177757263183594,
+      "learning_rate": 7.307692307692307e-05,
+      "loss": 1.1249,
+      "step": 57
+    },
+    {
+      "epoch": 0.011158137745286649,
+      "grad_norm": 1.7299762964248657,
+      "learning_rate": 7.435897435897436e-05,
+      "loss": 1.1632,
+      "step": 58
+    },
+    {
+      "epoch": 0.011350519430550212,
+      "grad_norm": 1.9382716417312622,
+      "learning_rate": 7.564102564102564e-05,
+      "loss": 1.1649,
+      "step": 59
+    },
+    {
+      "epoch": 0.011542901115813775,
+      "grad_norm": 2.0657081604003906,
+      "learning_rate": 7.692307692307693e-05,
+      "loss": 1.0948,
+      "step": 60
+    },
+    {
+      "epoch": 0.011735282801077338,
+      "grad_norm": 1.4956806898117065,
+      "learning_rate": 7.820512820512821e-05,
+      "loss": 1.1189,
+      "step": 61
+    },
+    {
+      "epoch": 0.0119276644863409,
+      "grad_norm": 1.809260368347168,
+      "learning_rate": 7.948717948717948e-05,
+      "loss": 1.1669,
+      "step": 62
+    },
+    {
+      "epoch": 0.012120046171604464,
+      "grad_norm": 1.8002903461456299,
+      "learning_rate": 8.076923076923078e-05,
+      "loss": 1.176,
+      "step": 63
+    },
+    {
+      "epoch": 0.012312427856868027,
+      "grad_norm": 1.8092522621154785,
+      "learning_rate": 8.205128205128205e-05,
+      "loss": 1.0735,
+      "step": 64
+    },
+    {
+      "epoch": 0.012504809542131588,
+      "grad_norm": 1.937788963317871,
+      "learning_rate": 8.333333333333334e-05,
+      "loss": 1.0497,
+      "step": 65
+    },
+    {
+      "epoch": 0.012697191227395151,
+      "grad_norm": 2.006169557571411,
+      "learning_rate": 8.461538461538461e-05,
+      "loss": 1.1037,
+      "step": 66
+    },
+    {
+      "epoch": 0.012889572912658714,
+      "grad_norm": 1.5862001180648804,
+      "learning_rate": 8.58974358974359e-05,
+      "loss": 1.0786,
+      "step": 67
+    },
+    {
+      "epoch": 0.013081954597922277,
+      "grad_norm": 1.3349323272705078,
+      "learning_rate": 8.717948717948718e-05,
+      "loss": 1.1243,
+      "step": 68
+    },
+    {
+      "epoch": 0.01327433628318584,
+      "grad_norm": 1.7694772481918335,
+      "learning_rate": 8.846153846153847e-05,
+      "loss": 1.1469,
+      "step": 69
+    },
+    {
+      "epoch": 0.013466717968449404,
+      "grad_norm": 1.1605740785598755,
+      "learning_rate": 8.974358974358975e-05,
+      "loss": 1.0911,
+      "step": 70
+    },
+    {
+      "epoch": 0.013659099653712967,
+      "grad_norm": 1.6868046522140503,
+      "learning_rate": 9.102564102564103e-05,
+      "loss": 1.0685,
+      "step": 71
+    },
+    {
+      "epoch": 0.01385148133897653,
+      "grad_norm": 1.7114582061767578,
+      "learning_rate": 9.230769230769232e-05,
+      "loss": 1.1205,
+      "step": 72
+    },
+    {
+      "epoch": 0.014043863024240093,
+      "grad_norm": 2.0132670402526855,
+      "learning_rate": 9.35897435897436e-05,
+      "loss": 1.0725,
+      "step": 73
+    },
+    {
+      "epoch": 0.014236244709503656,
+      "grad_norm": 1.4529597759246826,
+      "learning_rate": 9.487179487179487e-05,
+      "loss": 1.0874,
+      "step": 74
+    },
+    {
+      "epoch": 0.014428626394767219,
+      "grad_norm": 2.0363399982452393,
+      "learning_rate": 9.615384615384617e-05,
+      "loss": 1.0638,
+      "step": 75
+    },
+    {
+      "epoch": 0.014621008080030782,
+      "grad_norm": 1.6701346635818481,
+      "learning_rate": 9.743589743589744e-05,
+      "loss": 1.1051,
+      "step": 76
+    },
+    {
+      "epoch": 0.014813389765294343,
+      "grad_norm": 1.8440680503845215,
+      "learning_rate": 9.871794871794872e-05,
+      "loss": 1.0728,
+      "step": 77
+    },
+    {
+      "epoch": 0.015005771450557906,
+      "grad_norm": 1.745774745941162,
+      "learning_rate": 0.0001,
+      "loss": 1.1327,
+      "step": 78
+    },
+    {
+      "epoch": 0.01519815313582147,
+      "grad_norm": 1.5274074077606201,
+      "learning_rate": 0.00010128205128205129,
+      "loss": 1.1151,
+      "step": 79
+    },
+    {
+      "epoch": 0.015390534821085032,
+      "grad_norm": 1.8766214847564697,
+      "learning_rate": 0.00010256410256410256,
+      "loss": 1.0534,
+      "step": 80
+    },
+    {
+      "epoch": 0.015582916506348595,
+      "grad_norm": 1.3353139162063599,
+      "learning_rate": 0.00010384615384615386,
+      "loss": 1.0969,
+      "step": 81
+    },
+    {
+      "epoch": 0.015775298191612157,
+      "grad_norm": 1.8122713565826416,
+      "learning_rate": 0.00010512820512820514,
+      "loss": 1.0663,
+      "step": 82
+    },
+    {
+      "epoch": 0.01596767987687572,
+      "grad_norm": 1.7894037961959839,
+      "learning_rate": 0.00010641025641025641,
+      "loss": 1.0608,
+      "step": 83
+    },
+    {
+      "epoch": 0.016160061562139283,
+      "grad_norm": 1.7295691967010498,
+      "learning_rate": 0.0001076923076923077,
+      "loss": 1.0875,
+      "step": 84
+    },
+    {
+      "epoch": 0.016352443247402846,
+      "grad_norm": 1.5073554515838623,
+      "learning_rate": 0.00010897435897435896,
+      "loss": 0.7774,
+      "step": 85
+    },
+    {
+      "epoch": 0.01654482493266641,
+      "grad_norm": 1.995965600013733,
+      "learning_rate": 0.00011025641025641027,
+      "loss": 1.0697,
+      "step": 86
+    },
+    {
+      "epoch": 0.016737206617929972,
+      "grad_norm": 1.623641014099121,
+      "learning_rate": 0.00011153846153846154,
+      "loss": 1.1065,
+      "step": 87
+    },
+    {
+      "epoch": 0.016929588303193535,
+      "grad_norm": 1.6261321306228638,
+      "learning_rate": 0.00011282051282051283,
+      "loss": 1.1025,
+      "step": 88
+    },
+    {
+      "epoch": 0.017121969988457098,
+      "grad_norm": 1.4713211059570312,
+      "learning_rate": 0.0001141025641025641,
+      "loss": 1.1379,
+      "step": 89
+    },
+    {
+      "epoch": 0.01731435167372066,
+      "grad_norm": 1.8331042528152466,
+      "learning_rate": 0.00011538461538461538,
+      "loss": 1.0997,
+      "step": 90
+    },
+    {
+      "epoch": 0.017506733358984224,
+      "grad_norm": 1.9172674417495728,
+      "learning_rate": 0.00011666666666666668,
+      "loss": 1.1341,
+      "step": 91
+    },
+    {
+      "epoch": 0.017699115044247787,
+      "grad_norm": 1.7967580556869507,
+      "learning_rate": 0.00011794871794871796,
+      "loss": 1.1013,
+      "step": 92
+    },
+    {
+      "epoch": 0.01789149672951135,
+      "grad_norm": 1.579447865486145,
+      "learning_rate": 0.00011923076923076923,
+      "loss": 1.1092,
+      "step": 93
+    },
+    {
+      "epoch": 0.018083878414774913,
+      "grad_norm": 1.8507041931152344,
+      "learning_rate": 0.00012051282051282052,
+      "loss": 1.0473,
+      "step": 94
+    },
+    {
+      "epoch": 0.018276260100038477,
+      "grad_norm": 1.7790749073028564,
+      "learning_rate": 0.00012179487179487179,
+      "loss": 1.0771,
+      "step": 95
+    },
+    {
+      "epoch": 0.01846864178530204,
+      "grad_norm": 2.1344738006591797,
+      "learning_rate": 0.0001230769230769231,
+      "loss": 1.0827,
+      "step": 96
+    },
+    {
+      "epoch": 0.018661023470565603,
+      "grad_norm": 1.6558130979537964,
+      "learning_rate": 0.00012435897435897437,
+      "loss": 1.0684,
+      "step": 97
+    },
+    {
+      "epoch": 0.018853405155829166,
+      "grad_norm": 1.7201650142669678,
+      "learning_rate": 0.00012564102564102564,
+      "loss": 1.0587,
+      "step": 98
+    },
+    {
+      "epoch": 0.01904578684109273,
+      "grad_norm": 1.570198893547058,
+      "learning_rate": 0.00012692307692307693,
+      "loss": 1.0515,
+      "step": 99
+    },
+    {
+      "epoch": 0.019238168526356292,
+      "grad_norm": 1.6664339303970337,
+      "learning_rate": 0.00012820512820512823,
+      "loss": 1.1093,
+      "step": 100
+    },
+    {
+      "epoch": 0.019430550211619855,
+      "grad_norm": 1.782614827156067,
+      "learning_rate": 0.0001294871794871795,
+      "loss": 1.1668,
+      "step": 101
+    },
+    {
+      "epoch": 0.019622931896883418,
+      "grad_norm": 1.528406023979187,
+      "learning_rate": 0.00013076923076923077,
+      "loss": 1.0915,
+      "step": 102
+    },
+    {
+      "epoch": 0.01981531358214698,
+      "grad_norm": 1.9418797492980957,
+      "learning_rate": 0.00013205128205128204,
+      "loss": 1.1065,
+      "step": 103
+    },
+    {
+      "epoch": 0.020007695267410544,
+      "grad_norm": 1.3045812845230103,
+      "learning_rate": 0.00013333333333333334,
+      "loss": 1.0306,
+      "step": 104
+    },
+    {
+      "epoch": 0.020200076952674107,
+      "grad_norm": 2.028512954711914,
+      "learning_rate": 0.00013461538461538464,
+      "loss": 1.1452,
+      "step": 105
+    },
+    {
+      "epoch": 0.020392458637937667,
+      "grad_norm": 1.8551981449127197,
+      "learning_rate": 0.0001358974358974359,
+      "loss": 1.0781,
+      "step": 106
+    },
+    {
+      "epoch": 0.02058484032320123,
+      "grad_norm": 1.667006254196167,
+      "learning_rate": 0.00013717948717948718,
+      "loss": 1.1327,
+      "step": 107
+    },
+    {
+      "epoch": 0.020777222008464793,
+      "grad_norm": 2.012650489807129,
+      "learning_rate": 0.00013846153846153847,
+      "loss": 1.0226,
+      "step": 108
+    },
+    {
+      "epoch": 0.020969603693728356,
+      "grad_norm": 1.8974764347076416,
+      "learning_rate": 0.00013974358974358974,
+      "loss": 1.0731,
+      "step": 109
+    },
+    {
+      "epoch": 0.02116198537899192,
+      "grad_norm": 1.2909404039382935,
+      "learning_rate": 0.00014102564102564104,
+      "loss": 1.0788,
+      "step": 110
+    },
+    {
+      "epoch": 0.021354367064255482,
+      "grad_norm": 1.736701488494873,
+      "learning_rate": 0.0001423076923076923,
+      "loss": 1.0975,
+      "step": 111
+    },
+    {
+      "epoch": 0.021546748749519045,
+      "grad_norm": 1.5784549713134766,
+      "learning_rate": 0.0001435897435897436,
+      "loss": 1.0945,
+      "step": 112
+    },
+    {
+      "epoch": 0.021739130434782608,
+      "grad_norm": 1.8834593296051025,
+      "learning_rate": 0.00014487179487179488,
+      "loss": 1.0212,
+      "step": 113
+    },
+    {
+      "epoch": 0.02193151212004617,
+      "grad_norm": 1.2115700244903564,
+      "learning_rate": 0.00014615384615384615,
+      "loss": 1.1276,
+      "step": 114
+    },
+    {
+      "epoch": 0.022123893805309734,
+      "grad_norm": 1.6120754480361938,
+      "learning_rate": 0.00014743589743589745,
+      "loss": 1.0652,
+      "step": 115
+    },
+    {
+      "epoch": 0.022316275490573297,
+      "grad_norm": 1.772003173828125,
+      "learning_rate": 0.00014871794871794872,
+      "loss": 1.0749,
+      "step": 116
+    },
+    {
+      "epoch": 0.02250865717583686,
+      "grad_norm": 1.5839918851852417,
+      "learning_rate": 0.00015000000000000001,
+      "loss": 1.0462,
+      "step": 117
+    },
+    {
+      "epoch": 0.022701038861100423,
+      "grad_norm": 1.8216899633407593,
+      "learning_rate": 0.00015128205128205128,
+      "loss": 1.0588,
+      "step": 118
+    },
+    {
+      "epoch": 0.022893420546363986,
+      "grad_norm": 1.6347311735153198,
+      "learning_rate": 0.00015256410256410255,
+      "loss": 1.112,
+      "step": 119
+    },
+    {
+      "epoch": 0.02308580223162755,
+      "grad_norm": 1.7707439661026,
+      "learning_rate": 0.00015384615384615385,
+      "loss": 1.0934,
+      "step": 120
+    },
+    {
+      "epoch": 0.023278183916891113,
+      "grad_norm": 1.673147439956665,
+      "learning_rate": 0.00015512820512820515,
+      "loss": 1.1108,
+      "step": 121
+    },
+    {
+      "epoch": 0.023470565602154676,
+      "grad_norm": 1.9476605653762817,
+      "learning_rate": 0.00015641025641025642,
+      "loss": 1.0569,
+      "step": 122
+    },
+    {
+      "epoch": 0.02366294728741824,
+      "grad_norm": 2.0730769634246826,
+      "learning_rate": 0.0001576923076923077,
+      "loss": 1.0953,
+      "step": 123
+    },
+    {
+      "epoch": 0.0238553289726818,
+      "grad_norm": 1.3350309133529663,
+      "learning_rate": 0.00015897435897435896,
+      "loss": 1.1222,
+      "step": 124
+    },
+    {
+      "epoch": 0.024047710657945365,
+      "grad_norm": 1.9896444082260132,
+      "learning_rate": 0.00016025641025641028,
+      "loss": 1.0818,
+      "step": 125
+    },
+    {
+      "epoch": 0.024240092343208928,
+      "grad_norm": 1.6628100872039795,
+      "learning_rate": 0.00016153846153846155,
+      "loss": 1.1104,
+      "step": 126
+    },
+    {
+      "epoch": 0.02443247402847249,
+      "grad_norm": 1.3486324548721313,
+      "learning_rate": 0.00016282051282051282,
+      "loss": 1.0604,
+      "step": 127
+    },
+    {
+      "epoch": 0.024624855713736054,
+      "grad_norm": 1.9224865436553955,
+      "learning_rate": 0.0001641025641025641,
+      "loss": 1.0336,
+      "step": 128
+    },
+    {
+      "epoch": 0.024817237398999614,
+      "grad_norm": 1.915023922920227,
+      "learning_rate": 0.0001653846153846154,
+      "loss": 1.0455,
+      "step": 129
+    },
+    {
+      "epoch": 0.025009619084263177,
+      "grad_norm": 2.017141342163086,
+      "learning_rate": 0.0001666666666666667,
+      "loss": 1.1106,
+      "step": 130
+    },
+    {
+      "epoch": 0.02520200076952674,
+      "grad_norm": 1.7894591093063354,
+      "learning_rate": 0.00016794871794871796,
+      "loss": 1.0414,
+      "step": 131
+    },
+    {
+      "epoch": 0.025394382454790303,
+      "grad_norm": 1.7869231700897217,
+      "learning_rate": 0.00016923076923076923,
+      "loss": 1.1145,
+      "step": 132
+    },
+    {
+      "epoch": 0.025586764140053866,
+      "grad_norm": 2.238743543624878,
+      "learning_rate": 0.00017051282051282053,
+      "loss": 1.0903,
+      "step": 133
+    },
+    {
+      "epoch": 0.02577914582531743,
+      "grad_norm": 1.583886742591858,
+      "learning_rate": 0.0001717948717948718,
+      "loss": 1.0873,
+      "step": 134
+    },
+    {
+      "epoch": 0.025971527510580992,
+      "grad_norm": 2.2115650177001953,
+      "learning_rate": 0.0001730769230769231,
+      "loss": 1.1398,
+      "step": 135
+    },
+    {
+      "epoch": 0.026163909195844555,
+      "grad_norm": 0.8913138508796692,
+      "learning_rate": 0.00017435897435897436,
+      "loss": 0.7903,
+      "step": 136
+    },
+    {
+      "epoch": 0.026356290881108118,
+      "grad_norm": 0.882841944694519,
+      "learning_rate": 0.00017564102564102566,
+      "loss": 0.7798,
+      "step": 137
+    },
+    {
+      "epoch": 0.02654867256637168,
+      "grad_norm": 1.5686968564987183,
+      "learning_rate": 0.00017692307692307693,
+      "loss": 1.1232,
+      "step": 138
+    },
+    {
+      "epoch": 0.026741054251635244,
+      "grad_norm": 1.4875348806381226,
+      "learning_rate": 0.00017820512820512823,
+      "loss": 1.1107,
+      "step": 139
+    },
+    {
+      "epoch": 0.026933435936898807,
+      "grad_norm": 2.2209794521331787,
+      "learning_rate": 0.0001794871794871795,
+      "loss": 1.1109,
+      "step": 140
+    },
+    {
+      "epoch": 0.02712581762216237,
+      "grad_norm": 1.4982985258102417,
+      "learning_rate": 0.00018076923076923077,
+      "loss": 1.0494,
+      "step": 141
+    },
+    {
+      "epoch": 0.027318199307425933,
+      "grad_norm": 1.2728239297866821,
+      "learning_rate": 0.00018205128205128207,
+      "loss": 1.0818,
+      "step": 142
+    },
+    {
+      "epoch": 0.027510580992689496,
+      "grad_norm": 1.8564809560775757,
+      "learning_rate": 0.00018333333333333334,
+      "loss": 1.0076,
+      "step": 143
+    },
+    {
+      "epoch": 0.02770296267795306,
+      "grad_norm": 1.7920169830322266,
+      "learning_rate": 0.00018461538461538463,
+      "loss": 1.058,
+      "step": 144
+    },
+    {
+      "epoch": 0.027895344363216622,
+      "grad_norm": 1.8399617671966553,
+      "learning_rate": 0.0001858974358974359,
+      "loss": 1.0318,
+      "step": 145
+    },
+    {
+      "epoch": 0.028087726048480186,
+      "grad_norm": 1.5305655002593994,
+      "learning_rate": 0.0001871794871794872,
+      "loss": 1.0371,
+      "step": 146
+    },
+    {
+      "epoch": 0.02828010773374375,
+      "grad_norm": 1.7517366409301758,
+      "learning_rate": 0.00018846153846153847,
+      "loss": 1.1069,
+      "step": 147
+    },
+    {
+      "epoch": 0.02847248941900731,
+      "grad_norm": 1.5834318399429321,
+      "learning_rate": 0.00018974358974358974,
+      "loss": 0.9846,
+      "step": 148
+    },
+    {
+      "epoch": 0.028664871104270875,
+      "grad_norm": 1.715209722518921,
+      "learning_rate": 0.00019102564102564104,
+      "loss": 1.138,
+      "step": 149
+    },
+    {
+      "epoch": 0.028857252789534438,
+      "grad_norm": 1.7746837139129639,
+      "learning_rate": 0.00019230769230769233,
+      "loss": 1.0977,
+      "step": 150
+    },
+    {
+      "epoch": 0.029049634474798,
+      "grad_norm": 1.8572345972061157,
+      "learning_rate": 0.0001935897435897436,
+      "loss": 1.0266,
+      "step": 151
+    },
+    {
+      "epoch": 0.029242016160061564,
+      "grad_norm": 1.4200849533081055,
+      "learning_rate": 0.00019487179487179487,
+      "loss": 1.0488,
+      "step": 152
+    },
+    {
+      "epoch": 0.029434397845325123,
+      "grad_norm": 1.444139838218689,
+      "learning_rate": 0.00019615384615384615,
+      "loss": 1.0961,
+      "step": 153
+    },
+    {
+      "epoch": 0.029626779530588686,
+      "grad_norm": 1.7651954889297485,
+      "learning_rate": 0.00019743589743589744,
+      "loss": 1.0818,
+      "step": 154
+    },
+    {
+      "epoch": 0.02981916121585225,
+      "grad_norm": 2.0080835819244385,
+      "learning_rate": 0.00019871794871794874,
+      "loss": 1.0703,
+      "step": 155
+    },
+    {
+      "epoch": 0.030011542901115813,
+      "grad_norm": 1.6216682195663452,
+      "learning_rate": 0.0002,
+      "loss": 1.0945,
+      "step": 156
+    },
+    {
+      "epoch": 0.030203924586379376,
+      "grad_norm": 1.644627332687378,
+      "learning_rate": 0.00019999998058827846,
+      "loss": 1.0774,
+      "step": 157
+    },
+    {
+      "epoch": 0.03039630627164294,
+      "grad_norm": 1.6524956226348877,
+      "learning_rate": 0.00019999992235312134,
+      "loss": 1.0462,
+      "step": 158
+    },
+    {
+      "epoch": 0.030588687956906502,
+      "grad_norm": 1.8573015928268433,
+      "learning_rate": 0.00019999982529455128,
+      "loss": 1.0123,
+      "step": 159
+    },
+    {
+      "epoch": 0.030781069642170065,
+      "grad_norm": 1.7543931007385254,
+      "learning_rate": 0.00019999968941260594,
+      "loss": 1.1208,
+      "step": 160
+    },
+    {
+      "epoch": 0.030973451327433628,
+      "grad_norm": 1.6778252124786377,
+      "learning_rate": 0.00019999951470733808,
+      "loss": 1.0165,
+      "step": 161
+    },
+    {
+      "epoch": 0.03116583301269719,
+      "grad_norm": 0.9228732585906982,
+      "learning_rate": 0.0001999993011788155,
+      "loss": 0.963,
+      "step": 162
+    },
+    {
+      "epoch": 0.031358214697960754,
+      "grad_norm": 1.6544078588485718,
+      "learning_rate": 0.00019999904882712116,
+      "loss": 1.0607,
+      "step": 163
+    },
+    {
+      "epoch": 0.031550596383224314,
+      "grad_norm": 1.1496496200561523,
+      "learning_rate": 0.000199998757652353,
+      "loss": 1.081,
+      "step": 164
+    },
+    {
+      "epoch": 0.03174297806848788,
+      "grad_norm": 1.7392216920852661,
+      "learning_rate": 0.000199998427654624,
+      "loss": 1.1091,
+      "step": 165
+    },
+    {
+      "epoch": 0.03193535975375144,
+      "grad_norm": 1.7842694520950317,
+      "learning_rate": 0.0001999980588340624,
+      "loss": 1.024,
+      "step": 166
+    },
+    {
+      "epoch": 0.032127741439015006,
+      "grad_norm": 1.6926536560058594,
+      "learning_rate": 0.0001999976511908113,
+      "loss": 1.0869,
+      "step": 167
+    },
+    {
+      "epoch": 0.032320123124278566,
+      "grad_norm": 2.088805675506592,
+      "learning_rate": 0.00019999720472502902,
+      "loss": 1.099,
+      "step": 168
+    },
+    {
+      "epoch": 0.03251250480954213,
+      "grad_norm": 0.9160146117210388,
+      "learning_rate": 0.00019999671943688883,
+      "loss": 0.7922,
+      "step": 169
+    },
+    {
+      "epoch": 0.03270488649480569,
+      "grad_norm": 1.7307684421539307,
+      "learning_rate": 0.00019999619532657916,
+      "loss": 1.0625,
+      "step": 170
+    },
+    {
+      "epoch": 0.03289726818006926,
+      "grad_norm": 1.764432668685913,
+      "learning_rate": 0.00019999563239430352,
+      "loss": 1.0627,
+      "step": 171
+    },
+    {
+      "epoch": 0.03308964986533282,
+      "grad_norm": 1.8162776231765747,
+      "learning_rate": 0.00019999503064028042,
+      "loss": 1.0773,
+      "step": 172
+    },
+    {
+      "epoch": 0.033282031550596385,
+      "grad_norm": 1.9939111471176147,
+      "learning_rate": 0.0001999943900647435,
+      "loss": 1.0412,
+      "step": 173
+    },
+    {
+      "epoch": 0.033474413235859944,
+      "grad_norm": 2.055025577545166,
+      "learning_rate": 0.00019999371066794147,
+      "loss": 1.1056,
+      "step": 174
+    },
+    {
+      "epoch": 0.03366679492112351,
+      "grad_norm": 1.9209600687026978,
+      "learning_rate": 0.00019999299245013804,
+      "loss": 1.031,
+      "step": 175
+    },
+    {
+      "epoch": 0.03385917660638707,
+      "grad_norm": 1.745395302772522,
+      "learning_rate": 0.0001999922354116121,
+      "loss": 1.0836,
+      "step": 176
+    },
+    {
+      "epoch": 0.03405155829165064,
+      "grad_norm": 1.5391285419464111,
+      "learning_rate": 0.0001999914395526575,
+      "loss": 1.0579,
+      "step": 177
+    },
+    {
+      "epoch": 0.034243939976914196,
+      "grad_norm": 1.8213422298431396,
+      "learning_rate": 0.0001999906048735833,
+      "loss": 1.0863,
+      "step": 178
+    },
+    {
+      "epoch": 0.03443632166217776,
+      "grad_norm": 1.4262068271636963,
+      "learning_rate": 0.00019998973137471353,
+      "loss": 1.0729,
+      "step": 179
+    },
+    {
+      "epoch": 0.03462870334744132,
+      "grad_norm": 0.9120514988899231,
+      "learning_rate": 0.00019998881905638726,
+      "loss": 1.0785,
+      "step": 180
+    },
+    {
+      "epoch": 0.03482108503270489,
+      "grad_norm": 1.7651628255844116,
+      "learning_rate": 0.0001999878679189587,
+      "loss": 1.0701,
+      "step": 181
+    },
+    {
+      "epoch": 0.03501346671796845,
+      "grad_norm": 2.043696880340576,
+      "learning_rate": 0.00019998687796279717,
+      "loss": 1.0621,
+      "step": 182
+    },
+    {
+      "epoch": 0.035205848403232015,
+      "grad_norm": 1.5956097841262817,
+      "learning_rate": 0.00019998584918828695,
+      "loss": 1.0998,
+      "step": 183
+    },
+    {
+      "epoch": 0.035398230088495575,
+      "grad_norm": 1.047623634338379,
+      "learning_rate": 0.00019998478159582748,
+      "loss": 1.1621,
+      "step": 184
+    },
+    {
+      "epoch": 0.03559061177375914,
+      "grad_norm": 1.8303773403167725,
+      "learning_rate": 0.00019998367518583317,
+      "loss": 1.0211,
+      "step": 185
+    },
+    {
+      "epoch": 0.0357829934590227,
+      "grad_norm": 1.3786566257476807,
+      "learning_rate": 0.00019998252995873366,
+      "loss": 1.0522,
+      "step": 186
+    },
+    {
+      "epoch": 0.03597537514428627,
+      "grad_norm": 2.338752508163452,
+      "learning_rate": 0.0001999813459149735,
+      "loss": 1.1103,
+      "step": 187
+    },
+    {
+      "epoch": 0.03616775682954983,
+      "grad_norm": 1.7846789360046387,
+      "learning_rate": 0.0001999801230550124,
+      "loss": 1.041,
+      "step": 188
+    },
+    {
+      "epoch": 0.03636013851481339,
+      "grad_norm": 1.3562257289886475,
+      "learning_rate": 0.00019997886137932512,
+      "loss": 1.0911,
+      "step": 189
+    },
+    {
+      "epoch": 0.03655252020007695,
+      "grad_norm": 1.64940345287323,
+      "learning_rate": 0.00019997756088840148,
+      "loss": 1.063,
+      "step": 190
+    },
+    {
+      "epoch": 0.03674490188534051,
+      "grad_norm": 1.5958200693130493,
+      "learning_rate": 0.00019997622158274636,
+      "loss": 1.043,
+      "step": 191
+    },
+    {
+      "epoch": 0.03693728357060408,
+      "grad_norm": 1.5559808015823364,
+      "learning_rate": 0.00019997484346287973,
+      "loss": 1.0581,
+      "step": 192
+    },
+    {
+      "epoch": 0.03712966525586764,
+      "grad_norm": 1.8303838968276978,
+      "learning_rate": 0.00019997342652933666,
+      "loss": 1.0422,
+      "step": 193
+    },
+    {
+      "epoch": 0.037322046941131205,
+      "grad_norm": 1.6936732530593872,
+      "learning_rate": 0.00019997197078266724,
+      "loss": 1.031,
+      "step": 194
+    },
+    {
+      "epoch": 0.037514428626394765,
+      "grad_norm": 0.8400489091873169,
+      "learning_rate": 0.00019997047622343658,
+      "loss": 1.1014,
+      "step": 195
+    },
+    {
+      "epoch": 0.03770681031165833,
+      "grad_norm": 1.9532846212387085,
+      "learning_rate": 0.00019996894285222497,
+      "loss": 1.0611,
+      "step": 196
+    },
+    {
+      "epoch": 0.03789919199692189,
+      "grad_norm": 1.6949049234390259,
+      "learning_rate": 0.00019996737066962775,
+      "loss": 0.9709,
+      "step": 197
+    },
+    {
+      "epoch": 0.03809157368218546,
+      "grad_norm": 1.6634737253189087,
+      "learning_rate": 0.00019996575967625523,
+      "loss": 1.0674,
+      "step": 198
+    },
+    {
+      "epoch": 0.03828395536744902,
+      "grad_norm": 1.8566974401474,
+      "learning_rate": 0.0001999641098727329,
+      "loss": 1.1037,
+      "step": 199
+    },
+    {
+      "epoch": 0.038476337052712584,
+      "grad_norm": 1.8973345756530762,
+      "learning_rate": 0.00019996242125970126,
+      "loss": 0.9974,
+      "step": 200
+    },
+    {
+      "epoch": 0.03866871873797614,
+      "grad_norm": 1.683227300643921,
+      "learning_rate": 0.00019996069383781587,
+      "loss": 1.0472,
+      "step": 201
+    },
+    {
+      "epoch": 0.03886110042323971,
+      "grad_norm": 1.7202367782592773,
+      "learning_rate": 0.00019995892760774737,
+      "loss": 1.0747,
+      "step": 202
+    },
+    {
+      "epoch": 0.03905348210850327,
+      "grad_norm": 1.5691304206848145,
+      "learning_rate": 0.0001999571225701815,
+      "loss": 1.119,
+      "step": 203
+    },
+    {
+      "epoch": 0.039245863793766836,
+      "grad_norm": 1.9726015329360962,
+      "learning_rate": 0.00019995527872581902,
+      "loss": 1.0435,
+      "step": 204
+    },
+    {
+      "epoch": 0.039438245479030395,
+      "grad_norm": 2.2409567832946777,
+      "learning_rate": 0.00019995339607537579,
+      "loss": 1.0586,
+      "step": 205
+    },
+    {
+      "epoch": 0.03963062716429396,
+      "grad_norm": 1.8972445726394653,
+      "learning_rate": 0.00019995147461958267,
+      "loss": 1.0799,
+      "step": 206
+    },
+    {
+      "epoch": 0.03982300884955752,
+      "grad_norm": 0.9581467509269714,
+      "learning_rate": 0.0001999495143591857,
+      "loss": 0.7462,
+      "step": 207
+    },
+    {
+      "epoch": 0.04001539053482109,
+      "grad_norm": 1.9798808097839355,
+      "learning_rate": 0.0001999475152949459,
+      "loss": 1.0399,
+      "step": 208
+    },
+    {
+      "epoch": 0.04020777222008465,
+      "grad_norm": 2.289254665374756,
+      "learning_rate": 0.00019994547742763934,
+      "loss": 1.0842,
+      "step": 209
+    },
+    {
+      "epoch": 0.040400153905348214,
+      "grad_norm": 1.9446367025375366,
+      "learning_rate": 0.00019994340075805725,
+      "loss": 1.1144,
+      "step": 210
+    },
+    {
+      "epoch": 0.040592535590611774,
+      "grad_norm": 1.5793346166610718,
+      "learning_rate": 0.00019994128528700583,
+      "loss": 1.1093,
+      "step": 211
+    },
+    {
+      "epoch": 0.04078491727587533,
+      "grad_norm": 1.7397487163543701,
+      "learning_rate": 0.00019993913101530635,
+      "loss": 1.0464,
+      "step": 212
+    },
+    {
+      "epoch": 0.0409772989611389,
+      "grad_norm": 1.9671047925949097,
+      "learning_rate": 0.00019993693794379525,
+      "loss": 0.9992,
+      "step": 213
+    },
+    {
+      "epoch": 0.04116968064640246,
+      "grad_norm": 1.6423097848892212,
+      "learning_rate": 0.0001999347060733239,
+      "loss": 1.1047,
+      "step": 214
+    },
+    {
+      "epoch": 0.041362062331666026,
+      "grad_norm": 1.6032108068466187,
+      "learning_rate": 0.00019993243540475877,
+      "loss": 1.086,
+      "step": 215
+    },
+    {
+      "epoch": 0.041554444016929586,
+      "grad_norm": 1.8130393028259277,
+      "learning_rate": 0.00019993012593898146,
+      "loss": 1.027,
+      "step": 216
+    },
+    {
+      "epoch": 0.04174682570219315,
+      "grad_norm": 2.2153468132019043,
+      "learning_rate": 0.00019992777767688854,
+      "loss": 1.024,
+      "step": 217
+    },
+    {
+      "epoch": 0.04193920738745671,
+      "grad_norm": 1.9197262525558472,
+      "learning_rate": 0.00019992539061939175,
+      "loss": 1.131,
+      "step": 218
+    },
+    {
+      "epoch": 0.04213158907272028,
+      "grad_norm": 2.1445183753967285,
+      "learning_rate": 0.00019992296476741776,
+      "loss": 1.0662,
+      "step": 219
+    },
+    {
+      "epoch": 0.04232397075798384,
+      "grad_norm": 1.4467809200286865,
+      "learning_rate": 0.00019992050012190843,
+      "loss": 1.0185,
+      "step": 220
+    },
+    {
+      "epoch": 0.042516352443247404,
+      "grad_norm": 0.8590517640113831,
+      "learning_rate": 0.00019991799668382058,
+      "loss": 0.7443,
+      "step": 221
+    },
+    {
+      "epoch": 0.042708734128510964,
+      "grad_norm": 1.814706802368164,
+      "learning_rate": 0.00019991545445412613,
+      "loss": 1.0738,
+      "step": 222
+    },
+    {
+      "epoch": 0.04290111581377453,
+      "grad_norm": 1.3723639249801636,
+      "learning_rate": 0.0001999128734338121,
+      "loss": 1.2236,
+      "step": 223
+    },
+    {
+      "epoch": 0.04309349749903809,
+      "grad_norm": 1.8626703023910522,
+      "learning_rate": 0.00019991025362388044,
+      "loss": 1.0328,
+      "step": 224
+    },
+    {
+      "epoch": 0.04328587918430166,
+      "grad_norm": 1.6433138847351074,
+      "learning_rate": 0.00019990759502534834,
+      "loss": 1.0657,
+      "step": 225
+    },
+    {
+      "epoch": 0.043478260869565216,
+      "grad_norm": 1.3124516010284424,
+      "learning_rate": 0.00019990489763924797,
+      "loss": 1.06,
+      "step": 226
+    },
+    {
+      "epoch": 0.04367064255482878,
+      "grad_norm": 2.049762487411499,
+      "learning_rate": 0.00019990216146662646,
+      "loss": 1.0517,
+      "step": 227
+    },
+    {
+      "epoch": 0.04386302424009234,
+      "grad_norm": 1.572428822517395,
+      "learning_rate": 0.00019989938650854616,
+      "loss": 1.112,
+      "step": 228
+    },
+    {
+      "epoch": 0.04405540592535591,
+      "grad_norm": 1.5736290216445923,
+      "learning_rate": 0.00019989657276608437,
+      "loss": 1.0434,
+      "step": 229
+    },
+    {
+      "epoch": 0.04424778761061947,
+      "grad_norm": 1.6951377391815186,
+      "learning_rate": 0.00019989372024033352,
+      "loss": 1.074,
+      "step": 230
+    },
+    {
+      "epoch": 0.044440169295883035,
+      "grad_norm": 2.084453582763672,
+      "learning_rate": 0.000199890828932401,
+      "loss": 1.0748,
+      "step": 231
+    },
+    {
+      "epoch": 0.044632550981146595,
+      "grad_norm": 1.6027588844299316,
+      "learning_rate": 0.00019988789884340936,
+      "loss": 0.9743,
+      "step": 232
+    },
+    {
+      "epoch": 0.04482493266641016,
+      "grad_norm": 1.7621992826461792,
+      "learning_rate": 0.00019988492997449615,
+      "loss": 1.0,
+      "step": 233
+    },
+    {
+      "epoch": 0.04501731435167372,
+      "grad_norm": 1.6983375549316406,
+      "learning_rate": 0.00019988192232681397,
+      "loss": 0.9818,
+      "step": 234
+    },
+    {
+      "epoch": 0.04520969603693728,
+      "grad_norm": 1.6871540546417236,
+      "learning_rate": 0.00019987887590153055,
+      "loss": 1.0904,
+      "step": 235
+    },
+    {
+      "epoch": 0.04540207772220085,
+      "grad_norm": 2.01318359375,
+      "learning_rate": 0.00019987579069982855,
+      "loss": 1.0391,
+      "step": 236
+    },
+    {
+      "epoch": 0.045594459407464406,
+      "grad_norm": 1.4290627241134644,
+      "learning_rate": 0.00019987266672290575,
+      "loss": 1.044,
+      "step": 237
+    },
+    {
+      "epoch": 0.04578684109272797,
+      "grad_norm": 1.7981986999511719,
+      "learning_rate": 0.00019986950397197503,
+      "loss": 1.0642,
+      "step": 238
+    },
+    {
+      "epoch": 0.04597922277799153,
+      "grad_norm": 1.7093929052352905,
+      "learning_rate": 0.00019986630244826426,
+      "loss": 1.0276,
+      "step": 239
+    },
+    {
+      "epoch": 0.0461716044632551,
+      "grad_norm": 1.85580575466156,
+      "learning_rate": 0.00019986306215301637,
+      "loss": 1.0397,
+      "step": 240
+    },
+    {
+      "epoch": 0.04636398614851866,
+      "grad_norm": 0.7930272221565247,
+      "learning_rate": 0.00019985978308748937,
+      "loss": 1.0109,
+      "step": 241
+    },
+    {
+      "epoch": 0.046556367833782225,
+      "grad_norm": 1.8546979427337646,
+      "learning_rate": 0.00019985646525295632,
+      "loss": 1.1186,
+      "step": 242
+    },
+    {
+      "epoch": 0.046748749519045785,
+      "grad_norm": 1.6254647970199585,
+      "learning_rate": 0.0001998531086507053,
+      "loss": 1.036,
+      "step": 243
+    },
+    {
+      "epoch": 0.04694113120430935,
+      "grad_norm": 1.6355758905410767,
+      "learning_rate": 0.00019984971328203946,
+      "loss": 1.0898,
+      "step": 244
+    },
+    {
+      "epoch": 0.04713351288957291,
+      "grad_norm": 1.7732716798782349,
+      "learning_rate": 0.00019984627914827697,
+      "loss": 1.0757,
+      "step": 245
+    },
+    {
+      "epoch": 0.04732589457483648,
+      "grad_norm": 1.7871285676956177,
+      "learning_rate": 0.00019984280625075113,
+      "loss": 1.0434,
+      "step": 246
+    },
+    {
+      "epoch": 0.04751827626010004,
+      "grad_norm": 1.6214029788970947,
+      "learning_rate": 0.0001998392945908102,
+      "loss": 0.9826,
+      "step": 247
+    },
+    {
+      "epoch": 0.0477106579453636,
+      "grad_norm": 0.9327989220619202,
+      "learning_rate": 0.00019983574416981755,
+      "loss": 0.7359,
+      "step": 248
+    },
+    {
+      "epoch": 0.04790303963062716,
+      "grad_norm": 1.4894016981124878,
+      "learning_rate": 0.00019983215498915158,
+      "loss": 1.0607,
+      "step": 249
+    },
+    {
+      "epoch": 0.04809542131589073,
+      "grad_norm": 1.4642492532730103,
+      "learning_rate": 0.0001998285270502057,
+      "loss": 1.1051,
+      "step": 250
+    },
+    {
+      "epoch": 0.04828780300115429,
+      "grad_norm": 1.6882504224777222,
+      "learning_rate": 0.00019982486035438846,
+      "loss": 1.0779,
+      "step": 251
+    },
+    {
+      "epoch": 0.048480184686417856,
+      "grad_norm": 1.644441843032837,
+      "learning_rate": 0.00019982115490312332,
+      "loss": 1.0283,
+      "step": 252
+    },
+    {
+      "epoch": 0.048672566371681415,
+      "grad_norm": 0.8184213638305664,
+      "learning_rate": 0.0001998174106978489,
+      "loss": 0.7265,
+      "step": 253
+    },
+    {
+      "epoch": 0.04886494805694498,
+      "grad_norm": 2.1976845264434814,
+      "learning_rate": 0.00019981362774001887,
+      "loss": 1.0824,
+      "step": 254
+    },
+    {
+      "epoch": 0.04905732974220854,
+      "grad_norm": 1.6611489057540894,
+      "learning_rate": 0.00019980980603110185,
+      "loss": 1.0956,
+      "step": 255
+    },
+    {
+      "epoch": 0.04924971142747211,
+      "grad_norm": 0.9055196046829224,
+      "learning_rate": 0.00019980594557258155,
+      "loss": 0.7521,
+      "step": 256
+    },
+    {
+      "epoch": 0.04944209311273567,
+      "grad_norm": 0.8672589063644409,
+      "learning_rate": 0.00019980204636595682,
+      "loss": 0.747,
+      "step": 257
+    },
+    {
+      "epoch": 0.04963447479799923,
+      "grad_norm": 0.8366668820381165,
+      "learning_rate": 0.00019979810841274132,
+      "loss": 0.6566,
+      "step": 258
+    },
+    {
+      "epoch": 0.049826856483262794,
+      "grad_norm": 1.9930254220962524,
+      "learning_rate": 0.00019979413171446403,
+      "loss": 1.0429,
+      "step": 259
+    },
+    {
+      "epoch": 0.05001923816852635,
+      "grad_norm": 1.8433520793914795,
+      "learning_rate": 0.00019979011627266883,
+      "loss": 1.0928,
+      "step": 260
+    },
+    {
+      "epoch": 0.05021161985378992,
+      "grad_norm": 1.6963917016983032,
+      "learning_rate": 0.00019978606208891458,
+      "loss": 1.0687,
+      "step": 261
+    },
+    {
+      "epoch": 0.05040400153905348,
+      "grad_norm": 2.0809738636016846,
+      "learning_rate": 0.0001997819691647753,
+      "loss": 1.0391,
+      "step": 262
+    },
+    {
+      "epoch": 0.050596383224317046,
+      "grad_norm": 1.992314338684082,
+      "learning_rate": 0.00019977783750183998,
+      "loss": 1.0341,
+      "step": 263
+    },
+    {
+      "epoch": 0.050788764909580605,
+      "grad_norm": 1.6958954334259033,
+      "learning_rate": 0.00019977366710171272,
+      "loss": 0.9958,
+      "step": 264
+    },
+    {
+      "epoch": 0.05098114659484417,
+      "grad_norm": 1.4711647033691406,
+      "learning_rate": 0.00019976945796601258,
+      "loss": 1.0811,
+      "step": 265
+    },
+    {
+      "epoch": 0.05117352828010773,
+      "grad_norm": 1.5001955032348633,
+      "learning_rate": 0.00019976521009637364,
+      "loss": 0.9945,
+      "step": 266
+    },
+    {
+      "epoch": 0.0513659099653713,
+      "grad_norm": 1.711107850074768,
+      "learning_rate": 0.0001997609234944452,
+      "loss": 1.0754,
+      "step": 267
+    },
+    {
+      "epoch": 0.05155829165063486,
+      "grad_norm": 1.730202317237854,
+      "learning_rate": 0.00019975659816189137,
+      "loss": 0.9587,
+      "step": 268
+    },
+    {
+      "epoch": 0.051750673335898424,
+      "grad_norm": 1.549720287322998,
+      "learning_rate": 0.0001997522341003914,
+      "loss": 1.0339,
+      "step": 269
+    },
+    {
+      "epoch": 0.051943055021161984,
+      "grad_norm": 1.4957996606826782,
+      "learning_rate": 0.00019974783131163957,
+      "loss": 1.0939,
+      "step": 270
+    },
+    {
+      "epoch": 0.05213543670642555,
+      "grad_norm": 1.9354572296142578,
+      "learning_rate": 0.00019974338979734523,
+      "loss": 0.9848,
+      "step": 271
+    },
+    {
+      "epoch": 0.05232781839168911,
+      "grad_norm": 2.1979103088378906,
+      "learning_rate": 0.00019973890955923268,
+      "loss": 1.0221,
+      "step": 272
+    },
+    {
+      "epoch": 0.052520200076952676,
+      "grad_norm": 1.7516456842422485,
+      "learning_rate": 0.00019973439059904133,
+      "loss": 1.0954,
+      "step": 273
+    },
+    {
+      "epoch": 0.052712581762216236,
+      "grad_norm": 1.9222300052642822,
+      "learning_rate": 0.00019972983291852563,
+      "loss": 1.0699,
+      "step": 274
+    },
+    {
+      "epoch": 0.0529049634474798,
+      "grad_norm": 1.784993052482605,
+      "learning_rate": 0.00019972523651945497,
+      "loss": 1.022,
+      "step": 275
+    },
+    {
+      "epoch": 0.05309734513274336,
+      "grad_norm": 1.6627116203308105,
+      "learning_rate": 0.00019972060140361383,
+      "loss": 1.0914,
+      "step": 276
+    },
+    {
+      "epoch": 0.05328972681800693,
+      "grad_norm": 1.9766520261764526,
+      "learning_rate": 0.00019971592757280177,
+      "loss": 1.042,
+      "step": 277
+    },
+    {
+      "epoch": 0.05348210850327049,
+      "grad_norm": 1.924538016319275,
+      "learning_rate": 0.0001997112150288333,
+      "loss": 1.002,
+      "step": 278
+    },
+    {
+      "epoch": 0.053674490188534055,
+      "grad_norm": 1.8875436782836914,
+      "learning_rate": 0.000199706463773538,
+      "loss": 1.0148,
+      "step": 279
+    },
+    {
+      "epoch": 0.053866871873797614,
+      "grad_norm": 1.8483328819274902,
+      "learning_rate": 0.0001997016738087605,
+      "loss": 1.0694,
+      "step": 280
+    },
+    {
+      "epoch": 0.054059253559061174,
+      "grad_norm": 1.4825000762939453,
+      "learning_rate": 0.00019969684513636033,
+      "loss": 0.9735,
+      "step": 281
+    },
+    {
+      "epoch": 0.05425163524432474,
+      "grad_norm": 1.9385045766830444,
+      "learning_rate": 0.00019969197775821225,
+      "loss": 1.0501,
+      "step": 282
+    },
+    {
+      "epoch": 0.0544440169295883,
+      "grad_norm": 1.7237247228622437,
+      "learning_rate": 0.00019968707167620592,
+      "loss": 1.0723,
+      "step": 283
+    },
+    {
+      "epoch": 0.05463639861485187,
+      "grad_norm": 2.4142708778381348,
+      "learning_rate": 0.00019968212689224602,
+      "loss": 1.0182,
+      "step": 284
+    },
+    {
+      "epoch": 0.054828780300115426,
+      "grad_norm": 1.522884726524353,
+      "learning_rate": 0.00019967714340825233,
+      "loss": 1.0136,
+      "step": 285
+    },
+    {
+      "epoch": 0.05502116198537899,
+      "grad_norm": 1.65798020362854,
+      "learning_rate": 0.00019967212122615956,
+      "loss": 1.0393,
+      "step": 286
+    },
+    {
+      "epoch": 0.05521354367064255,
+      "grad_norm": 0.8446666598320007,
+      "learning_rate": 0.00019966706034791752,
+      "loss": 0.6628,
+      "step": 287
+    },
+    {
+      "epoch": 0.05540592535590612,
+      "grad_norm": 1.8884879350662231,
+      "learning_rate": 0.00019966196077549105,
+      "loss": 1.0829,
+      "step": 288
+    },
+    {
+      "epoch": 0.05559830704116968,
+      "grad_norm": 0.8589560985565186,
+      "learning_rate": 0.0001996568225108599,
+      "loss": 0.7185,
+      "step": 289
+    },
+    {
+      "epoch": 0.055790688726433245,
+      "grad_norm": 1.486825942993164,
+      "learning_rate": 0.00019965164555601898,
+      "loss": 1.0411,
+      "step": 290
+    },
+    {
+      "epoch": 0.055983070411696804,
+      "grad_norm": 1.921863317489624,
+      "learning_rate": 0.00019964642991297815,
+      "loss": 1.0917,
+      "step": 291
+    },
+    {
+      "epoch": 0.05617545209696037,
+      "grad_norm": 1.822689175605774,
+      "learning_rate": 0.0001996411755837623,
+      "loss": 1.0501,
+      "step": 292
+    },
+    {
+      "epoch": 0.05636783378222393,
+      "grad_norm": 1.7954978942871094,
+      "learning_rate": 0.00019963588257041135,
+      "loss": 1.1045,
+      "step": 293
+    },
+    {
+      "epoch": 0.0565602154674875,
+      "grad_norm": 1.6871027946472168,
+      "learning_rate": 0.00019963055087498022,
+      "loss": 1.1007,
+      "step": 294
+    },
+    {
+      "epoch": 0.05675259715275106,
+      "grad_norm": 1.6988451480865479,
+      "learning_rate": 0.00019962518049953886,
+      "loss": 1.0606,
+      "step": 295
+    },
+    {
+      "epoch": 0.05694497883801462,
+      "grad_norm": 1.7476335763931274,
+      "learning_rate": 0.00019961977144617224,
+      "loss": 0.9733,
+      "step": 296
+    },
+    {
+      "epoch": 0.05713736052327818,
+      "grad_norm": 1.7625575065612793,
+      "learning_rate": 0.0001996143237169803,
+      "loss": 1.0374,
+      "step": 297
+    },
+    {
+      "epoch": 0.05732974220854175,
+      "grad_norm": 1.7636874914169312,
+      "learning_rate": 0.00019960883731407812,
+      "loss": 1.006,
+      "step": 298
+    },
+    {
+      "epoch": 0.05752212389380531,
+      "grad_norm": 2.2348296642303467,
+      "learning_rate": 0.00019960331223959564,
+      "loss": 1.0707,
+      "step": 299
+    },
+    {
+      "epoch": 0.057714505579068875,
+      "grad_norm": 1.6358139514923096,
+      "learning_rate": 0.0001995977484956779,
+      "loss": 1.0043,
+      "step": 300
+    },
+    {
+      "epoch": 0.057906887264332435,
+      "grad_norm": 1.8958913087844849,
+      "learning_rate": 0.00019959214608448496,
+      "loss": 1.0072,
+      "step": 301
+    },
+    {
+      "epoch": 0.058099268949596,
+      "grad_norm": 2.036186695098877,
+      "learning_rate": 0.00019958650500819184,
+      "loss": 1.0412,
+      "step": 302
+    },
+    {
+      "epoch": 0.05829165063485956,
+      "grad_norm": 2.2211945056915283,
+      "learning_rate": 0.0001995808252689886,
+      "loss": 1.0951,
+      "step": 303
+    },
+    {
+      "epoch": 0.05848403232012313,
+      "grad_norm": 1.8488328456878662,
+      "learning_rate": 0.00019957510686908033,
+      "loss": 1.0099,
+      "step": 304
+    },
+    {
+      "epoch": 0.05867641400538669,
+      "grad_norm": 1.8776886463165283,
+      "learning_rate": 0.00019956934981068712,
+      "loss": 1.0739,
+      "step": 305
+    },
+    {
+      "epoch": 0.05886879569065025,
+      "grad_norm": 1.9558100700378418,
+      "learning_rate": 0.00019956355409604402,
+      "loss": 1.05,
+      "step": 306
+    },
+    {
+      "epoch": 0.05906117737591381,
+      "grad_norm": 1.8915047645568848,
+      "learning_rate": 0.00019955771972740116,
+      "loss": 1.0549,
+      "step": 307
+    },
+    {
+      "epoch": 0.05925355906117737,
+      "grad_norm": 1.5240460634231567,
+      "learning_rate": 0.0001995518467070236,
+      "loss": 1.044,
+      "step": 308
+    },
+    {
+      "epoch": 0.05944594074644094,
+      "grad_norm": 2.0224764347076416,
+      "learning_rate": 0.00019954593503719153,
+      "loss": 1.0578,
+      "step": 309
+    },
+    {
+      "epoch": 0.0596383224317045,
+      "grad_norm": 2.010066032409668,
+      "learning_rate": 0.00019953998472019995,
+      "loss": 1.13,
+      "step": 310
+    },
+    {
+      "epoch": 0.059830704116968066,
+      "grad_norm": 2.2116637229919434,
+      "learning_rate": 0.0001995339957583591,
+      "loss": 1.1072,
+      "step": 311
+    },
+    {
+      "epoch": 0.060023085802231625,
+      "grad_norm": 0.7816396951675415,
+      "learning_rate": 0.00019952796815399403,
+      "loss": 0.7374,
+      "step": 312
+    },
+    {
+      "epoch": 0.06021546748749519,
+      "grad_norm": 1.952818751335144,
+      "learning_rate": 0.00019952190190944484,
+      "loss": 0.9872,
+      "step": 313
+    },
+    {
+      "epoch": 0.06040784917275875,
+      "grad_norm": 1.946341872215271,
+      "learning_rate": 0.00019951579702706668,
+      "loss": 1.0385,
+      "step": 314
+    },
+    {
+      "epoch": 0.06060023085802232,
+      "grad_norm": 1.6838220357894897,
+      "learning_rate": 0.00019950965350922975,
+      "loss": 1.0586,
+      "step": 315
+    },
+    {
+      "epoch": 0.06079261254328588,
+      "grad_norm": 2.1243534088134766,
+      "learning_rate": 0.00019950347135831907,
+      "loss": 1.0568,
+      "step": 316
+    },
+    {
+      "epoch": 0.060984994228549444,
+      "grad_norm": 1.9448307752609253,
+      "learning_rate": 0.0001994972505767348,
+      "loss": 0.9277,
+      "step": 317
+    },
+    {
+      "epoch": 0.061177375913813004,
+      "grad_norm": 1.0634533166885376,
+      "learning_rate": 0.00019949099116689208,
+      "loss": 1.0151,
+      "step": 318
+    },
+    {
+      "epoch": 0.06136975759907657,
+      "grad_norm": 2.0178449153900146,
+      "learning_rate": 0.000199484693131221,
+      "loss": 1.0835,
+      "step": 319
+    },
+    {
+      "epoch": 0.06156213928434013,
+      "grad_norm": 1.8600071668624878,
+      "learning_rate": 0.00019947835647216668,
+      "loss": 1.1008,
+      "step": 320
+    },
+    {
+      "epoch": 0.061754520969603696,
+      "grad_norm": 1.387091875076294,
+      "learning_rate": 0.00019947198119218924,
+      "loss": 1.0618,
+      "step": 321
+    },
+    {
+      "epoch": 0.061946902654867256,
+      "grad_norm": 1.7352557182312012,
+      "learning_rate": 0.0001994655672937638,
+      "loss": 0.9916,
+      "step": 322
+    },
+    {
+      "epoch": 0.06213928434013082,
+      "grad_norm": 2.0627620220184326,
+      "learning_rate": 0.00019945911477938042,
+      "loss": 1.0509,
+      "step": 323
+    },
+    {
+      "epoch": 0.06233166602539438,
+      "grad_norm": 0.9057157635688782,
+      "learning_rate": 0.0001994526236515442,
+      "loss": 0.7634,
+      "step": 324
+    },
+    {
+      "epoch": 0.06252404771065795,
+      "grad_norm": 0.8613125681877136,
+      "learning_rate": 0.00019944609391277528,
+      "loss": 0.7346,
+      "step": 325
+    },
+    {
+      "epoch": 0.06271642939592151,
+      "grad_norm": 1.6417115926742554,
+      "learning_rate": 0.0001994395255656086,
+      "loss": 1.0101,
+      "step": 326
+    },
+    {
+      "epoch": 0.06290881108118507,
+      "grad_norm": 1.5522997379302979,
+      "learning_rate": 0.00019943291861259431,
+      "loss": 0.9407,
+      "step": 327
+    },
+    {
+      "epoch": 0.06310119276644863,
+      "grad_norm": 1.7593917846679688,
+      "learning_rate": 0.00019942627305629746,
+      "loss": 1.1029,
+      "step": 328
+    },
+    {
+      "epoch": 0.0632935744517122,
+      "grad_norm": 2.1477913856506348,
+      "learning_rate": 0.00019941958889929806,
+      "loss": 1.0659,
+      "step": 329
+    },
+    {
+      "epoch": 0.06348595613697576,
+      "grad_norm": 2.273244619369507,
+      "learning_rate": 0.0001994128661441911,
+      "loss": 1.0821,
+      "step": 330
+    },
+    {
+      "epoch": 0.06367833782223932,
+      "grad_norm": 2.5667707920074463,
+      "learning_rate": 0.00019940610479358669,
+      "loss": 1.0065,
+      "step": 331
+    },
+    {
+      "epoch": 0.06387071950750288,
+      "grad_norm": 2.6487419605255127,
+      "learning_rate": 0.00019939930485010968,
+      "loss": 1.1011,
+      "step": 332
+    },
+    {
+      "epoch": 0.06406310119276645,
+      "grad_norm": 2.4940288066864014,
+      "learning_rate": 0.00019939246631640013,
+      "loss": 1.0727,
+      "step": 333
+    },
+    {
+      "epoch": 0.06425548287803001,
+      "grad_norm": 1.2685796022415161,
+      "learning_rate": 0.00019938558919511298,
+      "loss": 1.0245,
+      "step": 334
+    },
+    {
+      "epoch": 0.06444786456329357,
+      "grad_norm": 1.8175230026245117,
+      "learning_rate": 0.00019937867348891815,
+      "loss": 0.9901,
+      "step": 335
+    },
+    {
+      "epoch": 0.06464024624855713,
+      "grad_norm": 1.7727982997894287,
+      "learning_rate": 0.00019937171920050056,
+      "loss": 1.0789,
+      "step": 336
+    },
+    {
+      "epoch": 0.0648326279338207,
+      "grad_norm": 1.0619704723358154,
+      "learning_rate": 0.00019936472633256012,
+      "loss": 0.7211,
+      "step": 337
+    },
+    {
+      "epoch": 0.06502500961908426,
+      "grad_norm": 1.4708467721939087,
+      "learning_rate": 0.00019935769488781168,
+      "loss": 1.0666,
+      "step": 338
+    },
+    {
+      "epoch": 0.06521739130434782,
+      "grad_norm": 2.2777631282806396,
+      "learning_rate": 0.0001993506248689851,
+      "loss": 1.0412,
+      "step": 339
+    },
+    {
+      "epoch": 0.06540977298961138,
+      "grad_norm": 2.392672061920166,
+      "learning_rate": 0.0001993435162788252,
+      "loss": 1.1575,
+      "step": 340
+    },
+    {
+      "epoch": 0.06560215467487496,
+      "grad_norm": 2.0187485218048096,
+      "learning_rate": 0.0001993363691200918,
+      "loss": 1.1453,
+      "step": 341
+    },
+    {
+      "epoch": 0.06579453636013852,
+      "grad_norm": 2.273244619369507,
+      "learning_rate": 0.00019932918339555963,
+      "loss": 1.0649,
+      "step": 342
+    },
+    {
+      "epoch": 0.06598691804540208,
+      "grad_norm": 2.070495128631592,
+      "learning_rate": 0.00019932195910801848,
+      "loss": 1.1678,
+      "step": 343
+    },
+    {
+      "epoch": 0.06617929973066564,
+      "grad_norm": 2.0118939876556396,
+      "learning_rate": 0.00019931469626027304,
+      "loss": 1.0858,
+      "step": 344
+    },
+    {
+      "epoch": 0.06637168141592921,
+      "grad_norm": 1.4493905305862427,
+      "learning_rate": 0.00019930739485514302,
+      "loss": 1.0984,
+      "step": 345
+    },
+    {
+      "epoch": 0.06656406310119277,
+      "grad_norm": 1.6027426719665527,
+      "learning_rate": 0.00019930005489546306,
+      "loss": 1.0658,
+      "step": 346
+    },
+    {
+      "epoch": 0.06675644478645633,
+      "grad_norm": 2.2023303508758545,
+      "learning_rate": 0.00019929267638408278,
+      "loss": 1.0352,
+      "step": 347
+    },
+    {
+      "epoch": 0.06694882647171989,
+      "grad_norm": 1.387244462966919,
+      "learning_rate": 0.00019928525932386677,
+      "loss": 0.769,
+      "step": 348
+    },
+    {
+      "epoch": 0.06714120815698346,
+      "grad_norm": 0.9762587547302246,
+      "learning_rate": 0.00019927780371769462,
+      "loss": 0.7309,
+      "step": 349
+    },
+    {
+      "epoch": 0.06733358984224702,
+      "grad_norm": 1.9229068756103516,
+      "learning_rate": 0.00019927030956846083,
+      "loss": 1.0224,
+      "step": 350
+    },
+    {
+      "epoch": 0.06752597152751058,
+      "grad_norm": 2.1374595165252686,
+      "learning_rate": 0.0001992627768790749,
+      "loss": 1.0039,
+      "step": 351
+    },
+    {
+      "epoch": 0.06771835321277414,
+      "grad_norm": 2.0982937812805176,
+      "learning_rate": 0.00019925520565246122,
+      "loss": 1.0848,
+      "step": 352
+    },
+    {
+      "epoch": 0.0679107348980377,
+      "grad_norm": 2.1090385913848877,
+      "learning_rate": 0.0001992475958915593,
+      "loss": 1.1081,
+      "step": 353
+    },
+    {
+      "epoch": 0.06810311658330127,
+      "grad_norm": 1.7833540439605713,
+      "learning_rate": 0.00019923994759932343,
+      "loss": 1.0203,
+      "step": 354
+    },
+    {
+      "epoch": 0.06829549826856483,
+      "grad_norm": 0.8505796194076538,
+      "learning_rate": 0.00019923226077872298,
+      "loss": 1.1049,
+      "step": 355
+    },
+    {
+      "epoch": 0.06848787995382839,
+      "grad_norm": 1.7475645542144775,
+      "learning_rate": 0.0001992245354327422,
+      "loss": 1.0787,
+      "step": 356
+    },
+    {
+      "epoch": 0.06868026163909195,
+      "grad_norm": 1.7338604927062988,
+      "learning_rate": 0.00019921677156438045,
+      "loss": 1.1016,
+      "step": 357
+    },
+    {
+      "epoch": 0.06887264332435553,
+      "grad_norm": 2.1728498935699463,
+      "learning_rate": 0.00019920896917665178,
+      "loss": 1.0689,
+      "step": 358
+    },
+    {
+      "epoch": 0.06906502500961909,
+      "grad_norm": 2.3959460258483887,
+      "learning_rate": 0.0001992011282725854,
+      "loss": 1.0672,
+      "step": 359
+    },
+    {
+      "epoch": 0.06925740669488265,
+      "grad_norm": 2.207045078277588,
+      "learning_rate": 0.00019919324885522545,
+      "loss": 1.1243,
+      "step": 360
+    },
+    {
+      "epoch": 0.0694497883801462,
+      "grad_norm": 1.7654908895492554,
+      "learning_rate": 0.000199185330927631,
+      "loss": 1.0086,
+      "step": 361
+    },
+    {
+      "epoch": 0.06964217006540978,
+      "grad_norm": 2.088961601257324,
+      "learning_rate": 0.00019917737449287598,
+      "loss": 1.0017,
+      "step": 362
+    },
+    {
+      "epoch": 0.06983455175067334,
+      "grad_norm": 1.998194694519043,
+      "learning_rate": 0.00019916937955404942,
+      "loss": 1.0954,
+      "step": 363
+    },
+    {
+      "epoch": 0.0700269334359369,
+      "grad_norm": 0.9245341420173645,
+      "learning_rate": 0.00019916134611425523,
+      "loss": 0.7132,
+      "step": 364
+    },
+    {
+      "epoch": 0.07021931512120046,
+      "grad_norm": 1.3984133005142212,
+      "learning_rate": 0.00019915327417661225,
+      "loss": 1.0182,
+      "step": 365
+    },
+    {
+      "epoch": 0.07041169680646403,
+      "grad_norm": 2.1753625869750977,
+      "learning_rate": 0.00019914516374425429,
+      "loss": 0.9751,
+      "step": 366
+    },
+    {
+      "epoch": 0.07060407849172759,
+      "grad_norm": 0.8639774918556213,
+      "learning_rate": 0.00019913701482033008,
+      "loss": 0.7411,
+      "step": 367
+    },
+    {
+      "epoch": 0.07079646017699115,
+      "grad_norm": 1.8092215061187744,
+      "learning_rate": 0.00019912882740800333,
+      "loss": 0.9889,
+      "step": 368
+    },
+    {
+      "epoch": 0.07098884186225471,
+      "grad_norm": 2.333197832107544,
+      "learning_rate": 0.0001991206015104527,
+      "loss": 1.0429,
+      "step": 369
+    },
+    {
+      "epoch": 0.07118122354751828,
+      "grad_norm": 2.335036039352417,
+      "learning_rate": 0.0001991123371308717,
+      "loss": 1.107,
+      "step": 370
+    },
+    {
+      "epoch": 0.07137360523278184,
+      "grad_norm": 1.9960662126541138,
+      "learning_rate": 0.00019910403427246895,
+      "loss": 0.9865,
+      "step": 371
+    },
+    {
+      "epoch": 0.0715659869180454,
+      "grad_norm": 0.8282999992370605,
+      "learning_rate": 0.0001990956929384678,
+      "loss": 0.7326,
+      "step": 372
+    },
+    {
+      "epoch": 0.07175836860330896,
+      "grad_norm": 1.3194504976272583,
+      "learning_rate": 0.0001990873131321067,
+      "loss": 1.1612,
+      "step": 373
+    },
+    {
+      "epoch": 0.07195075028857253,
+      "grad_norm": 1.6644527912139893,
+      "learning_rate": 0.00019907889485663895,
+      "loss": 1.0513,
+      "step": 374
+    },
+    {
+      "epoch": 0.0721431319738361,
+      "grad_norm": 2.3779428005218506,
+      "learning_rate": 0.00019907043811533283,
+      "loss": 1.0809,
+      "step": 375
+    },
+    {
+      "epoch": 0.07233551365909965,
+      "grad_norm": 2.166625499725342,
+      "learning_rate": 0.00019906194291147156,
+      "loss": 1.1015,
+      "step": 376
+    },
+    {
+      "epoch": 0.07252789534436321,
+      "grad_norm": 1.9422602653503418,
+      "learning_rate": 0.00019905340924835323,
+      "loss": 1.0105,
+      "step": 377
+    },
+    {
+      "epoch": 0.07272027702962677,
+      "grad_norm": 1.9412823915481567,
+      "learning_rate": 0.00019904483712929094,
+      "loss": 1.0298,
+      "step": 378
+    },
+    {
+      "epoch": 0.07291265871489035,
+      "grad_norm": 2.030547618865967,
+      "learning_rate": 0.00019903622655761266,
+      "loss": 1.1065,
+      "step": 379
+    },
+    {
+      "epoch": 0.0731050404001539,
+      "grad_norm": 1.6042516231536865,
+      "learning_rate": 0.0001990275775366613,
+      "loss": 1.0748,
+      "step": 380
+    },
+    {
+      "epoch": 0.07329742208541747,
+      "grad_norm": 2.295745611190796,
+      "learning_rate": 0.00019901889006979472,
+      "loss": 1.0839,
+      "step": 381
+    },
+    {
+      "epoch": 0.07348980377068103,
+      "grad_norm": 1.9287062883377075,
+      "learning_rate": 0.0001990101641603857,
+      "loss": 1.0441,
+      "step": 382
+    },
+    {
+      "epoch": 0.0736821854559446,
+      "grad_norm": 1.8921302556991577,
+      "learning_rate": 0.00019900139981182194,
+      "loss": 1.1143,
+      "step": 383
+    },
+    {
+      "epoch": 0.07387456714120816,
+      "grad_norm": 0.8464456796646118,
+      "learning_rate": 0.00019899259702750604,
+      "loss": 0.6867,
+      "step": 384
+    },
+    {
+      "epoch": 0.07406694882647172,
+      "grad_norm": 1.5495660305023193,
+      "learning_rate": 0.00019898375581085557,
+      "loss": 0.9989,
+      "step": 385
+    },
+    {
+      "epoch": 0.07425933051173528,
+      "grad_norm": 1.8430824279785156,
+      "learning_rate": 0.00019897487616530297,
+      "loss": 1.0,
+      "step": 386
+    },
+    {
+      "epoch": 0.07445171219699885,
+      "grad_norm": 1.5126748085021973,
+      "learning_rate": 0.00019896595809429566,
+      "loss": 1.1011,
+      "step": 387
+    },
+    {
+      "epoch": 0.07464409388226241,
+      "grad_norm": 1.7694623470306396,
+      "learning_rate": 0.0001989570016012959,
+      "loss": 1.0354,
+      "step": 388
+    },
+    {
+      "epoch": 0.07483647556752597,
+      "grad_norm": 1.6930314302444458,
+      "learning_rate": 0.00019894800668978095,
+      "loss": 1.0855,
+      "step": 389
+    },
+    {
+      "epoch": 0.07502885725278953,
+      "grad_norm": 1.7760292291641235,
+      "learning_rate": 0.0001989389733632429,
+      "loss": 0.9783,
+      "step": 390
+    },
+    {
+      "epoch": 0.0752212389380531,
+      "grad_norm": 1.683892011642456,
+      "learning_rate": 0.00019892990162518882,
+      "loss": 1.012,
+      "step": 391
+    },
+    {
+      "epoch": 0.07541362062331666,
+      "grad_norm": 2.0871386528015137,
+      "learning_rate": 0.00019892079147914072,
+      "loss": 1.0304,
+      "step": 392
+    },
+    {
+      "epoch": 0.07560600230858022,
+      "grad_norm": 1.811582088470459,
+      "learning_rate": 0.00019891164292863537,
+      "loss": 1.0488,
+      "step": 393
+    },
+    {
+      "epoch": 0.07579838399384378,
+      "grad_norm": 1.763335943222046,
+      "learning_rate": 0.00019890245597722463,
+      "loss": 1.0343,
+      "step": 394
+    },
+    {
+      "epoch": 0.07599076567910736,
+      "grad_norm": 1.3949939012527466,
+      "learning_rate": 0.00019889323062847515,
+      "loss": 1.0725,
+      "step": 395
+    },
+    {
+      "epoch": 0.07618314736437092,
+      "grad_norm": 2.1831634044647217,
+      "learning_rate": 0.0001988839668859686,
+      "loss": 1.0842,
+      "step": 396
+    },
+    {
+      "epoch": 0.07637552904963447,
+      "grad_norm": 2.1750335693359375,
+      "learning_rate": 0.00019887466475330138,
+      "loss": 1.0125,
+      "step": 397
+    },
+    {
+      "epoch": 0.07656791073489803,
+      "grad_norm": 1.977915644645691,
+      "learning_rate": 0.00019886532423408495,
+      "loss": 1.0936,
+      "step": 398
+    },
+    {
+      "epoch": 0.0767602924201616,
+      "grad_norm": 2.3975865840911865,
+      "learning_rate": 0.00019885594533194562,
+      "loss": 1.1082,
+      "step": 399
+    },
+    {
+      "epoch": 0.07695267410542517,
+      "grad_norm": 1.853113055229187,
+      "learning_rate": 0.00019884652805052464,
+      "loss": 1.0599,
+      "step": 400
+    },
+    {
+      "epoch": 0.07714505579068873,
+      "grad_norm": 1.9065228700637817,
+      "learning_rate": 0.00019883707239347803,
+      "loss": 1.053,
+      "step": 401
+    },
+    {
+      "epoch": 0.07733743747595229,
+      "grad_norm": 1.8549379110336304,
+      "learning_rate": 0.0001988275783644769,
+      "loss": 1.0485,
+      "step": 402
+    },
+    {
+      "epoch": 0.07752981916121585,
+      "grad_norm": 2.1358065605163574,
+      "learning_rate": 0.0001988180459672071,
+      "loss": 1.0008,
+      "step": 403
+    },
+    {
+      "epoch": 0.07772220084647942,
+      "grad_norm": 2.0919485092163086,
+      "learning_rate": 0.00019880847520536946,
+      "loss": 1.0981,
+      "step": 404
+    },
+    {
+      "epoch": 0.07791458253174298,
+      "grad_norm": 1.814460277557373,
+      "learning_rate": 0.00019879886608267967,
+      "loss": 0.9917,
+      "step": 405
+    },
+    {
+      "epoch": 0.07810696421700654,
+      "grad_norm": 1.9914829730987549,
+      "learning_rate": 0.0001987892186028683,
+      "loss": 1.1255,
+      "step": 406
+    },
+    {
+      "epoch": 0.0782993459022701,
+      "grad_norm": 2.201082229614258,
+      "learning_rate": 0.00019877953276968086,
+      "loss": 1.0664,
+      "step": 407
+    },
+    {
+      "epoch": 0.07849172758753367,
+      "grad_norm": 2.0426323413848877,
+      "learning_rate": 0.00019876980858687775,
+      "loss": 0.977,
+      "step": 408
+    },
+    {
+      "epoch": 0.07868410927279723,
+      "grad_norm": 1.772365689277649,
+      "learning_rate": 0.00019876004605823419,
+      "loss": 1.0554,
+      "step": 409
+    },
+    {
+      "epoch": 0.07887649095806079,
+      "grad_norm": 2.6195316314697266,
+      "learning_rate": 0.0001987502451875403,
+      "loss": 1.0504,
+      "step": 410
+    },
+    {
+      "epoch": 0.07906887264332435,
+      "grad_norm": 2.27302622795105,
+      "learning_rate": 0.0001987404059786012,
+      "loss": 1.1081,
+      "step": 411
+    },
+    {
+      "epoch": 0.07926125432858792,
+      "grad_norm": 1.7451651096343994,
+      "learning_rate": 0.00019873052843523677,
+      "loss": 1.0095,
+      "step": 412
+    },
+    {
+      "epoch": 0.07945363601385148,
+      "grad_norm": 2.4049737453460693,
+      "learning_rate": 0.0001987206125612818,
+      "loss": 1.0828,
+      "step": 413
+    },
+    {
+      "epoch": 0.07964601769911504,
+      "grad_norm": 2.3564841747283936,
+      "learning_rate": 0.00019871065836058596,
+      "loss": 1.0717,
+      "step": 414
+    },
+    {
+      "epoch": 0.0798383993843786,
+      "grad_norm": 2.3972725868225098,
+      "learning_rate": 0.00019870066583701387,
+      "loss": 1.041,
+      "step": 415
+    },
+    {
+      "epoch": 0.08003078106964218,
+      "grad_norm": 1.932621717453003,
+      "learning_rate": 0.00019869063499444495,
+      "loss": 1.0603,
+      "step": 416
+    },
+    {
+      "epoch": 0.08022316275490574,
+      "grad_norm": 2.1673390865325928,
+      "learning_rate": 0.00019868056583677345,
+      "loss": 1.0273,
+      "step": 417
+    },
+    {
+      "epoch": 0.0804155444401693,
+      "grad_norm": 2.1673779487609863,
+      "learning_rate": 0.00019867045836790867,
+      "loss": 1.02,
+      "step": 418
+    },
+    {
+      "epoch": 0.08060792612543285,
+      "grad_norm": 0.8491818308830261,
+      "learning_rate": 0.00019866031259177462,
+      "loss": 0.7233,
+      "step": 419
+    },
+    {
+      "epoch": 0.08080030781069643,
+      "grad_norm": 2.090510129928589,
+      "learning_rate": 0.00019865012851231023,
+      "loss": 1.1273,
+      "step": 420
+    },
+    {
+      "epoch": 0.08099268949595999,
+      "grad_norm": 1.9813342094421387,
+      "learning_rate": 0.00019863990613346933,
+      "loss": 1.0491,
+      "step": 421
+    },
+    {
+      "epoch": 0.08118507118122355,
+      "grad_norm": 1.6830374002456665,
+      "learning_rate": 0.0001986296454592206,
+      "loss": 1.0697,
+      "step": 422
+    },
+    {
+      "epoch": 0.08137745286648711,
+      "grad_norm": 2.0291988849639893,
+      "learning_rate": 0.00019861934649354762,
+      "loss": 0.9799,
+      "step": 423
+    },
+    {
+      "epoch": 0.08156983455175067,
+      "grad_norm": 2.0100879669189453,
+      "learning_rate": 0.00019860900924044872,
+      "loss": 1.0038,
+      "step": 424
+    },
+    {
+      "epoch": 0.08176221623701424,
+      "grad_norm": 0.6497704982757568,
+      "learning_rate": 0.00019859863370393725,
+      "loss": 0.7603,
+      "step": 425
+    },
+    {
+      "epoch": 0.0819545979222778,
+      "grad_norm": 1.3792794942855835,
+      "learning_rate": 0.00019858821988804132,
+      "loss": 1.0063,
+      "step": 426
+    },
+    {
+      "epoch": 0.08214697960754136,
+      "grad_norm": 2.618529796600342,
+      "learning_rate": 0.00019857776779680394,
+      "loss": 1.097,
+      "step": 427
+    },
+    {
+      "epoch": 0.08233936129280492,
+      "grad_norm": 2.5826332569122314,
+      "learning_rate": 0.00019856727743428296,
+      "loss": 1.1257,
+      "step": 428
+    },
+    {
+      "epoch": 0.08253174297806849,
+      "grad_norm": 1.876307725906372,
+      "learning_rate": 0.00019855674880455112,
+      "loss": 1.0493,
+      "step": 429
+    },
+    {
+      "epoch": 0.08272412466333205,
+      "grad_norm": 1.8525056838989258,
+      "learning_rate": 0.00019854618191169599,
+      "loss": 1.0434,
+      "step": 430
+    },
+    {
+      "epoch": 0.08291650634859561,
+      "grad_norm": 2.0910542011260986,
+      "learning_rate": 0.00019853557675982,
+      "loss": 1.0931,
+      "step": 431
+    },
+    {
+      "epoch": 0.08310888803385917,
+      "grad_norm": 2.1633503437042236,
+      "learning_rate": 0.0001985249333530404,
+      "loss": 1.0992,
+      "step": 432
+    },
+    {
+      "epoch": 0.08330126971912274,
+      "grad_norm": 2.1211047172546387,
+      "learning_rate": 0.00019851425169548938,
+      "loss": 1.0878,
+      "step": 433
+    },
+    {
+      "epoch": 0.0834936514043863,
+      "grad_norm": 2.041811227798462,
+      "learning_rate": 0.0001985035317913139,
+      "loss": 1.0296,
+      "step": 434
+    },
+    {
+      "epoch": 0.08368603308964986,
+      "grad_norm": 2.3062193393707275,
+      "learning_rate": 0.0001984927736446758,
+      "loss": 1.1134,
+      "step": 435
+    },
+    {
+      "epoch": 0.08387841477491342,
+      "grad_norm": 0.9969825744628906,
+      "learning_rate": 0.0001984819772597518,
+      "loss": 0.7644,
+      "step": 436
+    },
+    {
+      "epoch": 0.084070796460177,
+      "grad_norm": 1.7924903631210327,
+      "learning_rate": 0.00019847114264073338,
+      "loss": 1.0151,
+      "step": 437
+    },
+    {
+      "epoch": 0.08426317814544056,
+      "grad_norm": 1.687875509262085,
+      "learning_rate": 0.0001984602697918269,
+      "loss": 1.0932,
+      "step": 438
+    },
+    {
+      "epoch": 0.08445555983070412,
+      "grad_norm": 1.5758436918258667,
+      "learning_rate": 0.0001984493587172536,
+      "loss": 1.0546,
+      "step": 439
+    },
+    {
+      "epoch": 0.08464794151596768,
+      "grad_norm": 2.164842128753662,
+      "learning_rate": 0.00019843840942124956,
+      "loss": 1.1105,
+      "step": 440
+    },
+    {
+      "epoch": 0.08484032320123125,
+      "grad_norm": 1.5914433002471924,
+      "learning_rate": 0.00019842742190806566,
+      "loss": 1.0402,
+      "step": 441
+    },
+    {
+      "epoch": 0.08503270488649481,
+      "grad_norm": 1.9680742025375366,
+      "learning_rate": 0.00019841639618196758,
+      "loss": 1.1226,
+      "step": 442
+    },
+    {
+      "epoch": 0.08522508657175837,
+      "grad_norm": 1.9345786571502686,
+      "learning_rate": 0.00019840533224723593,
+      "loss": 1.0541,
+      "step": 443
+    },
+    {
+      "epoch": 0.08541746825702193,
+      "grad_norm": 1.9476826190948486,
+      "learning_rate": 0.00019839423010816615,
+      "loss": 1.0506,
+      "step": 444
+    },
+    {
+      "epoch": 0.08560984994228549,
+      "grad_norm": 1.760840892791748,
+      "learning_rate": 0.0001983830897690684,
+      "loss": 0.9935,
+      "step": 445
+    },
+    {
+      "epoch": 0.08580223162754906,
+      "grad_norm": 2.0714244842529297,
+      "learning_rate": 0.00019837191123426776,
+      "loss": 1.1086,
+      "step": 446
+    },
+    {
+      "epoch": 0.08599461331281262,
+      "grad_norm": 2.367671012878418,
+      "learning_rate": 0.00019836069450810418,
+      "loss": 1.1061,
+      "step": 447
+    },
+    {
+      "epoch": 0.08618699499807618,
+      "grad_norm": 1.7732300758361816,
+      "learning_rate": 0.0001983494395949323,
+      "loss": 1.0722,
+      "step": 448
+    },
+    {
+      "epoch": 0.08637937668333974,
+      "grad_norm": 2.557396411895752,
+      "learning_rate": 0.0001983381464991217,
+      "loss": 1.1235,
+      "step": 449
+    },
+    {
+      "epoch": 0.08657175836860331,
+      "grad_norm": 0.9706582427024841,
+      "learning_rate": 0.00019832681522505676,
+      "loss": 1.0483,
+      "step": 450
+    },
+    {
+      "epoch": 0.08676414005386687,
+      "grad_norm": 1.8328685760498047,
+      "learning_rate": 0.00019831544577713663,
+      "loss": 1.0022,
+      "step": 451
+    },
+    {
+      "epoch": 0.08695652173913043,
+      "grad_norm": 1.5880244970321655,
+      "learning_rate": 0.0001983040381597754,
+      "loss": 1.0455,
+      "step": 452
+    },
+    {
+      "epoch": 0.08714890342439399,
+      "grad_norm": 1.938947081565857,
+      "learning_rate": 0.00019829259237740177,
+      "loss": 1.0882,
+      "step": 453
+    },
+    {
+      "epoch": 0.08734128510965757,
+      "grad_norm": 0.9265813827514648,
+      "learning_rate": 0.00019828110843445954,
+      "loss": 0.7089,
+      "step": 454
+    },
+    {
+      "epoch": 0.08753366679492113,
+      "grad_norm": 0.8548473119735718,
+      "learning_rate": 0.00019826958633540706,
+      "loss": 0.7598,
+      "step": 455
+    },
+    {
+      "epoch": 0.08772604848018468,
+      "grad_norm": 2.152094602584839,
+      "learning_rate": 0.00019825802608471765,
+      "loss": 1.0565,
+      "step": 456
+    },
+    {
+      "epoch": 0.08791843016544824,
+      "grad_norm": 1.98738694190979,
+      "learning_rate": 0.0001982464276868794,
+      "loss": 1.0039,
+      "step": 457
+    },
+    {
+      "epoch": 0.08811081185071182,
+      "grad_norm": 2.5002996921539307,
+      "learning_rate": 0.0001982347911463952,
+      "loss": 1.0867,
+      "step": 458
+    },
+    {
+      "epoch": 0.08830319353597538,
+      "grad_norm": 2.2744555473327637,
+      "learning_rate": 0.00019822311646778277,
+      "loss": 1.1631,
+      "step": 459
+    },
+    {
+      "epoch": 0.08849557522123894,
+      "grad_norm": 1.7732540369033813,
+      "learning_rate": 0.00019821140365557457,
+      "loss": 1.1161,
+      "step": 460
+    },
+    {
+      "epoch": 0.0886879569065025,
+      "grad_norm": 2.2270076274871826,
+      "learning_rate": 0.00019819965271431796,
+      "loss": 1.0572,
+      "step": 461
+    },
+    {
+      "epoch": 0.08888033859176607,
+      "grad_norm": 1.7358040809631348,
+      "learning_rate": 0.00019818786364857505,
+      "loss": 1.0703,
+      "step": 462
+    },
+    {
+      "epoch": 0.08907272027702963,
+      "grad_norm": 2.114112615585327,
+      "learning_rate": 0.00019817603646292276,
+      "loss": 1.0393,
+      "step": 463
+    },
+    {
+      "epoch": 0.08926510196229319,
+      "grad_norm": 1.7979570627212524,
+      "learning_rate": 0.00019816417116195285,
+      "loss": 1.0212,
+      "step": 464
+    },
+    {
+      "epoch": 0.08945748364755675,
+      "grad_norm": 1.3968870639801025,
+      "learning_rate": 0.0001981522677502718,
+      "loss": 1.045,
+      "step": 465
+    },
+    {
+      "epoch": 0.08964986533282032,
+      "grad_norm": 0.9796537756919861,
+      "learning_rate": 0.00019814032623250092,
+      "loss": 0.6854,
+      "step": 466
+    },
+    {
+      "epoch": 0.08984224701808388,
+      "grad_norm": 1.974333643913269,
+      "learning_rate": 0.0001981283466132763,
+      "loss": 1.0738,
+      "step": 467
+    },
+    {
+      "epoch": 0.09003462870334744,
+      "grad_norm": 2.404754161834717,
+      "learning_rate": 0.00019811632889724887,
+      "loss": 1.0386,
+      "step": 468
+    },
+    {
+      "epoch": 0.090227010388611,
+      "grad_norm": 2.1973717212677,
+      "learning_rate": 0.00019810427308908436,
+      "loss": 0.9929,
+      "step": 469
+    },
+    {
+      "epoch": 0.09041939207387456,
+      "grad_norm": 2.124910354614258,
+      "learning_rate": 0.00019809217919346315,
+      "loss": 1.0659,
+      "step": 470
+    },
+    {
+      "epoch": 0.09061177375913813,
+      "grad_norm": 2.4899840354919434,
+      "learning_rate": 0.0001980800472150806,
+      "loss": 1.0622,
+      "step": 471
+    },
+    {
+      "epoch": 0.0908041554444017,
+      "grad_norm": 2.5782594680786133,
+      "learning_rate": 0.00019806787715864673,
+      "loss": 1.0561,
+      "step": 472
+    },
+    {
+      "epoch": 0.09099653712966525,
+      "grad_norm": 2.032315254211426,
+      "learning_rate": 0.00019805566902888634,
+      "loss": 1.0293,
+      "step": 473
+    },
+    {
+      "epoch": 0.09118891881492881,
+      "grad_norm": 1.196146011352539,
+      "learning_rate": 0.00019804342283053915,
+      "loss": 1.0995,
+      "step": 474
+    },
+    {
+      "epoch": 0.09138130050019239,
+      "grad_norm": 1.9662435054779053,
+      "learning_rate": 0.00019803113856835938,
+      "loss": 0.9634,
+      "step": 475
+    },
+    {
+      "epoch": 0.09157368218545595,
+      "grad_norm": 1.663698673248291,
+      "learning_rate": 0.00019801881624711637,
+      "loss": 1.0654,
+      "step": 476
+    },
+    {
+      "epoch": 0.0917660638707195,
+      "grad_norm": 2.0356335639953613,
+      "learning_rate": 0.000198006455871594,
+      "loss": 0.9793,
+      "step": 477
+    },
+    {
+      "epoch": 0.09195844555598306,
+      "grad_norm": 2.205167531967163,
+      "learning_rate": 0.000197994057446591,
+      "loss": 1.0688,
+      "step": 478
+    },
+    {
+      "epoch": 0.09215082724124664,
+      "grad_norm": 2.5868940353393555,
+      "learning_rate": 0.00019798162097692089,
+      "loss": 1.0625,
+      "step": 479
+    },
+    {
+      "epoch": 0.0923432089265102,
+      "grad_norm": 2.0621633529663086,
+      "learning_rate": 0.00019796914646741186,
+      "loss": 1.0844,
+      "step": 480
+    },
+    {
+      "epoch": 0.09253559061177376,
+      "grad_norm": 2.055061101913452,
+      "learning_rate": 0.000197956633922907,
+      "loss": 1.0663,
+      "step": 481
+    },
+    {
+      "epoch": 0.09272797229703732,
+      "grad_norm": 0.9432541131973267,
+      "learning_rate": 0.00019794408334826414,
+      "loss": 0.721,
+      "step": 482
+    },
+    {
+      "epoch": 0.09292035398230089,
+      "grad_norm": 2.1164863109588623,
+      "learning_rate": 0.0001979314947483558,
+      "loss": 1.0727,
+      "step": 483
+    },
+    {
+      "epoch": 0.09311273566756445,
+      "grad_norm": 1.7463334798812866,
+      "learning_rate": 0.0001979188681280693,
+      "loss": 1.0689,
+      "step": 484
+    },
+    {
+      "epoch": 0.09330511735282801,
+      "grad_norm": 2.430405378341675,
+      "learning_rate": 0.00019790620349230674,
+      "loss": 1.012,
+      "step": 485
+    },
+    {
+      "epoch": 0.09349749903809157,
+      "grad_norm": 0.8223225474357605,
+      "learning_rate": 0.000197893500845985,
+      "loss": 0.7111,
+      "step": 486
+    },
+    {
+      "epoch": 0.09368988072335514,
+      "grad_norm": 2.442974805831909,
+      "learning_rate": 0.00019788076019403565,
+      "loss": 1.086,
+      "step": 487
+    },
+    {
+      "epoch": 0.0938822624086187,
+      "grad_norm": 1.6417433023452759,
+      "learning_rate": 0.00019786798154140507,
+      "loss": 1.0305,
+      "step": 488
+    },
+    {
+      "epoch": 0.09407464409388226,
+      "grad_norm": 2.0929977893829346,
+      "learning_rate": 0.00019785516489305435,
+      "loss": 1.0311,
+      "step": 489
+    },
+    {
+      "epoch": 0.09426702577914582,
+      "grad_norm": 2.137655019760132,
+      "learning_rate": 0.00019784231025395937,
+      "loss": 1.0884,
+      "step": 490
+    },
+    {
+      "epoch": 0.0944594074644094,
+      "grad_norm": 1.1784093379974365,
+      "learning_rate": 0.00019782941762911074,
+      "loss": 1.0926,
+      "step": 491
+    },
+    {
+      "epoch": 0.09465178914967295,
+      "grad_norm": 1.9071825742721558,
+      "learning_rate": 0.00019781648702351382,
+      "loss": 1.0472,
+      "step": 492
+    },
+    {
+      "epoch": 0.09484417083493651,
+      "grad_norm": 1.8211740255355835,
+      "learning_rate": 0.0001978035184421887,
+      "loss": 1.1071,
+      "step": 493
+    },
+    {
+      "epoch": 0.09503655252020007,
+      "grad_norm": 1.724429965019226,
+      "learning_rate": 0.0001977905118901703,
+      "loss": 1.0749,
+      "step": 494
+    },
+    {
+      "epoch": 0.09522893420546363,
+      "grad_norm": 0.7985373139381409,
+      "learning_rate": 0.00019777746737250807,
+      "loss": 1.0265,
+      "step": 495
+    },
+    {
+      "epoch": 0.0954213158907272,
+      "grad_norm": 2.027189016342163,
+      "learning_rate": 0.0001977643848942665,
+      "loss": 1.0808,
+      "step": 496
+    },
+    {
+      "epoch": 0.09561369757599077,
+      "grad_norm": 2.0485899448394775,
+      "learning_rate": 0.00019775126446052458,
+      "loss": 1.0419,
+      "step": 497
+    },
+    {
+      "epoch": 0.09580607926125433,
+      "grad_norm": 0.82000732421875,
+      "learning_rate": 0.0001977381060763761,
+      "loss": 0.7208,
+      "step": 498
+    },
+    {
+      "epoch": 0.09599846094651789,
+      "grad_norm": 1.9050426483154297,
+      "learning_rate": 0.00019772490974692962,
+      "loss": 1.0436,
+      "step": 499
+    },
+    {
+      "epoch": 0.09619084263178146,
+      "grad_norm": 1.7683523893356323,
+      "learning_rate": 0.00019771167547730844,
+      "loss": 1.0729,
+      "step": 500
+    },
+    {
+      "epoch": 0.09638322431704502,
+      "grad_norm": 2.499324083328247,
+      "learning_rate": 0.0001976984032726505,
+      "loss": 1.0409,
+      "step": 501
+    },
+    {
+      "epoch": 0.09657560600230858,
+      "grad_norm": 2.4197275638580322,
+      "learning_rate": 0.0001976850931381086,
+      "loss": 1.062,
+      "step": 502
+    },
+    {
+      "epoch": 0.09676798768757214,
+      "grad_norm": 2.1558237075805664,
+      "learning_rate": 0.00019767174507885008,
+      "loss": 1.0431,
+      "step": 503
+    },
+    {
+      "epoch": 0.09696036937283571,
+      "grad_norm": 2.250523805618286,
+      "learning_rate": 0.00019765835910005722,
+      "loss": 1.1185,
+      "step": 504
+    },
+    {
+      "epoch": 0.09715275105809927,
+      "grad_norm": 2.14961838722229,
+      "learning_rate": 0.00019764493520692687,
+      "loss": 1.0216,
+      "step": 505
+    },
+    {
+      "epoch": 0.09734513274336283,
+      "grad_norm": 2.312534809112549,
+      "learning_rate": 0.00019763147340467067,
+      "loss": 1.0686,
+      "step": 506
+    },
+    {
+      "epoch": 0.09753751442862639,
+      "grad_norm": 2.0099306106567383,
+      "learning_rate": 0.00019761797369851497,
+      "loss": 1.0289,
+      "step": 507
+    },
+    {
+      "epoch": 0.09772989611388996,
+      "grad_norm": 2.170553207397461,
+      "learning_rate": 0.00019760443609370074,
+      "loss": 1.0073,
+      "step": 508
+    },
+    {
+      "epoch": 0.09792227779915352,
+      "grad_norm": 1.0866827964782715,
+      "learning_rate": 0.00019759086059548383,
+      "loss": 0.753,
+      "step": 509
+    },
+    {
+      "epoch": 0.09811465948441708,
+      "grad_norm": 1.855659008026123,
+      "learning_rate": 0.00019757724720913467,
+      "loss": 1.1037,
+      "step": 510
+    },
+    {
+      "epoch": 0.09830704116968064,
+      "grad_norm": 2.1122050285339355,
+      "learning_rate": 0.00019756359593993845,
+      "loss": 1.0378,
+      "step": 511
+    },
+    {
+      "epoch": 0.09849942285494422,
+      "grad_norm": 1.1042943000793457,
+      "learning_rate": 0.0001975499067931951,
+      "loss": 1.0536,
+      "step": 512
+    },
+    {
+      "epoch": 0.09869180454020778,
+      "grad_norm": 1.8844605684280396,
+      "learning_rate": 0.0001975361797742192,
+      "loss": 0.987,
+      "step": 513
+    },
+    {
+      "epoch": 0.09888418622547133,
+      "grad_norm": 1.9729514122009277,
+      "learning_rate": 0.00019752241488834002,
+      "loss": 0.9964,
+      "step": 514
+    },
+    {
+      "epoch": 0.0990765679107349,
+      "grad_norm": 2.1976277828216553,
+      "learning_rate": 0.0001975086121409016,
+      "loss": 1.0935,
+      "step": 515
+    },
+    {
+      "epoch": 0.09926894959599845,
+      "grad_norm": 2.121910810470581,
+      "learning_rate": 0.0001974947715372626,
+      "loss": 1.0412,
+      "step": 516
+    },
+    {
+      "epoch": 0.09946133128126203,
+      "grad_norm": 2.4691405296325684,
+      "learning_rate": 0.00019748089308279648,
+      "loss": 1.06,
+      "step": 517
+    },
+    {
+      "epoch": 0.09965371296652559,
+      "grad_norm": 1.9663869142532349,
+      "learning_rate": 0.0001974669767828913,
+      "loss": 1.0416,
+      "step": 518
+    },
+    {
+      "epoch": 0.09984609465178915,
+      "grad_norm": 2.54974102973938,
+      "learning_rate": 0.0001974530226429498,
+      "loss": 1.1344,
+      "step": 519
+    },
+    {
+      "epoch": 0.1000384763370527,
+      "grad_norm": 2.1954617500305176,
+      "learning_rate": 0.00019743903066838953,
+      "loss": 1.0561,
+      "step": 520
+    },
+    {
+      "epoch": 0.10023085802231628,
+      "grad_norm": 1.3314032554626465,
+      "learning_rate": 0.00019742500086464263,
+      "loss": 1.0294,
+      "step": 521
+    },
+    {
+      "epoch": 0.10042323970757984,
+      "grad_norm": 2.2481703758239746,
+      "learning_rate": 0.00019741093323715595,
+      "loss": 0.9633,
+      "step": 522
+    },
+    {
+      "epoch": 0.1006156213928434,
+      "grad_norm": 1.8195617198944092,
+      "learning_rate": 0.00019739682779139106,
+      "loss": 1.0524,
+      "step": 523
+    },
+    {
+      "epoch": 0.10080800307810696,
+      "grad_norm": 1.2284661531448364,
+      "learning_rate": 0.0001973826845328241,
+      "loss": 1.0368,
+      "step": 524
+    },
+    {
+      "epoch": 0.10100038476337053,
+      "grad_norm": 2.610520124435425,
+      "learning_rate": 0.0001973685034669461,
+      "loss": 1.017,
+      "step": 525
+    },
+    {
+      "epoch": 0.10119276644863409,
+      "grad_norm": 1.6822082996368408,
+      "learning_rate": 0.00019735428459926249,
+      "loss": 1.1316,
+      "step": 526
+    },
+    {
+      "epoch": 0.10138514813389765,
+      "grad_norm": 1.5387839078903198,
+      "learning_rate": 0.0001973400279352936,
+      "loss": 1.051,
+      "step": 527
+    },
+    {
+      "epoch": 0.10157752981916121,
+      "grad_norm": 1.0294824838638306,
+      "learning_rate": 0.00019732573348057436,
+      "loss": 1.0156,
+      "step": 528
+    },
+    {
+      "epoch": 0.10176991150442478,
+      "grad_norm": 1.8559293746948242,
+      "learning_rate": 0.00019731140124065437,
+      "loss": 0.9777,
+      "step": 529
+    },
+    {
+      "epoch": 0.10196229318968834,
+      "grad_norm": 1.86829674243927,
+      "learning_rate": 0.00019729703122109787,
+      "loss": 1.0716,
+      "step": 530
+    },
+    {
+      "epoch": 0.1021546748749519,
+      "grad_norm": 2.041043758392334,
+      "learning_rate": 0.00019728262342748383,
+      "loss": 1.0486,
+      "step": 531
+    },
+    {
+      "epoch": 0.10234705656021546,
+      "grad_norm": 2.3273279666900635,
+      "learning_rate": 0.00019726817786540584,
+      "loss": 1.0617,
+      "step": 532
+    },
+    {
+      "epoch": 0.10253943824547904,
+      "grad_norm": 2.125852584838867,
+      "learning_rate": 0.00019725369454047214,
+      "loss": 0.9966,
+      "step": 533
+    },
+    {
+      "epoch": 0.1027318199307426,
+      "grad_norm": 2.270155429840088,
+      "learning_rate": 0.00019723917345830568,
+      "loss": 1.0117,
+      "step": 534
+    },
+    {
+      "epoch": 0.10292420161600616,
+      "grad_norm": 2.1638431549072266,
+      "learning_rate": 0.00019722461462454404,
+      "loss": 1.0244,
+      "step": 535
+    },
+    {
+      "epoch": 0.10311658330126972,
+      "grad_norm": 1.5873651504516602,
+      "learning_rate": 0.00019721001804483948,
+      "loss": 1.0437,
+      "step": 536
+    },
+    {
+      "epoch": 0.10330896498653329,
+      "grad_norm": 2.3359193801879883,
+      "learning_rate": 0.00019719538372485885,
+      "loss": 1.0858,
+      "step": 537
+    },
+    {
+      "epoch": 0.10350134667179685,
+      "grad_norm": 2.625332832336426,
+      "learning_rate": 0.00019718071167028375,
+      "loss": 1.0025,
+      "step": 538
+    },
+    {
+      "epoch": 0.10369372835706041,
+      "grad_norm": 2.4390347003936768,
+      "learning_rate": 0.00019716600188681037,
+      "loss": 1.0054,
+      "step": 539
+    },
+    {
+      "epoch": 0.10388611004232397,
+      "grad_norm": 1.7317779064178467,
+      "learning_rate": 0.00019715125438014947,
+      "loss": 1.0181,
+      "step": 540
+    },
+    {
+      "epoch": 0.10407849172758753,
+      "grad_norm": 2.22090482711792,
+      "learning_rate": 0.00019713646915602663,
+      "loss": 0.9876,
+      "step": 541
+    },
+    {
+      "epoch": 0.1042708734128511,
+      "grad_norm": 1.274558424949646,
+      "learning_rate": 0.00019712164622018197,
+      "loss": 1.0888,
+      "step": 542
+    },
+    {
+      "epoch": 0.10446325509811466,
+      "grad_norm": 1.8960118293762207,
+      "learning_rate": 0.00019710678557837024,
+      "loss": 1.1041,
+      "step": 543
+    },
+    {
+      "epoch": 0.10465563678337822,
+      "grad_norm": 2.052793025970459,
+      "learning_rate": 0.00019709188723636086,
+      "loss": 1.0603,
+      "step": 544
+    },
+    {
+      "epoch": 0.10484801846864178,
+      "grad_norm": 2.1776952743530273,
+      "learning_rate": 0.00019707695119993793,
+      "loss": 1.0643,
+      "step": 545
+    },
+    {
+      "epoch": 0.10504040015390535,
+      "grad_norm": 2.0108580589294434,
+      "learning_rate": 0.00019706197747490001,
+      "loss": 1.0466,
+      "step": 546
+    },
+    {
+      "epoch": 0.10523278183916891,
+      "grad_norm": 1.9747531414031982,
+      "learning_rate": 0.00019704696606706053,
+      "loss": 1.0642,
+      "step": 547
+    },
+    {
+      "epoch": 0.10542516352443247,
+      "grad_norm": 1.6803699731826782,
+      "learning_rate": 0.00019703191698224742,
+      "loss": 0.954,
+      "step": 548
+    },
+    {
+      "epoch": 0.10561754520969603,
+      "grad_norm": 2.376168727874756,
+      "learning_rate": 0.0001970168302263032,
+      "loss": 1.0529,
+      "step": 549
+    },
+    {
+      "epoch": 0.1058099268949596,
+      "grad_norm": 2.2781989574432373,
+      "learning_rate": 0.00019700170580508512,
+      "loss": 1.0639,
+      "step": 550
+    },
+    {
+      "epoch": 0.10600230858022316,
+      "grad_norm": 1.9951698780059814,
+      "learning_rate": 0.00019698654372446493,
+      "loss": 1.0591,
+      "step": 551
+    },
+    {
+      "epoch": 0.10619469026548672,
+      "grad_norm": 1.175567388534546,
+      "learning_rate": 0.0001969713439903292,
+      "loss": 1.1148,
+      "step": 552
+    },
+    {
+      "epoch": 0.10638707195075028,
+      "grad_norm": 1.9796366691589355,
+      "learning_rate": 0.00019695610660857887,
+      "loss": 1.0773,
+      "step": 553
+    },
+    {
+      "epoch": 0.10657945363601386,
+      "grad_norm": 2.450127363204956,
+      "learning_rate": 0.00019694083158512964,
+      "loss": 1.0812,
+      "step": 554
+    },
+    {
+      "epoch": 0.10677183532127742,
+      "grad_norm": 1.0738441944122314,
+      "learning_rate": 0.00019692551892591185,
+      "loss": 1.0961,
+      "step": 555
+    },
+    {
+      "epoch": 0.10696421700654098,
+      "grad_norm": 1.941198706626892,
+      "learning_rate": 0.00019691016863687035,
+      "loss": 1.0279,
+      "step": 556
+    },
+    {
+      "epoch": 0.10715659869180454,
+      "grad_norm": 2.229801893234253,
+      "learning_rate": 0.00019689478072396467,
+      "loss": 1.0323,
+      "step": 557
+    },
+    {
+      "epoch": 0.10734898037706811,
+      "grad_norm": 1.9423714876174927,
+      "learning_rate": 0.00019687935519316897,
+      "loss": 1.0971,
+      "step": 558
+    },
+    {
+      "epoch": 0.10754136206233167,
+      "grad_norm": 0.8468499779701233,
+      "learning_rate": 0.00019686389205047188,
+      "loss": 0.6977,
+      "step": 559
+    },
+    {
+      "epoch": 0.10773374374759523,
+      "grad_norm": 2.0276622772216797,
+      "learning_rate": 0.0001968483913018768,
+      "loss": 1.1284,
+      "step": 560
+    },
+    {
+      "epoch": 0.10792612543285879,
+      "grad_norm": 1.2585800886154175,
+      "learning_rate": 0.0001968328529534016,
+      "loss": 1.0496,
+      "step": 561
+    },
+    {
+      "epoch": 0.10811850711812235,
+      "grad_norm": 2.4516706466674805,
+      "learning_rate": 0.00019681727701107885,
+      "loss": 1.019,
+      "step": 562
+    },
+    {
+      "epoch": 0.10831088880338592,
+      "grad_norm": 2.3377721309661865,
+      "learning_rate": 0.00019680166348095564,
+      "loss": 1.0665,
+      "step": 563
+    },
+    {
+      "epoch": 0.10850327048864948,
+      "grad_norm": 2.145272731781006,
+      "learning_rate": 0.00019678601236909371,
+      "loss": 1.0744,
+      "step": 564
+    },
+    {
+      "epoch": 0.10869565217391304,
+      "grad_norm": 2.0796902179718018,
+      "learning_rate": 0.00019677032368156931,
+      "loss": 1.0782,
+      "step": 565
+    },
+    {
+      "epoch": 0.1088880338591766,
+      "grad_norm": 2.352766990661621,
+      "learning_rate": 0.00019675459742447338,
+      "loss": 1.03,
+      "step": 566
+    },
+    {
+      "epoch": 0.10908041554444017,
+      "grad_norm": 1.9927926063537598,
+      "learning_rate": 0.00019673883360391136,
+      "loss": 1.0528,
+      "step": 567
+    },
+    {
+      "epoch": 0.10927279722970373,
+      "grad_norm": 1.8637689352035522,
+      "learning_rate": 0.00019672303222600333,
+      "loss": 0.9361,
+      "step": 568
+    },
+    {
+      "epoch": 0.10946517891496729,
+      "grad_norm": 1.370914340019226,
+      "learning_rate": 0.00019670719329688388,
+      "loss": 1.1167,
+      "step": 569
+    },
+    {
+      "epoch": 0.10965756060023085,
+      "grad_norm": 0.9067789912223816,
+      "learning_rate": 0.0001966913168227023,
+      "loss": 0.7092,
+      "step": 570
+    },
+    {
+      "epoch": 0.10984994228549443,
+      "grad_norm": 2.089810848236084,
+      "learning_rate": 0.00019667540280962235,
+      "loss": 1.0936,
+      "step": 571
+    },
+    {
+      "epoch": 0.11004232397075799,
+      "grad_norm": 1.501318335533142,
+      "learning_rate": 0.0001966594512638224,
+      "loss": 0.9699,
+      "step": 572
+    },
+    {
+      "epoch": 0.11023470565602154,
+      "grad_norm": 2.2103958129882812,
+      "learning_rate": 0.0001966434621914954,
+      "loss": 1.0246,
+      "step": 573
+    },
+    {
+      "epoch": 0.1104270873412851,
+      "grad_norm": 2.118896484375,
+      "learning_rate": 0.0001966274355988488,
+      "loss": 1.0678,
+      "step": 574
+    },
+    {
+      "epoch": 0.11061946902654868,
+      "grad_norm": 2.0669400691986084,
+      "learning_rate": 0.00019661137149210473,
+      "loss": 1.0382,
+      "step": 575
+    },
+    {
+      "epoch": 0.11081185071181224,
+      "grad_norm": 2.1090965270996094,
+      "learning_rate": 0.00019659526987749986,
+      "loss": 1.0475,
+      "step": 576
+    },
+    {
+      "epoch": 0.1110042323970758,
+      "grad_norm": 2.3913626670837402,
+      "learning_rate": 0.00019657913076128532,
+      "loss": 1.0923,
+      "step": 577
+    },
+    {
+      "epoch": 0.11119661408233936,
+      "grad_norm": 2.104969024658203,
+      "learning_rate": 0.0001965629541497269,
+      "loss": 1.074,
+      "step": 578
+    },
+    {
+      "epoch": 0.11138899576760293,
+      "grad_norm": 2.352670431137085,
+      "learning_rate": 0.00019654674004910492,
+      "loss": 1.0071,
+      "step": 579
+    },
+    {
+      "epoch": 0.11158137745286649,
+      "grad_norm": 1.8953325748443604,
+      "learning_rate": 0.00019653048846571426,
+      "loss": 1.0672,
+      "step": 580
+    },
+    {
+      "epoch": 0.11177375913813005,
+      "grad_norm": 1.6592098474502563,
+      "learning_rate": 0.00019651419940586436,
+      "loss": 1.0287,
+      "step": 581
+    },
+    {
+      "epoch": 0.11196614082339361,
+      "grad_norm": 1.872137188911438,
+      "learning_rate": 0.00019649787287587912,
+      "loss": 1.1299,
+      "step": 582
+    },
+    {
+      "epoch": 0.11215852250865718,
+      "grad_norm": 2.2436118125915527,
+      "learning_rate": 0.00019648150888209714,
+      "loss": 1.0198,
+      "step": 583
+    },
+    {
+      "epoch": 0.11235090419392074,
+      "grad_norm": 2.041688919067383,
+      "learning_rate": 0.00019646510743087143,
+      "loss": 1.1032,
+      "step": 584
+    },
+    {
+      "epoch": 0.1125432858791843,
+      "grad_norm": 1.7055717706680298,
+      "learning_rate": 0.00019644866852856966,
+      "loss": 1.0917,
+      "step": 585
+    },
+    {
+      "epoch": 0.11273566756444786,
+      "grad_norm": 1.7182458639144897,
+      "learning_rate": 0.00019643219218157392,
+      "loss": 0.9833,
+      "step": 586
+    },
+    {
+      "epoch": 0.11292804924971142,
+      "grad_norm": 2.09781551361084,
+      "learning_rate": 0.00019641567839628093,
+      "loss": 1.0417,
+      "step": 587
+    },
+    {
+      "epoch": 0.113120430934975,
+      "grad_norm": 2.653714418411255,
+      "learning_rate": 0.0001963991271791019,
+      "loss": 0.966,
+      "step": 588
+    },
+    {
+      "epoch": 0.11331281262023855,
+      "grad_norm": 1.5758801698684692,
+      "learning_rate": 0.00019638253853646257,
+      "loss": 0.9726,
+      "step": 589
+    },
+    {
+      "epoch": 0.11350519430550211,
+      "grad_norm": 2.0875661373138428,
+      "learning_rate": 0.00019636591247480323,
+      "loss": 1.014,
+      "step": 590
+    },
+    {
+      "epoch": 0.11369757599076567,
+      "grad_norm": 2.6447112560272217,
+      "learning_rate": 0.00019634924900057866,
+      "loss": 1.1428,
+      "step": 591
+    },
+    {
+      "epoch": 0.11388995767602925,
+      "grad_norm": 2.236492395401001,
+      "learning_rate": 0.0001963325481202583,
+      "loss": 1.004,
+      "step": 592
+    },
+    {
+      "epoch": 0.1140823393612928,
+      "grad_norm": 1.9506890773773193,
+      "learning_rate": 0.0001963158098403259,
+      "loss": 0.9754,
+      "step": 593
+    },
+    {
+      "epoch": 0.11427472104655637,
+      "grad_norm": 1.9617393016815186,
+      "learning_rate": 0.00019629903416727986,
+      "loss": 1.0259,
+      "step": 594
+    },
+    {
+      "epoch": 0.11446710273181993,
+      "grad_norm": 0.860497236251831,
+      "learning_rate": 0.0001962822211076331,
+      "loss": 0.717,
+      "step": 595
+    },
+    {
+      "epoch": 0.1146594844170835,
+      "grad_norm": 2.1541543006896973,
+      "learning_rate": 0.00019626537066791297,
+      "loss": 1.042,
+      "step": 596
+    },
+    {
+      "epoch": 0.11485186610234706,
+      "grad_norm": 2.803661346435547,
+      "learning_rate": 0.00019624848285466147,
+      "loss": 1.0225,
+      "step": 597
+    },
+    {
+      "epoch": 0.11504424778761062,
+      "grad_norm": 2.0338869094848633,
+      "learning_rate": 0.00019623155767443498,
+      "loss": 1.0435,
+      "step": 598
+    },
+    {
+      "epoch": 0.11523662947287418,
+      "grad_norm": 1.7751929759979248,
+      "learning_rate": 0.00019621459513380444,
+      "loss": 1.0639,
+      "step": 599
+    },
+    {
+      "epoch": 0.11542901115813775,
+      "grad_norm": 1.7959973812103271,
+      "learning_rate": 0.00019619759523935532,
+      "loss": 1.0765,
+      "step": 600
+    },
+    {
+      "epoch": 0.11562139284340131,
+      "grad_norm": 2.1445131301879883,
+      "learning_rate": 0.00019618055799768756,
+      "loss": 0.9783,
+      "step": 601
+    },
+    {
+      "epoch": 0.11581377452866487,
+      "grad_norm": 2.3026556968688965,
+      "learning_rate": 0.00019616348341541557,
+      "loss": 1.094,
+      "step": 602
+    },
+    {
+      "epoch": 0.11600615621392843,
+      "grad_norm": 0.9957302808761597,
+      "learning_rate": 0.00019614637149916836,
+      "loss": 0.7977,
+      "step": 603
+    },
+    {
+      "epoch": 0.116198537899192,
+      "grad_norm": 2.460299015045166,
+      "learning_rate": 0.00019612922225558925,
+      "loss": 1.1297,
+      "step": 604
+    },
+    {
+      "epoch": 0.11639091958445556,
+      "grad_norm": 1.7452443838119507,
+      "learning_rate": 0.00019611203569133627,
+      "loss": 1.0203,
+      "step": 605
+    },
+    {
+      "epoch": 0.11658330126971912,
+      "grad_norm": 1.8132526874542236,
+      "learning_rate": 0.00019609481181308177,
+      "loss": 1.1307,
+      "step": 606
+    },
+    {
+      "epoch": 0.11677568295498268,
+      "grad_norm": 2.2195136547088623,
+      "learning_rate": 0.00019607755062751273,
+      "loss": 1.0378,
+      "step": 607
+    },
+    {
+      "epoch": 0.11696806464024626,
+      "grad_norm": 1.892532467842102,
+      "learning_rate": 0.00019606025214133046,
+      "loss": 1.0616,
+      "step": 608
+    },
+    {
+      "epoch": 0.11716044632550981,
+      "grad_norm": 2.2749569416046143,
+      "learning_rate": 0.00019604291636125085,
+      "loss": 1.0723,
+      "step": 609
+    },
+    {
+      "epoch": 0.11735282801077337,
+      "grad_norm": 2.16485333442688,
+      "learning_rate": 0.00019602554329400428,
+      "loss": 1.0746,
+      "step": 610
+    },
+    {
+      "epoch": 0.11754520969603693,
+      "grad_norm": 2.2999167442321777,
+      "learning_rate": 0.00019600813294633554,
+      "loss": 1.0185,
+      "step": 611
+    },
+    {
+      "epoch": 0.1177375913813005,
+      "grad_norm": 1.0940454006195068,
+      "learning_rate": 0.00019599068532500394,
+      "loss": 0.7159,
+      "step": 612
+    },
+    {
+      "epoch": 0.11792997306656407,
+      "grad_norm": 2.843454360961914,
+      "learning_rate": 0.00019597320043678322,
+      "loss": 1.0293,
+      "step": 613
+    },
+    {
+      "epoch": 0.11812235475182763,
+      "grad_norm": 2.264554500579834,
+      "learning_rate": 0.00019595567828846163,
+      "loss": 1.0351,
+      "step": 614
+    },
+    {
+      "epoch": 0.11831473643709119,
+      "grad_norm": 2.616274118423462,
+      "learning_rate": 0.0001959381188868419,
+      "loss": 0.977,
+      "step": 615
+    },
+    {
+      "epoch": 0.11850711812235475,
+      "grad_norm": 1.9295307397842407,
+      "learning_rate": 0.00019592052223874117,
+      "loss": 1.0713,
+      "step": 616
+    },
+    {
+      "epoch": 0.11869949980761832,
+      "grad_norm": 1.968209981918335,
+      "learning_rate": 0.00019590288835099107,
+      "loss": 1.0621,
+      "step": 617
+    },
+    {
+      "epoch": 0.11889188149288188,
+      "grad_norm": 1.928808331489563,
+      "learning_rate": 0.00019588521723043766,
+      "loss": 1.0716,
+      "step": 618
+    },
+    {
+      "epoch": 0.11908426317814544,
+      "grad_norm": 2.1827564239501953,
+      "learning_rate": 0.0001958675088839415,
+      "loss": 1.023,
+      "step": 619
+    },
+    {
+      "epoch": 0.119276644863409,
+      "grad_norm": 2.0973799228668213,
+      "learning_rate": 0.00019584976331837758,
+      "loss": 0.9995,
+      "step": 620
+    },
+    {
+      "epoch": 0.11946902654867257,
+      "grad_norm": 2.1291584968566895,
+      "learning_rate": 0.00019583198054063533,
+      "loss": 1.0312,
+      "step": 621
+    },
+    {
+      "epoch": 0.11966140823393613,
+      "grad_norm": 2.1214654445648193,
+      "learning_rate": 0.00019581416055761865,
+      "loss": 1.0362,
+      "step": 622
+    },
+    {
+      "epoch": 0.11985378991919969,
+      "grad_norm": 1.049235224723816,
+      "learning_rate": 0.00019579630337624587,
+      "loss": 0.7402,
+      "step": 623
+    },
+    {
+      "epoch": 0.12004617160446325,
+      "grad_norm": 2.229199171066284,
+      "learning_rate": 0.00019577840900344976,
+      "loss": 1.1179,
+      "step": 624
+    },
+    {
+      "epoch": 0.12023855328972682,
+      "grad_norm": 1.3378239870071411,
+      "learning_rate": 0.00019576047744617752,
+      "loss": 1.0625,
+      "step": 625
+    },
+    {
+      "epoch": 0.12043093497499038,
+      "grad_norm": 1.0555399656295776,
+      "learning_rate": 0.0001957425087113908,
+      "loss": 0.7049,
+      "step": 626
+    },
+    {
+      "epoch": 0.12062331666025394,
+      "grad_norm": 1.3788373470306396,
+      "learning_rate": 0.0001957245028060657,
+      "loss": 1.1015,
+      "step": 627
+    },
+    {
+      "epoch": 0.1208156983455175,
+      "grad_norm": 2.3713624477386475,
+      "learning_rate": 0.0001957064597371927,
+      "loss": 1.0744,
+      "step": 628
+    },
+    {
+      "epoch": 0.12100808003078108,
+      "grad_norm": 3.2305057048797607,
+      "learning_rate": 0.00019568837951177676,
+      "loss": 1.1092,
+      "step": 629
+    },
+    {
+      "epoch": 0.12120046171604464,
+      "grad_norm": 2.117847442626953,
+      "learning_rate": 0.00019567026213683725,
+      "loss": 1.0979,
+      "step": 630
+    },
+    {
+      "epoch": 0.1213928434013082,
+      "grad_norm": 0.7957298159599304,
+      "learning_rate": 0.00019565210761940798,
+      "loss": 0.7147,
+      "step": 631
+    },
+    {
+      "epoch": 0.12158522508657175,
+      "grad_norm": 1.2993072271347046,
+      "learning_rate": 0.00019563391596653712,
+      "loss": 1.0347,
+      "step": 632
+    },
+    {
+      "epoch": 0.12177760677183531,
+      "grad_norm": 2.0916178226470947,
+      "learning_rate": 0.0001956156871852873,
+      "loss": 1.0323,
+      "step": 633
+    },
+    {
+      "epoch": 0.12196998845709889,
+      "grad_norm": 0.9948853254318237,
+      "learning_rate": 0.00019559742128273558,
+      "loss": 0.7028,
+      "step": 634
+    },
+    {
+      "epoch": 0.12216237014236245,
+      "grad_norm": 1.2554380893707275,
+      "learning_rate": 0.00019557911826597338,
+      "loss": 1.0481,
+      "step": 635
+    },
+    {
+      "epoch": 0.12235475182762601,
+      "grad_norm": 2.024512529373169,
+      "learning_rate": 0.0001955607781421066,
+      "loss": 1.0295,
+      "step": 636
+    },
+    {
+      "epoch": 0.12254713351288957,
+      "grad_norm": 1.0955697298049927,
+      "learning_rate": 0.00019554240091825548,
+      "loss": 1.1088,
+      "step": 637
+    },
+    {
+      "epoch": 0.12273951519815314,
+      "grad_norm": 2.4762513637542725,
+      "learning_rate": 0.0001955239866015547,
+      "loss": 1.0665,
+      "step": 638
+    },
+    {
+      "epoch": 0.1229318968834167,
+      "grad_norm": 2.2404592037200928,
+      "learning_rate": 0.00019550553519915337,
+      "loss": 1.0927,
+      "step": 639
+    },
+    {
+      "epoch": 0.12312427856868026,
+      "grad_norm": 1.9989155530929565,
+      "learning_rate": 0.00019548704671821493,
+      "loss": 0.9991,
+      "step": 640
+    },
+    {
+      "epoch": 0.12331666025394382,
+      "grad_norm": 2.237001419067383,
+      "learning_rate": 0.0001954685211659172,
+      "loss": 1.1044,
+      "step": 641
+    },
+    {
+      "epoch": 0.12350904193920739,
+      "grad_norm": 2.2847378253936768,
+      "learning_rate": 0.00019544995854945248,
+      "loss": 0.9973,
+      "step": 642
+    },
+    {
+      "epoch": 0.12370142362447095,
+      "grad_norm": 2.17987060546875,
+      "learning_rate": 0.0001954313588760274,
+      "loss": 1.0226,
+      "step": 643
+    },
+    {
+      "epoch": 0.12389380530973451,
+      "grad_norm": 2.3213963508605957,
+      "learning_rate": 0.00019541272215286304,
+      "loss": 1.0602,
+      "step": 644
+    },
+    {
+      "epoch": 0.12408618699499807,
+      "grad_norm": 1.330147624015808,
+      "learning_rate": 0.00019539404838719477,
+      "loss": 1.0335,
+      "step": 645
+    },
+    {
+      "epoch": 0.12427856868026164,
+      "grad_norm": 1.8381214141845703,
+      "learning_rate": 0.0001953753375862724,
+      "loss": 1.0595,
+      "step": 646
+    },
+    {
+      "epoch": 0.1244709503655252,
+      "grad_norm": 2.5730574131011963,
+      "learning_rate": 0.0001953565897573601,
+      "loss": 1.0028,
+      "step": 647
+    },
+    {
+      "epoch": 0.12466333205078876,
+      "grad_norm": 2.46988844871521,
+      "learning_rate": 0.00019533780490773643,
+      "loss": 1.0339,
+      "step": 648
+    },
+    {
+      "epoch": 0.12485571373605232,
+      "grad_norm": 2.1902990341186523,
+      "learning_rate": 0.00019531898304469433,
+      "loss": 1.0135,
+      "step": 649
+    },
+    {
+      "epoch": 0.1250480954213159,
+      "grad_norm": 1.1375412940979004,
+      "learning_rate": 0.0001953001241755411,
+      "loss": 0.7337,
+      "step": 650
+    },
+    {
+      "epoch": 0.12524047710657946,
+      "grad_norm": 1.8239678144454956,
+      "learning_rate": 0.0001952812283075984,
+      "loss": 0.9529,
+      "step": 651
+    },
+    {
+      "epoch": 0.12543285879184302,
+      "grad_norm": 1.9380667209625244,
+      "learning_rate": 0.00019526229544820217,
+      "loss": 1.0646,
+      "step": 652
+    },
+    {
+      "epoch": 0.12562524047710658,
+      "grad_norm": 2.004977226257324,
+      "learning_rate": 0.00019524332560470292,
+      "loss": 1.0285,
+      "step": 653
+    },
+    {
+      "epoch": 0.12581762216237014,
+      "grad_norm": 2.0921833515167236,
+      "learning_rate": 0.00019522431878446535,
+      "loss": 1.0216,
+      "step": 654
+    },
+    {
+      "epoch": 0.1260100038476337,
+      "grad_norm": 2.197117567062378,
+      "learning_rate": 0.00019520527499486856,
+      "loss": 1.0372,
+      "step": 655
+    },
+    {
+      "epoch": 0.12620238553289725,
+      "grad_norm": 1.887397289276123,
+      "learning_rate": 0.00019518619424330598,
+      "loss": 1.0472,
+      "step": 656
+    },
+    {
+      "epoch": 0.12639476721816084,
+      "grad_norm": 1.6989096403121948,
+      "learning_rate": 0.00019516707653718544,
+      "loss": 1.0865,
+      "step": 657
+    },
+    {
+      "epoch": 0.1265871489034244,
+      "grad_norm": 1.9916104078292847,
+      "learning_rate": 0.00019514792188392912,
+      "loss": 1.037,
+      "step": 658
+    },
+    {
+      "epoch": 0.12677953058868796,
+      "grad_norm": 1.9844062328338623,
+      "learning_rate": 0.00019512873029097345,
+      "loss": 0.9706,
+      "step": 659
+    },
+    {
+      "epoch": 0.12697191227395152,
+      "grad_norm": 2.256995439529419,
+      "learning_rate": 0.00019510950176576932,
+      "loss": 1.0425,
+      "step": 660
+    },
+    {
+      "epoch": 0.12716429395921508,
+      "grad_norm": 2.2489213943481445,
+      "learning_rate": 0.0001950902363157819,
+      "loss": 1.0634,
+      "step": 661
+    },
+    {
+      "epoch": 0.12735667564447864,
+      "grad_norm": 2.6758158206939697,
+      "learning_rate": 0.00019507093394849066,
+      "loss": 1.0231,
+      "step": 662
+    },
+    {
+      "epoch": 0.1275490573297422,
+      "grad_norm": 2.5774011611938477,
+      "learning_rate": 0.0001950515946713895,
+      "loss": 1.0672,
+      "step": 663
+    },
+    {
+      "epoch": 0.12774143901500576,
+      "grad_norm": 1.1645704507827759,
+      "learning_rate": 0.00019503221849198656,
+      "loss": 0.7086,
+      "step": 664
+    },
+    {
+      "epoch": 0.12793382070026935,
+      "grad_norm": 2.6528027057647705,
+      "learning_rate": 0.00019501280541780433,
+      "loss": 1.0345,
+      "step": 665
+    },
+    {
+      "epoch": 0.1281262023855329,
+      "grad_norm": 2.3895649909973145,
+      "learning_rate": 0.00019499335545637966,
+      "loss": 1.1277,
+      "step": 666
+    },
+    {
+      "epoch": 0.12831858407079647,
+      "grad_norm": 1.0713979005813599,
+      "learning_rate": 0.0001949738686152637,
+      "loss": 0.7238,
+      "step": 667
+    },
+    {
+      "epoch": 0.12851096575606002,
+      "grad_norm": 1.515957236289978,
+      "learning_rate": 0.00019495434490202188,
+      "loss": 1.122,
+      "step": 668
+    },
+    {
+      "epoch": 0.12870334744132358,
+      "grad_norm": 1.3128505945205688,
+      "learning_rate": 0.000194934784324234,
+      "loss": 0.9813,
+      "step": 669
+    },
+    {
+      "epoch": 0.12889572912658714,
+      "grad_norm": 1.9267456531524658,
+      "learning_rate": 0.00019491518688949416,
+      "loss": 1.072,
+      "step": 670
+    },
+    {
+      "epoch": 0.1290881108118507,
+      "grad_norm": 2.3292107582092285,
+      "learning_rate": 0.00019489555260541073,
+      "loss": 1.0661,
+      "step": 671
+    },
+    {
+      "epoch": 0.12928049249711426,
+      "grad_norm": 2.545958995819092,
+      "learning_rate": 0.0001948758814796064,
+      "loss": 1.0787,
+      "step": 672
+    },
+    {
+      "epoch": 0.12947287418237785,
+      "grad_norm": 1.797191858291626,
+      "learning_rate": 0.00019485617351971827,
+      "loss": 0.9685,
+      "step": 673
+    },
+    {
+      "epoch": 0.1296652558676414,
+      "grad_norm": 2.29461407661438,
+      "learning_rate": 0.00019483642873339751,
+      "loss": 1.0674,
+      "step": 674
+    },
+    {
+      "epoch": 0.12985763755290497,
+      "grad_norm": 1.7076826095581055,
+      "learning_rate": 0.00019481664712830984,
+      "loss": 1.0963,
+      "step": 675
+    },
+    {
+      "epoch": 0.13005001923816853,
+      "grad_norm": 1.9503265619277954,
+      "learning_rate": 0.00019479682871213514,
+      "loss": 1.0338,
+      "step": 676
+    },
+    {
+      "epoch": 0.1302424009234321,
+      "grad_norm": 1.81754469871521,
+      "learning_rate": 0.00019477697349256756,
+      "loss": 1.0655,
+      "step": 677
+    },
+    {
+      "epoch": 0.13043478260869565,
+      "grad_norm": 2.0934247970581055,
+      "learning_rate": 0.0001947570814773156,
+      "loss": 1.069,
+      "step": 678
+    },
+    {
+      "epoch": 0.1306271642939592,
+      "grad_norm": 1.9358630180358887,
+      "learning_rate": 0.00019473715267410207,
+      "loss": 1.0371,
+      "step": 679
+    },
+    {
+      "epoch": 0.13081954597922277,
+      "grad_norm": 1.4743399620056152,
+      "learning_rate": 0.00019471718709066394,
+      "loss": 1.099,
+      "step": 680
+    },
+    {
+      "epoch": 0.13101192766448633,
+      "grad_norm": 2.7317259311676025,
+      "learning_rate": 0.00019469718473475256,
+      "loss": 1.0789,
+      "step": 681
+    },
+    {
+      "epoch": 0.13120430934974991,
+      "grad_norm": 1.3007100820541382,
+      "learning_rate": 0.00019467714561413358,
+      "loss": 1.0844,
+      "step": 682
+    },
+    {
+      "epoch": 0.13139669103501347,
+      "grad_norm": 2.1076440811157227,
+      "learning_rate": 0.00019465706973658683,
+      "loss": 0.9659,
+      "step": 683
+    },
+    {
+      "epoch": 0.13158907272027703,
+      "grad_norm": 1.0914628505706787,
+      "learning_rate": 0.00019463695710990645,
+      "loss": 0.7216,
+      "step": 684
+    },
+    {
+      "epoch": 0.1317814544055406,
+      "grad_norm": 1.9552632570266724,
+      "learning_rate": 0.00019461680774190092,
+      "loss": 0.989,
+      "step": 685
+    },
+    {
+      "epoch": 0.13197383609080415,
+      "grad_norm": 2.428699254989624,
+      "learning_rate": 0.00019459662164039284,
+      "loss": 1.08,
+      "step": 686
+    },
+    {
+      "epoch": 0.1321662177760677,
+      "grad_norm": 2.2287731170654297,
+      "learning_rate": 0.00019457639881321917,
+      "loss": 1.0253,
+      "step": 687
+    },
+    {
+      "epoch": 0.13235859946133127,
+      "grad_norm": 2.4568309783935547,
+      "learning_rate": 0.00019455613926823114,
+      "loss": 0.9919,
+      "step": 688
+    },
+    {
+      "epoch": 0.13255098114659483,
+      "grad_norm": 1.7906675338745117,
+      "learning_rate": 0.0001945358430132942,
+      "loss": 1.0531,
+      "step": 689
+    },
+    {
+      "epoch": 0.13274336283185842,
+      "grad_norm": 2.2335593700408936,
+      "learning_rate": 0.000194515510056288,
+      "loss": 1.1603,
+      "step": 690
+    },
+    {
+      "epoch": 0.13293574451712198,
+      "grad_norm": 2.0844922065734863,
+      "learning_rate": 0.00019449514040510655,
+      "loss": 1.0053,
+      "step": 691
+    },
+    {
+      "epoch": 0.13312812620238554,
+      "grad_norm": 2.1246752738952637,
+      "learning_rate": 0.00019447473406765804,
+      "loss": 1.0179,
+      "step": 692
+    },
+    {
+      "epoch": 0.1333205078876491,
+      "grad_norm": 1.852514386177063,
+      "learning_rate": 0.00019445429105186487,
+      "loss": 0.9767,
+      "step": 693
+    },
+    {
+      "epoch": 0.13351288957291266,
+      "grad_norm": 1.803996205329895,
+      "learning_rate": 0.0001944338113656638,
+      "loss": 1.0929,
+      "step": 694
+    },
+    {
+      "epoch": 0.13370527125817622,
+      "grad_norm": 1.9504947662353516,
+      "learning_rate": 0.00019441329501700567,
+      "loss": 1.0118,
+      "step": 695
+    },
+    {
+      "epoch": 0.13389765294343978,
+      "grad_norm": 2.4281539916992188,
+      "learning_rate": 0.00019439274201385568,
+      "loss": 1.0833,
+      "step": 696
+    },
+    {
+      "epoch": 0.13409003462870334,
+      "grad_norm": 2.492562770843506,
+      "learning_rate": 0.0001943721523641932,
+      "loss": 0.9548,
+      "step": 697
+    },
+    {
+      "epoch": 0.13428241631396692,
+      "grad_norm": 2.266679286956787,
+      "learning_rate": 0.00019435152607601188,
+      "loss": 1.0563,
+      "step": 698
+    },
+    {
+      "epoch": 0.13447479799923048,
+      "grad_norm": 2.2576918601989746,
+      "learning_rate": 0.0001943308631573195,
+      "loss": 1.0038,
+      "step": 699
+    },
+    {
+      "epoch": 0.13466717968449404,
+      "grad_norm": 1.9259835481643677,
+      "learning_rate": 0.00019431016361613814,
+      "loss": 0.9701,
+      "step": 700
+    },
+    {
+      "epoch": 0.1348595613697576,
+      "grad_norm": 1.771523118019104,
+      "learning_rate": 0.00019428942746050406,
+      "loss": 1.0204,
+      "step": 701
+    },
+    {
+      "epoch": 0.13505194305502116,
+      "grad_norm": 2.4382894039154053,
+      "learning_rate": 0.00019426865469846774,
+      "loss": 1.0508,
+      "step": 702
+    },
+    {
+      "epoch": 0.13524432474028472,
+      "grad_norm": 2.4874427318573,
+      "learning_rate": 0.00019424784533809392,
+      "loss": 1.0327,
+      "step": 703
+    },
+    {
+      "epoch": 0.13543670642554828,
+      "grad_norm": 1.5919235944747925,
+      "learning_rate": 0.00019422699938746152,
+      "loss": 1.0077,
+      "step": 704
+    },
+    {
+      "epoch": 0.13562908811081184,
+      "grad_norm": 2.1816132068634033,
+      "learning_rate": 0.00019420611685466358,
+      "loss": 0.9467,
+      "step": 705
+    },
+    {
+      "epoch": 0.1358214697960754,
+      "grad_norm": 2.4410836696624756,
+      "learning_rate": 0.00019418519774780748,
+      "loss": 1.0676,
+      "step": 706
+    },
+    {
+      "epoch": 0.136013851481339,
+      "grad_norm": 1.4520705938339233,
+      "learning_rate": 0.00019416424207501474,
+      "loss": 1.0719,
+      "step": 707
+    },
+    {
+      "epoch": 0.13620623316660255,
+      "grad_norm": 2.0262439250946045,
+      "learning_rate": 0.00019414324984442103,
+      "loss": 1.0231,
+      "step": 708
+    },
+    {
+      "epoch": 0.1363986148518661,
+      "grad_norm": 2.3686747550964355,
+      "learning_rate": 0.0001941222210641763,
+      "loss": 1.0494,
+      "step": 709
+    },
+    {
+      "epoch": 0.13659099653712967,
+      "grad_norm": 1.4251441955566406,
+      "learning_rate": 0.00019410115574244462,
+      "loss": 1.0198,
+      "step": 710
+    },
+    {
+      "epoch": 0.13678337822239323,
+      "grad_norm": 1.4239583015441895,
+      "learning_rate": 0.00019408005388740432,
+      "loss": 1.0726,
+      "step": 711
+    },
+    {
+      "epoch": 0.13697575990765679,
+      "grad_norm": 1.8954933881759644,
+      "learning_rate": 0.00019405891550724778,
+      "loss": 1.0025,
+      "step": 712
+    },
+    {
+      "epoch": 0.13716814159292035,
+      "grad_norm": 2.2760331630706787,
+      "learning_rate": 0.00019403774061018173,
+      "loss": 1.0039,
+      "step": 713
+    },
+    {
+      "epoch": 0.1373605232781839,
+      "grad_norm": 2.588055372238159,
+      "learning_rate": 0.00019401652920442692,
+      "loss": 1.0726,
+      "step": 714
+    },
+    {
+      "epoch": 0.1375529049634475,
+      "grad_norm": 2.085352897644043,
+      "learning_rate": 0.00019399528129821842,
+      "loss": 1.0898,
+      "step": 715
+    },
+    {
+      "epoch": 0.13774528664871105,
+      "grad_norm": 2.7718844413757324,
+      "learning_rate": 0.0001939739968998054,
+      "loss": 1.0196,
+      "step": 716
+    },
+    {
+      "epoch": 0.1379376683339746,
+      "grad_norm": 2.0528547763824463,
+      "learning_rate": 0.0001939526760174511,
+      "loss": 1.0655,
+      "step": 717
+    },
+    {
+      "epoch": 0.13813005001923817,
+      "grad_norm": 2.206932544708252,
+      "learning_rate": 0.0001939313186594331,
+      "loss": 1.1056,
+      "step": 718
+    },
+    {
+      "epoch": 0.13832243170450173,
+      "grad_norm": 2.0666933059692383,
+      "learning_rate": 0.00019390992483404306,
+      "loss": 1.0732,
+      "step": 719
+    },
+    {
+      "epoch": 0.1385148133897653,
+      "grad_norm": 2.5182764530181885,
+      "learning_rate": 0.0001938884945495868,
+      "loss": 1.0364,
+      "step": 720
+    },
+    {
+      "epoch": 0.13870719507502885,
+      "grad_norm": 2.0749943256378174,
+      "learning_rate": 0.00019386702781438425,
+      "loss": 1.0521,
+      "step": 721
+    },
+    {
+      "epoch": 0.1388995767602924,
+      "grad_norm": 2.059662103652954,
+      "learning_rate": 0.0001938455246367696,
+      "loss": 1.0931,
+      "step": 722
+    },
+    {
+      "epoch": 0.139091958445556,
+      "grad_norm": 2.178807258605957,
+      "learning_rate": 0.00019382398502509107,
+      "loss": 1.0941,
+      "step": 723
+    },
+    {
+      "epoch": 0.13928434013081956,
+      "grad_norm": 1.9291013479232788,
+      "learning_rate": 0.0001938024089877111,
+      "loss": 1.0182,
+      "step": 724
+    },
+    {
+      "epoch": 0.13947672181608312,
+      "grad_norm": 2.2059435844421387,
+      "learning_rate": 0.00019378079653300625,
+      "loss": 1.0768,
+      "step": 725
+    },
+    {
+      "epoch": 0.13966910350134668,
+      "grad_norm": 1.1393191814422607,
+      "learning_rate": 0.0001937591476693672,
+      "loss": 0.7233,
+      "step": 726
+    },
+    {
+      "epoch": 0.13986148518661023,
+      "grad_norm": 1.7406129837036133,
+      "learning_rate": 0.00019373746240519883,
+      "loss": 1.0113,
+      "step": 727
+    },
+    {
+      "epoch": 0.1400538668718738,
+      "grad_norm": 2.461513042449951,
+      "learning_rate": 0.00019371574074892006,
+      "loss": 1.0233,
+      "step": 728
+    },
+    {
+      "epoch": 0.14024624855713735,
+      "grad_norm": 2.333397150039673,
+      "learning_rate": 0.00019369398270896403,
+      "loss": 1.0208,
+      "step": 729
+    },
+    {
+      "epoch": 0.1404386302424009,
+      "grad_norm": 2.1566736698150635,
+      "learning_rate": 0.0001936721882937779,
+      "loss": 1.1024,
+      "step": 730
+    },
+    {
+      "epoch": 0.14063101192766447,
+      "grad_norm": 2.2222979068756104,
+      "learning_rate": 0.0001936503575118231,
+      "loss": 1.026,
+      "step": 731
+    },
+    {
+      "epoch": 0.14082339361292806,
+      "grad_norm": 2.1811416149139404,
+      "learning_rate": 0.00019362849037157497,
+      "loss": 1.0402,
+      "step": 732
+    },
+    {
+      "epoch": 0.14101577529819162,
+      "grad_norm": 1.1387676000595093,
+      "learning_rate": 0.0001936065868815232,
+      "loss": 1.0632,
+      "step": 733
+    },
+    {
+      "epoch": 0.14120815698345518,
+      "grad_norm": 2.1770312786102295,
+      "learning_rate": 0.00019358464705017143,
+      "loss": 0.959,
+      "step": 734
+    },
+    {
+      "epoch": 0.14140053866871874,
+      "grad_norm": 1.678392767906189,
+      "learning_rate": 0.00019356267088603744,
+      "loss": 0.9501,
+      "step": 735
+    },
+    {
+      "epoch": 0.1415929203539823,
+      "grad_norm": 1.2155213356018066,
+      "learning_rate": 0.00019354065839765315,
+      "loss": 1.1381,
+      "step": 736
+    },
+    {
+      "epoch": 0.14178530203924586,
+      "grad_norm": 1.5049333572387695,
+      "learning_rate": 0.0001935186095935646,
+      "loss": 1.0793,
+      "step": 737
+    },
+    {
+      "epoch": 0.14197768372450942,
+      "grad_norm": 2.082159996032715,
+      "learning_rate": 0.00019349652448233187,
+      "loss": 1.0324,
+      "step": 738
+    },
+    {
+      "epoch": 0.14217006540977298,
+      "grad_norm": 2.2204439640045166,
+      "learning_rate": 0.00019347440307252911,
+      "loss": 1.0296,
+      "step": 739
+    },
+    {
+      "epoch": 0.14236244709503657,
+      "grad_norm": 2.348345994949341,
+      "learning_rate": 0.0001934522453727447,
+      "loss": 1.0063,
+      "step": 740
+    },
+    {
+      "epoch": 0.14255482878030012,
+      "grad_norm": 2.331740617752075,
+      "learning_rate": 0.00019343005139158096,
+      "loss": 0.9818,
+      "step": 741
+    },
+    {
+      "epoch": 0.14274721046556368,
+      "grad_norm": 1.599328875541687,
+      "learning_rate": 0.00019340782113765439,
+      "loss": 1.099,
+      "step": 742
+    },
+    {
+      "epoch": 0.14293959215082724,
+      "grad_norm": 1.5397690534591675,
+      "learning_rate": 0.00019338555461959552,
+      "loss": 1.0454,
+      "step": 743
+    },
+    {
+      "epoch": 0.1431319738360908,
+      "grad_norm": 2.1968979835510254,
+      "learning_rate": 0.000193363251846049,
+      "loss": 1.0023,
+      "step": 744
+    },
+    {
+      "epoch": 0.14332435552135436,
+      "grad_norm": 2.0155386924743652,
+      "learning_rate": 0.00019334091282567352,
+      "loss": 1.0304,
+      "step": 745
+    },
+    {
+      "epoch": 0.14351673720661792,
+      "grad_norm": 2.4428703784942627,
+      "learning_rate": 0.00019331853756714184,
+      "loss": 1.0802,
+      "step": 746
+    },
+    {
+      "epoch": 0.14370911889188148,
+      "grad_norm": 1.9965702295303345,
+      "learning_rate": 0.00019329612607914086,
+      "loss": 1.0822,
+      "step": 747
+    },
+    {
+      "epoch": 0.14390150057714507,
+      "grad_norm": 2.672553777694702,
+      "learning_rate": 0.00019327367837037142,
+      "loss": 1.0287,
+      "step": 748
+    },
+    {
+      "epoch": 0.14409388226240863,
+      "grad_norm": 2.200561285018921,
+      "learning_rate": 0.00019325119444954855,
+      "loss": 1.0125,
+      "step": 749
+    },
+    {
+      "epoch": 0.1442862639476722,
+      "grad_norm": 0.9463424682617188,
+      "learning_rate": 0.00019322867432540125,
+      "loss": 0.7108,
+      "step": 750
+    },
+    {
+      "epoch": 0.14447864563293575,
+      "grad_norm": 1.95294189453125,
+      "learning_rate": 0.00019320611800667265,
+      "loss": 1.0457,
+      "step": 751
+    },
+    {
+      "epoch": 0.1446710273181993,
+      "grad_norm": 1.9982850551605225,
+      "learning_rate": 0.00019318352550211986,
+      "loss": 0.9928,
+      "step": 752
+    },
+    {
+      "epoch": 0.14486340900346287,
+      "grad_norm": 1.888887643814087,
+      "learning_rate": 0.00019316089682051404,
+      "loss": 1.021,
+      "step": 753
+    },
+    {
+      "epoch": 0.14505579068872643,
+      "grad_norm": 2.5148606300354004,
+      "learning_rate": 0.00019313823197064043,
+      "loss": 0.9782,
+      "step": 754
+    },
+    {
+      "epoch": 0.14524817237399,
+      "grad_norm": 2.2383780479431152,
+      "learning_rate": 0.00019311553096129836,
+      "loss": 1.0831,
+      "step": 755
+    },
+    {
+      "epoch": 0.14544055405925355,
+      "grad_norm": 2.2670137882232666,
+      "learning_rate": 0.0001930927938013011,
+      "loss": 1.0488,
+      "step": 756
+    },
+    {
+      "epoch": 0.14563293574451713,
+      "grad_norm": 2.33215069770813,
+      "learning_rate": 0.000193070020499476,
+      "loss": 1.0528,
+      "step": 757
+    },
+    {
+      "epoch": 0.1458253174297807,
+      "grad_norm": 1.9963504076004028,
+      "learning_rate": 0.00019304721106466448,
+      "loss": 0.9693,
+      "step": 758
+    },
+    {
+      "epoch": 0.14601769911504425,
+      "grad_norm": 1.0200390815734863,
+      "learning_rate": 0.00019302436550572188,
+      "loss": 0.6819,
+      "step": 759
+    },
+    {
+      "epoch": 0.1462100808003078,
+      "grad_norm": 1.8697434663772583,
+      "learning_rate": 0.00019300148383151767,
+      "loss": 0.9854,
+      "step": 760
+    },
+    {
+      "epoch": 0.14640246248557137,
+      "grad_norm": 2.360827684402466,
+      "learning_rate": 0.00019297856605093533,
+      "loss": 1.0438,
+      "step": 761
+    },
+    {
+      "epoch": 0.14659484417083493,
+      "grad_norm": 1.7935510873794556,
+      "learning_rate": 0.00019295561217287226,
+      "loss": 1.035,
+      "step": 762
+    },
+    {
+      "epoch": 0.1467872258560985,
+      "grad_norm": 2.71183705329895,
+      "learning_rate": 0.00019293262220624002,
+      "loss": 1.057,
+      "step": 763
+    },
+    {
+      "epoch": 0.14697960754136205,
+      "grad_norm": 2.346677541732788,
+      "learning_rate": 0.00019290959615996407,
+      "loss": 1.0666,
+      "step": 764
+    },
+    {
+      "epoch": 0.14717198922662564,
+      "grad_norm": 2.1040353775024414,
+      "learning_rate": 0.0001928865340429839,
+      "loss": 1.1057,
+      "step": 765
+    },
+    {
+      "epoch": 0.1473643709118892,
+      "grad_norm": 2.0906779766082764,
+      "learning_rate": 0.00019286343586425307,
+      "loss": 1.0134,
+      "step": 766
+    },
+    {
+      "epoch": 0.14755675259715276,
+      "grad_norm": 2.4059629440307617,
+      "learning_rate": 0.00019284030163273905,
+      "loss": 1.054,
+      "step": 767
+    },
+    {
+      "epoch": 0.14774913428241632,
+      "grad_norm": 1.82550847530365,
+      "learning_rate": 0.00019281713135742333,
+      "loss": 1.0404,
+      "step": 768
+    },
+    {
+      "epoch": 0.14794151596767988,
+      "grad_norm": 1.9847731590270996,
+      "learning_rate": 0.00019279392504730145,
+      "loss": 1.0448,
+      "step": 769
+    },
+    {
+      "epoch": 0.14813389765294344,
+      "grad_norm": 2.140960454940796,
+      "learning_rate": 0.00019277068271138288,
+      "loss": 0.987,
+      "step": 770
+    },
+    {
+      "epoch": 0.148326279338207,
+      "grad_norm": 2.486142635345459,
+      "learning_rate": 0.00019274740435869106,
+      "loss": 1.0723,
+      "step": 771
+    },
+    {
+      "epoch": 0.14851866102347056,
+      "grad_norm": 2.4124739170074463,
+      "learning_rate": 0.0001927240899982635,
+      "loss": 1.0146,
+      "step": 772
+    },
+    {
+      "epoch": 0.14871104270873411,
+      "grad_norm": 1.9632935523986816,
+      "learning_rate": 0.00019270073963915162,
+      "loss": 1.0543,
+      "step": 773
+    },
+    {
+      "epoch": 0.1489034243939977,
+      "grad_norm": 2.3346195220947266,
+      "learning_rate": 0.00019267735329042086,
+      "loss": 1.0625,
+      "step": 774
+    },
+    {
+      "epoch": 0.14909580607926126,
+      "grad_norm": 1.5415757894515991,
+      "learning_rate": 0.00019265393096115055,
+      "loss": 1.0226,
+      "step": 775
+    },
+    {
+      "epoch": 0.14928818776452482,
+      "grad_norm": 2.3834292888641357,
+      "learning_rate": 0.00019263047266043407,
+      "loss": 1.0058,
+      "step": 776
+    },
+    {
+      "epoch": 0.14948056944978838,
+      "grad_norm": 2.533355474472046,
+      "learning_rate": 0.00019260697839737875,
+      "loss": 1.0652,
+      "step": 777
+    },
+    {
+      "epoch": 0.14967295113505194,
+      "grad_norm": 2.4399399757385254,
+      "learning_rate": 0.00019258344818110588,
+      "loss": 0.9541,
+      "step": 778
+    },
+    {
+      "epoch": 0.1498653328203155,
+      "grad_norm": 1.8456133604049683,
+      "learning_rate": 0.00019255988202075065,
+      "loss": 1.06,
+      "step": 779
+    },
+    {
+      "epoch": 0.15005771450557906,
+      "grad_norm": 2.3437342643737793,
+      "learning_rate": 0.0001925362799254623,
+      "loss": 1.0573,
+      "step": 780
+    },
+    {
+      "epoch": 0.15025009619084262,
+      "grad_norm": 2.710301399230957,
+      "learning_rate": 0.00019251264190440397,
+      "loss": 1.0951,
+      "step": 781
+    },
+    {
+      "epoch": 0.1504424778761062,
+      "grad_norm": 2.535231351852417,
+      "learning_rate": 0.00019248896796675275,
+      "loss": 1.0499,
+      "step": 782
+    },
+    {
+      "epoch": 0.15063485956136977,
+      "grad_norm": 1.9606608152389526,
+      "learning_rate": 0.0001924652581216997,
+      "loss": 1.0647,
+      "step": 783
+    },
+    {
+      "epoch": 0.15082724124663333,
+      "grad_norm": 2.816706895828247,
+      "learning_rate": 0.00019244151237844974,
+      "loss": 1.0915,
+      "step": 784
+    },
+    {
+      "epoch": 0.15101962293189689,
+      "grad_norm": 2.2878031730651855,
+      "learning_rate": 0.00019241773074622182,
+      "loss": 1.0542,
+      "step": 785
+    },
+    {
+      "epoch": 0.15121200461716044,
+      "grad_norm": 2.7728383541107178,
+      "learning_rate": 0.00019239391323424881,
+      "loss": 1.0049,
+      "step": 786
+    },
+    {
+      "epoch": 0.151404386302424,
+      "grad_norm": 2.7083451747894287,
+      "learning_rate": 0.00019237005985177748,
+      "loss": 1.089,
+      "step": 787
+    },
+    {
+      "epoch": 0.15159676798768756,
+      "grad_norm": 2.2165157794952393,
+      "learning_rate": 0.00019234617060806848,
+      "loss": 1.0554,
+      "step": 788
+    },
+    {
+      "epoch": 0.15178914967295112,
+      "grad_norm": 1.979903221130371,
+      "learning_rate": 0.0001923222455123965,
+      "loss": 1.0583,
+      "step": 789
+    },
+    {
+      "epoch": 0.1519815313582147,
+      "grad_norm": 1.7565046548843384,
+      "learning_rate": 0.00019229828457405007,
+      "loss": 1.0114,
+      "step": 790
+    },
+    {
+      "epoch": 0.15217391304347827,
+      "grad_norm": 1.8657946586608887,
+      "learning_rate": 0.00019227428780233165,
+      "loss": 1.0772,
+      "step": 791
+    },
+    {
+      "epoch": 0.15236629472874183,
+      "grad_norm": 2.202068567276001,
+      "learning_rate": 0.00019225025520655758,
+      "loss": 1.0348,
+      "step": 792
+    },
+    {
+      "epoch": 0.1525586764140054,
+      "grad_norm": 2.0710840225219727,
+      "learning_rate": 0.0001922261867960582,
+      "loss": 1.0055,
+      "step": 793
+    },
+    {
+      "epoch": 0.15275105809926895,
+      "grad_norm": 2.1057894229888916,
+      "learning_rate": 0.00019220208258017763,
+      "loss": 1.0381,
+      "step": 794
+    },
+    {
+      "epoch": 0.1529434397845325,
+      "grad_norm": 2.326620101928711,
+      "learning_rate": 0.00019217794256827403,
+      "loss": 1.072,
+      "step": 795
+    },
+    {
+      "epoch": 0.15313582146979607,
+      "grad_norm": 2.0878610610961914,
+      "learning_rate": 0.0001921537667697193,
+      "loss": 1.0619,
+      "step": 796
+    },
+    {
+      "epoch": 0.15332820315505963,
+      "grad_norm": 2.2210214138031006,
+      "learning_rate": 0.0001921295551938994,
+      "loss": 1.0806,
+      "step": 797
+    },
+    {
+      "epoch": 0.1535205848403232,
+      "grad_norm": 2.0698788166046143,
+      "learning_rate": 0.00019210530785021405,
+      "loss": 1.0719,
+      "step": 798
+    },
+    {
+      "epoch": 0.15371296652558677,
+      "grad_norm": 1.8600722551345825,
+      "learning_rate": 0.0001920810247480769,
+      "loss": 1.0814,
+      "step": 799
+    },
+    {
+      "epoch": 0.15390534821085033,
+      "grad_norm": 1.7580174207687378,
+      "learning_rate": 0.0001920567058969155,
+      "loss": 1.0268,
+      "step": 800
+    },
+    {
+      "epoch": 0.1540977298961139,
+      "grad_norm": 1.581208348274231,
+      "learning_rate": 0.0001920323513061713,
+      "loss": 1.0964,
+      "step": 801
+    },
+    {
+      "epoch": 0.15429011158137745,
+      "grad_norm": 2.035865306854248,
+      "learning_rate": 0.00019200796098529956,
+      "loss": 1.0313,
+      "step": 802
+    },
+    {
+      "epoch": 0.154482493266641,
+      "grad_norm": 2.494929790496826,
+      "learning_rate": 0.00019198353494376938,
+      "loss": 1.072,
+      "step": 803
+    },
+    {
+      "epoch": 0.15467487495190457,
+      "grad_norm": 2.6849095821380615,
+      "learning_rate": 0.00019195907319106392,
+      "loss": 0.9587,
+      "step": 804
+    },
+    {
+      "epoch": 0.15486725663716813,
+      "grad_norm": 2.106247663497925,
+      "learning_rate": 0.00019193457573667995,
+      "loss": 1.0654,
+      "step": 805
+    },
+    {
+      "epoch": 0.1550596383224317,
+      "grad_norm": 1.8343539237976074,
+      "learning_rate": 0.0001919100425901283,
+      "loss": 1.0783,
+      "step": 806
+    },
+    {
+      "epoch": 0.15525202000769528,
+      "grad_norm": 2.307262897491455,
+      "learning_rate": 0.00019188547376093354,
+      "loss": 1.0212,
+      "step": 807
+    },
+    {
+      "epoch": 0.15544440169295884,
+      "grad_norm": 2.469435214996338,
+      "learning_rate": 0.0001918608692586342,
+      "loss": 1.0256,
+      "step": 808
+    },
+    {
+      "epoch": 0.1556367833782224,
+      "grad_norm": 2.3533775806427,
+      "learning_rate": 0.0001918362290927825,
+      "loss": 1.0574,
+      "step": 809
+    },
+    {
+      "epoch": 0.15582916506348596,
+      "grad_norm": 2.31184983253479,
+      "learning_rate": 0.0001918115532729447,
+      "loss": 1.0947,
+      "step": 810
+    },
+    {
+      "epoch": 0.15602154674874952,
+      "grad_norm": 2.3351171016693115,
+      "learning_rate": 0.00019178684180870073,
+      "loss": 1.0534,
+      "step": 811
+    },
+    {
+      "epoch": 0.15621392843401308,
+      "grad_norm": 1.9652353525161743,
+      "learning_rate": 0.00019176209470964445,
+      "loss": 1.0532,
+      "step": 812
+    },
+    {
+      "epoch": 0.15640631011927664,
+      "grad_norm": 2.4892051219940186,
+      "learning_rate": 0.00019173731198538353,
+      "loss": 1.0066,
+      "step": 813
+    },
+    {
+      "epoch": 0.1565986918045402,
+      "grad_norm": 2.045156478881836,
+      "learning_rate": 0.00019171249364553954,
+      "loss": 1.0291,
+      "step": 814
+    },
+    {
+      "epoch": 0.15679107348980378,
+      "grad_norm": 2.171950101852417,
+      "learning_rate": 0.00019168763969974772,
+      "loss": 1.0366,
+      "step": 815
+    },
+    {
+      "epoch": 0.15698345517506734,
+      "grad_norm": 1.3550125360488892,
+      "learning_rate": 0.0001916627501576573,
+      "loss": 1.1019,
+      "step": 816
+    },
+    {
+      "epoch": 0.1571758368603309,
+      "grad_norm": 1.9300423860549927,
+      "learning_rate": 0.0001916378250289312,
+      "loss": 0.9413,
+      "step": 817
+    },
+    {
+      "epoch": 0.15736821854559446,
+      "grad_norm": 1.3997080326080322,
+      "learning_rate": 0.00019161286432324624,
+      "loss": 1.0615,
+      "step": 818
+    },
+    {
+      "epoch": 0.15756060023085802,
+      "grad_norm": 2.1858975887298584,
+      "learning_rate": 0.00019158786805029306,
+      "loss": 0.9553,
+      "step": 819
+    },
+    {
+      "epoch": 0.15775298191612158,
+      "grad_norm": 2.0630860328674316,
+      "learning_rate": 0.00019156283621977603,
+      "loss": 1.087,
+      "step": 820
+    },
+    {
+      "epoch": 0.15794536360138514,
+      "grad_norm": 2.501772880554199,
+      "learning_rate": 0.00019153776884141336,
+      "loss": 1.06,
+      "step": 821
+    },
+    {
+      "epoch": 0.1581377452866487,
+      "grad_norm": 2.3481626510620117,
+      "learning_rate": 0.00019151266592493712,
+      "loss": 1.1193,
+      "step": 822
+    },
+    {
+      "epoch": 0.15833012697191226,
+      "grad_norm": 2.1778578758239746,
+      "learning_rate": 0.00019148752748009306,
+      "loss": 1.0132,
+      "step": 823
+    },
+    {
+      "epoch": 0.15852250865717585,
+      "grad_norm": 1.7775094509124756,
+      "learning_rate": 0.0001914623535166408,
+      "loss": 0.9387,
+      "step": 824
+    },
+    {
+      "epoch": 0.1587148903424394,
+      "grad_norm": 1.1277555227279663,
+      "learning_rate": 0.00019143714404435381,
+      "loss": 0.7813,
+      "step": 825
+    },
+    {
+      "epoch": 0.15890727202770297,
+      "grad_norm": 2.8418593406677246,
+      "learning_rate": 0.00019141189907301921,
+      "loss": 1.0781,
+      "step": 826
+    },
+    {
+      "epoch": 0.15909965371296653,
+      "grad_norm": 2.066009521484375,
+      "learning_rate": 0.000191386618612438,
+      "loss": 0.9447,
+      "step": 827
+    },
+    {
+      "epoch": 0.1592920353982301,
+      "grad_norm": 1.5063756704330444,
+      "learning_rate": 0.00019136130267242488,
+      "loss": 1.0248,
+      "step": 828
+    },
+    {
+      "epoch": 0.15948441708349365,
+      "grad_norm": 2.1444802284240723,
+      "learning_rate": 0.00019133595126280846,
+      "loss": 1.0245,
+      "step": 829
+    },
+    {
+      "epoch": 0.1596767987687572,
+      "grad_norm": 2.159559726715088,
+      "learning_rate": 0.00019131056439343093,
+      "loss": 1.0674,
+      "step": 830
+    },
+    {
+      "epoch": 0.15986918045402077,
+      "grad_norm": 1.086140513420105,
+      "learning_rate": 0.0001912851420741484,
+      "loss": 1.0169,
+      "step": 831
+    },
+    {
+      "epoch": 0.16006156213928435,
+      "grad_norm": 1.685892939567566,
+      "learning_rate": 0.00019125968431483065,
+      "loss": 1.0799,
+      "step": 832
+    },
+    {
+      "epoch": 0.1602539438245479,
+      "grad_norm": 2.9055469036102295,
+      "learning_rate": 0.0001912341911253613,
+      "loss": 0.9926,
+      "step": 833
+    },
+    {
+      "epoch": 0.16044632550981147,
+      "grad_norm": 2.1495256423950195,
+      "learning_rate": 0.0001912086625156377,
+      "loss": 1.1149,
+      "step": 834
+    },
+    {
+      "epoch": 0.16063870719507503,
+      "grad_norm": 2.7917542457580566,
+      "learning_rate": 0.00019118309849557087,
+      "loss": 1.1064,
+      "step": 835
+    },
+    {
+      "epoch": 0.1608310888803386,
+      "grad_norm": 1.934350848197937,
+      "learning_rate": 0.0001911574990750857,
+      "loss": 0.9863,
+      "step": 836
+    },
+    {
+      "epoch": 0.16102347056560215,
+      "grad_norm": 2.171539068222046,
+      "learning_rate": 0.00019113186426412073,
+      "loss": 0.9778,
+      "step": 837
+    },
+    {
+      "epoch": 0.1612158522508657,
+      "grad_norm": 2.367551565170288,
+      "learning_rate": 0.0001911061940726283,
+      "loss": 1.0405,
+      "step": 838
+    },
+    {
+      "epoch": 0.16140823393612927,
+      "grad_norm": 2.2148897647857666,
+      "learning_rate": 0.00019108048851057446,
+      "loss": 1.0546,
+      "step": 839
+    },
+    {
+      "epoch": 0.16160061562139286,
+      "grad_norm": 1.8189516067504883,
+      "learning_rate": 0.00019105474758793895,
+      "loss": 1.0635,
+      "step": 840
+    },
+    {
+      "epoch": 0.16179299730665642,
+      "grad_norm": 2.425858497619629,
+      "learning_rate": 0.00019102897131471536,
+      "loss": 1.0616,
+      "step": 841
+    },
+    {
+      "epoch": 0.16198537899191998,
+      "grad_norm": 2.0700628757476807,
+      "learning_rate": 0.00019100315970091086,
+      "loss": 1.015,
+      "step": 842
+    },
+    {
+      "epoch": 0.16217776067718354,
+      "grad_norm": 1.871505856513977,
+      "learning_rate": 0.00019097731275654643,
+      "loss": 1.05,
+      "step": 843
+    },
+    {
+      "epoch": 0.1623701423624471,
+      "grad_norm": 2.0348622798919678,
+      "learning_rate": 0.00019095143049165678,
+      "loss": 1.0421,
+      "step": 844
+    },
+    {
+      "epoch": 0.16256252404771065,
+      "grad_norm": 2.674696207046509,
+      "learning_rate": 0.00019092551291629027,
+      "loss": 1.0788,
+      "step": 845
+    },
+    {
+      "epoch": 0.16275490573297421,
+      "grad_norm": 2.392015218734741,
+      "learning_rate": 0.00019089956004050894,
+      "loss": 1.0498,
+      "step": 846
+    },
+    {
+      "epoch": 0.16294728741823777,
+      "grad_norm": 2.4924919605255127,
+      "learning_rate": 0.00019087357187438868,
+      "loss": 1.0705,
+      "step": 847
+    },
+    {
+      "epoch": 0.16313966910350133,
+      "grad_norm": 2.112170934677124,
+      "learning_rate": 0.0001908475484280189,
+      "loss": 1.0652,
+      "step": 848
+    },
+    {
+      "epoch": 0.16333205078876492,
+      "grad_norm": 2.090059995651245,
+      "learning_rate": 0.0001908214897115029,
+      "loss": 1.0154,
+      "step": 849
+    },
+    {
+      "epoch": 0.16352443247402848,
+      "grad_norm": 2.358523368835449,
+      "learning_rate": 0.00019079539573495748,
+      "loss": 1.0776,
+      "step": 850
+    },
+    {
+      "epoch": 0.16371681415929204,
+      "grad_norm": 2.3671445846557617,
+      "learning_rate": 0.0001907692665085133,
+      "loss": 0.9816,
+      "step": 851
+    },
+    {
+      "epoch": 0.1639091958445556,
+      "grad_norm": 2.538585901260376,
+      "learning_rate": 0.00019074310204231457,
+      "loss": 1.0401,
+      "step": 852
+    },
+    {
+      "epoch": 0.16410157752981916,
+      "grad_norm": 1.2981934547424316,
+      "learning_rate": 0.0001907169023465192,
+      "loss": 0.9892,
+      "step": 853
+    },
+    {
+      "epoch": 0.16429395921508272,
+      "grad_norm": 1.668711543083191,
+      "learning_rate": 0.00019069066743129893,
+      "loss": 1.1255,
+      "step": 854
+    },
+    {
+      "epoch": 0.16448634090034628,
+      "grad_norm": 2.354081630706787,
+      "learning_rate": 0.000190664397306839,
+      "loss": 1.0605,
+      "step": 855
+    },
+    {
+      "epoch": 0.16467872258560984,
+      "grad_norm": 1.2690718173980713,
+      "learning_rate": 0.00019063809198333832,
+      "loss": 1.0339,
+      "step": 856
+    },
+    {
+      "epoch": 0.16487110427087343,
+      "grad_norm": 2.5924296379089355,
+      "learning_rate": 0.0001906117514710096,
+      "loss": 1.0166,
+      "step": 857
+    },
+    {
+      "epoch": 0.16506348595613698,
+      "grad_norm": 1.0021717548370361,
+      "learning_rate": 0.00019058537578007907,
+      "loss": 0.7129,
+      "step": 858
+    },
+    {
+      "epoch": 0.16525586764140054,
+      "grad_norm": 2.0402424335479736,
+      "learning_rate": 0.00019055896492078676,
+      "loss": 1.0785,
+      "step": 859
+    },
+    {
+      "epoch": 0.1654482493266641,
+      "grad_norm": 1.9293748140335083,
+      "learning_rate": 0.0001905325189033862,
+      "loss": 1.0107,
+      "step": 860
+    },
+    {
+      "epoch": 0.16564063101192766,
+      "grad_norm": 2.1741909980773926,
+      "learning_rate": 0.00019050603773814468,
+      "loss": 1.0523,
+      "step": 861
+    },
+    {
+      "epoch": 0.16583301269719122,
+      "grad_norm": 1.660751223564148,
+      "learning_rate": 0.0001904795214353431,
+      "loss": 1.0741,
+      "step": 862
+    },
+    {
+      "epoch": 0.16602539438245478,
+      "grad_norm": 2.036715269088745,
+      "learning_rate": 0.000190452970005276,
+      "loss": 1.0795,
+      "step": 863
+    },
+    {
+      "epoch": 0.16621777606771834,
+      "grad_norm": 2.4203226566314697,
+      "learning_rate": 0.00019042638345825156,
+      "loss": 1.0792,
+      "step": 864
+    },
+    {
+      "epoch": 0.16641015775298193,
+      "grad_norm": 0.9645435810089111,
+      "learning_rate": 0.00019039976180459158,
+      "loss": 1.1254,
+      "step": 865
+    },
+    {
+      "epoch": 0.1666025394382455,
+      "grad_norm": 1.7508660554885864,
+      "learning_rate": 0.00019037310505463153,
+      "loss": 1.0349,
+      "step": 866
+    },
+    {
+      "epoch": 0.16679492112350905,
+      "grad_norm": 2.098126173019409,
+      "learning_rate": 0.00019034641321872045,
+      "loss": 1.0926,
+      "step": 867
+    },
+    {
+      "epoch": 0.1669873028087726,
+      "grad_norm": 1.9512134790420532,
+      "learning_rate": 0.00019031968630722103,
+      "loss": 0.9759,
+      "step": 868
+    },
+    {
+      "epoch": 0.16717968449403617,
+      "grad_norm": 1.4637136459350586,
+      "learning_rate": 0.0001902929243305096,
+      "loss": 1.0461,
+      "step": 869
+    },
+    {
+      "epoch": 0.16737206617929973,
+      "grad_norm": 2.370793581008911,
+      "learning_rate": 0.00019026612729897607,
+      "loss": 0.9615,
+      "step": 870
+    },
+    {
+      "epoch": 0.1675644478645633,
+      "grad_norm": 2.4272987842559814,
+      "learning_rate": 0.00019023929522302396,
+      "loss": 1.0354,
+      "step": 871
+    },
+    {
+      "epoch": 0.16775682954982685,
+      "grad_norm": 2.1404869556427,
+      "learning_rate": 0.00019021242811307043,
+      "loss": 1.0989,
+      "step": 872
+    },
+    {
+      "epoch": 0.1679492112350904,
+      "grad_norm": 2.2725977897644043,
+      "learning_rate": 0.0001901855259795462,
+      "loss": 0.9888,
+      "step": 873
+    },
+    {
+      "epoch": 0.168141592920354,
+      "grad_norm": 1.5183085203170776,
+      "learning_rate": 0.00019015858883289556,
+      "loss": 0.9729,
+      "step": 874
+    },
+    {
+      "epoch": 0.16833397460561755,
+      "grad_norm": 2.10738205909729,
+      "learning_rate": 0.00019013161668357653,
+      "loss": 1.031,
+      "step": 875
+    },
+    {
+      "epoch": 0.1685263562908811,
+      "grad_norm": 1.7453025579452515,
+      "learning_rate": 0.00019010460954206056,
+      "loss": 1.1123,
+      "step": 876
+    },
+    {
+      "epoch": 0.16871873797614467,
+      "grad_norm": 2.3236074447631836,
+      "learning_rate": 0.00019007756741883282,
+      "loss": 1.0156,
+      "step": 877
+    },
+    {
+      "epoch": 0.16891111966140823,
+      "grad_norm": 2.997732639312744,
+      "learning_rate": 0.0001900504903243919,
+      "loss": 1.0449,
+      "step": 878
+    },
+    {
+      "epoch": 0.1691035013466718,
+      "grad_norm": 2.0996599197387695,
+      "learning_rate": 0.00019002337826925012,
+      "loss": 1.0681,
+      "step": 879
+    },
+    {
+      "epoch": 0.16929588303193535,
+      "grad_norm": 1.0544583797454834,
+      "learning_rate": 0.0001899962312639333,
+      "loss": 0.7653,
+      "step": 880
+    },
+    {
+      "epoch": 0.1694882647171989,
+      "grad_norm": 1.7731493711471558,
+      "learning_rate": 0.00018996904931898082,
+      "loss": 0.9776,
+      "step": 881
+    },
+    {
+      "epoch": 0.1696806464024625,
+      "grad_norm": 2.4669175148010254,
+      "learning_rate": 0.0001899418324449457,
+      "loss": 1.1207,
+      "step": 882
+    },
+    {
+      "epoch": 0.16987302808772606,
+      "grad_norm": 2.1910853385925293,
+      "learning_rate": 0.00018991458065239445,
+      "loss": 1.1214,
+      "step": 883
+    },
+    {
+      "epoch": 0.17006540977298962,
+      "grad_norm": 2.115065574645996,
+      "learning_rate": 0.0001898872939519071,
+      "loss": 1.0221,
+      "step": 884
+    },
+    {
+      "epoch": 0.17025779145825318,
+      "grad_norm": 2.187981605529785,
+      "learning_rate": 0.00018985997235407734,
+      "loss": 1.0631,
+      "step": 885
+    },
+    {
+      "epoch": 0.17045017314351674,
+      "grad_norm": 2.674283027648926,
+      "learning_rate": 0.00018983261586951237,
+      "loss": 1.0269,
+      "step": 886
+    },
+    {
+      "epoch": 0.1706425548287803,
+      "grad_norm": 2.3584272861480713,
+      "learning_rate": 0.00018980522450883286,
+      "loss": 1.0105,
+      "step": 887
+    },
+    {
+      "epoch": 0.17083493651404386,
+      "grad_norm": 2.323780059814453,
+      "learning_rate": 0.00018977779828267311,
+      "loss": 1.1024,
+      "step": 888
+    },
+    {
+      "epoch": 0.17102731819930742,
+      "grad_norm": 1.73771333694458,
+      "learning_rate": 0.00018975033720168094,
+      "loss": 1.0709,
+      "step": 889
+    },
+    {
+      "epoch": 0.17121969988457098,
+      "grad_norm": 1.7440807819366455,
+      "learning_rate": 0.00018972284127651768,
+      "loss": 1.0347,
+      "step": 890
+    },
+    {
+      "epoch": 0.17141208156983456,
+      "grad_norm": 0.8926966190338135,
+      "learning_rate": 0.00018969531051785818,
+      "loss": 0.7244,
+      "step": 891
+    },
+    {
+      "epoch": 0.17160446325509812,
+      "grad_norm": 2.244826555252075,
+      "learning_rate": 0.00018966774493639084,
+      "loss": 1.0541,
+      "step": 892
+    },
+    {
+      "epoch": 0.17179684494036168,
+      "grad_norm": 1.157251238822937,
+      "learning_rate": 0.00018964014454281758,
+      "loss": 0.7671,
+      "step": 893
+    },
+    {
+      "epoch": 0.17198922662562524,
+      "grad_norm": 1.1734495162963867,
+      "learning_rate": 0.00018961250934785377,
+      "loss": 1.0806,
+      "step": 894
+    },
+    {
+      "epoch": 0.1721816083108888,
+      "grad_norm": 1.179213523864746,
+      "learning_rate": 0.00018958483936222842,
+      "loss": 0.7551,
+      "step": 895
+    },
+    {
+      "epoch": 0.17237398999615236,
+      "grad_norm": 2.381864070892334,
+      "learning_rate": 0.0001895571345966839,
+      "loss": 1.0433,
+      "step": 896
+    },
+    {
+      "epoch": 0.17256637168141592,
+      "grad_norm": 2.217658042907715,
+      "learning_rate": 0.0001895293950619762,
+      "loss": 1.1009,
+      "step": 897
+    },
+    {
+      "epoch": 0.17275875336667948,
+      "grad_norm": 1.350746989250183,
+      "learning_rate": 0.00018950162076887477,
+      "loss": 1.0388,
+      "step": 898
+    },
+    {
+      "epoch": 0.17295113505194307,
+      "grad_norm": 2.1843059062957764,
+      "learning_rate": 0.00018947381172816248,
+      "loss": 1.0014,
+      "step": 899
+    },
+    {
+      "epoch": 0.17314351673720663,
+      "grad_norm": 2.5313618183135986,
+      "learning_rate": 0.00018944596795063582,
+      "loss": 1.0602,
+      "step": 900
+    },
+    {
+      "epoch": 0.17333589842247019,
+      "grad_norm": 2.009249448776245,
+      "learning_rate": 0.0001894180894471047,
+      "loss": 1.0167,
+      "step": 901
+    },
+    {
+      "epoch": 0.17352828010773375,
+      "grad_norm": 2.2480275630950928,
+      "learning_rate": 0.00018939017622839253,
+      "loss": 0.9758,
+      "step": 902
+    },
+    {
+      "epoch": 0.1737206617929973,
+      "grad_norm": 1.1013227701187134,
+      "learning_rate": 0.0001893622283053361,
+      "loss": 0.6714,
+      "step": 903
+    },
+    {
+      "epoch": 0.17391304347826086,
+      "grad_norm": 2.1999173164367676,
+      "learning_rate": 0.00018933424568878584,
+      "loss": 0.9874,
+      "step": 904
+    },
+    {
+      "epoch": 0.17410542516352442,
+      "grad_norm": 2.791902780532837,
+      "learning_rate": 0.00018930622838960555,
+      "loss": 1.0608,
+      "step": 905
+    },
+    {
+      "epoch": 0.17429780684878798,
+      "grad_norm": 2.152021884918213,
+      "learning_rate": 0.00018927817641867245,
+      "loss": 1.0826,
+      "step": 906
+    },
+    {
+      "epoch": 0.17449018853405157,
+      "grad_norm": 2.624910354614258,
+      "learning_rate": 0.00018925008978687736,
+      "loss": 1.0208,
+      "step": 907
+    },
+    {
+      "epoch": 0.17468257021931513,
+      "grad_norm": 2.2706263065338135,
+      "learning_rate": 0.00018922196850512446,
+      "loss": 1.0359,
+      "step": 908
+    },
+    {
+      "epoch": 0.1748749519045787,
+      "grad_norm": 2.2981173992156982,
+      "learning_rate": 0.00018919381258433133,
+      "loss": 1.0391,
+      "step": 909
+    },
+    {
+      "epoch": 0.17506733358984225,
+      "grad_norm": 2.0417447090148926,
+      "learning_rate": 0.00018916562203542917,
+      "loss": 1.0633,
+      "step": 910
+    },
+    {
+      "epoch": 0.1752597152751058,
+      "grad_norm": 2.7648067474365234,
+      "learning_rate": 0.00018913739686936246,
+      "loss": 0.9963,
+      "step": 911
+    },
+    {
+      "epoch": 0.17545209696036937,
+      "grad_norm": 2.741966485977173,
+      "learning_rate": 0.00018910913709708918,
+      "loss": 0.9375,
+      "step": 912
+    },
+    {
+      "epoch": 0.17564447864563293,
+      "grad_norm": 2.2619497776031494,
+      "learning_rate": 0.00018908084272958078,
+      "loss": 1.0243,
+      "step": 913
+    },
+    {
+      "epoch": 0.1758368603308965,
+      "grad_norm": 1.7858800888061523,
+      "learning_rate": 0.00018905251377782206,
+      "loss": 1.01,
+      "step": 914
+    },
+    {
+      "epoch": 0.17602924201616005,
+      "grad_norm": 1.1395015716552734,
+      "learning_rate": 0.00018902415025281134,
+      "loss": 0.8043,
+      "step": 915
+    },
+    {
+      "epoch": 0.17622162370142364,
+      "grad_norm": 2.292489767074585,
+      "learning_rate": 0.00018899575216556032,
+      "loss": 1.1011,
+      "step": 916
+    },
+    {
+      "epoch": 0.1764140053866872,
+      "grad_norm": 2.0286691188812256,
+      "learning_rate": 0.00018896731952709408,
+      "loss": 1.1303,
+      "step": 917
+    },
+    {
+      "epoch": 0.17660638707195075,
+      "grad_norm": 1.531886100769043,
+      "learning_rate": 0.00018893885234845116,
+      "loss": 1.0674,
+      "step": 918
+    },
+    {
+      "epoch": 0.17679876875721431,
+      "grad_norm": 2.733804941177368,
+      "learning_rate": 0.0001889103506406835,
+      "loss": 1.0207,
+      "step": 919
+    },
+    {
+      "epoch": 0.17699115044247787,
+      "grad_norm": 2.147122621536255,
+      "learning_rate": 0.00018888181441485647,
+      "loss": 1.0399,
+      "step": 920
+    },
+    {
+      "epoch": 0.17718353212774143,
+      "grad_norm": 2.1837103366851807,
+      "learning_rate": 0.0001888532436820488,
+      "loss": 1.028,
+      "step": 921
+    },
+    {
+      "epoch": 0.177375913813005,
+      "grad_norm": 1.9035612344741821,
+      "learning_rate": 0.0001888246384533526,
+      "loss": 1.0424,
+      "step": 922
+    },
+    {
+      "epoch": 0.17756829549826855,
+      "grad_norm": 2.0806283950805664,
+      "learning_rate": 0.00018879599873987345,
+      "loss": 0.9986,
+      "step": 923
+    },
+    {
+      "epoch": 0.17776067718353214,
+      "grad_norm": 2.1577892303466797,
+      "learning_rate": 0.00018876732455273024,
+      "loss": 1.023,
+      "step": 924
+    },
+    {
+      "epoch": 0.1779530588687957,
+      "grad_norm": 2.2184789180755615,
+      "learning_rate": 0.00018873861590305526,
+      "loss": 1.0573,
+      "step": 925
+    },
+    {
+      "epoch": 0.17814544055405926,
+      "grad_norm": 2.0347344875335693,
+      "learning_rate": 0.00018870987280199426,
+      "loss": 1.0255,
+      "step": 926
+    },
+    {
+      "epoch": 0.17833782223932282,
+      "grad_norm": 1.4399735927581787,
+      "learning_rate": 0.00018868109526070628,
+      "loss": 0.9907,
+      "step": 927
+    },
+    {
+      "epoch": 0.17853020392458638,
+      "grad_norm": 1.8984928131103516,
+      "learning_rate": 0.00018865228329036372,
+      "loss": 1.0244,
+      "step": 928
+    },
+    {
+      "epoch": 0.17872258560984994,
+      "grad_norm": 1.9499598741531372,
+      "learning_rate": 0.0001886234369021524,
+      "loss": 1.0075,
+      "step": 929
+    },
+    {
+      "epoch": 0.1789149672951135,
+      "grad_norm": 1.1615753173828125,
+      "learning_rate": 0.0001885945561072715,
+      "loss": 0.7744,
+      "step": 930
+    },
+    {
+      "epoch": 0.17910734898037706,
+      "grad_norm": 2.1331787109375,
+      "learning_rate": 0.0001885656409169335,
+      "loss": 1.0775,
+      "step": 931
+    },
+    {
+      "epoch": 0.17929973066564064,
+      "grad_norm": 2.1473729610443115,
+      "learning_rate": 0.0001885366913423643,
+      "loss": 1.089,
+      "step": 932
+    },
+    {
+      "epoch": 0.1794921123509042,
+      "grad_norm": 2.159719944000244,
+      "learning_rate": 0.0001885077073948031,
+      "loss": 1.0129,
+      "step": 933
+    },
+    {
+      "epoch": 0.17968449403616776,
+      "grad_norm": 2.2321410179138184,
+      "learning_rate": 0.0001884786890855025,
+      "loss": 1.0632,
+      "step": 934
+    },
+    {
+      "epoch": 0.17987687572143132,
+      "grad_norm": 2.750988245010376,
+      "learning_rate": 0.00018844963642572837,
+      "loss": 1.1293,
+      "step": 935
+    },
+    {
+      "epoch": 0.18006925740669488,
+      "grad_norm": 1.7408156394958496,
+      "learning_rate": 0.00018842054942675996,
+      "loss": 1.0551,
+      "step": 936
+    },
+    {
+      "epoch": 0.18026163909195844,
+      "grad_norm": 1.6306977272033691,
+      "learning_rate": 0.00018839142809988987,
+      "loss": 0.9784,
+      "step": 937
+    },
+    {
+      "epoch": 0.180454020777222,
+      "grad_norm": 2.212202787399292,
+      "learning_rate": 0.00018836227245642398,
+      "loss": 1.079,
+      "step": 938
+    },
+    {
+      "epoch": 0.18064640246248556,
+      "grad_norm": 2.2630717754364014,
+      "learning_rate": 0.00018833308250768154,
+      "loss": 0.9975,
+      "step": 939
+    },
+    {
+      "epoch": 0.18083878414774912,
+      "grad_norm": 1.15382719039917,
+      "learning_rate": 0.00018830385826499506,
+      "loss": 0.7531,
+      "step": 940
+    },
+    {
+      "epoch": 0.1810311658330127,
+      "grad_norm": 2.393907308578491,
+      "learning_rate": 0.0001882745997397104,
+      "loss": 1.0157,
+      "step": 941
+    },
+    {
+      "epoch": 0.18122354751827627,
+      "grad_norm": 2.8717808723449707,
+      "learning_rate": 0.00018824530694318674,
+      "loss": 0.9606,
+      "step": 942
+    },
+    {
+      "epoch": 0.18141592920353983,
+      "grad_norm": 2.8729214668273926,
+      "learning_rate": 0.00018821597988679658,
+      "loss": 1.0522,
+      "step": 943
+    },
+    {
+      "epoch": 0.1816083108888034,
+      "grad_norm": 1.4063445329666138,
+      "learning_rate": 0.0001881866185819256,
+      "loss": 1.0649,
+      "step": 944
+    },
+    {
+      "epoch": 0.18180069257406695,
+      "grad_norm": 0.9870543479919434,
+      "learning_rate": 0.000188157223039973,
+      "loss": 1.0183,
+      "step": 945
+    },
+    {
+      "epoch": 0.1819930742593305,
+      "grad_norm": 1.3965227603912354,
+      "learning_rate": 0.00018812779327235103,
+      "loss": 1.0525,
+      "step": 946
+    },
+    {
+      "epoch": 0.18218545594459407,
+      "grad_norm": 2.246839761734009,
+      "learning_rate": 0.0001880983292904854,
+      "loss": 1.0233,
+      "step": 947
+    },
+    {
+      "epoch": 0.18237783762985763,
+      "grad_norm": 1.137377142906189,
+      "learning_rate": 0.000188068831105815,
+      "loss": 1.1333,
+      "step": 948
+    },
+    {
+      "epoch": 0.1825702193151212,
+      "grad_norm": 1.3836743831634521,
+      "learning_rate": 0.00018803929872979213,
+      "loss": 1.0447,
+      "step": 949
+    },
+    {
+      "epoch": 0.18276260100038477,
+      "grad_norm": 1.8202382326126099,
+      "learning_rate": 0.00018800973217388215,
+      "loss": 1.0021,
+      "step": 950
+    },
+    {
+      "epoch": 0.18295498268564833,
+      "grad_norm": 1.7602394819259644,
+      "learning_rate": 0.0001879801314495639,
+      "loss": 1.0637,
+      "step": 951
+    },
+    {
+      "epoch": 0.1831473643709119,
+      "grad_norm": 2.3875410556793213,
+      "learning_rate": 0.00018795049656832936,
+      "loss": 1.1076,
+      "step": 952
+    },
+    {
+      "epoch": 0.18333974605617545,
+      "grad_norm": 1.834930419921875,
+      "learning_rate": 0.00018792082754168386,
+      "loss": 1.1426,
+      "step": 953
+    },
+    {
+      "epoch": 0.183532127741439,
+      "grad_norm": 2.6041758060455322,
+      "learning_rate": 0.00018789112438114588,
+      "loss": 1.0098,
+      "step": 954
+    },
+    {
+      "epoch": 0.18372450942670257,
+      "grad_norm": 2.5503475666046143,
+      "learning_rate": 0.00018786138709824725,
+      "loss": 0.98,
+      "step": 955
+    },
+    {
+      "epoch": 0.18391689111196613,
+      "grad_norm": 2.220343589782715,
+      "learning_rate": 0.00018783161570453297,
+      "loss": 0.9714,
+      "step": 956
+    },
+    {
+      "epoch": 0.18410927279722972,
+      "grad_norm": 1.1839293241500854,
+      "learning_rate": 0.0001878018102115614,
+      "loss": 0.771,
+      "step": 957
+    },
+    {
+      "epoch": 0.18430165448249328,
+      "grad_norm": 2.093596935272217,
+      "learning_rate": 0.00018777197063090393,
+      "loss": 1.0519,
+      "step": 958
+    },
+    {
+      "epoch": 0.18449403616775684,
+      "grad_norm": 1.7083114385604858,
+      "learning_rate": 0.0001877420969741454,
+      "loss": 1.0304,
+      "step": 959
+    },
+    {
+      "epoch": 0.1846864178530204,
+      "grad_norm": 2.0280098915100098,
+      "learning_rate": 0.0001877121892528838,
+      "loss": 1.0594,
+      "step": 960
+    },
+    {
+      "epoch": 0.18487879953828396,
+      "grad_norm": 2.7876853942871094,
+      "learning_rate": 0.0001876822474787303,
+      "loss": 1.0304,
+      "step": 961
+    },
+    {
+      "epoch": 0.18507118122354752,
+      "grad_norm": 2.1303255558013916,
+      "learning_rate": 0.00018765227166330935,
+      "loss": 1.1134,
+      "step": 962
+    },
+    {
+      "epoch": 0.18526356290881107,
+      "grad_norm": 2.255577802658081,
+      "learning_rate": 0.00018762226181825856,
+      "loss": 1.0928,
+      "step": 963
+    },
+    {
+      "epoch": 0.18545594459407463,
+      "grad_norm": 2.5859456062316895,
+      "learning_rate": 0.0001875922179552288,
+      "loss": 1.079,
+      "step": 964
+    },
+    {
+      "epoch": 0.1856483262793382,
+      "grad_norm": 2.076875925064087,
+      "learning_rate": 0.00018756214008588417,
+      "loss": 0.971,
+      "step": 965
+    },
+    {
+      "epoch": 0.18584070796460178,
+      "grad_norm": 2.2438061237335205,
+      "learning_rate": 0.0001875320282219019,
+      "loss": 1.0884,
+      "step": 966
+    },
+    {
+      "epoch": 0.18603308964986534,
+      "grad_norm": 1.1319012641906738,
+      "learning_rate": 0.00018750188237497245,
+      "loss": 1.0485,
+      "step": 967
+    },
+    {
+      "epoch": 0.1862254713351289,
+      "grad_norm": 2.102001190185547,
+      "learning_rate": 0.0001874717025567995,
+      "loss": 1.0799,
+      "step": 968
+    },
+    {
+      "epoch": 0.18641785302039246,
+      "grad_norm": 1.51614248752594,
+      "learning_rate": 0.00018744148877909987,
+      "loss": 1.0783,
+      "step": 969
+    },
+    {
+      "epoch": 0.18661023470565602,
+      "grad_norm": 1.6770238876342773,
+      "learning_rate": 0.0001874112410536036,
+      "loss": 1.0695,
+      "step": 970
+    },
+    {
+      "epoch": 0.18680261639091958,
+      "grad_norm": 1.6965899467468262,
+      "learning_rate": 0.0001873809593920539,
+      "loss": 1.0799,
+      "step": 971
+    },
+    {
+      "epoch": 0.18699499807618314,
+      "grad_norm": 2.154007911682129,
+      "learning_rate": 0.00018735064380620716,
+      "loss": 1.0278,
+      "step": 972
+    },
+    {
+      "epoch": 0.1871873797614467,
+      "grad_norm": 1.687626600265503,
+      "learning_rate": 0.0001873202943078329,
+      "loss": 1.0679,
+      "step": 973
+    },
+    {
+      "epoch": 0.18737976144671029,
+      "grad_norm": 2.050696849822998,
+      "learning_rate": 0.00018728991090871387,
+      "loss": 1.0185,
+      "step": 974
+    },
+    {
+      "epoch": 0.18757214313197385,
+      "grad_norm": 1.6873154640197754,
+      "learning_rate": 0.00018725949362064594,
+      "loss": 1.055,
+      "step": 975
+    },
+    {
+      "epoch": 0.1877645248172374,
+      "grad_norm": 2.1564433574676514,
+      "learning_rate": 0.00018722904245543816,
+      "loss": 1.0701,
+      "step": 976
+    },
+    {
+      "epoch": 0.18795690650250096,
+      "grad_norm": 1.8510510921478271,
+      "learning_rate": 0.0001871985574249127,
+      "loss": 0.9605,
+      "step": 977
+    },
+    {
+      "epoch": 0.18814928818776452,
+      "grad_norm": 2.3141815662384033,
+      "learning_rate": 0.00018716803854090493,
+      "loss": 1.0007,
+      "step": 978
+    },
+    {
+      "epoch": 0.18834166987302808,
+      "grad_norm": 1.5177409648895264,
+      "learning_rate": 0.00018713748581526332,
+      "loss": 1.0631,
+      "step": 979
+    },
+    {
+      "epoch": 0.18853405155829164,
+      "grad_norm": 2.884469747543335,
+      "learning_rate": 0.00018710689925984948,
+      "loss": 1.0678,
+      "step": 980
+    },
+    {
+      "epoch": 0.1887264332435552,
+      "grad_norm": 2.3163034915924072,
+      "learning_rate": 0.00018707627888653815,
+      "loss": 1.0407,
+      "step": 981
+    },
+    {
+      "epoch": 0.1889188149288188,
+      "grad_norm": 2.531386137008667,
+      "learning_rate": 0.00018704562470721727,
+      "loss": 1.0018,
+      "step": 982
+    },
+    {
+      "epoch": 0.18911119661408235,
+      "grad_norm": 2.631882429122925,
+      "learning_rate": 0.00018701493673378778,
+      "loss": 1.0214,
+      "step": 983
+    },
+    {
+      "epoch": 0.1893035782993459,
+      "grad_norm": 1.3075729608535767,
+      "learning_rate": 0.00018698421497816386,
+      "loss": 0.7881,
+      "step": 984
+    },
+    {
+      "epoch": 0.18949595998460947,
+      "grad_norm": 2.8190805912017822,
+      "learning_rate": 0.0001869534594522727,
+      "loss": 0.9912,
+      "step": 985
+    },
+    {
+      "epoch": 0.18968834166987303,
+      "grad_norm": 2.0970537662506104,
+      "learning_rate": 0.00018692267016805471,
+      "loss": 1.0188,
+      "step": 986
+    },
+    {
+      "epoch": 0.1898807233551366,
+      "grad_norm": 1.3758864402770996,
+      "learning_rate": 0.00018689184713746332,
+      "loss": 1.0078,
+      "step": 987
+    },
+    {
+      "epoch": 0.19007310504040015,
+      "grad_norm": 2.205094337463379,
+      "learning_rate": 0.0001868609903724651,
+      "loss": 0.9252,
+      "step": 988
+    },
+    {
+      "epoch": 0.1902654867256637,
+      "grad_norm": 2.1029763221740723,
+      "learning_rate": 0.0001868300998850397,
+      "loss": 1.0013,
+      "step": 989
+    },
+    {
+      "epoch": 0.19045786841092727,
+      "grad_norm": 0.769168496131897,
+      "learning_rate": 0.0001867991756871799,
+      "loss": 0.7194,
+      "step": 990
+    },
+    {
+      "epoch": 0.19065025009619085,
+      "grad_norm": 2.2458767890930176,
+      "learning_rate": 0.0001867682177908915,
+      "loss": 0.9994,
+      "step": 991
+    },
+    {
+      "epoch": 0.1908426317814544,
+      "grad_norm": 2.5786895751953125,
+      "learning_rate": 0.0001867372262081934,
+      "loss": 1.08,
+      "step": 992
+    },
+    {
+      "epoch": 0.19103501346671797,
+      "grad_norm": 1.4976664781570435,
+      "learning_rate": 0.00018670620095111767,
+      "loss": 1.0784,
+      "step": 993
+    },
+    {
+      "epoch": 0.19122739515198153,
+      "grad_norm": 2.277066946029663,
+      "learning_rate": 0.00018667514203170936,
+      "loss": 1.0493,
+      "step": 994
+    },
+    {
+      "epoch": 0.1914197768372451,
+      "grad_norm": 1.173574447631836,
+      "learning_rate": 0.00018664404946202657,
+      "loss": 1.1411,
+      "step": 995
+    },
+    {
+      "epoch": 0.19161215852250865,
+      "grad_norm": 2.0875654220581055,
+      "learning_rate": 0.00018661292325414057,
+      "loss": 0.9949,
+      "step": 996
+    },
+    {
+      "epoch": 0.1918045402077722,
+      "grad_norm": 1.8674207925796509,
+      "learning_rate": 0.00018658176342013557,
+      "loss": 1.0119,
+      "step": 997
+    },
+    {
+      "epoch": 0.19199692189303577,
+      "grad_norm": 2.6191811561584473,
+      "learning_rate": 0.0001865505699721089,
+      "loss": 1.0458,
+      "step": 998
+    },
+    {
+      "epoch": 0.19218930357829936,
+      "grad_norm": 2.7400975227355957,
+      "learning_rate": 0.00018651934292217097,
+      "loss": 1.0789,
+      "step": 999
+    },
+    {
+      "epoch": 0.19238168526356292,
+      "grad_norm": 2.3012349605560303,
+      "learning_rate": 0.00018648808228244516,
+      "loss": 1.0421,
+      "step": 1000
+    },
+    {
+      "epoch": 0.19257406694882648,
+      "grad_norm": 2.2889530658721924,
+      "learning_rate": 0.00018645678806506795,
+      "loss": 1.0314,
+      "step": 1001
+    },
+    {
+      "epoch": 0.19276644863409004,
+      "grad_norm": 2.3252811431884766,
+      "learning_rate": 0.0001864254602821888,
+      "loss": 1.0229,
+      "step": 1002
+    },
+    {
+      "epoch": 0.1929588303193536,
+      "grad_norm": 2.133779287338257,
+      "learning_rate": 0.00018639409894597025,
+      "loss": 1.0189,
+      "step": 1003
+    },
+    {
+      "epoch": 0.19315121200461716,
+      "grad_norm": 2.3881304264068604,
+      "learning_rate": 0.00018636270406858786,
+      "loss": 1.0578,
+      "step": 1004
+    },
+    {
+      "epoch": 0.19334359368988072,
+      "grad_norm": 2.3048503398895264,
+      "learning_rate": 0.00018633127566223021,
+      "loss": 1.0341,
+      "step": 1005
+    },
+    {
+      "epoch": 0.19353597537514428,
+      "grad_norm": 2.0837533473968506,
+      "learning_rate": 0.00018629981373909885,
+      "loss": 0.9883,
+      "step": 1006
+    },
+    {
+      "epoch": 0.19372835706040784,
+      "grad_norm": 1.5290104150772095,
+      "learning_rate": 0.00018626831831140844,
+      "loss": 0.9813,
+      "step": 1007
+    },
+    {
+      "epoch": 0.19392073874567142,
+      "grad_norm": 2.408082962036133,
+      "learning_rate": 0.0001862367893913865,
+      "loss": 1.0419,
+      "step": 1008
+    },
+    {
+      "epoch": 0.19411312043093498,
+      "grad_norm": 2.103527307510376,
+      "learning_rate": 0.00018620522699127373,
+      "loss": 1.0638,
+      "step": 1009
+    },
+    {
+      "epoch": 0.19430550211619854,
+      "grad_norm": 2.216693878173828,
+      "learning_rate": 0.00018617363112332375,
+      "loss": 1.0579,
+      "step": 1010
+    },
+    {
+      "epoch": 0.1944978838014621,
+      "grad_norm": 2.393205404281616,
+      "learning_rate": 0.00018614200179980307,
+      "loss": 1.0562,
+      "step": 1011
+    },
+    {
+      "epoch": 0.19469026548672566,
+      "grad_norm": 2.300666332244873,
+      "learning_rate": 0.00018611033903299134,
+      "loss": 1.0054,
+      "step": 1012
+    },
+    {
+      "epoch": 0.19488264717198922,
+      "grad_norm": 1.9840978384017944,
+      "learning_rate": 0.00018607864283518114,
+      "loss": 1.0657,
+      "step": 1013
+    },
+    {
+      "epoch": 0.19507502885725278,
+      "grad_norm": 2.201632022857666,
+      "learning_rate": 0.00018604691321867805,
+      "loss": 1.0154,
+      "step": 1014
+    },
+    {
+      "epoch": 0.19526741054251634,
+      "grad_norm": 2.232729434967041,
+      "learning_rate": 0.00018601515019580053,
+      "loss": 1.0635,
+      "step": 1015
+    },
+    {
+      "epoch": 0.19545979222777993,
+      "grad_norm": 1.1075389385223389,
+      "learning_rate": 0.00018598335377888013,
+      "loss": 1.1013,
+      "step": 1016
+    },
+    {
+      "epoch": 0.1956521739130435,
+      "grad_norm": 2.345454454421997,
+      "learning_rate": 0.0001859515239802613,
+      "loss": 1.0258,
+      "step": 1017
+    },
+    {
+      "epoch": 0.19584455559830705,
+      "grad_norm": 1.3347303867340088,
+      "learning_rate": 0.00018591966081230144,
+      "loss": 0.7157,
+      "step": 1018
+    },
+    {
+      "epoch": 0.1960369372835706,
+      "grad_norm": 1.0910418033599854,
+      "learning_rate": 0.00018588776428737097,
+      "loss": 0.7556,
+      "step": 1019
+    },
+    {
+      "epoch": 0.19622931896883417,
+      "grad_norm": 2.386136531829834,
+      "learning_rate": 0.0001858558344178532,
+      "loss": 1.0384,
+      "step": 1020
+    },
+    {
+      "epoch": 0.19642170065409773,
+      "grad_norm": 2.209826707839966,
+      "learning_rate": 0.00018582387121614437,
+      "loss": 1.0621,
+      "step": 1021
+    },
+    {
+      "epoch": 0.19661408233936128,
+      "grad_norm": 1.610894799232483,
+      "learning_rate": 0.00018579187469465377,
+      "loss": 1.0667,
+      "step": 1022
+    },
+    {
+      "epoch": 0.19680646402462484,
+      "grad_norm": 1.0762776136398315,
+      "learning_rate": 0.0001857598448658035,
+      "loss": 1.1319,
+      "step": 1023
+    },
+    {
+      "epoch": 0.19699884570988843,
+      "grad_norm": 2.0586037635803223,
+      "learning_rate": 0.0001857277817420287,
+      "loss": 1.0434,
+      "step": 1024
+    },
+    {
+      "epoch": 0.197191227395152,
+      "grad_norm": 2.3307480812072754,
+      "learning_rate": 0.00018569568533577727,
+      "loss": 1.0655,
+      "step": 1025
+    },
+    {
+      "epoch": 0.19738360908041555,
+      "grad_norm": 2.127830982208252,
+      "learning_rate": 0.00018566355565951024,
+      "loss": 1.117,
+      "step": 1026
+    },
+    {
+      "epoch": 0.1975759907656791,
+      "grad_norm": 1.84950852394104,
+      "learning_rate": 0.0001856313927257014,
+      "loss": 1.0454,
+      "step": 1027
+    },
+    {
+      "epoch": 0.19776837245094267,
+      "grad_norm": 2.2244958877563477,
+      "learning_rate": 0.00018559919654683756,
+      "loss": 1.0709,
+      "step": 1028
+    },
+    {
+      "epoch": 0.19796075413620623,
+      "grad_norm": 1.8004997968673706,
+      "learning_rate": 0.00018556696713541832,
+      "loss": 1.0685,
+      "step": 1029
+    },
+    {
+      "epoch": 0.1981531358214698,
+      "grad_norm": 2.4836461544036865,
+      "learning_rate": 0.0001855347045039563,
+      "loss": 1.0764,
+      "step": 1030
+    },
+    {
+      "epoch": 0.19834551750673335,
+      "grad_norm": 1.5049992799758911,
+      "learning_rate": 0.00018550240866497695,
+      "loss": 1.0316,
+      "step": 1031
+    },
+    {
+      "epoch": 0.1985378991919969,
+      "grad_norm": 2.009117603302002,
+      "learning_rate": 0.0001854700796310186,
+      "loss": 1.0238,
+      "step": 1032
+    },
+    {
+      "epoch": 0.1987302808772605,
+      "grad_norm": 2.829646348953247,
+      "learning_rate": 0.00018543771741463255,
+      "loss": 1.0209,
+      "step": 1033
+    },
+    {
+      "epoch": 0.19892266256252406,
+      "grad_norm": 2.1958870887756348,
+      "learning_rate": 0.00018540532202838286,
+      "loss": 1.0593,
+      "step": 1034
+    },
+    {
+      "epoch": 0.19911504424778761,
+      "grad_norm": 2.2011876106262207,
+      "learning_rate": 0.00018537289348484657,
+      "loss": 1.0205,
+      "step": 1035
+    },
+    {
+      "epoch": 0.19930742593305117,
+      "grad_norm": 2.3444974422454834,
+      "learning_rate": 0.00018534043179661356,
+      "loss": 0.9599,
+      "step": 1036
+    },
+    {
+      "epoch": 0.19949980761831473,
+      "grad_norm": 1.3066757917404175,
+      "learning_rate": 0.00018530793697628658,
+      "loss": 0.7488,
+      "step": 1037
+    },
+    {
+      "epoch": 0.1996921893035783,
+      "grad_norm": 2.410810947418213,
+      "learning_rate": 0.00018527540903648122,
+      "loss": 1.134,
+      "step": 1038
+    },
+    {
+      "epoch": 0.19988457098884185,
+      "grad_norm": 2.4184865951538086,
+      "learning_rate": 0.00018524284798982594,
+      "loss": 1.082,
+      "step": 1039
+    },
+    {
+      "epoch": 0.2000769526741054,
+      "grad_norm": 2.3008995056152344,
+      "learning_rate": 0.00018521025384896208,
+      "loss": 1.0326,
+      "step": 1040
+    },
+    {
+      "epoch": 0.200269334359369,
+      "grad_norm": 2.0151100158691406,
+      "learning_rate": 0.0001851776266265438,
+      "loss": 0.9956,
+      "step": 1041
+    },
+    {
+      "epoch": 0.20046171604463256,
+      "grad_norm": 0.9155529141426086,
+      "learning_rate": 0.00018514496633523813,
+      "loss": 0.7878,
+      "step": 1042
+    },
+    {
+      "epoch": 0.20065409772989612,
+      "grad_norm": 2.863663673400879,
+      "learning_rate": 0.0001851122729877249,
+      "loss": 1.0005,
+      "step": 1043
+    },
+    {
+      "epoch": 0.20084647941515968,
+      "grad_norm": 2.4747300148010254,
+      "learning_rate": 0.00018507954659669677,
+      "loss": 0.9776,
+      "step": 1044
+    },
+    {
+      "epoch": 0.20103886110042324,
+      "grad_norm": 2.478010654449463,
+      "learning_rate": 0.0001850467871748593,
+      "loss": 1.0186,
+      "step": 1045
+    },
+    {
+      "epoch": 0.2012312427856868,
+      "grad_norm": 1.985946536064148,
+      "learning_rate": 0.0001850139947349308,
+      "loss": 1.0137,
+      "step": 1046
+    },
+    {
+      "epoch": 0.20142362447095036,
+      "grad_norm": 2.119316339492798,
+      "learning_rate": 0.0001849811692896424,
+      "loss": 1.0851,
+      "step": 1047
+    },
+    {
+      "epoch": 0.20161600615621392,
+      "grad_norm": 2.495323896408081,
+      "learning_rate": 0.0001849483108517381,
+      "loss": 1.0452,
+      "step": 1048
+    },
+    {
+      "epoch": 0.2018083878414775,
+      "grad_norm": 2.114518165588379,
+      "learning_rate": 0.0001849154194339747,
+      "loss": 1.0295,
+      "step": 1049
+    },
+    {
+      "epoch": 0.20200076952674106,
+      "grad_norm": 2.1689465045928955,
+      "learning_rate": 0.00018488249504912174,
+      "loss": 1.0462,
+      "step": 1050
+    },
+    {
+      "epoch": 0.20219315121200462,
+      "grad_norm": 2.3133955001831055,
+      "learning_rate": 0.00018484953770996163,
+      "loss": 1.0524,
+      "step": 1051
+    },
+    {
+      "epoch": 0.20238553289726818,
+      "grad_norm": 0.9804697632789612,
+      "learning_rate": 0.00018481654742928948,
+      "loss": 0.9635,
+      "step": 1052
+    },
+    {
+      "epoch": 0.20257791458253174,
+      "grad_norm": 1.5462021827697754,
+      "learning_rate": 0.00018478352421991335,
+      "loss": 1.0298,
+      "step": 1053
+    },
+    {
+      "epoch": 0.2027702962677953,
+      "grad_norm": 2.3460171222686768,
+      "learning_rate": 0.0001847504680946539,
+      "loss": 1.0973,
+      "step": 1054
+    },
+    {
+      "epoch": 0.20296267795305886,
+      "grad_norm": 2.5708367824554443,
+      "learning_rate": 0.0001847173790663447,
+      "loss": 1.021,
+      "step": 1055
+    },
+    {
+      "epoch": 0.20315505963832242,
+      "grad_norm": 1.9747501611709595,
+      "learning_rate": 0.00018468425714783204,
+      "loss": 0.993,
+      "step": 1056
+    },
+    {
+      "epoch": 0.20334744132358598,
+      "grad_norm": 1.6529561281204224,
+      "learning_rate": 0.000184651102351975,
+      "loss": 1.0854,
+      "step": 1057
+    },
+    {
+      "epoch": 0.20353982300884957,
+      "grad_norm": 3.0383362770080566,
+      "learning_rate": 0.0001846179146916454,
+      "loss": 1.05,
+      "step": 1058
+    },
+    {
+      "epoch": 0.20373220469411313,
+      "grad_norm": 2.1850438117980957,
+      "learning_rate": 0.00018458469417972784,
+      "loss": 1.0152,
+      "step": 1059
+    },
+    {
+      "epoch": 0.2039245863793767,
+      "grad_norm": 1.1899678707122803,
+      "learning_rate": 0.00018455144082911966,
+      "loss": 0.7729,
+      "step": 1060
+    },
+    {
+      "epoch": 0.20411696806464025,
+      "grad_norm": 2.708326578140259,
+      "learning_rate": 0.00018451815465273094,
+      "loss": 1.0356,
+      "step": 1061
+    },
+    {
+      "epoch": 0.2043093497499038,
+      "grad_norm": 2.149033308029175,
+      "learning_rate": 0.00018448483566348457,
+      "loss": 1.0323,
+      "step": 1062
+    },
+    {
+      "epoch": 0.20450173143516737,
+      "grad_norm": 2.58823561668396,
+      "learning_rate": 0.00018445148387431605,
+      "loss": 1.0207,
+      "step": 1063
+    },
+    {
+      "epoch": 0.20469411312043093,
+      "grad_norm": 1.465346336364746,
+      "learning_rate": 0.0001844180992981738,
+      "loss": 1.039,
+      "step": 1064
+    },
+    {
+      "epoch": 0.20488649480569449,
+      "grad_norm": 2.1820597648620605,
+      "learning_rate": 0.00018438468194801875,
+      "loss": 1.0612,
+      "step": 1065
+    },
+    {
+      "epoch": 0.20507887649095807,
+      "grad_norm": 1.8158183097839355,
+      "learning_rate": 0.00018435123183682475,
+      "loss": 0.99,
+      "step": 1066
+    },
+    {
+      "epoch": 0.20527125817622163,
+      "grad_norm": 1.7176711559295654,
+      "learning_rate": 0.00018431774897757826,
+      "loss": 0.9645,
+      "step": 1067
+    },
+    {
+      "epoch": 0.2054636398614852,
+      "grad_norm": 2.331918716430664,
+      "learning_rate": 0.00018428423338327844,
+      "loss": 1.0323,
+      "step": 1068
+    },
+    {
+      "epoch": 0.20565602154674875,
+      "grad_norm": 1.5375444889068604,
+      "learning_rate": 0.00018425068506693727,
+      "loss": 1.1276,
+      "step": 1069
+    },
+    {
+      "epoch": 0.2058484032320123,
+      "grad_norm": 1.674390435218811,
+      "learning_rate": 0.00018421710404157928,
+      "loss": 1.0095,
+      "step": 1070
+    },
+    {
+      "epoch": 0.20604078491727587,
+      "grad_norm": 2.7592508792877197,
+      "learning_rate": 0.00018418349032024183,
+      "loss": 1.0229,
+      "step": 1071
+    },
+    {
+      "epoch": 0.20623316660253943,
+      "grad_norm": 2.2855546474456787,
+      "learning_rate": 0.0001841498439159749,
+      "loss": 1.0688,
+      "step": 1072
+    },
+    {
+      "epoch": 0.206425548287803,
+      "grad_norm": 2.362215280532837,
+      "learning_rate": 0.00018411616484184125,
+      "loss": 1.021,
+      "step": 1073
+    },
+    {
+      "epoch": 0.20661792997306658,
+      "grad_norm": 1.9094370603561401,
+      "learning_rate": 0.00018408245311091616,
+      "loss": 0.9887,
+      "step": 1074
+    },
+    {
+      "epoch": 0.20681031165833014,
+      "grad_norm": 2.176823616027832,
+      "learning_rate": 0.00018404870873628775,
+      "loss": 0.9892,
+      "step": 1075
+    },
+    {
+      "epoch": 0.2070026933435937,
+      "grad_norm": 2.4544410705566406,
+      "learning_rate": 0.00018401493173105673,
+      "loss": 1.0926,
+      "step": 1076
+    },
+    {
+      "epoch": 0.20719507502885726,
+      "grad_norm": 1.9796204566955566,
+      "learning_rate": 0.0001839811221083365,
+      "loss": 1.0042,
+      "step": 1077
+    },
+    {
+      "epoch": 0.20738745671412082,
+      "grad_norm": 1.998678207397461,
+      "learning_rate": 0.00018394727988125308,
+      "loss": 1.0508,
+      "step": 1078
+    },
+    {
+      "epoch": 0.20757983839938438,
+      "grad_norm": 2.2559316158294678,
+      "learning_rate": 0.00018391340506294522,
+      "loss": 1.0477,
+      "step": 1079
+    },
+    {
+      "epoch": 0.20777222008464794,
+      "grad_norm": 2.67911434173584,
+      "learning_rate": 0.00018387949766656434,
+      "loss": 0.9838,
+      "step": 1080
+    },
+    {
+      "epoch": 0.2079646017699115,
+      "grad_norm": 1.5213801860809326,
+      "learning_rate": 0.00018384555770527435,
+      "loss": 1.0589,
+      "step": 1081
+    },
+    {
+      "epoch": 0.20815698345517505,
+      "grad_norm": 2.0479016304016113,
+      "learning_rate": 0.00018381158519225203,
+      "loss": 1.0372,
+      "step": 1082
+    },
+    {
+      "epoch": 0.20834936514043864,
+      "grad_norm": 2.5023577213287354,
+      "learning_rate": 0.0001837775801406866,
+      "loss": 1.0734,
+      "step": 1083
+    },
+    {
+      "epoch": 0.2085417468257022,
+      "grad_norm": 1.693630337715149,
+      "learning_rate": 0.00018374354256378,
+      "loss": 1.098,
+      "step": 1084
+    },
+    {
+      "epoch": 0.20873412851096576,
+      "grad_norm": 1.2053192853927612,
+      "learning_rate": 0.0001837094724747468,
+      "loss": 0.7606,
+      "step": 1085
+    },
+    {
+      "epoch": 0.20892651019622932,
+      "grad_norm": 1.359632968902588,
+      "learning_rate": 0.0001836753698868142,
+      "loss": 1.0516,
+      "step": 1086
+    },
+    {
+      "epoch": 0.20911889188149288,
+      "grad_norm": 2.8057937622070312,
+      "learning_rate": 0.00018364123481322197,
+      "loss": 1.0895,
+      "step": 1087
+    },
+    {
+      "epoch": 0.20931127356675644,
+      "grad_norm": 1.8939366340637207,
+      "learning_rate": 0.00018360706726722252,
+      "loss": 1.026,
+      "step": 1088
+    },
+    {
+      "epoch": 0.20950365525202,
+      "grad_norm": 2.353043556213379,
+      "learning_rate": 0.0001835728672620809,
+      "loss": 0.9366,
+      "step": 1089
+    },
+    {
+      "epoch": 0.20969603693728356,
+      "grad_norm": 2.2262446880340576,
+      "learning_rate": 0.0001835386348110747,
+      "loss": 1.0982,
+      "step": 1090
+    },
+    {
+      "epoch": 0.20988841862254715,
+      "grad_norm": 2.030750036239624,
+      "learning_rate": 0.00018350436992749418,
+      "loss": 1.0194,
+      "step": 1091
+    },
+    {
+      "epoch": 0.2100808003078107,
+      "grad_norm": 1.870241641998291,
+      "learning_rate": 0.00018347007262464206,
+      "loss": 1.0466,
+      "step": 1092
+    },
+    {
+      "epoch": 0.21027318199307427,
+      "grad_norm": 1.922324776649475,
+      "learning_rate": 0.00018343574291583383,
+      "loss": 0.9677,
+      "step": 1093
+    },
+    {
+      "epoch": 0.21046556367833782,
+      "grad_norm": 1.6290240287780762,
+      "learning_rate": 0.0001834013808143974,
+      "loss": 0.9872,
+      "step": 1094
+    },
+    {
+      "epoch": 0.21065794536360138,
+      "grad_norm": 2.1090519428253174,
+      "learning_rate": 0.00018336698633367337,
+      "loss": 1.0425,
+      "step": 1095
+    },
+    {
+      "epoch": 0.21085032704886494,
+      "grad_norm": 2.132826328277588,
+      "learning_rate": 0.00018333255948701482,
+      "loss": 0.9857,
+      "step": 1096
+    },
+    {
+      "epoch": 0.2110427087341285,
+      "grad_norm": 2.1436119079589844,
+      "learning_rate": 0.00018329810028778747,
+      "loss": 1.0757,
+      "step": 1097
+    },
+    {
+      "epoch": 0.21123509041939206,
+      "grad_norm": 2.555048942565918,
+      "learning_rate": 0.0001832636087493695,
+      "loss": 1.0745,
+      "step": 1098
+    },
+    {
+      "epoch": 0.21142747210465565,
+      "grad_norm": 2.386772632598877,
+      "learning_rate": 0.0001832290848851518,
+      "loss": 1.0375,
+      "step": 1099
+    },
+    {
+      "epoch": 0.2116198537899192,
+      "grad_norm": 2.133273124694824,
+      "learning_rate": 0.0001831945287085377,
+      "loss": 1.0717,
+      "step": 1100
+    },
+    {
+      "epoch": 0.21181223547518277,
+      "grad_norm": 1.8994041681289673,
+      "learning_rate": 0.00018315994023294306,
+      "loss": 1.0926,
+      "step": 1101
+    },
+    {
+      "epoch": 0.21200461716044633,
+      "grad_norm": 2.4985532760620117,
+      "learning_rate": 0.00018312531947179634,
+      "loss": 1.0647,
+      "step": 1102
+    },
+    {
+      "epoch": 0.2121969988457099,
+      "grad_norm": 1.7739999294281006,
+      "learning_rate": 0.00018309066643853853,
+      "loss": 1.0778,
+      "step": 1103
+    },
+    {
+      "epoch": 0.21238938053097345,
+      "grad_norm": 2.1154944896698,
+      "learning_rate": 0.00018305598114662312,
+      "loss": 1.0559,
+      "step": 1104
+    },
+    {
+      "epoch": 0.212581762216237,
+      "grad_norm": 1.4314388036727905,
+      "learning_rate": 0.00018302126360951611,
+      "loss": 1.0894,
+      "step": 1105
+    },
+    {
+      "epoch": 0.21277414390150057,
+      "grad_norm": 2.3400816917419434,
+      "learning_rate": 0.00018298651384069604,
+      "loss": 1.019,
+      "step": 1106
+    },
+    {
+      "epoch": 0.21296652558676413,
+      "grad_norm": 1.5140652656555176,
+      "learning_rate": 0.00018295173185365403,
+      "loss": 0.7626,
+      "step": 1107
+    },
+    {
+      "epoch": 0.21315890727202771,
+      "grad_norm": 1.9267557859420776,
+      "learning_rate": 0.00018291691766189358,
+      "loss": 1.0183,
+      "step": 1108
+    },
+    {
+      "epoch": 0.21335128895729127,
+      "grad_norm": 2.2177510261535645,
+      "learning_rate": 0.0001828820712789308,
+      "loss": 1.0249,
+      "step": 1109
+    },
+    {
+      "epoch": 0.21354367064255483,
+      "grad_norm": 2.0770859718322754,
+      "learning_rate": 0.0001828471927182942,
+      "loss": 1.0129,
+      "step": 1110
+    },
+    {
+      "epoch": 0.2137360523278184,
+      "grad_norm": 2.5951449871063232,
+      "learning_rate": 0.0001828122819935249,
+      "loss": 1.032,
+      "step": 1111
+    },
+    {
+      "epoch": 0.21392843401308195,
+      "grad_norm": 2.311044692993164,
+      "learning_rate": 0.0001827773391181764,
+      "loss": 0.9925,
+      "step": 1112
+    },
+    {
+      "epoch": 0.2141208156983455,
+      "grad_norm": 2.1006081104278564,
+      "learning_rate": 0.0001827423641058148,
+      "loss": 0.9714,
+      "step": 1113
+    },
+    {
+      "epoch": 0.21431319738360907,
+      "grad_norm": 1.2790133953094482,
+      "learning_rate": 0.00018270735697001848,
+      "loss": 1.0629,
+      "step": 1114
+    },
+    {
+      "epoch": 0.21450557906887263,
+      "grad_norm": 2.4742965698242188,
+      "learning_rate": 0.00018267231772437853,
+      "loss": 1.033,
+      "step": 1115
+    },
+    {
+      "epoch": 0.21469796075413622,
+      "grad_norm": 2.3977818489074707,
+      "learning_rate": 0.00018263724638249834,
+      "loss": 1.0457,
+      "step": 1116
+    },
+    {
+      "epoch": 0.21489034243939978,
+      "grad_norm": 2.336143970489502,
+      "learning_rate": 0.00018260214295799383,
+      "loss": 1.005,
+      "step": 1117
+    },
+    {
+      "epoch": 0.21508272412466334,
+      "grad_norm": 2.7737388610839844,
+      "learning_rate": 0.0001825670074644933,
+      "loss": 1.0304,
+      "step": 1118
+    },
+    {
+      "epoch": 0.2152751058099269,
+      "grad_norm": 1.2106578350067139,
+      "learning_rate": 0.00018253183991563768,
+      "loss": 0.7857,
+      "step": 1119
+    },
+    {
+      "epoch": 0.21546748749519046,
+      "grad_norm": 1.739078402519226,
+      "learning_rate": 0.0001824966403250801,
+      "loss": 1.0509,
+      "step": 1120
+    },
+    {
+      "epoch": 0.21565986918045402,
+      "grad_norm": 2.648611307144165,
+      "learning_rate": 0.00018246140870648633,
+      "loss": 1.0762,
+      "step": 1121
+    },
+    {
+      "epoch": 0.21585225086571758,
+      "grad_norm": 1.2581489086151123,
+      "learning_rate": 0.00018242614507353445,
+      "loss": 0.7622,
+      "step": 1122
+    },
+    {
+      "epoch": 0.21604463255098114,
+      "grad_norm": 2.1422009468078613,
+      "learning_rate": 0.00018239084943991505,
+      "loss": 1.0385,
+      "step": 1123
+    },
+    {
+      "epoch": 0.2162370142362447,
+      "grad_norm": 10.27823257446289,
+      "learning_rate": 0.00018235552181933108,
+      "loss": 1.0328,
+      "step": 1124
+    },
+    {
+      "epoch": 0.21642939592150828,
+      "grad_norm": 2.191406726837158,
+      "learning_rate": 0.00018232016222549796,
+      "loss": 1.034,
+      "step": 1125
+    },
+    {
+      "epoch": 0.21662177760677184,
+      "grad_norm": 1.0778344869613647,
+      "learning_rate": 0.0001822847706721435,
+      "loss": 0.7686,
+      "step": 1126
+    },
+    {
+      "epoch": 0.2168141592920354,
+      "grad_norm": 1.9736641645431519,
+      "learning_rate": 0.00018224934717300792,
+      "loss": 1.0582,
+      "step": 1127
+    },
+    {
+      "epoch": 0.21700654097729896,
+      "grad_norm": 2.540222644805908,
+      "learning_rate": 0.00018221389174184386,
+      "loss": 1.0897,
+      "step": 1128
+    },
+    {
+      "epoch": 0.21719892266256252,
+      "grad_norm": 2.4644381999969482,
+      "learning_rate": 0.00018217840439241633,
+      "loss": 0.967,
+      "step": 1129
+    },
+    {
+      "epoch": 0.21739130434782608,
+      "grad_norm": 2.440218210220337,
+      "learning_rate": 0.00018214288513850266,
+      "loss": 1.0942,
+      "step": 1130
+    },
+    {
+      "epoch": 0.21758368603308964,
+      "grad_norm": 2.2348904609680176,
+      "learning_rate": 0.00018210733399389277,
+      "loss": 1.0071,
+      "step": 1131
+    },
+    {
+      "epoch": 0.2177760677183532,
+      "grad_norm": 1.8997058868408203,
+      "learning_rate": 0.00018207175097238878,
+      "loss": 0.9947,
+      "step": 1132
+    },
+    {
+      "epoch": 0.2179684494036168,
+      "grad_norm": 2.165062189102173,
+      "learning_rate": 0.00018203613608780523,
+      "loss": 1.0465,
+      "step": 1133
+    },
+    {
+      "epoch": 0.21816083108888035,
+      "grad_norm": 2.305433511734009,
+      "learning_rate": 0.00018200048935396908,
+      "loss": 1.0967,
+      "step": 1134
+    },
+    {
+      "epoch": 0.2183532127741439,
+      "grad_norm": 2.078404426574707,
+      "learning_rate": 0.00018196481078471962,
+      "loss": 0.9856,
+      "step": 1135
+    },
+    {
+      "epoch": 0.21854559445940747,
+      "grad_norm": 2.0693931579589844,
+      "learning_rate": 0.00018192910039390844,
+      "loss": 1.0714,
+      "step": 1136
+    },
+    {
+      "epoch": 0.21873797614467103,
+      "grad_norm": 1.7110666036605835,
+      "learning_rate": 0.0001818933581953996,
+      "loss": 1.065,
+      "step": 1137
+    },
+    {
+      "epoch": 0.21893035782993459,
+      "grad_norm": 1.5230662822723389,
+      "learning_rate": 0.00018185758420306947,
+      "loss": 0.9816,
+      "step": 1138
+    },
+    {
+      "epoch": 0.21912273951519814,
+      "grad_norm": 2.575566291809082,
+      "learning_rate": 0.0001818217784308067,
+      "loss": 1.0958,
+      "step": 1139
+    },
+    {
+      "epoch": 0.2193151212004617,
+      "grad_norm": 1.3658833503723145,
+      "learning_rate": 0.0001817859408925123,
+      "loss": 0.7364,
+      "step": 1140
+    },
+    {
+      "epoch": 0.2195075028857253,
+      "grad_norm": 2.6197457313537598,
+      "learning_rate": 0.00018175007160209968,
+      "loss": 0.9983,
+      "step": 1141
+    },
+    {
+      "epoch": 0.21969988457098885,
+      "grad_norm": 2.421739339828491,
+      "learning_rate": 0.00018171417057349457,
+      "loss": 0.9514,
+      "step": 1142
+    },
+    {
+      "epoch": 0.2198922662562524,
+      "grad_norm": 2.266059398651123,
+      "learning_rate": 0.00018167823782063488,
+      "loss": 1.0355,
+      "step": 1143
+    },
+    {
+      "epoch": 0.22008464794151597,
+      "grad_norm": 2.235506772994995,
+      "learning_rate": 0.00018164227335747105,
+      "loss": 1.1301,
+      "step": 1144
+    },
+    {
+      "epoch": 0.22027702962677953,
+      "grad_norm": 1.5237572193145752,
+      "learning_rate": 0.00018160627719796568,
+      "loss": 1.0696,
+      "step": 1145
+    },
+    {
+      "epoch": 0.2204694113120431,
+      "grad_norm": 1.4860085248947144,
+      "learning_rate": 0.0001815702493560937,
+      "loss": 1.1016,
+      "step": 1146
+    },
+    {
+      "epoch": 0.22066179299730665,
+      "grad_norm": 2.781360387802124,
+      "learning_rate": 0.00018153418984584238,
+      "loss": 1.0498,
+      "step": 1147
+    },
+    {
+      "epoch": 0.2208541746825702,
+      "grad_norm": 2.21144437789917,
+      "learning_rate": 0.00018149809868121125,
+      "loss": 1.0649,
+      "step": 1148
+    },
+    {
+      "epoch": 0.22104655636783377,
+      "grad_norm": 2.374551773071289,
+      "learning_rate": 0.00018146197587621217,
+      "loss": 1.0119,
+      "step": 1149
+    },
+    {
+      "epoch": 0.22123893805309736,
+      "grad_norm": 2.1132407188415527,
+      "learning_rate": 0.0001814258214448692,
+      "loss": 1.0771,
+      "step": 1150
+    },
+    {
+      "epoch": 0.22143131973836092,
+      "grad_norm": 2.3171133995056152,
+      "learning_rate": 0.0001813896354012188,
+      "loss": 1.0758,
+      "step": 1151
+    },
+    {
+      "epoch": 0.22162370142362448,
+      "grad_norm": 1.6915555000305176,
+      "learning_rate": 0.0001813534177593096,
+      "loss": 0.9993,
+      "step": 1152
+    },
+    {
+      "epoch": 0.22181608310888803,
+      "grad_norm": 1.1079490184783936,
+      "learning_rate": 0.00018131716853320254,
+      "loss": 1.0585,
+      "step": 1153
+    },
+    {
+      "epoch": 0.2220084647941516,
+      "grad_norm": 2.649815082550049,
+      "learning_rate": 0.00018128088773697085,
+      "loss": 1.013,
+      "step": 1154
+    },
+    {
+      "epoch": 0.22220084647941515,
+      "grad_norm": 1.3279314041137695,
+      "learning_rate": 0.00018124457538469996,
+      "loss": 1.0735,
+      "step": 1155
+    },
+    {
+      "epoch": 0.2223932281646787,
+      "grad_norm": 2.027423858642578,
+      "learning_rate": 0.00018120823149048754,
+      "loss": 0.9723,
+      "step": 1156
+    },
+    {
+      "epoch": 0.22258560984994227,
+      "grad_norm": 2.110697031021118,
+      "learning_rate": 0.00018117185606844358,
+      "loss": 1.0586,
+      "step": 1157
+    },
+    {
+      "epoch": 0.22277799153520586,
+      "grad_norm": 1.1766694784164429,
+      "learning_rate": 0.00018113544913269026,
+      "loss": 0.7946,
+      "step": 1158
+    },
+    {
+      "epoch": 0.22297037322046942,
+      "grad_norm": 1.7141345739364624,
+      "learning_rate": 0.00018109901069736202,
+      "loss": 1.0831,
+      "step": 1159
+    },
+    {
+      "epoch": 0.22316275490573298,
+      "grad_norm": 2.552920341491699,
+      "learning_rate": 0.00018106254077660552,
+      "loss": 1.014,
+      "step": 1160
+    },
+    {
+      "epoch": 0.22335513659099654,
+      "grad_norm": 1.6646509170532227,
+      "learning_rate": 0.0001810260393845796,
+      "loss": 1.0311,
+      "step": 1161
+    },
+    {
+      "epoch": 0.2235475182762601,
+      "grad_norm": 1.725799798965454,
+      "learning_rate": 0.0001809895065354554,
+      "loss": 1.0609,
+      "step": 1162
+    },
+    {
+      "epoch": 0.22373989996152366,
+      "grad_norm": 1.9704070091247559,
+      "learning_rate": 0.00018095294224341622,
+      "loss": 1.0193,
+      "step": 1163
+    },
+    {
+      "epoch": 0.22393228164678722,
+      "grad_norm": 1.5823508501052856,
+      "learning_rate": 0.00018091634652265753,
+      "loss": 1.0379,
+      "step": 1164
+    },
+    {
+      "epoch": 0.22412466333205078,
+      "grad_norm": 2.3071742057800293,
+      "learning_rate": 0.00018087971938738714,
+      "loss": 0.9892,
+      "step": 1165
+    },
+    {
+      "epoch": 0.22431704501731436,
+      "grad_norm": 2.066230058670044,
+      "learning_rate": 0.00018084306085182488,
+      "loss": 1.0076,
+      "step": 1166
+    },
+    {
+      "epoch": 0.22450942670257792,
+      "grad_norm": 2.0005757808685303,
+      "learning_rate": 0.00018080637093020288,
+      "loss": 1.0238,
+      "step": 1167
+    },
+    {
+      "epoch": 0.22470180838784148,
+      "grad_norm": 2.9497032165527344,
+      "learning_rate": 0.0001807696496367655,
+      "loss": 1.0487,
+      "step": 1168
+    },
+    {
+      "epoch": 0.22489419007310504,
+      "grad_norm": 1.7876861095428467,
+      "learning_rate": 0.00018073289698576913,
+      "loss": 1.0697,
+      "step": 1169
+    },
+    {
+      "epoch": 0.2250865717583686,
+      "grad_norm": 2.535524606704712,
+      "learning_rate": 0.00018069611299148237,
+      "loss": 1.0282,
+      "step": 1170
+    },
+    {
+      "epoch": 0.22527895344363216,
+      "grad_norm": 2.3567421436309814,
+      "learning_rate": 0.00018065929766818617,
+      "loss": 1.0275,
+      "step": 1171
+    },
+    {
+      "epoch": 0.22547133512889572,
+      "grad_norm": 1.1814204454421997,
+      "learning_rate": 0.0001806224510301734,
+      "loss": 0.6948,
+      "step": 1172
+    },
+    {
+      "epoch": 0.22566371681415928,
+      "grad_norm": 2.5134947299957275,
+      "learning_rate": 0.00018058557309174925,
+      "loss": 1.051,
+      "step": 1173
+    },
+    {
+      "epoch": 0.22585609849942284,
+      "grad_norm": 2.1561102867126465,
+      "learning_rate": 0.00018054866386723096,
+      "loss": 0.9866,
+      "step": 1174
+    },
+    {
+      "epoch": 0.22604848018468643,
+      "grad_norm": 2.644540548324585,
+      "learning_rate": 0.000180511723370948,
+      "loss": 1.0594,
+      "step": 1175
+    },
+    {
+      "epoch": 0.22624086186995,
+      "grad_norm": 2.971217393875122,
+      "learning_rate": 0.0001804747516172419,
+      "loss": 1.1058,
+      "step": 1176
+    },
+    {
+      "epoch": 0.22643324355521355,
+      "grad_norm": 2.172950029373169,
+      "learning_rate": 0.00018043774862046643,
+      "loss": 0.9808,
+      "step": 1177
+    },
+    {
+      "epoch": 0.2266256252404771,
+      "grad_norm": 2.2045700550079346,
+      "learning_rate": 0.0001804007143949874,
+      "loss": 1.0355,
+      "step": 1178
+    },
+    {
+      "epoch": 0.22681800692574067,
+      "grad_norm": 2.608950138092041,
+      "learning_rate": 0.00018036364895518273,
+      "loss": 0.9324,
+      "step": 1179
+    },
+    {
+      "epoch": 0.22701038861100423,
+      "grad_norm": 2.2471365928649902,
+      "learning_rate": 0.00018032655231544253,
+      "loss": 0.9707,
+      "step": 1180
+    },
+    {
+      "epoch": 0.2272027702962678,
+      "grad_norm": 2.513303279876709,
+      "learning_rate": 0.00018028942449016904,
+      "loss": 1.0566,
+      "step": 1181
+    },
+    {
+      "epoch": 0.22739515198153135,
+      "grad_norm": 2.8760597705841064,
+      "learning_rate": 0.00018025226549377648,
+      "loss": 1.0089,
+      "step": 1182
+    },
+    {
+      "epoch": 0.22758753366679493,
+      "grad_norm": 2.5678937435150146,
+      "learning_rate": 0.0001802150753406913,
+      "loss": 1.0567,
+      "step": 1183
+    },
+    {
+      "epoch": 0.2277799153520585,
+      "grad_norm": 2.192810535430908,
+      "learning_rate": 0.000180177854045352,
+      "loss": 0.9603,
+      "step": 1184
+    },
+    {
+      "epoch": 0.22797229703732205,
+      "grad_norm": 2.597440004348755,
+      "learning_rate": 0.00018014060162220913,
+      "loss": 1.049,
+      "step": 1185
+    },
+    {
+      "epoch": 0.2281646787225856,
+      "grad_norm": 2.5057008266448975,
+      "learning_rate": 0.0001801033180857254,
+      "loss": 1.0134,
+      "step": 1186
+    },
+    {
+      "epoch": 0.22835706040784917,
+      "grad_norm": 1.5720535516738892,
+      "learning_rate": 0.00018006600345037556,
+      "loss": 1.0806,
+      "step": 1187
+    },
+    {
+      "epoch": 0.22854944209311273,
+      "grad_norm": 2.126380443572998,
+      "learning_rate": 0.00018002865773064644,
+      "loss": 1.0063,
+      "step": 1188
+    },
+    {
+      "epoch": 0.2287418237783763,
+      "grad_norm": 1.431265115737915,
+      "learning_rate": 0.00017999128094103688,
+      "loss": 1.0441,
+      "step": 1189
+    },
+    {
+      "epoch": 0.22893420546363985,
+      "grad_norm": 1.2462618350982666,
+      "learning_rate": 0.00017995387309605788,
+      "loss": 1.0391,
+      "step": 1190
+    },
+    {
+      "epoch": 0.22912658714890344,
+      "grad_norm": 2.1262803077697754,
+      "learning_rate": 0.0001799164342102325,
+      "loss": 1.0439,
+      "step": 1191
+    },
+    {
+      "epoch": 0.229318968834167,
+      "grad_norm": 1.1519289016723633,
+      "learning_rate": 0.0001798789642980957,
+      "loss": 1.0239,
+      "step": 1192
+    },
+    {
+      "epoch": 0.22951135051943056,
+      "grad_norm": 1.105398178100586,
+      "learning_rate": 0.00017984146337419467,
+      "loss": 0.7501,
+      "step": 1193
+    },
+    {
+      "epoch": 0.22970373220469412,
+      "grad_norm": 2.241414785385132,
+      "learning_rate": 0.00017980393145308857,
+      "loss": 0.9496,
+      "step": 1194
+    },
+    {
+      "epoch": 0.22989611388995768,
+      "grad_norm": 2.034768581390381,
+      "learning_rate": 0.00017976636854934852,
+      "loss": 1.0716,
+      "step": 1195
+    },
+    {
+      "epoch": 0.23008849557522124,
+      "grad_norm": 1.8366289138793945,
+      "learning_rate": 0.00017972877467755776,
+      "loss": 1.0534,
+      "step": 1196
+    },
+    {
+      "epoch": 0.2302808772604848,
+      "grad_norm": 2.357349157333374,
+      "learning_rate": 0.00017969114985231152,
+      "loss": 1.0243,
+      "step": 1197
+    },
+    {
+      "epoch": 0.23047325894574835,
+      "grad_norm": 2.1695549488067627,
+      "learning_rate": 0.00017965349408821709,
+      "loss": 1.0933,
+      "step": 1198
+    },
+    {
+      "epoch": 0.23066564063101191,
+      "grad_norm": 2.0856971740722656,
+      "learning_rate": 0.00017961580739989365,
+      "loss": 0.9387,
+      "step": 1199
+    },
+    {
+      "epoch": 0.2308580223162755,
+      "grad_norm": 2.325125217437744,
+      "learning_rate": 0.0001795780898019726,
+      "loss": 1.1027,
+      "step": 1200
+    },
+    {
+      "epoch": 0.23105040400153906,
+      "grad_norm": 1.6141457557678223,
+      "learning_rate": 0.00017954034130909707,
+      "loss": 1.0986,
+      "step": 1201
+    },
+    {
+      "epoch": 0.23124278568680262,
+      "grad_norm": 2.047602653503418,
+      "learning_rate": 0.00017950256193592244,
+      "loss": 0.9974,
+      "step": 1202
+    },
+    {
+      "epoch": 0.23143516737206618,
+      "grad_norm": 2.2844059467315674,
+      "learning_rate": 0.00017946475169711588,
+      "loss": 1.0369,
+      "step": 1203
+    },
+    {
+      "epoch": 0.23162754905732974,
+      "grad_norm": 2.9441487789154053,
+      "learning_rate": 0.00017942691060735666,
+      "loss": 1.0038,
+      "step": 1204
+    },
+    {
+      "epoch": 0.2318199307425933,
+      "grad_norm": 2.2085440158843994,
+      "learning_rate": 0.00017938903868133598,
+      "loss": 1.0322,
+      "step": 1205
+    },
+    {
+      "epoch": 0.23201231242785686,
+      "grad_norm": 2.71103572845459,
+      "learning_rate": 0.00017935113593375707,
+      "loss": 0.9543,
+      "step": 1206
+    },
+    {
+      "epoch": 0.23220469411312042,
+      "grad_norm": 2.7186548709869385,
+      "learning_rate": 0.00017931320237933503,
+      "loss": 1.0343,
+      "step": 1207
+    },
+    {
+      "epoch": 0.232397075798384,
+      "grad_norm": 2.017627000808716,
+      "learning_rate": 0.000179275238032797,
+      "loss": 1.0999,
+      "step": 1208
+    },
+    {
+      "epoch": 0.23258945748364757,
+      "grad_norm": 2.463900566101074,
+      "learning_rate": 0.00017923724290888206,
+      "loss": 1.0129,
+      "step": 1209
+    },
+    {
+      "epoch": 0.23278183916891113,
+      "grad_norm": 2.5354909896850586,
+      "learning_rate": 0.00017919921702234118,
+      "loss": 1.0418,
+      "step": 1210
+    },
+    {
+      "epoch": 0.23297422085417469,
+      "grad_norm": 2.327570676803589,
+      "learning_rate": 0.0001791611603879374,
+      "loss": 1.0154,
+      "step": 1211
+    },
+    {
+      "epoch": 0.23316660253943824,
+      "grad_norm": 2.317172050476074,
+      "learning_rate": 0.00017912307302044548,
+      "loss": 1.0644,
+      "step": 1212
+    },
+    {
+      "epoch": 0.2333589842247018,
+      "grad_norm": 1.230281114578247,
+      "learning_rate": 0.00017908495493465236,
+      "loss": 1.0362,
+      "step": 1213
+    },
+    {
+      "epoch": 0.23355136590996536,
+      "grad_norm": 2.1708242893218994,
+      "learning_rate": 0.00017904680614535675,
+      "loss": 1.1298,
+      "step": 1214
+    },
+    {
+      "epoch": 0.23374374759522892,
+      "grad_norm": 1.5918229818344116,
+      "learning_rate": 0.00017900862666736934,
+      "loss": 1.002,
+      "step": 1215
+    },
+    {
+      "epoch": 0.2339361292804925,
+      "grad_norm": 1.2032103538513184,
+      "learning_rate": 0.0001789704165155127,
+      "loss": 1.0116,
+      "step": 1216
+    },
+    {
+      "epoch": 0.23412851096575607,
+      "grad_norm": 2.1854023933410645,
+      "learning_rate": 0.00017893217570462134,
+      "loss": 1.0156,
+      "step": 1217
+    },
+    {
+      "epoch": 0.23432089265101963,
+      "grad_norm": 1.0488488674163818,
+      "learning_rate": 0.00017889390424954166,
+      "loss": 1.0284,
+      "step": 1218
+    },
+    {
+      "epoch": 0.2345132743362832,
+      "grad_norm": 2.528773307800293,
+      "learning_rate": 0.00017885560216513197,
+      "loss": 1.04,
+      "step": 1219
+    },
+    {
+      "epoch": 0.23470565602154675,
+      "grad_norm": 1.436869502067566,
+      "learning_rate": 0.00017881726946626243,
+      "loss": 1.0464,
+      "step": 1220
+    },
+    {
+      "epoch": 0.2348980377068103,
+      "grad_norm": 2.2210469245910645,
+      "learning_rate": 0.00017877890616781512,
+      "loss": 1.0411,
+      "step": 1221
+    },
+    {
+      "epoch": 0.23509041939207387,
+      "grad_norm": 2.654679775238037,
+      "learning_rate": 0.000178740512284684,
+      "loss": 1.0521,
+      "step": 1222
+    },
+    {
+      "epoch": 0.23528280107733743,
+      "grad_norm": 2.3937325477600098,
+      "learning_rate": 0.0001787020878317749,
+      "loss": 1.0303,
+      "step": 1223
+    },
+    {
+      "epoch": 0.235475182762601,
+      "grad_norm": 1.8900139331817627,
+      "learning_rate": 0.00017866363282400554,
+      "loss": 1.0029,
+      "step": 1224
+    },
+    {
+      "epoch": 0.23566756444786457,
+      "grad_norm": 2.116438627243042,
+      "learning_rate": 0.00017862514727630543,
+      "loss": 1.0379,
+      "step": 1225
+    },
+    {
+      "epoch": 0.23585994613312813,
+      "grad_norm": 2.2265512943267822,
+      "learning_rate": 0.00017858663120361596,
+      "loss": 1.0776,
+      "step": 1226
+    },
+    {
+      "epoch": 0.2360523278183917,
+      "grad_norm": 2.128690481185913,
+      "learning_rate": 0.00017854808462089048,
+      "loss": 1.0843,
+      "step": 1227
+    },
+    {
+      "epoch": 0.23624470950365525,
+      "grad_norm": 1.8086310625076294,
+      "learning_rate": 0.00017850950754309407,
+      "loss": 1.0004,
+      "step": 1228
+    },
+    {
+      "epoch": 0.2364370911889188,
+      "grad_norm": 2.247142791748047,
+      "learning_rate": 0.00017847089998520365,
+      "loss": 1.0168,
+      "step": 1229
+    },
+    {
+      "epoch": 0.23662947287418237,
+      "grad_norm": 2.3121066093444824,
+      "learning_rate": 0.00017843226196220802,
+      "loss": 1.1086,
+      "step": 1230
+    },
+    {
+      "epoch": 0.23682185455944593,
+      "grad_norm": 2.0860531330108643,
+      "learning_rate": 0.0001783935934891078,
+      "loss": 1.0128,
+      "step": 1231
+    },
+    {
+      "epoch": 0.2370142362447095,
+      "grad_norm": 2.119245767593384,
+      "learning_rate": 0.00017835489458091543,
+      "loss": 1.0827,
+      "step": 1232
+    },
+    {
+      "epoch": 0.23720661792997308,
+      "grad_norm": 2.149806261062622,
+      "learning_rate": 0.00017831616525265513,
+      "loss": 1.0282,
+      "step": 1233
+    },
+    {
+      "epoch": 0.23739899961523664,
+      "grad_norm": 1.3275519609451294,
+      "learning_rate": 0.00017827740551936295,
+      "loss": 1.1392,
+      "step": 1234
+    },
+    {
+      "epoch": 0.2375913813005002,
+      "grad_norm": 2.11102032661438,
+      "learning_rate": 0.00017823861539608684,
+      "loss": 1.1036,
+      "step": 1235
+    },
+    {
+      "epoch": 0.23778376298576376,
+      "grad_norm": 2.63999080657959,
+      "learning_rate": 0.00017819979489788637,
+      "loss": 1.1314,
+      "step": 1236
+    },
+    {
+      "epoch": 0.23797614467102732,
+      "grad_norm": 2.2188775539398193,
+      "learning_rate": 0.00017816094403983298,
+      "loss": 0.9727,
+      "step": 1237
+    },
+    {
+      "epoch": 0.23816852635629088,
+      "grad_norm": 2.0501976013183594,
+      "learning_rate": 0.00017812206283701002,
+      "loss": 1.0649,
+      "step": 1238
+    },
+    {
+      "epoch": 0.23836090804155444,
+      "grad_norm": 1.5250564813613892,
+      "learning_rate": 0.00017808315130451243,
+      "loss": 1.0516,
+      "step": 1239
+    },
+    {
+      "epoch": 0.238553289726818,
+      "grad_norm": 2.0089285373687744,
+      "learning_rate": 0.000178044209457447,
+      "loss": 0.9768,
+      "step": 1240
+    },
+    {
+      "epoch": 0.23874567141208156,
+      "grad_norm": 1.5823297500610352,
+      "learning_rate": 0.00017800523731093234,
+      "loss": 1.1003,
+      "step": 1241
+    },
+    {
+      "epoch": 0.23893805309734514,
+      "grad_norm": 2.5209779739379883,
+      "learning_rate": 0.00017796623488009874,
+      "loss": 1.0228,
+      "step": 1242
+    },
+    {
+      "epoch": 0.2391304347826087,
+      "grad_norm": 1.590396523475647,
+      "learning_rate": 0.00017792720218008827,
+      "loss": 1.0458,
+      "step": 1243
+    },
+    {
+      "epoch": 0.23932281646787226,
+      "grad_norm": 1.2493937015533447,
+      "learning_rate": 0.00017788813922605488,
+      "loss": 1.0293,
+      "step": 1244
+    },
+    {
+      "epoch": 0.23951519815313582,
+      "grad_norm": 2.1971514225006104,
+      "learning_rate": 0.000177849046033164,
+      "loss": 0.9955,
+      "step": 1245
+    },
+    {
+      "epoch": 0.23970757983839938,
+      "grad_norm": 2.3416659832000732,
+      "learning_rate": 0.00017780992261659306,
+      "loss": 1.003,
+      "step": 1246
+    },
+    {
+      "epoch": 0.23989996152366294,
+      "grad_norm": 1.6771947145462036,
+      "learning_rate": 0.00017777076899153112,
+      "loss": 1.0184,
+      "step": 1247
+    },
+    {
+      "epoch": 0.2400923432089265,
+      "grad_norm": 2.4952237606048584,
+      "learning_rate": 0.00017773158517317888,
+      "loss": 0.996,
+      "step": 1248
+    },
+    {
+      "epoch": 0.24028472489419006,
+      "grad_norm": 1.9115630388259888,
+      "learning_rate": 0.00017769237117674892,
+      "loss": 1.0324,
+      "step": 1249
+    },
+    {
+      "epoch": 0.24047710657945365,
+      "grad_norm": 1.9835323095321655,
+      "learning_rate": 0.00017765312701746545,
+      "loss": 1.0354,
+      "step": 1250
+    },
+    {
+      "epoch": 0.2406694882647172,
+      "grad_norm": 2.302229642868042,
+      "learning_rate": 0.00017761385271056437,
+      "loss": 1.055,
+      "step": 1251
+    },
+    {
+      "epoch": 0.24086186994998077,
+      "grad_norm": 2.347031354904175,
+      "learning_rate": 0.0001775745482712934,
+      "loss": 1.0716,
+      "step": 1252
+    },
+    {
+      "epoch": 0.24105425163524433,
+      "grad_norm": 1.3683199882507324,
+      "learning_rate": 0.00017753521371491175,
+      "loss": 0.8074,
+      "step": 1253
+    },
+    {
+      "epoch": 0.24124663332050789,
+      "grad_norm": 2.1878585815429688,
+      "learning_rate": 0.00017749584905669058,
+      "loss": 0.9352,
+      "step": 1254
+    },
+    {
+      "epoch": 0.24143901500577145,
+      "grad_norm": 2.431112289428711,
+      "learning_rate": 0.00017745645431191252,
+      "loss": 1.066,
+      "step": 1255
+    },
+    {
+      "epoch": 0.241631396691035,
+      "grad_norm": 2.8352198600769043,
+      "learning_rate": 0.00017741702949587194,
+      "loss": 1.0638,
+      "step": 1256
+    },
+    {
+      "epoch": 0.24182377837629856,
+      "grad_norm": 2.51709246635437,
+      "learning_rate": 0.00017737757462387506,
+      "loss": 1.0723,
+      "step": 1257
+    },
+    {
+      "epoch": 0.24201616006156215,
+      "grad_norm": 2.622459888458252,
+      "learning_rate": 0.00017733808971123947,
+      "loss": 1.094,
+      "step": 1258
+    },
+    {
+      "epoch": 0.2422085417468257,
+      "grad_norm": 2.2875192165374756,
+      "learning_rate": 0.00017729857477329463,
+      "loss": 1.0228,
+      "step": 1259
+    },
+    {
+      "epoch": 0.24240092343208927,
+      "grad_norm": 2.6404519081115723,
+      "learning_rate": 0.00017725902982538162,
+      "loss": 0.9921,
+      "step": 1260
+    },
+    {
+      "epoch": 0.24259330511735283,
+      "grad_norm": 1.8993335962295532,
+      "learning_rate": 0.0001772194548828531,
+      "loss": 1.0213,
+      "step": 1261
+    },
+    {
+      "epoch": 0.2427856868026164,
+      "grad_norm": 2.2974517345428467,
+      "learning_rate": 0.00017717984996107344,
+      "loss": 0.9945,
+      "step": 1262
+    },
+    {
+      "epoch": 0.24297806848787995,
+      "grad_norm": 1.206773281097412,
+      "learning_rate": 0.00017714021507541866,
+      "loss": 0.7567,
+      "step": 1263
+    },
+    {
+      "epoch": 0.2431704501731435,
+      "grad_norm": 1.7663559913635254,
+      "learning_rate": 0.00017710055024127637,
+      "loss": 0.9944,
+      "step": 1264
+    },
+    {
+      "epoch": 0.24336283185840707,
+      "grad_norm": 2.63303542137146,
+      "learning_rate": 0.00017706085547404584,
+      "loss": 1.0416,
+      "step": 1265
+    },
+    {
+      "epoch": 0.24355521354367063,
+      "grad_norm": 1.3399397134780884,
+      "learning_rate": 0.0001770211307891379,
+      "loss": 1.0674,
+      "step": 1266
+    },
+    {
+      "epoch": 0.24374759522893422,
+      "grad_norm": 1.8678351640701294,
+      "learning_rate": 0.0001769813762019751,
+      "loss": 0.9822,
+      "step": 1267
+    },
+    {
+      "epoch": 0.24393997691419778,
+      "grad_norm": 2.6744906902313232,
+      "learning_rate": 0.00017694159172799152,
+      "loss": 0.9794,
+      "step": 1268
+    },
+    {
+      "epoch": 0.24413235859946134,
+      "grad_norm": 2.15325927734375,
+      "learning_rate": 0.00017690177738263283,
+      "loss": 1.039,
+      "step": 1269
+    },
+    {
+      "epoch": 0.2443247402847249,
+      "grad_norm": 2.158799171447754,
+      "learning_rate": 0.00017686193318135635,
+      "loss": 1.09,
+      "step": 1270
+    },
+    {
+      "epoch": 0.24451712196998845,
+      "grad_norm": 2.0430867671966553,
+      "learning_rate": 0.00017682205913963103,
+      "loss": 1.04,
+      "step": 1271
+    },
+    {
+      "epoch": 0.24470950365525201,
+      "grad_norm": 1.3784855604171753,
+      "learning_rate": 0.00017678215527293722,
+      "loss": 1.0528,
+      "step": 1272
+    },
+    {
+      "epoch": 0.24490188534051557,
+      "grad_norm": 2.3015739917755127,
+      "learning_rate": 0.00017674222159676708,
+      "loss": 1.0555,
+      "step": 1273
+    },
+    {
+      "epoch": 0.24509426702577913,
+      "grad_norm": 1.3986289501190186,
+      "learning_rate": 0.0001767022581266242,
+      "loss": 1.0861,
+      "step": 1274
+    },
+    {
+      "epoch": 0.24528664871104272,
+      "grad_norm": 1.4476958513259888,
+      "learning_rate": 0.0001766622648780238,
+      "loss": 1.0986,
+      "step": 1275
+    },
+    {
+      "epoch": 0.24547903039630628,
+      "grad_norm": 2.301908493041992,
+      "learning_rate": 0.00017662224186649258,
+      "loss": 0.9753,
+      "step": 1276
+    },
+    {
+      "epoch": 0.24567141208156984,
+      "grad_norm": 1.5684734582901,
+      "learning_rate": 0.00017658218910756893,
+      "loss": 1.0774,
+      "step": 1277
+    },
+    {
+      "epoch": 0.2458637937668334,
+      "grad_norm": 1.1958787441253662,
+      "learning_rate": 0.00017654210661680262,
+      "loss": 1.0338,
+      "step": 1278
+    },
+    {
+      "epoch": 0.24605617545209696,
+      "grad_norm": 1.15227472782135,
+      "learning_rate": 0.0001765019944097551,
+      "loss": 1.0933,
+      "step": 1279
+    },
+    {
+      "epoch": 0.24624855713736052,
+      "grad_norm": 1.098379373550415,
+      "learning_rate": 0.00017646185250199935,
+      "loss": 0.7616,
+      "step": 1280
+    },
+    {
+      "epoch": 0.24644093882262408,
+      "grad_norm": 1.2993512153625488,
+      "learning_rate": 0.00017642168090911973,
+      "loss": 0.7536,
+      "step": 1281
+    },
+    {
+      "epoch": 0.24663332050788764,
+      "grad_norm": 2.544689893722534,
+      "learning_rate": 0.00017638147964671238,
+      "loss": 1.1096,
+      "step": 1282
+    },
+    {
+      "epoch": 0.24682570219315123,
+      "grad_norm": 2.68131422996521,
+      "learning_rate": 0.0001763412487303847,
+      "loss": 1.0169,
+      "step": 1283
+    },
+    {
+      "epoch": 0.24701808387841478,
+      "grad_norm": 1.1457610130310059,
+      "learning_rate": 0.00017630098817575577,
+      "loss": 1.102,
+      "step": 1284
+    },
+    {
+      "epoch": 0.24721046556367834,
+      "grad_norm": 1.0906016826629639,
+      "learning_rate": 0.0001762606979984561,
+      "loss": 1.055,
+      "step": 1285
+    },
+    {
+      "epoch": 0.2474028472489419,
+      "grad_norm": 2.5457074642181396,
+      "learning_rate": 0.00017622037821412776,
+      "loss": 1.0555,
+      "step": 1286
+    },
+    {
+      "epoch": 0.24759522893420546,
+      "grad_norm": 2.6482160091400146,
+      "learning_rate": 0.00017618002883842423,
+      "loss": 0.9556,
+      "step": 1287
+    },
+    {
+      "epoch": 0.24778761061946902,
+      "grad_norm": 1.5438766479492188,
+      "learning_rate": 0.00017613964988701057,
+      "loss": 1.0435,
+      "step": 1288
+    },
+    {
+      "epoch": 0.24797999230473258,
+      "grad_norm": 2.06453537940979,
+      "learning_rate": 0.00017609924137556324,
+      "loss": 1.1069,
+      "step": 1289
+    },
+    {
+      "epoch": 0.24817237398999614,
+      "grad_norm": 2.455542802810669,
+      "learning_rate": 0.00017605880331977023,
+      "loss": 0.9613,
+      "step": 1290
+    },
+    {
+      "epoch": 0.2483647556752597,
+      "grad_norm": 1.5138311386108398,
+      "learning_rate": 0.000176018335735331,
+      "loss": 1.079,
+      "step": 1291
+    },
+    {
+      "epoch": 0.2485571373605233,
+      "grad_norm": 2.5072662830352783,
+      "learning_rate": 0.00017597783863795643,
+      "loss": 1.0925,
+      "step": 1292
+    },
+    {
+      "epoch": 0.24874951904578685,
+      "grad_norm": 2.5864195823669434,
+      "learning_rate": 0.00017593731204336893,
+      "loss": 0.9974,
+      "step": 1293
+    },
+    {
+      "epoch": 0.2489419007310504,
+      "grad_norm": 2.3414435386657715,
+      "learning_rate": 0.0001758967559673023,
+      "loss": 1.0753,
+      "step": 1294
+    },
+    {
+      "epoch": 0.24913428241631397,
+      "grad_norm": 2.2752461433410645,
+      "learning_rate": 0.0001758561704255018,
+      "loss": 1.0812,
+      "step": 1295
+    },
+    {
+      "epoch": 0.24932666410157753,
+      "grad_norm": 1.477254033088684,
+      "learning_rate": 0.00017581555543372412,
+      "loss": 1.0693,
+      "step": 1296
+    },
+    {
+      "epoch": 0.2495190457868411,
+      "grad_norm": 0.8532500267028809,
+      "learning_rate": 0.00017577491100773744,
+      "loss": 1.0218,
+      "step": 1297
+    },
+    {
+      "epoch": 0.24971142747210465,
+      "grad_norm": 2.206050157546997,
+      "learning_rate": 0.00017573423716332127,
+      "loss": 0.9974,
+      "step": 1298
+    },
+    {
+      "epoch": 0.2499038091573682,
+      "grad_norm": 2.388279914855957,
+      "learning_rate": 0.00017569353391626663,
+      "loss": 1.0076,
+      "step": 1299
+    },
+    {
+      "epoch": 0.2500961908426318,
+      "grad_norm": 2.104302406311035,
+      "learning_rate": 0.00017565280128237595,
+      "loss": 1.0418,
+      "step": 1300
+    },
+    {
+      "epoch": 0.2502885725278953,
+      "grad_norm": 0.9650487303733826,
+      "learning_rate": 0.000175612039277463,
+      "loss": 0.7721,
+      "step": 1301
+    },
+    {
+      "epoch": 0.2504809542131589,
+      "grad_norm": 2.598994731903076,
+      "learning_rate": 0.00017557124791735298,
+      "loss": 1.0643,
+      "step": 1302
+    },
+    {
+      "epoch": 0.25067333589842244,
+      "grad_norm": 1.3419384956359863,
+      "learning_rate": 0.00017553042721788254,
+      "loss": 1.0282,
+      "step": 1303
+    },
+    {
+      "epoch": 0.25086571758368603,
+      "grad_norm": 0.9640091061592102,
+      "learning_rate": 0.00017548957719489968,
+      "loss": 0.771,
+      "step": 1304
+    },
+    {
+      "epoch": 0.2510580992689496,
+      "grad_norm": 2.685713529586792,
+      "learning_rate": 0.00017544869786426372,
+      "loss": 1.0808,
+      "step": 1305
+    },
+    {
+      "epoch": 0.25125048095421315,
+      "grad_norm": 2.195932626724243,
+      "learning_rate": 0.00017540778924184553,
+      "loss": 1.0271,
+      "step": 1306
+    },
+    {
+      "epoch": 0.25144286263947674,
+      "grad_norm": 2.195891857147217,
+      "learning_rate": 0.00017536685134352716,
+      "loss": 1.0586,
+      "step": 1307
+    },
+    {
+      "epoch": 0.25163524432474027,
+      "grad_norm": 1.154550552368164,
+      "learning_rate": 0.00017532588418520215,
+      "loss": 1.086,
+      "step": 1308
+    },
+    {
+      "epoch": 0.25182762601000386,
+      "grad_norm": 2.162400245666504,
+      "learning_rate": 0.00017528488778277535,
+      "loss": 1.1033,
+      "step": 1309
+    },
+    {
+      "epoch": 0.2520200076952674,
+      "grad_norm": 2.220500946044922,
+      "learning_rate": 0.000175243862152163,
+      "loss": 0.9728,
+      "step": 1310
+    },
+    {
+      "epoch": 0.252212389380531,
+      "grad_norm": 1.2025529146194458,
+      "learning_rate": 0.0001752028073092926,
+      "loss": 0.9757,
+      "step": 1311
+    },
+    {
+      "epoch": 0.2524047710657945,
+      "grad_norm": 2.7478811740875244,
+      "learning_rate": 0.0001751617232701031,
+      "loss": 1.0634,
+      "step": 1312
+    },
+    {
+      "epoch": 0.2525971527510581,
+      "grad_norm": 2.1777567863464355,
+      "learning_rate": 0.0001751206100505448,
+      "loss": 1.0701,
+      "step": 1313
+    },
+    {
+      "epoch": 0.2527895344363217,
+      "grad_norm": 1.9557124376296997,
+      "learning_rate": 0.00017507946766657912,
+      "loss": 1.0926,
+      "step": 1314
+    },
+    {
+      "epoch": 0.2529819161215852,
+      "grad_norm": 2.5790274143218994,
+      "learning_rate": 0.00017503829613417905,
+      "loss": 1.0147,
+      "step": 1315
+    },
+    {
+      "epoch": 0.2531742978068488,
+      "grad_norm": 2.549546480178833,
+      "learning_rate": 0.0001749970954693288,
+      "loss": 1.0016,
+      "step": 1316
+    },
+    {
+      "epoch": 0.25336667949211233,
+      "grad_norm": 2.2119216918945312,
+      "learning_rate": 0.00017495586568802384,
+      "loss": 1.0276,
+      "step": 1317
+    },
+    {
+      "epoch": 0.2535590611773759,
+      "grad_norm": 2.4197356700897217,
+      "learning_rate": 0.00017491460680627103,
+      "loss": 1.0438,
+      "step": 1318
+    },
+    {
+      "epoch": 0.25375144286263945,
+      "grad_norm": 2.891319990158081,
+      "learning_rate": 0.00017487331884008845,
+      "loss": 1.0422,
+      "step": 1319
+    },
+    {
+      "epoch": 0.25394382454790304,
+      "grad_norm": 2.1061878204345703,
+      "learning_rate": 0.00017483200180550554,
+      "loss": 1.1359,
+      "step": 1320
+    },
+    {
+      "epoch": 0.25413620623316663,
+      "grad_norm": 2.0467493534088135,
+      "learning_rate": 0.00017479065571856302,
+      "loss": 1.0014,
+      "step": 1321
+    },
+    {
+      "epoch": 0.25432858791843016,
+      "grad_norm": 1.3461437225341797,
+      "learning_rate": 0.0001747492805953128,
+      "loss": 0.7614,
+      "step": 1322
+    },
+    {
+      "epoch": 0.25452096960369375,
+      "grad_norm": 2.3887431621551514,
+      "learning_rate": 0.00017470787645181815,
+      "loss": 1.0007,
+      "step": 1323
+    },
+    {
+      "epoch": 0.2547133512889573,
+      "grad_norm": 2.7009894847869873,
+      "learning_rate": 0.00017466644330415364,
+      "loss": 1.0424,
+      "step": 1324
+    },
+    {
+      "epoch": 0.25490573297422087,
+      "grad_norm": 1.9325355291366577,
+      "learning_rate": 0.00017462498116840495,
+      "loss": 0.9965,
+      "step": 1325
+    },
+    {
+      "epoch": 0.2550981146594844,
+      "grad_norm": 2.3231923580169678,
+      "learning_rate": 0.00017458349006066917,
+      "loss": 1.1046,
+      "step": 1326
+    },
+    {
+      "epoch": 0.255290496344748,
+      "grad_norm": 2.42737078666687,
+      "learning_rate": 0.00017454196999705456,
+      "loss": 0.9701,
+      "step": 1327
+    },
+    {
+      "epoch": 0.2554828780300115,
+      "grad_norm": 2.4193291664123535,
+      "learning_rate": 0.00017450042099368066,
+      "loss": 1.0974,
+      "step": 1328
+    },
+    {
+      "epoch": 0.2556752597152751,
+      "grad_norm": 2.055560827255249,
+      "learning_rate": 0.0001744588430666782,
+      "loss": 1.0257,
+      "step": 1329
+    },
+    {
+      "epoch": 0.2558676414005387,
+      "grad_norm": 2.146850109100342,
+      "learning_rate": 0.00017441723623218915,
+      "loss": 1.0047,
+      "step": 1330
+    },
+    {
+      "epoch": 0.2560600230858022,
+      "grad_norm": 2.0058836936950684,
+      "learning_rate": 0.00017437560050636677,
+      "loss": 1.0018,
+      "step": 1331
+    },
+    {
+      "epoch": 0.2562524047710658,
+      "grad_norm": 2.548206329345703,
+      "learning_rate": 0.0001743339359053754,
+      "loss": 0.9804,
+      "step": 1332
+    },
+    {
+      "epoch": 0.25644478645632934,
+      "grad_norm": 2.3021647930145264,
+      "learning_rate": 0.00017429224244539077,
+      "loss": 1.0527,
+      "step": 1333
+    },
+    {
+      "epoch": 0.25663716814159293,
+      "grad_norm": 2.4729061126708984,
+      "learning_rate": 0.00017425052014259963,
+      "loss": 1.06,
+      "step": 1334
+    },
+    {
+      "epoch": 0.25682954982685646,
+      "grad_norm": 2.7245595455169678,
+      "learning_rate": 0.00017420876901320007,
+      "loss": 1.0827,
+      "step": 1335
+    },
+    {
+      "epoch": 0.25702193151212005,
+      "grad_norm": 1.823163628578186,
+      "learning_rate": 0.00017416698907340126,
+      "loss": 1.0603,
+      "step": 1336
+    },
+    {
+      "epoch": 0.2572143131973836,
+      "grad_norm": 1.9595435857772827,
+      "learning_rate": 0.00017412518033942372,
+      "loss": 1.0737,
+      "step": 1337
+    },
+    {
+      "epoch": 0.25740669488264717,
+      "grad_norm": 2.444624423980713,
+      "learning_rate": 0.0001740833428274989,
+      "loss": 0.9972,
+      "step": 1338
+    },
+    {
+      "epoch": 0.25759907656791076,
+      "grad_norm": 2.540358066558838,
+      "learning_rate": 0.00017404147655386963,
+      "loss": 1.0421,
+      "step": 1339
+    },
+    {
+      "epoch": 0.2577914582531743,
+      "grad_norm": 1.977329969406128,
+      "learning_rate": 0.0001739995815347899,
+      "loss": 1.0828,
+      "step": 1340
+    },
+    {
+      "epoch": 0.2579838399384379,
+      "grad_norm": 1.520342469215393,
+      "learning_rate": 0.0001739576577865247,
+      "loss": 1.0976,
+      "step": 1341
+    },
+    {
+      "epoch": 0.2581762216237014,
+      "grad_norm": 1.9534211158752441,
+      "learning_rate": 0.0001739157053253503,
+      "loss": 1.0623,
+      "step": 1342
+    },
+    {
+      "epoch": 0.258368603308965,
+      "grad_norm": 2.3777732849121094,
+      "learning_rate": 0.0001738737241675541,
+      "loss": 1.0377,
+      "step": 1343
+    },
+    {
+      "epoch": 0.2585609849942285,
+      "grad_norm": 2.137897491455078,
+      "learning_rate": 0.00017383171432943466,
+      "loss": 1.066,
+      "step": 1344
+    },
+    {
+      "epoch": 0.2587533666794921,
+      "grad_norm": 2.1926071643829346,
+      "learning_rate": 0.00017378967582730158,
+      "loss": 1.0318,
+      "step": 1345
+    },
+    {
+      "epoch": 0.2589457483647557,
+      "grad_norm": 1.5116124153137207,
+      "learning_rate": 0.00017374760867747572,
+      "loss": 1.1194,
+      "step": 1346
+    },
+    {
+      "epoch": 0.25913813005001923,
+      "grad_norm": 2.8381106853485107,
+      "learning_rate": 0.00017370551289628896,
+      "loss": 1.0087,
+      "step": 1347
+    },
+    {
+      "epoch": 0.2593305117352828,
+      "grad_norm": 2.30118989944458,
+      "learning_rate": 0.00017366338850008432,
+      "loss": 1.0035,
+      "step": 1348
+    },
+    {
+      "epoch": 0.25952289342054635,
+      "grad_norm": 2.242506980895996,
+      "learning_rate": 0.000173621235505216,
+      "loss": 0.9834,
+      "step": 1349
+    },
+    {
+      "epoch": 0.25971527510580994,
+      "grad_norm": 2.3881704807281494,
+      "learning_rate": 0.00017357905392804916,
+      "loss": 1.1534,
+      "step": 1350
+    },
+    {
+      "epoch": 0.25990765679107347,
+      "grad_norm": 2.267690896987915,
+      "learning_rate": 0.0001735368437849602,
+      "loss": 1.0515,
+      "step": 1351
+    },
+    {
+      "epoch": 0.26010003847633706,
+      "grad_norm": 1.2961304187774658,
+      "learning_rate": 0.00017349460509233654,
+      "loss": 0.7712,
+      "step": 1352
+    },
+    {
+      "epoch": 0.2602924201616006,
+      "grad_norm": 2.303182601928711,
+      "learning_rate": 0.0001734523378665767,
+      "loss": 1.0696,
+      "step": 1353
+    },
+    {
+      "epoch": 0.2604848018468642,
+      "grad_norm": 2.274118661880493,
+      "learning_rate": 0.00017341004212409025,
+      "loss": 1.1122,
+      "step": 1354
+    },
+    {
+      "epoch": 0.26067718353212777,
+      "grad_norm": 2.6739158630371094,
+      "learning_rate": 0.00017336771788129785,
+      "loss": 0.9968,
+      "step": 1355
+    },
+    {
+      "epoch": 0.2608695652173913,
+      "grad_norm": 2.0355589389801025,
+      "learning_rate": 0.00017332536515463125,
+      "loss": 1.1069,
+      "step": 1356
+    },
+    {
+      "epoch": 0.2610619469026549,
+      "grad_norm": 1.905661702156067,
+      "learning_rate": 0.00017328298396053325,
+      "loss": 1.0534,
+      "step": 1357
+    },
+    {
+      "epoch": 0.2612543285879184,
+      "grad_norm": 2.6489291191101074,
+      "learning_rate": 0.00017324057431545765,
+      "loss": 1.0948,
+      "step": 1358
+    },
+    {
+      "epoch": 0.261446710273182,
+      "grad_norm": 2.3785314559936523,
+      "learning_rate": 0.00017319813623586935,
+      "loss": 1.0658,
+      "step": 1359
+    },
+    {
+      "epoch": 0.26163909195844554,
+      "grad_norm": 2.2833800315856934,
+      "learning_rate": 0.0001731556697382443,
+      "loss": 1.1292,
+      "step": 1360
+    },
+    {
+      "epoch": 0.2618314736437091,
+      "grad_norm": 1.9547423124313354,
+      "learning_rate": 0.00017311317483906943,
+      "loss": 1.1028,
+      "step": 1361
+    },
+    {
+      "epoch": 0.26202385532897265,
+      "grad_norm": 1.6821837425231934,
+      "learning_rate": 0.00017307065155484273,
+      "loss": 1.0818,
+      "step": 1362
+    },
+    {
+      "epoch": 0.26221623701423624,
+      "grad_norm": 0.9899166822433472,
+      "learning_rate": 0.00017302809990207317,
+      "loss": 1.059,
+      "step": 1363
+    },
+    {
+      "epoch": 0.26240861869949983,
+      "grad_norm": 2.402493953704834,
+      "learning_rate": 0.00017298551989728084,
+      "loss": 1.0477,
+      "step": 1364
+    },
+    {
+      "epoch": 0.26260100038476336,
+      "grad_norm": 2.4694020748138428,
+      "learning_rate": 0.0001729429115569967,
+      "loss": 1.0028,
+      "step": 1365
+    },
+    {
+      "epoch": 0.26279338207002695,
+      "grad_norm": 2.243969678878784,
+      "learning_rate": 0.00017290027489776278,
+      "loss": 1.0345,
+      "step": 1366
+    },
+    {
+      "epoch": 0.2629857637552905,
+      "grad_norm": 2.239915370941162,
+      "learning_rate": 0.00017285760993613215,
+      "loss": 1.0381,
+      "step": 1367
+    },
+    {
+      "epoch": 0.26317814544055407,
+      "grad_norm": 2.3187098503112793,
+      "learning_rate": 0.00017281491668866874,
+      "loss": 1.0303,
+      "step": 1368
+    },
+    {
+      "epoch": 0.2633705271258176,
+      "grad_norm": 2.463526964187622,
+      "learning_rate": 0.0001727721951719476,
+      "loss": 1.0406,
+      "step": 1369
+    },
+    {
+      "epoch": 0.2635629088110812,
+      "grad_norm": 1.8342443704605103,
+      "learning_rate": 0.0001727294454025547,
+      "loss": 1.0585,
+      "step": 1370
+    },
+    {
+      "epoch": 0.2637552904963448,
+      "grad_norm": 2.324680805206299,
+      "learning_rate": 0.0001726866673970869,
+      "loss": 1.0797,
+      "step": 1371
+    },
+    {
+      "epoch": 0.2639476721816083,
+      "grad_norm": 2.6571972370147705,
+      "learning_rate": 0.00017264386117215216,
+      "loss": 1.061,
+      "step": 1372
+    },
+    {
+      "epoch": 0.2641400538668719,
+      "grad_norm": 2.1576762199401855,
+      "learning_rate": 0.00017260102674436932,
+      "loss": 0.9648,
+      "step": 1373
+    },
+    {
+      "epoch": 0.2643324355521354,
+      "grad_norm": 2.1332836151123047,
+      "learning_rate": 0.0001725581641303682,
+      "loss": 0.9771,
+      "step": 1374
+    },
+    {
+      "epoch": 0.264524817237399,
+      "grad_norm": 1.057054042816162,
+      "learning_rate": 0.00017251527334678944,
+      "loss": 0.7249,
+      "step": 1375
+    },
+    {
+      "epoch": 0.26471719892266254,
+      "grad_norm": 1.5178654193878174,
+      "learning_rate": 0.00017247235441028485,
+      "loss": 1.111,
+      "step": 1376
+    },
+    {
+      "epoch": 0.26490958060792613,
+      "grad_norm": 1.61638605594635,
+      "learning_rate": 0.00017242940733751695,
+      "loss": 1.0225,
+      "step": 1377
+    },
+    {
+      "epoch": 0.26510196229318966,
+      "grad_norm": 2.3886115550994873,
+      "learning_rate": 0.00017238643214515932,
+      "loss": 1.0446,
+      "step": 1378
+    },
+    {
+      "epoch": 0.26529434397845325,
+      "grad_norm": 2.365663528442383,
+      "learning_rate": 0.00017234342884989642,
+      "loss": 0.979,
+      "step": 1379
+    },
+    {
+      "epoch": 0.26548672566371684,
+      "grad_norm": 2.4050042629241943,
+      "learning_rate": 0.00017230039746842353,
+      "loss": 1.0616,
+      "step": 1380
+    },
+    {
+      "epoch": 0.26567910734898037,
+      "grad_norm": 1.8654429912567139,
+      "learning_rate": 0.000172257338017447,
+      "loss": 1.0205,
+      "step": 1381
+    },
+    {
+      "epoch": 0.26587148903424396,
+      "grad_norm": 2.4190709590911865,
+      "learning_rate": 0.00017221425051368395,
+      "loss": 1.0777,
+      "step": 1382
+    },
+    {
+      "epoch": 0.2660638707195075,
+      "grad_norm": 1.6328825950622559,
+      "learning_rate": 0.00017217113497386242,
+      "loss": 1.0063,
+      "step": 1383
+    },
+    {
+      "epoch": 0.2662562524047711,
+      "grad_norm": 2.2495176792144775,
+      "learning_rate": 0.0001721279914147214,
+      "loss": 1.0603,
+      "step": 1384
+    },
+    {
+      "epoch": 0.2664486340900346,
+      "grad_norm": 2.403923511505127,
+      "learning_rate": 0.00017208481985301065,
+      "loss": 1.0425,
+      "step": 1385
+    },
+    {
+      "epoch": 0.2666410157752982,
+      "grad_norm": 2.4303441047668457,
+      "learning_rate": 0.00017204162030549092,
+      "loss": 1.0771,
+      "step": 1386
+    },
+    {
+      "epoch": 0.2668333974605617,
+      "grad_norm": 2.126521110534668,
+      "learning_rate": 0.00017199839278893368,
+      "loss": 1.0262,
+      "step": 1387
+    },
+    {
+      "epoch": 0.2670257791458253,
+      "grad_norm": 3.0165412425994873,
+      "learning_rate": 0.0001719551373201214,
+      "loss": 0.975,
+      "step": 1388
+    },
+    {
+      "epoch": 0.2672181608310889,
+      "grad_norm": 2.410034656524658,
+      "learning_rate": 0.00017191185391584733,
+      "loss": 1.0121,
+      "step": 1389
+    },
+    {
+      "epoch": 0.26741054251635243,
+      "grad_norm": 1.2418285608291626,
+      "learning_rate": 0.00017186854259291557,
+      "loss": 1.1224,
+      "step": 1390
+    },
+    {
+      "epoch": 0.267602924201616,
+      "grad_norm": 2.3707518577575684,
+      "learning_rate": 0.00017182520336814106,
+      "loss": 1.0124,
+      "step": 1391
+    },
+    {
+      "epoch": 0.26779530588687955,
+      "grad_norm": 1.229952335357666,
+      "learning_rate": 0.00017178183625834957,
+      "loss": 1.0159,
+      "step": 1392
+    },
+    {
+      "epoch": 0.26798768757214314,
+      "grad_norm": 2.779414653778076,
+      "learning_rate": 0.00017173844128037775,
+      "loss": 1.0279,
+      "step": 1393
+    },
+    {
+      "epoch": 0.2681800692574067,
+      "grad_norm": 2.0206186771392822,
+      "learning_rate": 0.000171695018451073,
+      "loss": 1.0392,
+      "step": 1394
+    },
+    {
+      "epoch": 0.26837245094267026,
+      "grad_norm": 1.816756010055542,
+      "learning_rate": 0.00017165156778729354,
+      "loss": 1.0052,
+      "step": 1395
+    },
+    {
+      "epoch": 0.26856483262793385,
+      "grad_norm": 2.403672456741333,
+      "learning_rate": 0.00017160808930590844,
+      "loss": 1.1009,
+      "step": 1396
+    },
+    {
+      "epoch": 0.2687572143131974,
+      "grad_norm": 2.326414108276367,
+      "learning_rate": 0.00017156458302379753,
+      "loss": 1.0849,
+      "step": 1397
+    },
+    {
+      "epoch": 0.26894959599846097,
+      "grad_norm": 1.672452449798584,
+      "learning_rate": 0.00017152104895785147,
+      "loss": 0.9546,
+      "step": 1398
+    },
+    {
+      "epoch": 0.2691419776837245,
+      "grad_norm": 1.4210208654403687,
+      "learning_rate": 0.00017147748712497164,
+      "loss": 1.0344,
+      "step": 1399
+    },
+    {
+      "epoch": 0.2693343593689881,
+      "grad_norm": 2.269484281539917,
+      "learning_rate": 0.00017143389754207026,
+      "loss": 1.1074,
+      "step": 1400
+    },
+    {
+      "epoch": 0.2695267410542516,
+      "grad_norm": 2.5031158924102783,
+      "learning_rate": 0.00017139028022607032,
+      "loss": 1.0734,
+      "step": 1401
+    },
+    {
+      "epoch": 0.2697191227395152,
+      "grad_norm": 2.597580671310425,
+      "learning_rate": 0.00017134663519390557,
+      "loss": 1.0647,
+      "step": 1402
+    },
+    {
+      "epoch": 0.26991150442477874,
+      "grad_norm": 2.19577693939209,
+      "learning_rate": 0.00017130296246252046,
+      "loss": 1.0456,
+      "step": 1403
+    },
+    {
+      "epoch": 0.2701038861100423,
+      "grad_norm": 2.221684217453003,
+      "learning_rate": 0.00017125926204887034,
+      "loss": 1.0503,
+      "step": 1404
+    },
+    {
+      "epoch": 0.2702962677953059,
+      "grad_norm": 2.5676352977752686,
+      "learning_rate": 0.00017121553396992113,
+      "loss": 1.0192,
+      "step": 1405
+    },
+    {
+      "epoch": 0.27048864948056944,
+      "grad_norm": 2.682461738586426,
+      "learning_rate": 0.00017117177824264962,
+      "loss": 1.052,
+      "step": 1406
+    },
+    {
+      "epoch": 0.27068103116583303,
+      "grad_norm": 1.3546051979064941,
+      "learning_rate": 0.00017112799488404326,
+      "loss": 0.7804,
+      "step": 1407
+    },
+    {
+      "epoch": 0.27087341285109656,
+      "grad_norm": 1.3363609313964844,
+      "learning_rate": 0.00017108418391110033,
+      "loss": 1.0081,
+      "step": 1408
+    },
+    {
+      "epoch": 0.27106579453636015,
+      "grad_norm": 1.6512356996536255,
+      "learning_rate": 0.00017104034534082968,
+      "loss": 1.0765,
+      "step": 1409
+    },
+    {
+      "epoch": 0.2712581762216237,
+      "grad_norm": 2.3677053451538086,
+      "learning_rate": 0.00017099647919025096,
+      "loss": 1.0028,
+      "step": 1410
+    },
+    {
+      "epoch": 0.27145055790688727,
+      "grad_norm": 1.815568208694458,
+      "learning_rate": 0.00017095258547639454,
+      "loss": 1.0535,
+      "step": 1411
+    },
+    {
+      "epoch": 0.2716429395921508,
+      "grad_norm": 1.549596905708313,
+      "learning_rate": 0.0001709086642163015,
+      "loss": 0.9981,
+      "step": 1412
+    },
+    {
+      "epoch": 0.2718353212774144,
+      "grad_norm": 2.2026844024658203,
+      "learning_rate": 0.00017086471542702354,
+      "loss": 1.0424,
+      "step": 1413
+    },
+    {
+      "epoch": 0.272027702962678,
+      "grad_norm": 2.0810914039611816,
+      "learning_rate": 0.00017082073912562309,
+      "loss": 1.0251,
+      "step": 1414
+    },
+    {
+      "epoch": 0.2722200846479415,
+      "grad_norm": 2.558828353881836,
+      "learning_rate": 0.0001707767353291733,
+      "loss": 0.9743,
+      "step": 1415
+    },
+    {
+      "epoch": 0.2724124663332051,
+      "grad_norm": 2.5498883724212646,
+      "learning_rate": 0.00017073270405475795,
+      "loss": 1.0242,
+      "step": 1416
+    },
+    {
+      "epoch": 0.2726048480184686,
+      "grad_norm": 2.480581521987915,
+      "learning_rate": 0.00017068864531947147,
+      "loss": 0.998,
+      "step": 1417
+    },
+    {
+      "epoch": 0.2727972297037322,
+      "grad_norm": 2.3526487350463867,
+      "learning_rate": 0.00017064455914041896,
+      "loss": 1.0752,
+      "step": 1418
+    },
+    {
+      "epoch": 0.27298961138899575,
+      "grad_norm": 1.8706908226013184,
+      "learning_rate": 0.00017060044553471627,
+      "loss": 1.0192,
+      "step": 1419
+    },
+    {
+      "epoch": 0.27318199307425933,
+      "grad_norm": 1.1634970903396606,
+      "learning_rate": 0.0001705563045194898,
+      "loss": 1.0572,
+      "step": 1420
+    },
+    {
+      "epoch": 0.2733743747595229,
+      "grad_norm": 1.6062870025634766,
+      "learning_rate": 0.00017051213611187654,
+      "loss": 1.0322,
+      "step": 1421
+    },
+    {
+      "epoch": 0.27356675644478645,
+      "grad_norm": 1.5084189176559448,
+      "learning_rate": 0.0001704679403290243,
+      "loss": 0.983,
+      "step": 1422
+    },
+    {
+      "epoch": 0.27375913813005004,
+      "grad_norm": 1.6595219373703003,
+      "learning_rate": 0.0001704237171880913,
+      "loss": 1.0531,
+      "step": 1423
+    },
+    {
+      "epoch": 0.27395151981531357,
+      "grad_norm": 1.8613473176956177,
+      "learning_rate": 0.00017037946670624653,
+      "loss": 1.0815,
+      "step": 1424
+    },
+    {
+      "epoch": 0.27414390150057716,
+      "grad_norm": 2.0711188316345215,
+      "learning_rate": 0.00017033518890066955,
+      "loss": 1.0265,
+      "step": 1425
+    },
+    {
+      "epoch": 0.2743362831858407,
+      "grad_norm": 2.651520252227783,
+      "learning_rate": 0.00017029088378855053,
+      "loss": 1.0353,
+      "step": 1426
+    },
+    {
+      "epoch": 0.2745286648711043,
+      "grad_norm": 1.1796756982803345,
+      "learning_rate": 0.00017024655138709024,
+      "loss": 1.0558,
+      "step": 1427
+    },
+    {
+      "epoch": 0.2747210465563678,
+      "grad_norm": 2.4815239906311035,
+      "learning_rate": 0.00017020219171350002,
+      "loss": 1.0061,
+      "step": 1428
+    },
+    {
+      "epoch": 0.2749134282416314,
+      "grad_norm": 1.3811341524124146,
+      "learning_rate": 0.00017015780478500187,
+      "loss": 0.7866,
+      "step": 1429
+    },
+    {
+      "epoch": 0.275105809926895,
+      "grad_norm": 2.2486257553100586,
+      "learning_rate": 0.0001701133906188283,
+      "loss": 0.9944,
+      "step": 1430
+    },
+    {
+      "epoch": 0.2752981916121585,
+      "grad_norm": 1.7694876194000244,
+      "learning_rate": 0.00017006894923222242,
+      "loss": 0.996,
+      "step": 1431
+    },
+    {
+      "epoch": 0.2754905732974221,
+      "grad_norm": 0.7516884207725525,
+      "learning_rate": 0.0001700244806424379,
+      "loss": 1.0879,
+      "step": 1432
+    },
+    {
+      "epoch": 0.27568295498268564,
+      "grad_norm": 2.6010098457336426,
+      "learning_rate": 0.00016997998486673895,
+      "loss": 1.0035,
+      "step": 1433
+    },
+    {
+      "epoch": 0.2758753366679492,
+      "grad_norm": 2.253753423690796,
+      "learning_rate": 0.00016993546192240041,
+      "loss": 1.0512,
+      "step": 1434
+    },
+    {
+      "epoch": 0.27606771835321275,
+      "grad_norm": 2.1556174755096436,
+      "learning_rate": 0.00016989091182670763,
+      "loss": 1.0057,
+      "step": 1435
+    },
+    {
+      "epoch": 0.27626010003847634,
+      "grad_norm": 1.5043907165527344,
+      "learning_rate": 0.00016984633459695645,
+      "loss": 1.0783,
+      "step": 1436
+    },
+    {
+      "epoch": 0.2764524817237399,
+      "grad_norm": 2.4058079719543457,
+      "learning_rate": 0.00016980173025045328,
+      "loss": 0.9525,
+      "step": 1437
+    },
+    {
+      "epoch": 0.27664486340900346,
+      "grad_norm": 1.7968467473983765,
+      "learning_rate": 0.0001697570988045151,
+      "loss": 1.0789,
+      "step": 1438
+    },
+    {
+      "epoch": 0.27683724509426705,
+      "grad_norm": 2.6519839763641357,
+      "learning_rate": 0.00016971244027646937,
+      "loss": 1.0929,
+      "step": 1439
+    },
+    {
+      "epoch": 0.2770296267795306,
+      "grad_norm": 2.536473274230957,
+      "learning_rate": 0.000169667754683654,
+      "loss": 0.9751,
+      "step": 1440
+    },
+    {
+      "epoch": 0.27722200846479417,
+      "grad_norm": 1.5594525337219238,
+      "learning_rate": 0.0001696230420434176,
+      "loss": 1.0039,
+      "step": 1441
+    },
+    {
+      "epoch": 0.2774143901500577,
+      "grad_norm": 2.3635408878326416,
+      "learning_rate": 0.00016957830237311904,
+      "loss": 0.9752,
+      "step": 1442
+    },
+    {
+      "epoch": 0.2776067718353213,
+      "grad_norm": 2.238832950592041,
+      "learning_rate": 0.00016953353569012784,
+      "loss": 1.0539,
+      "step": 1443
+    },
+    {
+      "epoch": 0.2777991535205848,
+      "grad_norm": 1.998901128768921,
+      "learning_rate": 0.00016948874201182403,
+      "loss": 0.9935,
+      "step": 1444
+    },
+    {
+      "epoch": 0.2779915352058484,
+      "grad_norm": 1.6751521825790405,
+      "learning_rate": 0.00016944392135559796,
+      "loss": 1.0536,
+      "step": 1445
+    },
+    {
+      "epoch": 0.278183916891112,
+      "grad_norm": 2.2686245441436768,
+      "learning_rate": 0.00016939907373885062,
+      "loss": 1.0081,
+      "step": 1446
+    },
+    {
+      "epoch": 0.2783762985763755,
+      "grad_norm": 2.406878709793091,
+      "learning_rate": 0.00016935419917899335,
+      "loss": 0.9994,
+      "step": 1447
+    },
+    {
+      "epoch": 0.2785686802616391,
+      "grad_norm": 2.474165678024292,
+      "learning_rate": 0.00016930929769344805,
+      "loss": 1.0405,
+      "step": 1448
+    },
+    {
+      "epoch": 0.27876106194690264,
+      "grad_norm": 2.21280574798584,
+      "learning_rate": 0.00016926436929964699,
+      "loss": 1.0558,
+      "step": 1449
+    },
+    {
+      "epoch": 0.27895344363216623,
+      "grad_norm": 1.9147347211837769,
+      "learning_rate": 0.0001692194140150329,
+      "loss": 1.0408,
+      "step": 1450
+    },
+    {
+      "epoch": 0.27914582531742976,
+      "grad_norm": 2.1440470218658447,
+      "learning_rate": 0.000169174431857059,
+      "loss": 1.07,
+      "step": 1451
+    },
+    {
+      "epoch": 0.27933820700269335,
+      "grad_norm": 1.0822981595993042,
+      "learning_rate": 0.00016912942284318894,
+      "loss": 0.8051,
+      "step": 1452
+    },
+    {
+      "epoch": 0.2795305886879569,
+      "grad_norm": 1.7865476608276367,
+      "learning_rate": 0.00016908438699089672,
+      "loss": 1.0233,
+      "step": 1453
+    },
+    {
+      "epoch": 0.27972297037322047,
+      "grad_norm": 2.4556214809417725,
+      "learning_rate": 0.0001690393243176668,
+      "loss": 1.0167,
+      "step": 1454
+    },
+    {
+      "epoch": 0.27991535205848406,
+      "grad_norm": 2.2799856662750244,
+      "learning_rate": 0.00016899423484099412,
+      "loss": 0.9492,
+      "step": 1455
+    },
+    {
+      "epoch": 0.2801077337437476,
+      "grad_norm": 1.9392930269241333,
+      "learning_rate": 0.00016894911857838393,
+      "loss": 1.0299,
+      "step": 1456
+    },
+    {
+      "epoch": 0.2803001154290112,
+      "grad_norm": 2.372624635696411,
+      "learning_rate": 0.00016890397554735192,
+      "loss": 1.0968,
+      "step": 1457
+    },
+    {
+      "epoch": 0.2804924971142747,
+      "grad_norm": 2.610835313796997,
+      "learning_rate": 0.00016885880576542417,
+      "loss": 0.9813,
+      "step": 1458
+    },
+    {
+      "epoch": 0.2806848787995383,
+      "grad_norm": 2.3700079917907715,
+      "learning_rate": 0.00016881360925013713,
+      "loss": 1.052,
+      "step": 1459
+    },
+    {
+      "epoch": 0.2808772604848018,
+      "grad_norm": 2.599966049194336,
+      "learning_rate": 0.00016876838601903767,
+      "loss": 1.0604,
+      "step": 1460
+    },
+    {
+      "epoch": 0.2810696421700654,
+      "grad_norm": 2.674349069595337,
+      "learning_rate": 0.00016872313608968294,
+      "loss": 1.0528,
+      "step": 1461
+    },
+    {
+      "epoch": 0.28126202385532895,
+      "grad_norm": 2.975950241088867,
+      "learning_rate": 0.00016867785947964064,
+      "loss": 1.0615,
+      "step": 1462
+    },
+    {
+      "epoch": 0.28145440554059253,
+      "grad_norm": 1.871152400970459,
+      "learning_rate": 0.00016863255620648863,
+      "loss": 1.0416,
+      "step": 1463
+    },
+    {
+      "epoch": 0.2816467872258561,
+      "grad_norm": 2.512075901031494,
+      "learning_rate": 0.0001685872262878152,
+      "loss": 0.9732,
+      "step": 1464
+    },
+    {
+      "epoch": 0.28183916891111965,
+      "grad_norm": 2.6508140563964844,
+      "learning_rate": 0.000168541869741219,
+      "loss": 1.0802,
+      "step": 1465
+    },
+    {
+      "epoch": 0.28203155059638324,
+      "grad_norm": 2.464792013168335,
+      "learning_rate": 0.000168496486584309,
+      "loss": 1.0576,
+      "step": 1466
+    },
+    {
+      "epoch": 0.28222393228164677,
+      "grad_norm": 1.1149930953979492,
+      "learning_rate": 0.00016845107683470452,
+      "loss": 1.0356,
+      "step": 1467
+    },
+    {
+      "epoch": 0.28241631396691036,
+      "grad_norm": 2.1875860691070557,
+      "learning_rate": 0.00016840564051003514,
+      "loss": 0.9991,
+      "step": 1468
+    },
+    {
+      "epoch": 0.2826086956521739,
+      "grad_norm": 2.3508529663085938,
+      "learning_rate": 0.00016836017762794086,
+      "loss": 1.0392,
+      "step": 1469
+    },
+    {
+      "epoch": 0.2828010773374375,
+      "grad_norm": 2.027266025543213,
+      "learning_rate": 0.0001683146882060719,
+      "loss": 1.0157,
+      "step": 1470
+    },
+    {
+      "epoch": 0.28299345902270107,
+      "grad_norm": 0.9561450481414795,
+      "learning_rate": 0.00016826917226208887,
+      "loss": 1.0997,
+      "step": 1471
+    },
+    {
+      "epoch": 0.2831858407079646,
+      "grad_norm": 2.5785396099090576,
+      "learning_rate": 0.00016822362981366257,
+      "loss": 0.982,
+      "step": 1472
+    },
+    {
+      "epoch": 0.2833782223932282,
+      "grad_norm": 2.364617109298706,
+      "learning_rate": 0.00016817806087847417,
+      "loss": 1.002,
+      "step": 1473
+    },
+    {
+      "epoch": 0.2835706040784917,
+      "grad_norm": 1.808104395866394,
+      "learning_rate": 0.0001681324654742151,
+      "loss": 0.9632,
+      "step": 1474
+    },
+    {
+      "epoch": 0.2837629857637553,
+      "grad_norm": 2.3287322521209717,
+      "learning_rate": 0.00016808684361858703,
+      "loss": 0.987,
+      "step": 1475
+    },
+    {
+      "epoch": 0.28395536744901884,
+      "grad_norm": 2.256842851638794,
+      "learning_rate": 0.000168041195329302,
+      "loss": 0.9858,
+      "step": 1476
+    },
+    {
+      "epoch": 0.2841477491342824,
+      "grad_norm": 2.9015562534332275,
+      "learning_rate": 0.00016799552062408225,
+      "loss": 1.0083,
+      "step": 1477
+    },
+    {
+      "epoch": 0.28434013081954596,
+      "grad_norm": 1.978420376777649,
+      "learning_rate": 0.00016794981952066018,
+      "loss": 0.9606,
+      "step": 1478
+    },
+    {
+      "epoch": 0.28453251250480954,
+      "grad_norm": 1.3773670196533203,
+      "learning_rate": 0.0001679040920367786,
+      "loss": 0.7885,
+      "step": 1479
+    },
+    {
+      "epoch": 0.28472489419007313,
+      "grad_norm": 1.9931249618530273,
+      "learning_rate": 0.0001678583381901905,
+      "loss": 1.0168,
+      "step": 1480
+    },
+    {
+      "epoch": 0.28491727587533666,
+      "grad_norm": 2.3492143154144287,
+      "learning_rate": 0.0001678125579986591,
+      "loss": 1.0362,
+      "step": 1481
+    },
+    {
+      "epoch": 0.28510965756060025,
+      "grad_norm": 2.273902654647827,
+      "learning_rate": 0.0001677667514799578,
+      "loss": 1.0647,
+      "step": 1482
+    },
+    {
+      "epoch": 0.2853020392458638,
+      "grad_norm": 2.1355292797088623,
+      "learning_rate": 0.0001677209186518703,
+      "loss": 0.9801,
+      "step": 1483
+    },
+    {
+      "epoch": 0.28549442093112737,
+      "grad_norm": 2.2714505195617676,
+      "learning_rate": 0.00016767505953219048,
+      "loss": 1.0327,
+      "step": 1484
+    },
+    {
+      "epoch": 0.2856868026163909,
+      "grad_norm": 1.7637529373168945,
+      "learning_rate": 0.00016762917413872246,
+      "loss": 0.9835,
+      "step": 1485
+    },
+    {
+      "epoch": 0.2858791843016545,
+      "grad_norm": 2.8837931156158447,
+      "learning_rate": 0.00016758326248928049,
+      "loss": 1.0291,
+      "step": 1486
+    },
+    {
+      "epoch": 0.286071565986918,
+      "grad_norm": 2.4129912853240967,
+      "learning_rate": 0.00016753732460168905,
+      "loss": 0.9016,
+      "step": 1487
+    },
+    {
+      "epoch": 0.2862639476721816,
+      "grad_norm": 2.529222011566162,
+      "learning_rate": 0.0001674913604937828,
+      "loss": 1.0259,
+      "step": 1488
+    },
+    {
+      "epoch": 0.2864563293574452,
+      "grad_norm": 2.343973398208618,
+      "learning_rate": 0.0001674453701834066,
+      "loss": 1.0446,
+      "step": 1489
+    },
+    {
+      "epoch": 0.2866487110427087,
+      "grad_norm": 1.877835750579834,
+      "learning_rate": 0.00016739935368841554,
+      "loss": 1.0453,
+      "step": 1490
+    },
+    {
+      "epoch": 0.2868410927279723,
+      "grad_norm": 1.7767246961593628,
+      "learning_rate": 0.00016735331102667473,
+      "loss": 1.0269,
+      "step": 1491
+    },
+    {
+      "epoch": 0.28703347441323585,
+      "grad_norm": 2.5047123432159424,
+      "learning_rate": 0.00016730724221605952,
+      "loss": 1.0017,
+      "step": 1492
+    },
+    {
+      "epoch": 0.28722585609849943,
+      "grad_norm": 2.4183034896850586,
+      "learning_rate": 0.00016726114727445544,
+      "loss": 1.0031,
+      "step": 1493
+    },
+    {
+      "epoch": 0.28741823778376296,
+      "grad_norm": 2.38810133934021,
+      "learning_rate": 0.00016721502621975813,
+      "loss": 1.0351,
+      "step": 1494
+    },
+    {
+      "epoch": 0.28761061946902655,
+      "grad_norm": 2.32948899269104,
+      "learning_rate": 0.00016716887906987333,
+      "loss": 1.0573,
+      "step": 1495
+    },
+    {
+      "epoch": 0.28780300115429014,
+      "grad_norm": 2.161437511444092,
+      "learning_rate": 0.000167122705842717,
+      "loss": 1.0618,
+      "step": 1496
+    },
+    {
+      "epoch": 0.28799538283955367,
+      "grad_norm": 2.0798885822296143,
+      "learning_rate": 0.00016707650655621516,
+      "loss": 1.086,
+      "step": 1497
+    },
+    {
+      "epoch": 0.28818776452481726,
+      "grad_norm": 2.2581613063812256,
+      "learning_rate": 0.000167030281228304,
+      "loss": 0.9859,
+      "step": 1498
+    },
+    {
+      "epoch": 0.2883801462100808,
+      "grad_norm": 2.4295785427093506,
+      "learning_rate": 0.00016698402987692966,
+      "loss": 1.0323,
+      "step": 1499
+    },
+    {
+      "epoch": 0.2885725278953444,
+      "grad_norm": 0.7178162932395935,
+      "learning_rate": 0.00016693775252004867,
+      "loss": 1.0698,
+      "step": 1500
+    },
+    {
+      "epoch": 0.2887649095806079,
+      "grad_norm": 1.9568020105361938,
+      "learning_rate": 0.0001668914491756274,
+      "loss": 1.0477,
+      "step": 1501
+    },
+    {
+      "epoch": 0.2889572912658715,
+      "grad_norm": 1.934302568435669,
+      "learning_rate": 0.0001668451198616424,
+      "loss": 0.9576,
+      "step": 1502
+    },
+    {
+      "epoch": 0.28914967295113503,
+      "grad_norm": 2.1525704860687256,
+      "learning_rate": 0.00016679876459608034,
+      "loss": 1.0335,
+      "step": 1503
+    },
+    {
+      "epoch": 0.2893420546363986,
+      "grad_norm": 1.245627999305725,
+      "learning_rate": 0.00016675238339693791,
+      "loss": 0.7562,
+      "step": 1504
+    },
+    {
+      "epoch": 0.2895344363216622,
+      "grad_norm": 2.8359084129333496,
+      "learning_rate": 0.00016670597628222188,
+      "loss": 1.0272,
+      "step": 1505
+    },
+    {
+      "epoch": 0.28972681800692573,
+      "grad_norm": 2.427338123321533,
+      "learning_rate": 0.0001666595432699491,
+      "loss": 1.0619,
+      "step": 1506
+    },
+    {
+      "epoch": 0.2899191996921893,
+      "grad_norm": 1.8673861026763916,
+      "learning_rate": 0.0001666130843781465,
+      "loss": 1.0121,
+      "step": 1507
+    },
+    {
+      "epoch": 0.29011158137745285,
+      "grad_norm": 2.1738505363464355,
+      "learning_rate": 0.00016656659962485098,
+      "loss": 1.0056,
+      "step": 1508
+    },
+    {
+      "epoch": 0.29030396306271644,
+      "grad_norm": 1.3233425617218018,
+      "learning_rate": 0.0001665200890281095,
+      "loss": 1.0612,
+      "step": 1509
+    },
+    {
+      "epoch": 0.29049634474798,
+      "grad_norm": 1.7807660102844238,
+      "learning_rate": 0.00016647355260597914,
+      "loss": 0.9735,
+      "step": 1510
+    },
+    {
+      "epoch": 0.29068872643324356,
+      "grad_norm": 2.8546807765960693,
+      "learning_rate": 0.00016642699037652692,
+      "loss": 1.0345,
+      "step": 1511
+    },
+    {
+      "epoch": 0.2908811081185071,
+      "grad_norm": 2.5237629413604736,
+      "learning_rate": 0.00016638040235782982,
+      "loss": 1.1298,
+      "step": 1512
+    },
+    {
+      "epoch": 0.2910734898037707,
+      "grad_norm": 2.301649332046509,
+      "learning_rate": 0.00016633378856797503,
+      "loss": 1.0835,
+      "step": 1513
+    },
+    {
+      "epoch": 0.29126587148903427,
+      "grad_norm": 1.451711893081665,
+      "learning_rate": 0.00016628714902505956,
+      "loss": 1.0872,
+      "step": 1514
+    },
+    {
+      "epoch": 0.2914582531742978,
+      "grad_norm": 2.301386594772339,
+      "learning_rate": 0.00016624048374719053,
+      "loss": 1.0642,
+      "step": 1515
+    },
+    {
+      "epoch": 0.2916506348595614,
+      "grad_norm": 2.180128812789917,
+      "learning_rate": 0.000166193792752485,
+      "loss": 1.071,
+      "step": 1516
+    },
+    {
+      "epoch": 0.2918430165448249,
+      "grad_norm": 2.2703309059143066,
+      "learning_rate": 0.00016614707605906996,
+      "loss": 1.0309,
+      "step": 1517
+    },
+    {
+      "epoch": 0.2920353982300885,
+      "grad_norm": 2.2813920974731445,
+      "learning_rate": 0.0001661003336850825,
+      "loss": 1.0514,
+      "step": 1518
+    },
+    {
+      "epoch": 0.29222777991535204,
+      "grad_norm": 1.1279929876327515,
+      "learning_rate": 0.00016605356564866958,
+      "loss": 1.057,
+      "step": 1519
+    },
+    {
+      "epoch": 0.2924201616006156,
+      "grad_norm": 1.2517904043197632,
+      "learning_rate": 0.00016600677196798821,
+      "loss": 1.0369,
+      "step": 1520
+    },
+    {
+      "epoch": 0.2926125432858792,
+      "grad_norm": 1.6124945878982544,
+      "learning_rate": 0.00016595995266120527,
+      "loss": 1.1408,
+      "step": 1521
+    },
+    {
+      "epoch": 0.29280492497114274,
+      "grad_norm": 2.2771055698394775,
+      "learning_rate": 0.00016591310774649766,
+      "loss": 1.0295,
+      "step": 1522
+    },
+    {
+      "epoch": 0.29299730665640633,
+      "grad_norm": 1.5582828521728516,
+      "learning_rate": 0.00016586623724205215,
+      "loss": 1.009,
+      "step": 1523
+    },
+    {
+      "epoch": 0.29318968834166986,
+      "grad_norm": 2.793675661087036,
+      "learning_rate": 0.00016581934116606552,
+      "loss": 1.0599,
+      "step": 1524
+    },
+    {
+      "epoch": 0.29338207002693345,
+      "grad_norm": 2.5479235649108887,
+      "learning_rate": 0.00016577241953674439,
+      "loss": 1.0382,
+      "step": 1525
+    },
+    {
+      "epoch": 0.293574451712197,
+      "grad_norm": 1.771727442741394,
+      "learning_rate": 0.0001657254723723054,
+      "loss": 0.9881,
+      "step": 1526
+    },
+    {
+      "epoch": 0.29376683339746057,
+      "grad_norm": 2.7877769470214844,
+      "learning_rate": 0.00016567849969097503,
+      "loss": 1.0192,
+      "step": 1527
+    },
+    {
+      "epoch": 0.2939592150827241,
+      "grad_norm": 1.9102647304534912,
+      "learning_rate": 0.0001656315015109897,
+      "loss": 1.0266,
+      "step": 1528
+    },
+    {
+      "epoch": 0.2941515967679877,
+      "grad_norm": 1.6697086095809937,
+      "learning_rate": 0.00016558447785059577,
+      "loss": 0.7666,
+      "step": 1529
+    },
+    {
+      "epoch": 0.2943439784532513,
+      "grad_norm": 2.454801321029663,
+      "learning_rate": 0.00016553742872804937,
+      "loss": 1.0964,
+      "step": 1530
+    },
+    {
+      "epoch": 0.2945363601385148,
+      "grad_norm": 2.327587842941284,
+      "learning_rate": 0.00016549035416161664,
+      "loss": 1.0782,
+      "step": 1531
+    },
+    {
+      "epoch": 0.2947287418237784,
+      "grad_norm": 1.1591378450393677,
+      "learning_rate": 0.0001654432541695735,
+      "loss": 1.0227,
+      "step": 1532
+    },
+    {
+      "epoch": 0.2949211235090419,
+      "grad_norm": 2.341610908508301,
+      "learning_rate": 0.00016539612877020582,
+      "loss": 1.0302,
+      "step": 1533
+    },
+    {
+      "epoch": 0.2951135051943055,
+      "grad_norm": 1.5264968872070312,
+      "learning_rate": 0.00016534897798180928,
+      "loss": 0.9983,
+      "step": 1534
+    },
+    {
+      "epoch": 0.29530588687956905,
+      "grad_norm": 2.2020068168640137,
+      "learning_rate": 0.00016530180182268946,
+      "loss": 0.969,
+      "step": 1535
+    },
+    {
+      "epoch": 0.29549826856483263,
+      "grad_norm": 2.0577402114868164,
+      "learning_rate": 0.00016525460031116177,
+      "loss": 0.9965,
+      "step": 1536
+    },
+    {
+      "epoch": 0.29569065025009617,
+      "grad_norm": 2.160315752029419,
+      "learning_rate": 0.00016520737346555149,
+      "loss": 0.9884,
+      "step": 1537
+    },
+    {
+      "epoch": 0.29588303193535975,
+      "grad_norm": 2.3820762634277344,
+      "learning_rate": 0.00016516012130419366,
+      "loss": 1.0064,
+      "step": 1538
+    },
+    {
+      "epoch": 0.29607541362062334,
+      "grad_norm": 2.2774720191955566,
+      "learning_rate": 0.00016511284384543317,
+      "loss": 0.979,
+      "step": 1539
+    },
+    {
+      "epoch": 0.29626779530588687,
+      "grad_norm": 2.3224332332611084,
+      "learning_rate": 0.00016506554110762482,
+      "loss": 1.0755,
+      "step": 1540
+    },
+    {
+      "epoch": 0.29646017699115046,
+      "grad_norm": 2.2901713848114014,
+      "learning_rate": 0.00016501821310913316,
+      "loss": 1.0402,
+      "step": 1541
+    },
+    {
+      "epoch": 0.296652558676414,
+      "grad_norm": 2.404642343521118,
+      "learning_rate": 0.00016497085986833253,
+      "loss": 1.0305,
+      "step": 1542
+    },
+    {
+      "epoch": 0.2968449403616776,
+      "grad_norm": 2.64076566696167,
+      "learning_rate": 0.00016492348140360706,
+      "loss": 1.0323,
+      "step": 1543
+    },
+    {
+      "epoch": 0.2970373220469411,
+      "grad_norm": 2.2320408821105957,
+      "learning_rate": 0.00016487607773335073,
+      "loss": 1.008,
+      "step": 1544
+    },
+    {
+      "epoch": 0.2972297037322047,
+      "grad_norm": 2.216697931289673,
+      "learning_rate": 0.0001648286488759673,
+      "loss": 0.9709,
+      "step": 1545
+    },
+    {
+      "epoch": 0.29742208541746823,
+      "grad_norm": 1.422713279724121,
+      "learning_rate": 0.00016478119484987025,
+      "loss": 1.0325,
+      "step": 1546
+    },
+    {
+      "epoch": 0.2976144671027318,
+      "grad_norm": 2.1282479763031006,
+      "learning_rate": 0.00016473371567348285,
+      "loss": 1.0248,
+      "step": 1547
+    },
+    {
+      "epoch": 0.2978068487879954,
+      "grad_norm": 1.4448484182357788,
+      "learning_rate": 0.00016468621136523823,
+      "loss": 0.7585,
+      "step": 1548
+    },
+    {
+      "epoch": 0.29799923047325894,
+      "grad_norm": 1.8351774215698242,
+      "learning_rate": 0.0001646386819435791,
+      "loss": 0.9991,
+      "step": 1549
+    },
+    {
+      "epoch": 0.2981916121585225,
+      "grad_norm": 2.227581024169922,
+      "learning_rate": 0.00016459112742695807,
+      "loss": 1.0162,
+      "step": 1550
+    },
+    {
+      "epoch": 0.29838399384378606,
+      "grad_norm": 2.0184073448181152,
+      "learning_rate": 0.00016454354783383747,
+      "loss": 0.954,
+      "step": 1551
+    },
+    {
+      "epoch": 0.29857637552904964,
+      "grad_norm": 2.473090648651123,
+      "learning_rate": 0.00016449594318268927,
+      "loss": 1.0829,
+      "step": 1552
+    },
+    {
+      "epoch": 0.2987687572143132,
+      "grad_norm": 2.045361280441284,
+      "learning_rate": 0.00016444831349199528,
+      "loss": 1.0115,
+      "step": 1553
+    },
+    {
+      "epoch": 0.29896113889957676,
+      "grad_norm": 1.2806012630462646,
+      "learning_rate": 0.00016440065878024697,
+      "loss": 1.0107,
+      "step": 1554
+    },
+    {
+      "epoch": 0.29915352058484035,
+      "grad_norm": 2.340334177017212,
+      "learning_rate": 0.00016435297906594555,
+      "loss": 1.0863,
+      "step": 1555
+    },
+    {
+      "epoch": 0.2993459022701039,
+      "grad_norm": 2.5942699909210205,
+      "learning_rate": 0.00016430527436760191,
+      "loss": 0.9393,
+      "step": 1556
+    },
+    {
+      "epoch": 0.29953828395536747,
+      "grad_norm": 2.058542013168335,
+      "learning_rate": 0.00016425754470373666,
+      "loss": 1.0269,
+      "step": 1557
+    },
+    {
+      "epoch": 0.299730665640631,
+      "grad_norm": 1.0915286540985107,
+      "learning_rate": 0.0001642097900928801,
+      "loss": 0.8258,
+      "step": 1558
+    },
+    {
+      "epoch": 0.2999230473258946,
+      "grad_norm": 1.7852873802185059,
+      "learning_rate": 0.00016416201055357224,
+      "loss": 0.9927,
+      "step": 1559
+    },
+    {
+      "epoch": 0.3001154290111581,
+      "grad_norm": 2.14880633354187,
+      "learning_rate": 0.0001641142061043627,
+      "loss": 0.9415,
+      "step": 1560
+    },
+    {
+      "epoch": 0.3003078106964217,
+      "grad_norm": 1.3924521207809448,
+      "learning_rate": 0.00016406637676381083,
+      "loss": 1.1133,
+      "step": 1561
+    },
+    {
+      "epoch": 0.30050019238168524,
+      "grad_norm": 1.8726837635040283,
+      "learning_rate": 0.00016401852255048564,
+      "loss": 1.031,
+      "step": 1562
+    },
+    {
+      "epoch": 0.3006925740669488,
+      "grad_norm": 2.436598777770996,
+      "learning_rate": 0.00016397064348296577,
+      "loss": 1.0049,
+      "step": 1563
+    },
+    {
+      "epoch": 0.3008849557522124,
+      "grad_norm": 1.4472073316574097,
+      "learning_rate": 0.00016392273957983956,
+      "loss": 0.8102,
+      "step": 1564
+    },
+    {
+      "epoch": 0.30107733743747594,
+      "grad_norm": 2.9171319007873535,
+      "learning_rate": 0.0001638748108597049,
+      "loss": 1.065,
+      "step": 1565
+    },
+    {
+      "epoch": 0.30126971912273953,
+      "grad_norm": 2.2946295738220215,
+      "learning_rate": 0.00016382685734116933,
+      "loss": 1.0434,
+      "step": 1566
+    },
+    {
+      "epoch": 0.30146210080800306,
+      "grad_norm": 2.304884433746338,
+      "learning_rate": 0.00016377887904285018,
+      "loss": 1.0301,
+      "step": 1567
+    },
+    {
+      "epoch": 0.30165448249326665,
+      "grad_norm": 1.8151887655258179,
+      "learning_rate": 0.00016373087598337422,
+      "loss": 0.9792,
+      "step": 1568
+    },
+    {
+      "epoch": 0.3018468641785302,
+      "grad_norm": 2.161468029022217,
+      "learning_rate": 0.00016368284818137787,
+      "loss": 1.0742,
+      "step": 1569
+    },
+    {
+      "epoch": 0.30203924586379377,
+      "grad_norm": 1.1271188259124756,
+      "learning_rate": 0.00016363479565550718,
+      "loss": 1.0582,
+      "step": 1570
+    },
+    {
+      "epoch": 0.3022316275490573,
+      "grad_norm": 2.6240782737731934,
+      "learning_rate": 0.0001635867184244178,
+      "loss": 1.017,
+      "step": 1571
+    },
+    {
+      "epoch": 0.3024240092343209,
+      "grad_norm": 1.6684452295303345,
+      "learning_rate": 0.000163538616506775,
+      "loss": 0.9931,
+      "step": 1572
+    },
+    {
+      "epoch": 0.3026163909195845,
+      "grad_norm": 2.4656317234039307,
+      "learning_rate": 0.00016349048992125357,
+      "loss": 1.0834,
+      "step": 1573
+    },
+    {
+      "epoch": 0.302808772604848,
+      "grad_norm": 2.787212610244751,
+      "learning_rate": 0.00016344233868653788,
+      "loss": 1.0263,
+      "step": 1574
+    },
+    {
+      "epoch": 0.3030011542901116,
+      "grad_norm": 1.3134548664093018,
+      "learning_rate": 0.00016339416282132196,
+      "loss": 1.0524,
+      "step": 1575
+    },
+    {
+      "epoch": 0.30319353597537513,
+      "grad_norm": 1.3846782445907593,
+      "learning_rate": 0.0001633459623443093,
+      "loss": 1.1013,
+      "step": 1576
+    },
+    {
+      "epoch": 0.3033859176606387,
+      "grad_norm": 2.356412649154663,
+      "learning_rate": 0.00016329773727421297,
+      "loss": 1.0489,
+      "step": 1577
+    },
+    {
+      "epoch": 0.30357829934590225,
+      "grad_norm": 1.9501971006393433,
+      "learning_rate": 0.00016324948762975566,
+      "loss": 1.0233,
+      "step": 1578
+    },
+    {
+      "epoch": 0.30377068103116583,
+      "grad_norm": 1.2132933139801025,
+      "learning_rate": 0.0001632012134296695,
+      "loss": 1.0201,
+      "step": 1579
+    },
+    {
+      "epoch": 0.3039630627164294,
+      "grad_norm": 2.3302407264709473,
+      "learning_rate": 0.0001631529146926962,
+      "loss": 1.0026,
+      "step": 1580
+    },
+    {
+      "epoch": 0.30415544440169295,
+      "grad_norm": 1.1951031684875488,
+      "learning_rate": 0.000163104591437587,
+      "loss": 0.7813,
+      "step": 1581
+    },
+    {
+      "epoch": 0.30434782608695654,
+      "grad_norm": 1.145522952079773,
+      "learning_rate": 0.00016305624368310263,
+      "loss": 0.7605,
+      "step": 1582
+    },
+    {
+      "epoch": 0.3045402077722201,
+      "grad_norm": 2.1174237728118896,
+      "learning_rate": 0.0001630078714480134,
+      "loss": 1.0325,
+      "step": 1583
+    },
+    {
+      "epoch": 0.30473258945748366,
+      "grad_norm": 2.199448347091675,
+      "learning_rate": 0.00016295947475109904,
+      "loss": 1.0298,
+      "step": 1584
+    },
+    {
+      "epoch": 0.3049249711427472,
+      "grad_norm": 2.2502939701080322,
+      "learning_rate": 0.0001629110536111488,
+      "loss": 1.0295,
+      "step": 1585
+    },
+    {
+      "epoch": 0.3051173528280108,
+      "grad_norm": 2.6069228649139404,
+      "learning_rate": 0.0001628626080469615,
+      "loss": 1.042,
+      "step": 1586
+    },
+    {
+      "epoch": 0.3053097345132743,
+      "grad_norm": 2.257227659225464,
+      "learning_rate": 0.0001628141380773453,
+      "loss": 0.9772,
+      "step": 1587
+    },
+    {
+      "epoch": 0.3055021161985379,
+      "grad_norm": 2.842010259628296,
+      "learning_rate": 0.00016276564372111796,
+      "loss": 1.0985,
+      "step": 1588
+    },
+    {
+      "epoch": 0.3056944978838015,
+      "grad_norm": 1.7872533798217773,
+      "learning_rate": 0.00016271712499710663,
+      "loss": 0.9829,
+      "step": 1589
+    },
+    {
+      "epoch": 0.305886879569065,
+      "grad_norm": 2.3818016052246094,
+      "learning_rate": 0.00016266858192414798,
+      "loss": 1.0744,
+      "step": 1590
+    },
+    {
+      "epoch": 0.3060792612543286,
+      "grad_norm": 1.285504698753357,
+      "learning_rate": 0.00016262001452108807,
+      "loss": 0.7599,
+      "step": 1591
+    },
+    {
+      "epoch": 0.30627164293959214,
+      "grad_norm": 1.254860281944275,
+      "learning_rate": 0.00016257142280678245,
+      "loss": 1.0483,
+      "step": 1592
+    },
+    {
+      "epoch": 0.3064640246248557,
+      "grad_norm": 1.3496578931808472,
+      "learning_rate": 0.00016252280680009612,
+      "loss": 1.0764,
+      "step": 1593
+    },
+    {
+      "epoch": 0.30665640631011926,
+      "grad_norm": 2.1964468955993652,
+      "learning_rate": 0.00016247416651990343,
+      "loss": 1.0022,
+      "step": 1594
+    },
+    {
+      "epoch": 0.30684878799538284,
+      "grad_norm": 1.929050087928772,
+      "learning_rate": 0.0001624255019850883,
+      "loss": 0.9607,
+      "step": 1595
+    },
+    {
+      "epoch": 0.3070411696806464,
+      "grad_norm": 2.450752019882202,
+      "learning_rate": 0.00016237681321454387,
+      "loss": 1.0764,
+      "step": 1596
+    },
+    {
+      "epoch": 0.30723355136590996,
+      "grad_norm": 1.972947359085083,
+      "learning_rate": 0.0001623281002271729,
+      "loss": 1.0657,
+      "step": 1597
+    },
+    {
+      "epoch": 0.30742593305117355,
+      "grad_norm": 1.4621106386184692,
+      "learning_rate": 0.0001622793630418874,
+      "loss": 0.9653,
+      "step": 1598
+    },
+    {
+      "epoch": 0.3076183147364371,
+      "grad_norm": 2.130143404006958,
+      "learning_rate": 0.0001622306016776088,
+      "loss": 0.9806,
+      "step": 1599
+    },
+    {
+      "epoch": 0.30781069642170067,
+      "grad_norm": 2.399383306503296,
+      "learning_rate": 0.00016218181615326795,
+      "loss": 1.0578,
+      "step": 1600
+    },
+    {
+      "epoch": 0.3080030781069642,
+      "grad_norm": 2.358139753341675,
+      "learning_rate": 0.00016213300648780513,
+      "loss": 1.0843,
+      "step": 1601
+    },
+    {
+      "epoch": 0.3081954597922278,
+      "grad_norm": 2.317319631576538,
+      "learning_rate": 0.0001620841727001699,
+      "loss": 0.9549,
+      "step": 1602
+    },
+    {
+      "epoch": 0.3083878414774913,
+      "grad_norm": 3.151749610900879,
+      "learning_rate": 0.00016203531480932115,
+      "loss": 0.9926,
+      "step": 1603
+    },
+    {
+      "epoch": 0.3085802231627549,
+      "grad_norm": 2.4561123847961426,
+      "learning_rate": 0.00016198643283422729,
+      "loss": 1.0304,
+      "step": 1604
+    },
+    {
+      "epoch": 0.3087726048480185,
+      "grad_norm": 2.7317025661468506,
+      "learning_rate": 0.00016193752679386593,
+      "loss": 1.0211,
+      "step": 1605
+    },
+    {
+      "epoch": 0.308964986533282,
+      "grad_norm": 2.4372620582580566,
+      "learning_rate": 0.0001618885967072241,
+      "loss": 0.9755,
+      "step": 1606
+    },
+    {
+      "epoch": 0.3091573682185456,
+      "grad_norm": 2.265374183654785,
+      "learning_rate": 0.00016183964259329815,
+      "loss": 1.0224,
+      "step": 1607
+    },
+    {
+      "epoch": 0.30934974990380915,
+      "grad_norm": 2.185959815979004,
+      "learning_rate": 0.00016179066447109377,
+      "loss": 0.997,
+      "step": 1608
+    },
+    {
+      "epoch": 0.30954213158907273,
+      "grad_norm": 2.664165496826172,
+      "learning_rate": 0.00016174166235962586,
+      "loss": 1.0236,
+      "step": 1609
+    },
+    {
+      "epoch": 0.30973451327433627,
+      "grad_norm": 2.341651201248169,
+      "learning_rate": 0.00016169263627791887,
+      "loss": 1.0706,
+      "step": 1610
+    },
+    {
+      "epoch": 0.30992689495959985,
+      "grad_norm": 2.7954154014587402,
+      "learning_rate": 0.0001616435862450063,
+      "loss": 1.0194,
+      "step": 1611
+    },
+    {
+      "epoch": 0.3101192766448634,
+      "grad_norm": 1.5539939403533936,
+      "learning_rate": 0.0001615945122799311,
+      "loss": 0.8054,
+      "step": 1612
+    },
+    {
+      "epoch": 0.31031165833012697,
+      "grad_norm": 1.202900767326355,
+      "learning_rate": 0.00016154541440174548,
+      "loss": 0.7813,
+      "step": 1613
+    },
+    {
+      "epoch": 0.31050404001539056,
+      "grad_norm": 2.1343038082122803,
+      "learning_rate": 0.0001614962926295109,
+      "loss": 0.9797,
+      "step": 1614
+    },
+    {
+      "epoch": 0.3106964217006541,
+      "grad_norm": 2.3647758960723877,
+      "learning_rate": 0.00016144714698229816,
+      "loss": 0.9455,
+      "step": 1615
+    },
+    {
+      "epoch": 0.3108888033859177,
+      "grad_norm": 2.4138777256011963,
+      "learning_rate": 0.00016139797747918725,
+      "loss": 1.0604,
+      "step": 1616
+    },
+    {
+      "epoch": 0.3110811850711812,
+      "grad_norm": 1.2279789447784424,
+      "learning_rate": 0.0001613487841392675,
+      "loss": 1.0329,
+      "step": 1617
+    },
+    {
+      "epoch": 0.3112735667564448,
+      "grad_norm": 2.1896681785583496,
+      "learning_rate": 0.00016129956698163745,
+      "loss": 1.0172,
+      "step": 1618
+    },
+    {
+      "epoch": 0.31146594844170833,
+      "grad_norm": 2.2433993816375732,
+      "learning_rate": 0.0001612503260254049,
+      "loss": 1.0752,
+      "step": 1619
+    },
+    {
+      "epoch": 0.3116583301269719,
+      "grad_norm": 2.160809278488159,
+      "learning_rate": 0.00016120106128968687,
+      "loss": 1.0936,
+      "step": 1620
+    },
+    {
+      "epoch": 0.31185071181223545,
+      "grad_norm": 2.119636058807373,
+      "learning_rate": 0.00016115177279360965,
+      "loss": 1.047,
+      "step": 1621
+    },
+    {
+      "epoch": 0.31204309349749904,
+      "grad_norm": 2.3325703144073486,
+      "learning_rate": 0.00016110246055630868,
+      "loss": 0.9669,
+      "step": 1622
+    },
+    {
+      "epoch": 0.3122354751827626,
+      "grad_norm": 2.290663480758667,
+      "learning_rate": 0.00016105312459692873,
+      "loss": 1.0433,
+      "step": 1623
+    },
+    {
+      "epoch": 0.31242785686802615,
+      "grad_norm": 2.285426139831543,
+      "learning_rate": 0.00016100376493462368,
+      "loss": 1.0904,
+      "step": 1624
+    },
+    {
+      "epoch": 0.31262023855328974,
+      "grad_norm": 2.012789249420166,
+      "learning_rate": 0.00016095438158855667,
+      "loss": 0.9774,
+      "step": 1625
+    },
+    {
+      "epoch": 0.3128126202385533,
+      "grad_norm": 1.3862818479537964,
+      "learning_rate": 0.00016090497457789998,
+      "loss": 0.7387,
+      "step": 1626
+    },
+    {
+      "epoch": 0.31300500192381686,
+      "grad_norm": 2.1518754959106445,
+      "learning_rate": 0.00016085554392183517,
+      "loss": 1.0395,
+      "step": 1627
+    },
+    {
+      "epoch": 0.3131973836090804,
+      "grad_norm": 2.4144937992095947,
+      "learning_rate": 0.00016080608963955288,
+      "loss": 0.996,
+      "step": 1628
+    },
+    {
+      "epoch": 0.313389765294344,
+      "grad_norm": 2.102864980697632,
+      "learning_rate": 0.00016075661175025298,
+      "loss": 1.0599,
+      "step": 1629
+    },
+    {
+      "epoch": 0.31358214697960757,
+      "grad_norm": 2.3571877479553223,
+      "learning_rate": 0.00016070711027314446,
+      "loss": 0.9746,
+      "step": 1630
+    },
+    {
+      "epoch": 0.3137745286648711,
+      "grad_norm": 1.5089541673660278,
+      "learning_rate": 0.00016065758522744558,
+      "loss": 1.145,
+      "step": 1631
+    },
+    {
+      "epoch": 0.3139669103501347,
+      "grad_norm": 1.2736268043518066,
+      "learning_rate": 0.0001606080366323836,
+      "loss": 0.805,
+      "step": 1632
+    },
+    {
+      "epoch": 0.3141592920353982,
+      "grad_norm": 2.290951728820801,
+      "learning_rate": 0.00016055846450719496,
+      "loss": 1.052,
+      "step": 1633
+    },
+    {
+      "epoch": 0.3143516737206618,
+      "grad_norm": 1.6116780042648315,
+      "learning_rate": 0.00016050886887112532,
+      "loss": 1.0549,
+      "step": 1634
+    },
+    {
+      "epoch": 0.31454405540592534,
+      "grad_norm": 2.03620982170105,
+      "learning_rate": 0.00016045924974342942,
+      "loss": 1.0449,
+      "step": 1635
+    },
+    {
+      "epoch": 0.3147364370911889,
+      "grad_norm": 2.3006699085235596,
+      "learning_rate": 0.0001604096071433711,
+      "loss": 0.9562,
+      "step": 1636
+    },
+    {
+      "epoch": 0.31492881877645246,
+      "grad_norm": 0.7473447918891907,
+      "learning_rate": 0.00016035994109022333,
+      "loss": 1.0434,
+      "step": 1637
+    },
+    {
+      "epoch": 0.31512120046171604,
+      "grad_norm": 2.235405206680298,
+      "learning_rate": 0.00016031025160326813,
+      "loss": 1.0258,
+      "step": 1638
+    },
+    {
+      "epoch": 0.31531358214697963,
+      "grad_norm": 2.409555435180664,
+      "learning_rate": 0.00016026053870179675,
+      "loss": 0.9565,
+      "step": 1639
+    },
+    {
+      "epoch": 0.31550596383224316,
+      "grad_norm": 2.653520345687866,
+      "learning_rate": 0.00016021080240510944,
+      "loss": 0.9978,
+      "step": 1640
+    },
+    {
+      "epoch": 0.31569834551750675,
+      "grad_norm": 2.6454713344573975,
+      "learning_rate": 0.0001601610427325155,
+      "loss": 1.0337,
+      "step": 1641
+    },
+    {
+      "epoch": 0.3158907272027703,
+      "grad_norm": 1.4742250442504883,
+      "learning_rate": 0.00016011125970333333,
+      "loss": 1.0632,
+      "step": 1642
+    },
+    {
+      "epoch": 0.31608310888803387,
+      "grad_norm": 2.190485954284668,
+      "learning_rate": 0.00016006145333689049,
+      "loss": 1.039,
+      "step": 1643
+    },
+    {
+      "epoch": 0.3162754905732974,
+      "grad_norm": 2.3585851192474365,
+      "learning_rate": 0.00016001162365252347,
+      "loss": 0.9948,
+      "step": 1644
+    },
+    {
+      "epoch": 0.316467872258561,
+      "grad_norm": 2.7076079845428467,
+      "learning_rate": 0.00015996177066957787,
+      "loss": 1.0275,
+      "step": 1645
+    },
+    {
+      "epoch": 0.3166602539438245,
+      "grad_norm": 2.700495958328247,
+      "learning_rate": 0.0001599118944074084,
+      "loss": 0.9822,
+      "step": 1646
+    },
+    {
+      "epoch": 0.3168526356290881,
+      "grad_norm": 2.2149248123168945,
+      "learning_rate": 0.00015986199488537866,
+      "loss": 1.0831,
+      "step": 1647
+    },
+    {
+      "epoch": 0.3170450173143517,
+      "grad_norm": 2.098933696746826,
+      "learning_rate": 0.0001598120721228614,
+      "loss": 1.0453,
+      "step": 1648
+    },
+    {
+      "epoch": 0.3172373989996152,
+      "grad_norm": 1.9501348733901978,
+      "learning_rate": 0.00015976212613923836,
+      "loss": 1.0367,
+      "step": 1649
+    },
+    {
+      "epoch": 0.3174297806848788,
+      "grad_norm": 2.8978190422058105,
+      "learning_rate": 0.00015971215695390027,
+      "loss": 1.0376,
+      "step": 1650
+    },
+    {
+      "epoch": 0.31762216237014235,
+      "grad_norm": 2.3352127075195312,
+      "learning_rate": 0.00015966216458624692,
+      "loss": 1.0107,
+      "step": 1651
+    },
+    {
+      "epoch": 0.31781454405540593,
+      "grad_norm": 2.484337568283081,
+      "learning_rate": 0.00015961214905568704,
+      "loss": 1.021,
+      "step": 1652
+    },
+    {
+      "epoch": 0.31800692574066947,
+      "grad_norm": 2.3851635456085205,
+      "learning_rate": 0.0001595621103816384,
+      "loss": 1.0811,
+      "step": 1653
+    },
+    {
+      "epoch": 0.31819930742593305,
+      "grad_norm": 2.043915033340454,
+      "learning_rate": 0.00015951204858352772,
+      "loss": 1.0521,
+      "step": 1654
+    },
+    {
+      "epoch": 0.31839168911119664,
+      "grad_norm": 1.9596813917160034,
+      "learning_rate": 0.0001594619636807907,
+      "loss": 0.979,
+      "step": 1655
+    },
+    {
+      "epoch": 0.3185840707964602,
+      "grad_norm": 2.3240394592285156,
+      "learning_rate": 0.00015941185569287204,
+      "loss": 1.0531,
+      "step": 1656
+    },
+    {
+      "epoch": 0.31877645248172376,
+      "grad_norm": 2.487884044647217,
+      "learning_rate": 0.00015936172463922542,
+      "loss": 0.9958,
+      "step": 1657
+    },
+    {
+      "epoch": 0.3189688341669873,
+      "grad_norm": 2.625814437866211,
+      "learning_rate": 0.0001593115705393134,
+      "loss": 0.9861,
+      "step": 1658
+    },
+    {
+      "epoch": 0.3191612158522509,
+      "grad_norm": 2.044440507888794,
+      "learning_rate": 0.00015926139341260756,
+      "loss": 1.0651,
+      "step": 1659
+    },
+    {
+      "epoch": 0.3193535975375144,
+      "grad_norm": 1.391687035560608,
+      "learning_rate": 0.00015921119327858834,
+      "loss": 1.0745,
+      "step": 1660
+    },
+    {
+      "epoch": 0.319545979222778,
+      "grad_norm": 2.091712474822998,
+      "learning_rate": 0.00015916097015674518,
+      "loss": 1.0517,
+      "step": 1661
+    },
+    {
+      "epoch": 0.31973836090804153,
+      "grad_norm": 2.143721580505371,
+      "learning_rate": 0.00015911072406657646,
+      "loss": 1.0352,
+      "step": 1662
+    },
+    {
+      "epoch": 0.3199307425933051,
+      "grad_norm": 2.495041608810425,
+      "learning_rate": 0.0001590604550275894,
+      "loss": 1.0111,
+      "step": 1663
+    },
+    {
+      "epoch": 0.3201231242785687,
+      "grad_norm": 2.3186514377593994,
+      "learning_rate": 0.0001590101630593002,
+      "loss": 0.9663,
+      "step": 1664
+    },
+    {
+      "epoch": 0.32031550596383224,
+      "grad_norm": 2.728496789932251,
+      "learning_rate": 0.0001589598481812339,
+      "loss": 1.1356,
+      "step": 1665
+    },
+    {
+      "epoch": 0.3205078876490958,
+      "grad_norm": 2.502781867980957,
+      "learning_rate": 0.0001589095104129245,
+      "loss": 1.0196,
+      "step": 1666
+    },
+    {
+      "epoch": 0.32070026933435936,
+      "grad_norm": 1.585044503211975,
+      "learning_rate": 0.00015885914977391486,
+      "loss": 1.0134,
+      "step": 1667
+    },
+    {
+      "epoch": 0.32089265101962294,
+      "grad_norm": 2.37007999420166,
+      "learning_rate": 0.00015880876628375666,
+      "loss": 0.9852,
+      "step": 1668
+    },
+    {
+      "epoch": 0.3210850327048865,
+      "grad_norm": 2.42063570022583,
+      "learning_rate": 0.00015875835996201058,
+      "loss": 0.9886,
+      "step": 1669
+    },
+    {
+      "epoch": 0.32127741439015006,
+      "grad_norm": 2.2740089893341064,
+      "learning_rate": 0.00015870793082824604,
+      "loss": 0.9512,
+      "step": 1670
+    },
+    {
+      "epoch": 0.3214697960754136,
+      "grad_norm": 1.3068575859069824,
+      "learning_rate": 0.00015865747890204138,
+      "loss": 1.0126,
+      "step": 1671
+    },
+    {
+      "epoch": 0.3216621777606772,
+      "grad_norm": 1.255905270576477,
+      "learning_rate": 0.00015860700420298378,
+      "loss": 0.79,
+      "step": 1672
+    },
+    {
+      "epoch": 0.32185455944594077,
+      "grad_norm": 2.245961904525757,
+      "learning_rate": 0.00015855650675066925,
+      "loss": 1.0553,
+      "step": 1673
+    },
+    {
+      "epoch": 0.3220469411312043,
+      "grad_norm": 1.9613453149795532,
+      "learning_rate": 0.00015850598656470263,
+      "loss": 1.0018,
+      "step": 1674
+    },
+    {
+      "epoch": 0.3222393228164679,
+      "grad_norm": 2.8318240642547607,
+      "learning_rate": 0.0001584554436646976,
+      "loss": 0.9875,
+      "step": 1675
+    },
+    {
+      "epoch": 0.3224317045017314,
+      "grad_norm": 1.54958975315094,
+      "learning_rate": 0.00015840487807027665,
+      "loss": 1.0488,
+      "step": 1676
+    },
+    {
+      "epoch": 0.322624086186995,
+      "grad_norm": 2.291625738143921,
+      "learning_rate": 0.00015835428980107112,
+      "loss": 1.0264,
+      "step": 1677
+    },
+    {
+      "epoch": 0.32281646787225854,
+      "grad_norm": 2.6585559844970703,
+      "learning_rate": 0.0001583036788767211,
+      "loss": 1.0444,
+      "step": 1678
+    },
+    {
+      "epoch": 0.3230088495575221,
+      "grad_norm": 2.1201868057250977,
+      "learning_rate": 0.0001582530453168755,
+      "loss": 1.0191,
+      "step": 1679
+    },
+    {
+      "epoch": 0.3232012312427857,
+      "grad_norm": 2.3504483699798584,
+      "learning_rate": 0.00015820238914119195,
+      "loss": 1.0009,
+      "step": 1680
+    },
+    {
+      "epoch": 0.32339361292804925,
+      "grad_norm": 2.2873616218566895,
+      "learning_rate": 0.000158151710369337,
+      "loss": 0.9149,
+      "step": 1681
+    },
+    {
+      "epoch": 0.32358599461331283,
+      "grad_norm": 1.0704349279403687,
+      "learning_rate": 0.00015810100902098582,
+      "loss": 1.0778,
+      "step": 1682
+    },
+    {
+      "epoch": 0.32377837629857636,
+      "grad_norm": 2.6844570636749268,
+      "learning_rate": 0.0001580502851158225,
+      "loss": 1.0526,
+      "step": 1683
+    },
+    {
+      "epoch": 0.32397075798383995,
+      "grad_norm": 1.9724894762039185,
+      "learning_rate": 0.00015799953867353975,
+      "loss": 1.0434,
+      "step": 1684
+    },
+    {
+      "epoch": 0.3241631396691035,
+      "grad_norm": 2.1440861225128174,
+      "learning_rate": 0.0001579487697138391,
+      "loss": 1.0477,
+      "step": 1685
+    },
+    {
+      "epoch": 0.32435552135436707,
+      "grad_norm": 2.490935802459717,
+      "learning_rate": 0.00015789797825643085,
+      "loss": 0.9889,
+      "step": 1686
+    },
+    {
+      "epoch": 0.3245479030396306,
+      "grad_norm": 1.7961583137512207,
+      "learning_rate": 0.00015784716432103393,
+      "loss": 1.0637,
+      "step": 1687
+    },
+    {
+      "epoch": 0.3247402847248942,
+      "grad_norm": 2.4223580360412598,
+      "learning_rate": 0.00015779632792737607,
+      "loss": 1.1111,
+      "step": 1688
+    },
+    {
+      "epoch": 0.3249326664101578,
+      "grad_norm": 1.944258213043213,
+      "learning_rate": 0.00015774546909519375,
+      "loss": 1.0383,
+      "step": 1689
+    },
+    {
+      "epoch": 0.3251250480954213,
+      "grad_norm": 1.7717947959899902,
+      "learning_rate": 0.00015769458784423206,
+      "loss": 1.0064,
+      "step": 1690
+    },
+    {
+      "epoch": 0.3253174297806849,
+      "grad_norm": 2.8095576763153076,
+      "learning_rate": 0.0001576436841942449,
+      "loss": 1.0337,
+      "step": 1691
+    },
+    {
+      "epoch": 0.32550981146594843,
+      "grad_norm": 2.0333824157714844,
+      "learning_rate": 0.0001575927581649948,
+      "loss": 1.054,
+      "step": 1692
+    },
+    {
+      "epoch": 0.325702193151212,
+      "grad_norm": 1.9423625469207764,
+      "learning_rate": 0.00015754180977625303,
+      "loss": 1.0796,
+      "step": 1693
+    },
+    {
+      "epoch": 0.32589457483647555,
+      "grad_norm": 2.402329206466675,
+      "learning_rate": 0.00015749083904779946,
+      "loss": 0.9939,
+      "step": 1694
+    },
+    {
+      "epoch": 0.32608695652173914,
+      "grad_norm": 1.1590107679367065,
+      "learning_rate": 0.00015743984599942272,
+      "loss": 1.0523,
+      "step": 1695
+    },
+    {
+      "epoch": 0.32627933820700267,
+      "grad_norm": 2.23703932762146,
+      "learning_rate": 0.00015738883065092004,
+      "loss": 1.0173,
+      "step": 1696
+    },
+    {
+      "epoch": 0.32647171989226625,
+      "grad_norm": 2.187486171722412,
+      "learning_rate": 0.00015733779302209736,
+      "loss": 0.9746,
+      "step": 1697
+    },
+    {
+      "epoch": 0.32666410157752984,
+      "grad_norm": 2.330338954925537,
+      "learning_rate": 0.0001572867331327692,
+      "loss": 0.9949,
+      "step": 1698
+    },
+    {
+      "epoch": 0.3268564832627934,
+      "grad_norm": 2.2853901386260986,
+      "learning_rate": 0.00015723565100275884,
+      "loss": 1.0496,
+      "step": 1699
+    },
+    {
+      "epoch": 0.32704886494805696,
+      "grad_norm": 2.120884418487549,
+      "learning_rate": 0.00015718454665189806,
+      "loss": 1.0532,
+      "step": 1700
+    },
+    {
+      "epoch": 0.3272412466333205,
+      "grad_norm": 1.993755578994751,
+      "learning_rate": 0.00015713342010002733,
+      "loss": 0.9933,
+      "step": 1701
+    },
+    {
+      "epoch": 0.3274336283185841,
+      "grad_norm": 2.2349724769592285,
+      "learning_rate": 0.00015708227136699575,
+      "loss": 0.9921,
+      "step": 1702
+    },
+    {
+      "epoch": 0.3276260100038476,
+      "grad_norm": 2.2307631969451904,
+      "learning_rate": 0.00015703110047266106,
+      "loss": 0.9703,
+      "step": 1703
+    },
+    {
+      "epoch": 0.3278183916891112,
+      "grad_norm": 2.3598129749298096,
+      "learning_rate": 0.0001569799074368895,
+      "loss": 1.0114,
+      "step": 1704
+    },
+    {
+      "epoch": 0.3280107733743748,
+      "grad_norm": 0.8789365887641907,
+      "learning_rate": 0.00015692869227955603,
+      "loss": 1.001,
+      "step": 1705
+    },
+    {
+      "epoch": 0.3282031550596383,
+      "grad_norm": 2.5519509315490723,
+      "learning_rate": 0.00015687745502054407,
+      "loss": 0.9318,
+      "step": 1706
+    },
+    {
+      "epoch": 0.3283955367449019,
+      "grad_norm": 3.712980031967163,
+      "learning_rate": 0.00015682619567974576,
+      "loss": 1.0653,
+      "step": 1707
+    },
+    {
+      "epoch": 0.32858791843016544,
+      "grad_norm": 2.8539199829101562,
+      "learning_rate": 0.00015677491427706168,
+      "loss": 1.0961,
+      "step": 1708
+    },
+    {
+      "epoch": 0.328780300115429,
+      "grad_norm": 2.933744192123413,
+      "learning_rate": 0.00015672361083240107,
+      "loss": 0.9969,
+      "step": 1709
+    },
+    {
+      "epoch": 0.32897268180069256,
+      "grad_norm": 1.8377469778060913,
+      "learning_rate": 0.00015667228536568167,
+      "loss": 1.0534,
+      "step": 1710
+    },
+    {
+      "epoch": 0.32916506348595614,
+      "grad_norm": 1.3225889205932617,
+      "learning_rate": 0.00015662093789682978,
+      "loss": 0.7875,
+      "step": 1711
+    },
+    {
+      "epoch": 0.3293574451712197,
+      "grad_norm": 2.7497472763061523,
+      "learning_rate": 0.0001565695684457803,
+      "loss": 0.9635,
+      "step": 1712
+    },
+    {
+      "epoch": 0.32954982685648326,
+      "grad_norm": 2.4529714584350586,
+      "learning_rate": 0.00015651817703247667,
+      "loss": 1.0812,
+      "step": 1713
+    },
+    {
+      "epoch": 0.32974220854174685,
+      "grad_norm": 2.4029667377471924,
+      "learning_rate": 0.00015646676367687067,
+      "loss": 1.0414,
+      "step": 1714
+    },
+    {
+      "epoch": 0.3299345902270104,
+      "grad_norm": 2.078246593475342,
+      "learning_rate": 0.0001564153283989228,
+      "loss": 0.9663,
+      "step": 1715
+    },
+    {
+      "epoch": 0.33012697191227397,
+      "grad_norm": 1.5466283559799194,
+      "learning_rate": 0.00015636387121860206,
+      "loss": 0.808,
+      "step": 1716
+    },
+    {
+      "epoch": 0.3303193535975375,
+      "grad_norm": 2.4247238636016846,
+      "learning_rate": 0.00015631239215588578,
+      "loss": 0.9022,
+      "step": 1717
+    },
+    {
+      "epoch": 0.3305117352828011,
+      "grad_norm": 2.2310853004455566,
+      "learning_rate": 0.00015626089123076,
+      "loss": 0.9709,
+      "step": 1718
+    },
+    {
+      "epoch": 0.3307041169680646,
+      "grad_norm": 2.2988712787628174,
+      "learning_rate": 0.00015620936846321917,
+      "loss": 1.0712,
+      "step": 1719
+    },
+    {
+      "epoch": 0.3308964986533282,
+      "grad_norm": 2.11505389213562,
+      "learning_rate": 0.0001561578238732661,
+      "loss": 0.9934,
+      "step": 1720
+    },
+    {
+      "epoch": 0.33108888033859174,
+      "grad_norm": 1.7904237508773804,
+      "learning_rate": 0.00015610625748091228,
+      "loss": 1.0098,
+      "step": 1721
+    },
+    {
+      "epoch": 0.3312812620238553,
+      "grad_norm": 1.424217939376831,
+      "learning_rate": 0.00015605466930617746,
+      "loss": 1.0491,
+      "step": 1722
+    },
+    {
+      "epoch": 0.3314736437091189,
+      "grad_norm": 2.346271514892578,
+      "learning_rate": 0.00015600305936909004,
+      "loss": 1.0262,
+      "step": 1723
+    },
+    {
+      "epoch": 0.33166602539438245,
+      "grad_norm": 2.768338680267334,
+      "learning_rate": 0.00015595142768968668,
+      "loss": 1.1213,
+      "step": 1724
+    },
+    {
+      "epoch": 0.33185840707964603,
+      "grad_norm": 1.582536220550537,
+      "learning_rate": 0.00015589977428801267,
+      "loss": 1.0928,
+      "step": 1725
+    },
+    {
+      "epoch": 0.33205078876490957,
+      "grad_norm": 2.348013162612915,
+      "learning_rate": 0.00015584809918412157,
+      "loss": 1.0225,
+      "step": 1726
+    },
+    {
+      "epoch": 0.33224317045017315,
+      "grad_norm": 1.726785659790039,
+      "learning_rate": 0.00015579640239807548,
+      "loss": 0.9582,
+      "step": 1727
+    },
+    {
+      "epoch": 0.3324355521354367,
+      "grad_norm": 1.471192717552185,
+      "learning_rate": 0.00015574468394994486,
+      "loss": 0.8196,
+      "step": 1728
+    },
+    {
+      "epoch": 0.3326279338207003,
+      "grad_norm": 2.5012784004211426,
+      "learning_rate": 0.00015569294385980856,
+      "loss": 1.0231,
+      "step": 1729
+    },
+    {
+      "epoch": 0.33282031550596386,
+      "grad_norm": 2.14742374420166,
+      "learning_rate": 0.0001556411821477539,
+      "loss": 0.9581,
+      "step": 1730
+    },
+    {
+      "epoch": 0.3330126971912274,
+      "grad_norm": 2.3684334754943848,
+      "learning_rate": 0.00015558939883387655,
+      "loss": 1.0447,
+      "step": 1731
+    },
+    {
+      "epoch": 0.333205078876491,
+      "grad_norm": 2.4995615482330322,
+      "learning_rate": 0.00015553759393828058,
+      "loss": 0.9264,
+      "step": 1732
+    },
+    {
+      "epoch": 0.3333974605617545,
+      "grad_norm": 2.763561248779297,
+      "learning_rate": 0.00015548576748107838,
+      "loss": 1.1001,
+      "step": 1733
+    },
+    {
+      "epoch": 0.3335898422470181,
+      "grad_norm": 1.553301215171814,
+      "learning_rate": 0.00015543391948239088,
+      "loss": 1.042,
+      "step": 1734
+    },
+    {
+      "epoch": 0.33378222393228163,
+      "grad_norm": 2.0217463970184326,
+      "learning_rate": 0.00015538204996234716,
+      "loss": 0.8934,
+      "step": 1735
+    },
+    {
+      "epoch": 0.3339746056175452,
+      "grad_norm": 1.9206043481826782,
+      "learning_rate": 0.0001553301589410848,
+      "loss": 0.9374,
+      "step": 1736
+    },
+    {
+      "epoch": 0.33416698730280875,
+      "grad_norm": 2.397955894470215,
+      "learning_rate": 0.00015527824643874966,
+      "loss": 1.0212,
+      "step": 1737
+    },
+    {
+      "epoch": 0.33435936898807234,
+      "grad_norm": 2.052290916442871,
+      "learning_rate": 0.00015522631247549598,
+      "loss": 0.9789,
+      "step": 1738
+    },
+    {
+      "epoch": 0.3345517506733359,
+      "grad_norm": 2.5327067375183105,
+      "learning_rate": 0.00015517435707148626,
+      "loss": 1.066,
+      "step": 1739
+    },
+    {
+      "epoch": 0.33474413235859946,
+      "grad_norm": 1.2953333854675293,
+      "learning_rate": 0.00015512238024689142,
+      "loss": 1.092,
+      "step": 1740
+    },
+    {
+      "epoch": 0.33493651404386304,
+      "grad_norm": 2.2163078784942627,
+      "learning_rate": 0.0001550703820218907,
+      "loss": 1.0302,
+      "step": 1741
+    },
+    {
+      "epoch": 0.3351288957291266,
+      "grad_norm": 1.7354991436004639,
+      "learning_rate": 0.0001550183624166715,
+      "loss": 1.0284,
+      "step": 1742
+    },
+    {
+      "epoch": 0.33532127741439016,
+      "grad_norm": 2.506408214569092,
+      "learning_rate": 0.0001549663214514297,
+      "loss": 1.0615,
+      "step": 1743
+    },
+    {
+      "epoch": 0.3355136590996537,
+      "grad_norm": 2.206111192703247,
+      "learning_rate": 0.00015491425914636935,
+      "loss": 0.964,
+      "step": 1744
+    },
+    {
+      "epoch": 0.3357060407849173,
+      "grad_norm": 2.505014419555664,
+      "learning_rate": 0.00015486217552170283,
+      "loss": 1.056,
+      "step": 1745
+    },
+    {
+      "epoch": 0.3358984224701808,
+      "grad_norm": 2.365098476409912,
+      "learning_rate": 0.00015481007059765082,
+      "loss": 1.036,
+      "step": 1746
+    },
+    {
+      "epoch": 0.3360908041554444,
+      "grad_norm": 2.2877280712127686,
+      "learning_rate": 0.00015475794439444226,
+      "loss": 1.0931,
+      "step": 1747
+    },
+    {
+      "epoch": 0.336283185840708,
+      "grad_norm": 2.429210662841797,
+      "learning_rate": 0.00015470579693231432,
+      "loss": 0.992,
+      "step": 1748
+    },
+    {
+      "epoch": 0.3364755675259715,
+      "grad_norm": 2.341491222381592,
+      "learning_rate": 0.00015465362823151245,
+      "loss": 1.0255,
+      "step": 1749
+    },
+    {
+      "epoch": 0.3366679492112351,
+      "grad_norm": 1.5152031183242798,
+      "learning_rate": 0.00015460143831229025,
+      "loss": 1.0567,
+      "step": 1750
+    },
+    {
+      "epoch": 0.33686033089649864,
+      "grad_norm": 2.6217751502990723,
+      "learning_rate": 0.00015454922719490978,
+      "loss": 1.0577,
+      "step": 1751
+    },
+    {
+      "epoch": 0.3370527125817622,
+      "grad_norm": 1.342531681060791,
+      "learning_rate": 0.0001544969948996411,
+      "loss": 1.0566,
+      "step": 1752
+    },
+    {
+      "epoch": 0.33724509426702576,
+      "grad_norm": 1.069663166999817,
+      "learning_rate": 0.0001544447414467626,
+      "loss": 0.963,
+      "step": 1753
+    },
+    {
+      "epoch": 0.33743747595228935,
+      "grad_norm": 2.8492870330810547,
+      "learning_rate": 0.00015439246685656092,
+      "loss": 1.013,
+      "step": 1754
+    },
+    {
+      "epoch": 0.33762985763755293,
+      "grad_norm": 2.7410707473754883,
+      "learning_rate": 0.0001543401711493308,
+      "loss": 0.9942,
+      "step": 1755
+    },
+    {
+      "epoch": 0.33782223932281646,
+      "grad_norm": 1.6646394729614258,
+      "learning_rate": 0.00015428785434537526,
+      "loss": 1.0797,
+      "step": 1756
+    },
+    {
+      "epoch": 0.33801462100808005,
+      "grad_norm": 2.720715045928955,
+      "learning_rate": 0.00015423551646500546,
+      "loss": 1.0262,
+      "step": 1757
+    },
+    {
+      "epoch": 0.3382070026933436,
+      "grad_norm": 2.1640377044677734,
+      "learning_rate": 0.0001541831575285408,
+      "loss": 0.9718,
+      "step": 1758
+    },
+    {
+      "epoch": 0.33839938437860717,
+      "grad_norm": 2.002558708190918,
+      "learning_rate": 0.00015413077755630877,
+      "loss": 1.0593,
+      "step": 1759
+    },
+    {
+      "epoch": 0.3385917660638707,
+      "grad_norm": 1.252556562423706,
+      "learning_rate": 0.0001540783765686452,
+      "loss": 1.0198,
+      "step": 1760
+    },
+    {
+      "epoch": 0.3387841477491343,
+      "grad_norm": 2.2940359115600586,
+      "learning_rate": 0.0001540259545858938,
+      "loss": 1.0693,
+      "step": 1761
+    },
+    {
+      "epoch": 0.3389765294343978,
+      "grad_norm": 2.374282121658325,
+      "learning_rate": 0.0001539735116284067,
+      "loss": 1.068,
+      "step": 1762
+    },
+    {
+      "epoch": 0.3391689111196614,
+      "grad_norm": 2.435203790664673,
+      "learning_rate": 0.000153921047716544,
+      "loss": 0.9314,
+      "step": 1763
+    },
+    {
+      "epoch": 0.339361292804925,
+      "grad_norm": 2.2659451961517334,
+      "learning_rate": 0.00015386856287067402,
+      "loss": 0.947,
+      "step": 1764
+    },
+    {
+      "epoch": 0.33955367449018853,
+      "grad_norm": 2.4021308422088623,
+      "learning_rate": 0.00015381605711117316,
+      "loss": 1.0518,
+      "step": 1765
+    },
+    {
+      "epoch": 0.3397460561754521,
+      "grad_norm": 1.4883792400360107,
+      "learning_rate": 0.00015376353045842603,
+      "loss": 0.7781,
+      "step": 1766
+    },
+    {
+      "epoch": 0.33993843786071565,
+      "grad_norm": 1.799773931503296,
+      "learning_rate": 0.00015371098293282524,
+      "loss": 1.1015,
+      "step": 1767
+    },
+    {
+      "epoch": 0.34013081954597923,
+      "grad_norm": 0.9405404329299927,
+      "learning_rate": 0.00015365841455477157,
+      "loss": 1.0787,
+      "step": 1768
+    },
+    {
+      "epoch": 0.34032320123124277,
+      "grad_norm": 1.6363595724105835,
+      "learning_rate": 0.0001536058253446738,
+      "loss": 1.011,
+      "step": 1769
+    },
+    {
+      "epoch": 0.34051558291650635,
+      "grad_norm": 1.05473792552948,
+      "learning_rate": 0.00015355321532294896,
+      "loss": 0.9932,
+      "step": 1770
+    },
+    {
+      "epoch": 0.3407079646017699,
+      "grad_norm": 1.142991304397583,
+      "learning_rate": 0.00015350058451002205,
+      "loss": 0.7239,
+      "step": 1771
+    },
+    {
+      "epoch": 0.3409003462870335,
+      "grad_norm": 1.641664981842041,
+      "learning_rate": 0.00015344793292632614,
+      "loss": 0.964,
+      "step": 1772
+    },
+    {
+      "epoch": 0.34109272797229706,
+      "grad_norm": 2.059943199157715,
+      "learning_rate": 0.0001533952605923024,
+      "loss": 0.9466,
+      "step": 1773
+    },
+    {
+      "epoch": 0.3412851096575606,
+      "grad_norm": 1.3908586502075195,
+      "learning_rate": 0.00015334256752840006,
+      "loss": 0.9909,
+      "step": 1774
+    },
+    {
+      "epoch": 0.3414774913428242,
+      "grad_norm": 2.306952953338623,
+      "learning_rate": 0.00015328985375507638,
+      "loss": 1.0345,
+      "step": 1775
+    },
+    {
+      "epoch": 0.3416698730280877,
+      "grad_norm": 2.3371009826660156,
+      "learning_rate": 0.0001532371192927966,
+      "loss": 1.0127,
+      "step": 1776
+    },
+    {
+      "epoch": 0.3418622547133513,
+      "grad_norm": 1.979836106300354,
+      "learning_rate": 0.00015318436416203413,
+      "loss": 0.9224,
+      "step": 1777
+    },
+    {
+      "epoch": 0.34205463639861483,
+      "grad_norm": 2.325639247894287,
+      "learning_rate": 0.0001531315883832703,
+      "loss": 1.0408,
+      "step": 1778
+    },
+    {
+      "epoch": 0.3422470180838784,
+      "grad_norm": 1.1084682941436768,
+      "learning_rate": 0.0001530787919769945,
+      "loss": 0.9406,
+      "step": 1779
+    },
+    {
+      "epoch": 0.34243939976914195,
+      "grad_norm": 1.3248711824417114,
+      "learning_rate": 0.00015302597496370408,
+      "loss": 0.8547,
+      "step": 1780
+    },
+    {
+      "epoch": 0.34263178145440554,
+      "grad_norm": 1.8264611959457397,
+      "learning_rate": 0.00015297313736390447,
+      "loss": 0.9635,
+      "step": 1781
+    },
+    {
+      "epoch": 0.3428241631396691,
+      "grad_norm": 1.5525778532028198,
+      "learning_rate": 0.00015292027919810896,
+      "loss": 1.0605,
+      "step": 1782
+    },
+    {
+      "epoch": 0.34301654482493266,
+      "grad_norm": 1.975189447402954,
+      "learning_rate": 0.00015286740048683898,
+      "loss": 0.9725,
+      "step": 1783
+    },
+    {
+      "epoch": 0.34320892651019624,
+      "grad_norm": 2.3917288780212402,
+      "learning_rate": 0.0001528145012506239,
+      "loss": 0.9902,
+      "step": 1784
+    },
+    {
+      "epoch": 0.3434013081954598,
+      "grad_norm": 1.9404232501983643,
+      "learning_rate": 0.00015276158151000094,
+      "loss": 1.0502,
+      "step": 1785
+    },
+    {
+      "epoch": 0.34359368988072336,
+      "grad_norm": 1.3491287231445312,
+      "learning_rate": 0.00015270864128551543,
+      "loss": 0.7566,
+      "step": 1786
+    },
+    {
+      "epoch": 0.3437860715659869,
+      "grad_norm": 1.9027959108352661,
+      "learning_rate": 0.00015265568059772054,
+      "loss": 1.0041,
+      "step": 1787
+    },
+    {
+      "epoch": 0.3439784532512505,
+      "grad_norm": 2.260106325149536,
+      "learning_rate": 0.00015260269946717746,
+      "loss": 1.0315,
+      "step": 1788
+    },
+    {
+      "epoch": 0.34417083493651407,
+      "grad_norm": 1.71160888671875,
+      "learning_rate": 0.00015254969791445526,
+      "loss": 1.0432,
+      "step": 1789
+    },
+    {
+      "epoch": 0.3443632166217776,
+      "grad_norm": 1.9653297662734985,
+      "learning_rate": 0.000152496675960131,
+      "loss": 0.9524,
+      "step": 1790
+    },
+    {
+      "epoch": 0.3445555983070412,
+      "grad_norm": 2.4355297088623047,
+      "learning_rate": 0.00015244363362478967,
+      "loss": 1.0087,
+      "step": 1791
+    },
+    {
+      "epoch": 0.3447479799923047,
+      "grad_norm": 1.3247562646865845,
+      "learning_rate": 0.00015239057092902405,
+      "loss": 1.0459,
+      "step": 1792
+    },
+    {
+      "epoch": 0.3449403616775683,
+      "grad_norm": 1.6985278129577637,
+      "learning_rate": 0.00015233748789343488,
+      "loss": 1.0071,
+      "step": 1793
+    },
+    {
+      "epoch": 0.34513274336283184,
+      "grad_norm": 2.7786402702331543,
+      "learning_rate": 0.00015228438453863095,
+      "loss": 1.0438,
+      "step": 1794
+    },
+    {
+      "epoch": 0.3453251250480954,
+      "grad_norm": 1.6278852224349976,
+      "learning_rate": 0.0001522312608852287,
+      "loss": 1.0023,
+      "step": 1795
+    },
+    {
+      "epoch": 0.34551750673335896,
+      "grad_norm": 1.3875616788864136,
+      "learning_rate": 0.00015217811695385263,
+      "loss": 1.0784,
+      "step": 1796
+    },
+    {
+      "epoch": 0.34570988841862255,
+      "grad_norm": 2.367966890335083,
+      "learning_rate": 0.000152124952765135,
+      "loss": 1.0553,
+      "step": 1797
+    },
+    {
+      "epoch": 0.34590227010388613,
+      "grad_norm": 1.535637378692627,
+      "learning_rate": 0.00015207176833971598,
+      "loss": 0.8342,
+      "step": 1798
+    },
+    {
+      "epoch": 0.34609465178914967,
+      "grad_norm": 1.093474268913269,
+      "learning_rate": 0.0001520185636982436,
+      "loss": 0.7975,
+      "step": 1799
+    },
+    {
+      "epoch": 0.34628703347441325,
+      "grad_norm": 2.811380386352539,
+      "learning_rate": 0.00015196533886137377,
+      "loss": 1.0551,
+      "step": 1800
+    },
+    {
+      "epoch": 0.3464794151596768,
+      "grad_norm": 1.383840799331665,
+      "learning_rate": 0.00015191209384977014,
+      "loss": 1.0015,
+      "step": 1801
+    },
+    {
+      "epoch": 0.34667179684494037,
+      "grad_norm": 2.2462167739868164,
+      "learning_rate": 0.0001518588286841043,
+      "loss": 1.0087,
+      "step": 1802
+    },
+    {
+      "epoch": 0.3468641785302039,
+      "grad_norm": 1.2599152326583862,
+      "learning_rate": 0.00015180554338505565,
+      "loss": 0.7817,
+      "step": 1803
+    },
+    {
+      "epoch": 0.3470565602154675,
+      "grad_norm": 1.8663216829299927,
+      "learning_rate": 0.0001517522379733113,
+      "loss": 0.9769,
+      "step": 1804
+    },
+    {
+      "epoch": 0.347248941900731,
+      "grad_norm": 2.2505054473876953,
+      "learning_rate": 0.0001516989124695663,
+      "loss": 1.017,
+      "step": 1805
+    },
+    {
+      "epoch": 0.3474413235859946,
+      "grad_norm": 1.9362179040908813,
+      "learning_rate": 0.00015164556689452346,
+      "loss": 0.985,
+      "step": 1806
+    },
+    {
+      "epoch": 0.3476337052712582,
+      "grad_norm": 2.707742929458618,
+      "learning_rate": 0.0001515922012688933,
+      "loss": 0.9566,
+      "step": 1807
+    },
+    {
+      "epoch": 0.34782608695652173,
+      "grad_norm": 2.1221394538879395,
+      "learning_rate": 0.00015153881561339426,
+      "loss": 1.0476,
+      "step": 1808
+    },
+    {
+      "epoch": 0.3480184686417853,
+      "grad_norm": 2.2644317150115967,
+      "learning_rate": 0.0001514854099487524,
+      "loss": 1.0041,
+      "step": 1809
+    },
+    {
+      "epoch": 0.34821085032704885,
+      "grad_norm": 2.529546022415161,
+      "learning_rate": 0.0001514319842957018,
+      "loss": 1.072,
+      "step": 1810
+    },
+    {
+      "epoch": 0.34840323201231244,
+      "grad_norm": 1.2526394128799438,
+      "learning_rate": 0.000151378538674984,
+      "loss": 0.7423,
+      "step": 1811
+    },
+    {
+      "epoch": 0.34859561369757597,
+      "grad_norm": 2.0346269607543945,
+      "learning_rate": 0.00015132507310734847,
+      "loss": 1.0653,
+      "step": 1812
+    },
+    {
+      "epoch": 0.34878799538283956,
+      "grad_norm": 2.2786552906036377,
+      "learning_rate": 0.0001512715876135524,
+      "loss": 1.0905,
+      "step": 1813
+    },
+    {
+      "epoch": 0.34898037706810314,
+      "grad_norm": 1.8580405712127686,
+      "learning_rate": 0.0001512180822143607,
+      "loss": 0.9891,
+      "step": 1814
+    },
+    {
+      "epoch": 0.3491727587533667,
+      "grad_norm": 2.422562599182129,
+      "learning_rate": 0.00015116455693054594,
+      "loss": 1.0093,
+      "step": 1815
+    },
+    {
+      "epoch": 0.34936514043863026,
+      "grad_norm": 1.8113282918930054,
+      "learning_rate": 0.00015111101178288857,
+      "loss": 0.9849,
+      "step": 1816
+    },
+    {
+      "epoch": 0.3495575221238938,
+      "grad_norm": 1.2271798849105835,
+      "learning_rate": 0.0001510574467921766,
+      "loss": 1.0163,
+      "step": 1817
+    },
+    {
+      "epoch": 0.3497499038091574,
+      "grad_norm": 2.256627321243286,
+      "learning_rate": 0.00015100386197920584,
+      "loss": 0.9101,
+      "step": 1818
+    },
+    {
+      "epoch": 0.3499422854944209,
+      "grad_norm": 2.108581066131592,
+      "learning_rate": 0.00015095025736477978,
+      "loss": 1.0601,
+      "step": 1819
+    },
+    {
+      "epoch": 0.3501346671796845,
+      "grad_norm": 1.4779902696609497,
+      "learning_rate": 0.0001508966329697095,
+      "loss": 1.0089,
+      "step": 1820
+    },
+    {
+      "epoch": 0.35032704886494803,
+      "grad_norm": 2.173335075378418,
+      "learning_rate": 0.00015084298881481387,
+      "loss": 0.973,
+      "step": 1821
+    },
+    {
+      "epoch": 0.3505194305502116,
+      "grad_norm": 2.591949701309204,
+      "learning_rate": 0.00015078932492091943,
+      "loss": 1.0372,
+      "step": 1822
+    },
+    {
+      "epoch": 0.3507118122354752,
+      "grad_norm": 1.3401998281478882,
+      "learning_rate": 0.0001507356413088603,
+      "loss": 1.0516,
+      "step": 1823
+    },
+    {
+      "epoch": 0.35090419392073874,
+      "grad_norm": 2.67911958694458,
+      "learning_rate": 0.00015068193799947839,
+      "loss": 0.9029,
+      "step": 1824
+    },
+    {
+      "epoch": 0.3510965756060023,
+      "grad_norm": 1.1708649396896362,
+      "learning_rate": 0.00015062821501362307,
+      "loss": 0.9971,
+      "step": 1825
+    },
+    {
+      "epoch": 0.35128895729126586,
+      "grad_norm": 1.988483190536499,
+      "learning_rate": 0.0001505744723721515,
+      "loss": 1.0408,
+      "step": 1826
+    },
+    {
+      "epoch": 0.35148133897652944,
+      "grad_norm": 2.272282123565674,
+      "learning_rate": 0.00015052071009592844,
+      "loss": 0.9552,
+      "step": 1827
+    },
+    {
+      "epoch": 0.351673720661793,
+      "grad_norm": 2.3391902446746826,
+      "learning_rate": 0.00015046692820582625,
+      "loss": 1.0259,
+      "step": 1828
+    },
+    {
+      "epoch": 0.35186610234705656,
+      "grad_norm": 2.1411263942718506,
+      "learning_rate": 0.0001504131267227249,
+      "loss": 1.0205,
+      "step": 1829
+    },
+    {
+      "epoch": 0.3520584840323201,
+      "grad_norm": 2.2304575443267822,
+      "learning_rate": 0.00015035930566751197,
+      "loss": 1.0687,
+      "step": 1830
+    },
+    {
+      "epoch": 0.3522508657175837,
+      "grad_norm": 1.3422719240188599,
+      "learning_rate": 0.00015030546506108268,
+      "loss": 1.0914,
+      "step": 1831
+    },
+    {
+      "epoch": 0.35244324740284727,
+      "grad_norm": 2.472715139389038,
+      "learning_rate": 0.00015025160492433977,
+      "loss": 0.998,
+      "step": 1832
+    },
+    {
+      "epoch": 0.3526356290881108,
+      "grad_norm": 1.6115520000457764,
+      "learning_rate": 0.0001501977252781936,
+      "loss": 1.0361,
+      "step": 1833
+    },
+    {
+      "epoch": 0.3528280107733744,
+      "grad_norm": 2.5772037506103516,
+      "learning_rate": 0.00015014382614356213,
+      "loss": 1.0494,
+      "step": 1834
+    },
+    {
+      "epoch": 0.3530203924586379,
+      "grad_norm": 2.4408538341522217,
+      "learning_rate": 0.00015008990754137086,
+      "loss": 0.8671,
+      "step": 1835
+    },
+    {
+      "epoch": 0.3532127741439015,
+      "grad_norm": 2.381021022796631,
+      "learning_rate": 0.00015003596949255283,
+      "loss": 0.9677,
+      "step": 1836
+    },
+    {
+      "epoch": 0.35340515582916504,
+      "grad_norm": 1.9839907884597778,
+      "learning_rate": 0.00014998201201804867,
+      "loss": 1.0184,
+      "step": 1837
+    },
+    {
+      "epoch": 0.35359753751442863,
+      "grad_norm": 2.1137781143188477,
+      "learning_rate": 0.0001499280351388065,
+      "loss": 0.9882,
+      "step": 1838
+    },
+    {
+      "epoch": 0.3537899191996922,
+      "grad_norm": 2.1574034690856934,
+      "learning_rate": 0.000149874038875782,
+      "loss": 1.0128,
+      "step": 1839
+    },
+    {
+      "epoch": 0.35398230088495575,
+      "grad_norm": 1.5967105627059937,
+      "learning_rate": 0.0001498200232499384,
+      "loss": 1.0073,
+      "step": 1840
+    },
+    {
+      "epoch": 0.35417468257021933,
+      "grad_norm": 1.931406855583191,
+      "learning_rate": 0.0001497659882822464,
+      "loss": 1.0583,
+      "step": 1841
+    },
+    {
+      "epoch": 0.35436706425548287,
+      "grad_norm": 2.2014951705932617,
+      "learning_rate": 0.0001497119339936843,
+      "loss": 0.9298,
+      "step": 1842
+    },
+    {
+      "epoch": 0.35455944594074645,
+      "grad_norm": 2.2766635417938232,
+      "learning_rate": 0.0001496578604052378,
+      "loss": 1.0253,
+      "step": 1843
+    },
+    {
+      "epoch": 0.35475182762601,
+      "grad_norm": 1.8644800186157227,
+      "learning_rate": 0.0001496037675379001,
+      "loss": 1.0238,
+      "step": 1844
+    },
+    {
+      "epoch": 0.3549442093112736,
+      "grad_norm": 1.2467238903045654,
+      "learning_rate": 0.00014954965541267193,
+      "loss": 0.7552,
+      "step": 1845
+    },
+    {
+      "epoch": 0.3551365909965371,
+      "grad_norm": 2.3253023624420166,
+      "learning_rate": 0.0001494955240505615,
+      "loss": 0.9334,
+      "step": 1846
+    },
+    {
+      "epoch": 0.3553289726818007,
+      "grad_norm": 1.95561683177948,
+      "learning_rate": 0.0001494413734725844,
+      "loss": 1.021,
+      "step": 1847
+    },
+    {
+      "epoch": 0.3555213543670643,
+      "grad_norm": 2.2173526287078857,
+      "learning_rate": 0.00014938720369976386,
+      "loss": 0.9909,
+      "step": 1848
+    },
+    {
+      "epoch": 0.3557137360523278,
+      "grad_norm": 1.2715378999710083,
+      "learning_rate": 0.00014933301475313036,
+      "loss": 1.057,
+      "step": 1849
+    },
+    {
+      "epoch": 0.3559061177375914,
+      "grad_norm": 2.606572389602661,
+      "learning_rate": 0.00014927880665372196,
+      "loss": 1.104,
+      "step": 1850
+    },
+    {
+      "epoch": 0.35609849942285493,
+      "grad_norm": 2.3754220008850098,
+      "learning_rate": 0.0001492245794225841,
+      "loss": 1.011,
+      "step": 1851
+    },
+    {
+      "epoch": 0.3562908811081185,
+      "grad_norm": 1.6918216943740845,
+      "learning_rate": 0.00014917033308076965,
+      "loss": 1.0443,
+      "step": 1852
+    },
+    {
+      "epoch": 0.35648326279338205,
+      "grad_norm": 2.376300096511841,
+      "learning_rate": 0.00014911606764933893,
+      "loss": 1.072,
+      "step": 1853
+    },
+    {
+      "epoch": 0.35667564447864564,
+      "grad_norm": 1.8126744031906128,
+      "learning_rate": 0.0001490617831493596,
+      "loss": 1.0511,
+      "step": 1854
+    },
+    {
+      "epoch": 0.35686802616390917,
+      "grad_norm": 1.831702709197998,
+      "learning_rate": 0.00014900747960190682,
+      "loss": 0.9953,
+      "step": 1855
+    },
+    {
+      "epoch": 0.35706040784917276,
+      "grad_norm": 1.4260374307632446,
+      "learning_rate": 0.0001489531570280631,
+      "loss": 1.0239,
+      "step": 1856
+    },
+    {
+      "epoch": 0.35725278953443634,
+      "grad_norm": 2.210874080657959,
+      "learning_rate": 0.0001488988154489183,
+      "loss": 0.9675,
+      "step": 1857
+    },
+    {
+      "epoch": 0.3574451712196999,
+      "grad_norm": 2.373155355453491,
+      "learning_rate": 0.0001488444548855697,
+      "loss": 0.9967,
+      "step": 1858
+    },
+    {
+      "epoch": 0.35763755290496346,
+      "grad_norm": 1.5005172491073608,
+      "learning_rate": 0.00014879007535912198,
+      "loss": 1.0435,
+      "step": 1859
+    },
+    {
+      "epoch": 0.357829934590227,
+      "grad_norm": 1.3598963022232056,
+      "learning_rate": 0.00014873567689068707,
+      "loss": 0.7862,
+      "step": 1860
+    },
+    {
+      "epoch": 0.3580223162754906,
+      "grad_norm": 2.0960216522216797,
+      "learning_rate": 0.00014868125950138442,
+      "loss": 1.0401,
+      "step": 1861
+    },
+    {
+      "epoch": 0.3582146979607541,
+      "grad_norm": 2.1839089393615723,
+      "learning_rate": 0.00014862682321234062,
+      "loss": 0.9771,
+      "step": 1862
+    },
+    {
+      "epoch": 0.3584070796460177,
+      "grad_norm": 1.7468359470367432,
+      "learning_rate": 0.00014857236804468983,
+      "loss": 1.0602,
+      "step": 1863
+    },
+    {
+      "epoch": 0.3585994613312813,
+      "grad_norm": 2.784757375717163,
+      "learning_rate": 0.00014851789401957338,
+      "loss": 0.9827,
+      "step": 1864
+    },
+    {
+      "epoch": 0.3587918430165448,
+      "grad_norm": 1.3139410018920898,
+      "learning_rate": 0.00014846340115813993,
+      "loss": 0.9516,
+      "step": 1865
+    },
+    {
+      "epoch": 0.3589842247018084,
+      "grad_norm": 2.594316005706787,
+      "learning_rate": 0.00014840888948154549,
+      "loss": 1.0367,
+      "step": 1866
+    },
+    {
+      "epoch": 0.35917660638707194,
+      "grad_norm": 2.3461267948150635,
+      "learning_rate": 0.0001483543590109534,
+      "loss": 1.1238,
+      "step": 1867
+    },
+    {
+      "epoch": 0.3593689880723355,
+      "grad_norm": 1.3676774501800537,
+      "learning_rate": 0.00014829980976753426,
+      "loss": 0.7942,
+      "step": 1868
+    },
+    {
+      "epoch": 0.35956136975759906,
+      "grad_norm": 1.8753057718276978,
+      "learning_rate": 0.00014824524177246598,
+      "loss": 1.0183,
+      "step": 1869
+    },
+    {
+      "epoch": 0.35975375144286265,
+      "grad_norm": 2.5065596103668213,
+      "learning_rate": 0.00014819065504693364,
+      "loss": 0.9479,
+      "step": 1870
+    },
+    {
+      "epoch": 0.3599461331281262,
+      "grad_norm": 1.5775113105773926,
+      "learning_rate": 0.00014813604961212983,
+      "loss": 1.0955,
+      "step": 1871
+    },
+    {
+      "epoch": 0.36013851481338977,
+      "grad_norm": 1.7840373516082764,
+      "learning_rate": 0.00014808142548925418,
+      "loss": 1.0878,
+      "step": 1872
+    },
+    {
+      "epoch": 0.36033089649865335,
+      "grad_norm": 2.5700676441192627,
+      "learning_rate": 0.00014802678269951365,
+      "loss": 0.9988,
+      "step": 1873
+    },
+    {
+      "epoch": 0.3605232781839169,
+      "grad_norm": 2.401923894882202,
+      "learning_rate": 0.00014797212126412243,
+      "loss": 1.0084,
+      "step": 1874
+    },
+    {
+      "epoch": 0.36071565986918047,
+      "grad_norm": 1.5055781602859497,
+      "learning_rate": 0.00014791744120430203,
+      "loss": 0.9438,
+      "step": 1875
+    },
+    {
+      "epoch": 0.360908041554444,
+      "grad_norm": 1.917465329170227,
+      "learning_rate": 0.0001478627425412811,
+      "loss": 1.0516,
+      "step": 1876
+    },
+    {
+      "epoch": 0.3611004232397076,
+      "grad_norm": 2.677349090576172,
+      "learning_rate": 0.00014780802529629558,
+      "loss": 1.0187,
+      "step": 1877
+    },
+    {
+      "epoch": 0.3612928049249711,
+      "grad_norm": 2.7106614112854004,
+      "learning_rate": 0.00014775328949058854,
+      "loss": 1.0102,
+      "step": 1878
+    },
+    {
+      "epoch": 0.3614851866102347,
+      "grad_norm": 2.41949462890625,
+      "learning_rate": 0.00014769853514541036,
+      "loss": 1.1053,
+      "step": 1879
+    },
+    {
+      "epoch": 0.36167756829549824,
+      "grad_norm": 2.2019076347351074,
+      "learning_rate": 0.0001476437622820185,
+      "loss": 0.9244,
+      "step": 1880
+    },
+    {
+      "epoch": 0.36186994998076183,
+      "grad_norm": 2.3630292415618896,
+      "learning_rate": 0.0001475889709216777,
+      "loss": 0.9534,
+      "step": 1881
+    },
+    {
+      "epoch": 0.3620623316660254,
+      "grad_norm": 2.0500335693359375,
+      "learning_rate": 0.00014753416108565984,
+      "loss": 1.0616,
+      "step": 1882
+    },
+    {
+      "epoch": 0.36225471335128895,
+      "grad_norm": 2.444465398788452,
+      "learning_rate": 0.000147479332795244,
+      "loss": 1.0157,
+      "step": 1883
+    },
+    {
+      "epoch": 0.36244709503655254,
+      "grad_norm": 1.7136799097061157,
+      "learning_rate": 0.00014742448607171644,
+      "loss": 1.0739,
+      "step": 1884
+    },
+    {
+      "epoch": 0.36263947672181607,
+      "grad_norm": 1.4231313467025757,
+      "learning_rate": 0.0001473696209363705,
+      "loss": 0.8549,
+      "step": 1885
+    },
+    {
+      "epoch": 0.36283185840707965,
+      "grad_norm": 2.247445821762085,
+      "learning_rate": 0.00014731473741050671,
+      "loss": 0.9863,
+      "step": 1886
+    },
+    {
+      "epoch": 0.3630242400923432,
+      "grad_norm": 1.6322393417358398,
+      "learning_rate": 0.00014725983551543278,
+      "loss": 0.9679,
+      "step": 1887
+    },
+    {
+      "epoch": 0.3632166217776068,
+      "grad_norm": 3.1925957202911377,
+      "learning_rate": 0.0001472049152724635,
+      "loss": 1.0107,
+      "step": 1888
+    },
+    {
+      "epoch": 0.36340900346287036,
+      "grad_norm": 2.2896840572357178,
+      "learning_rate": 0.00014714997670292078,
+      "loss": 1.0722,
+      "step": 1889
+    },
+    {
+      "epoch": 0.3636013851481339,
+      "grad_norm": 2.3834798336029053,
+      "learning_rate": 0.00014709501982813368,
+      "loss": 1.0464,
+      "step": 1890
+    },
+    {
+      "epoch": 0.3637937668333975,
+      "grad_norm": 1.9034366607666016,
+      "learning_rate": 0.0001470400446694384,
+      "loss": 0.9751,
+      "step": 1891
+    },
+    {
+      "epoch": 0.363986148518661,
+      "grad_norm": 3.1833677291870117,
+      "learning_rate": 0.00014698505124817812,
+      "loss": 0.9856,
+      "step": 1892
+    },
+    {
+      "epoch": 0.3641785302039246,
+      "grad_norm": 2.7991302013397217,
+      "learning_rate": 0.00014693003958570318,
+      "loss": 1.0009,
+      "step": 1893
+    },
+    {
+      "epoch": 0.36437091188918813,
+      "grad_norm": 1.4610553979873657,
+      "learning_rate": 0.00014687500970337103,
+      "loss": 0.8244,
+      "step": 1894
+    },
+    {
+      "epoch": 0.3645632935744517,
+      "grad_norm": 2.3026859760284424,
+      "learning_rate": 0.00014681996162254618,
+      "loss": 1.0413,
+      "step": 1895
+    },
+    {
+      "epoch": 0.36475567525971525,
+      "grad_norm": 2.136388063430786,
+      "learning_rate": 0.00014676489536460014,
+      "loss": 1.0897,
+      "step": 1896
+    },
+    {
+      "epoch": 0.36494805694497884,
+      "grad_norm": 2.1729350090026855,
+      "learning_rate": 0.0001467098109509116,
+      "loss": 1.0851,
+      "step": 1897
+    },
+    {
+      "epoch": 0.3651404386302424,
+      "grad_norm": 2.1727094650268555,
+      "learning_rate": 0.00014665470840286615,
+      "loss": 0.9917,
+      "step": 1898
+    },
+    {
+      "epoch": 0.36533282031550596,
+      "grad_norm": 2.5488390922546387,
+      "learning_rate": 0.00014659958774185654,
+      "loss": 1.0369,
+      "step": 1899
+    },
+    {
+      "epoch": 0.36552520200076954,
+      "grad_norm": 1.8945949077606201,
+      "learning_rate": 0.0001465444489892825,
+      "loss": 1.0334,
+      "step": 1900
+    },
+    {
+      "epoch": 0.3657175836860331,
+      "grad_norm": 1.5437698364257812,
+      "learning_rate": 0.00014648929216655077,
+      "loss": 1.0638,
+      "step": 1901
+    },
+    {
+      "epoch": 0.36590996537129666,
+      "grad_norm": 2.5042154788970947,
+      "learning_rate": 0.00014643411729507516,
+      "loss": 1.0786,
+      "step": 1902
+    },
+    {
+      "epoch": 0.3661023470565602,
+      "grad_norm": 2.0564630031585693,
+      "learning_rate": 0.00014637892439627642,
+      "loss": 1.028,
+      "step": 1903
+    },
+    {
+      "epoch": 0.3662947287418238,
+      "grad_norm": 2.233368158340454,
+      "learning_rate": 0.0001463237134915824,
+      "loss": 1.0883,
+      "step": 1904
+    },
+    {
+      "epoch": 0.3664871104270873,
+      "grad_norm": 2.4524013996124268,
+      "learning_rate": 0.0001462684846024278,
+      "loss": 1.076,
+      "step": 1905
+    },
+    {
+      "epoch": 0.3666794921123509,
+      "grad_norm": 2.265178918838501,
+      "learning_rate": 0.00014621323775025445,
+      "loss": 0.9916,
+      "step": 1906
+    },
+    {
+      "epoch": 0.3668718737976145,
+      "grad_norm": 2.8307313919067383,
+      "learning_rate": 0.000146157972956511,
+      "loss": 1.0064,
+      "step": 1907
+    },
+    {
+      "epoch": 0.367064255482878,
+      "grad_norm": 2.4624857902526855,
+      "learning_rate": 0.00014610269024265316,
+      "loss": 1.09,
+      "step": 1908
+    },
+    {
+      "epoch": 0.3672566371681416,
+      "grad_norm": 0.9381546974182129,
+      "learning_rate": 0.00014604738963014364,
+      "loss": 0.9925,
+      "step": 1909
+    },
+    {
+      "epoch": 0.36744901885340514,
+      "grad_norm": 1.1692792177200317,
+      "learning_rate": 0.00014599207114045202,
+      "loss": 0.7957,
+      "step": 1910
+    },
+    {
+      "epoch": 0.36764140053866873,
+      "grad_norm": 1.6477652788162231,
+      "learning_rate": 0.00014593673479505482,
+      "loss": 1.0454,
+      "step": 1911
+    },
+    {
+      "epoch": 0.36783378222393226,
+      "grad_norm": 1.5695788860321045,
+      "learning_rate": 0.0001458813806154355,
+      "loss": 0.9869,
+      "step": 1912
+    },
+    {
+      "epoch": 0.36802616390919585,
+      "grad_norm": 2.5645623207092285,
+      "learning_rate": 0.00014582600862308447,
+      "loss": 1.0871,
+      "step": 1913
+    },
+    {
+      "epoch": 0.36821854559445943,
+      "grad_norm": 1.1208261251449585,
+      "learning_rate": 0.00014577061883949913,
+      "loss": 0.7263,
+      "step": 1914
+    },
+    {
+      "epoch": 0.36841092727972297,
+      "grad_norm": 2.05623197555542,
+      "learning_rate": 0.00014571521128618357,
+      "loss": 0.9827,
+      "step": 1915
+    },
+    {
+      "epoch": 0.36860330896498655,
+      "grad_norm": 2.3050057888031006,
+      "learning_rate": 0.00014565978598464894,
+      "loss": 1.0153,
+      "step": 1916
+    },
+    {
+      "epoch": 0.3687956906502501,
+      "grad_norm": 2.530029773712158,
+      "learning_rate": 0.00014560434295641337,
+      "loss": 1.015,
+      "step": 1917
+    },
+    {
+      "epoch": 0.3689880723355137,
+      "grad_norm": 1.831270456314087,
+      "learning_rate": 0.0001455488822230016,
+      "loss": 0.9421,
+      "step": 1918
+    },
+    {
+      "epoch": 0.3691804540207772,
+      "grad_norm": 1.5288769006729126,
+      "learning_rate": 0.00014549340380594545,
+      "loss": 0.976,
+      "step": 1919
+    },
+    {
+      "epoch": 0.3693728357060408,
+      "grad_norm": 1.9744091033935547,
+      "learning_rate": 0.00014543790772678357,
+      "loss": 0.9948,
+      "step": 1920
+    },
+    {
+      "epoch": 0.3695652173913043,
+      "grad_norm": 2.9557251930236816,
+      "learning_rate": 0.00014538239400706146,
+      "loss": 1.0113,
+      "step": 1921
+    },
+    {
+      "epoch": 0.3697575990765679,
+      "grad_norm": 2.6687474250793457,
+      "learning_rate": 0.00014532686266833142,
+      "loss": 1.026,
+      "step": 1922
+    },
+    {
+      "epoch": 0.3699499807618315,
+      "grad_norm": 1.3764855861663818,
+      "learning_rate": 0.00014527131373215263,
+      "loss": 0.8017,
+      "step": 1923
+    },
+    {
+      "epoch": 0.37014236244709503,
+      "grad_norm": 1.6365402936935425,
+      "learning_rate": 0.00014521574722009114,
+      "loss": 1.0138,
+      "step": 1924
+    },
+    {
+      "epoch": 0.3703347441323586,
+      "grad_norm": 0.8941077589988708,
+      "learning_rate": 0.00014516016315371975,
+      "loss": 0.7289,
+      "step": 1925
+    },
+    {
+      "epoch": 0.37052712581762215,
+      "grad_norm": 2.0027785301208496,
+      "learning_rate": 0.00014510456155461806,
+      "loss": 1.0621,
+      "step": 1926
+    },
+    {
+      "epoch": 0.37071950750288574,
+      "grad_norm": 2.154841423034668,
+      "learning_rate": 0.00014504894244437265,
+      "loss": 0.9969,
+      "step": 1927
+    },
+    {
+      "epoch": 0.37091188918814927,
+      "grad_norm": 2.186342239379883,
+      "learning_rate": 0.00014499330584457666,
+      "loss": 1.0404,
+      "step": 1928
+    },
+    {
+      "epoch": 0.37110427087341286,
+      "grad_norm": 2.3161327838897705,
+      "learning_rate": 0.00014493765177683016,
+      "loss": 1.018,
+      "step": 1929
+    },
+    {
+      "epoch": 0.3712966525586764,
+      "grad_norm": 1.914618968963623,
+      "learning_rate": 0.00014488198026274006,
+      "loss": 1.0475,
+      "step": 1930
+    },
+    {
+      "epoch": 0.37148903424394,
+      "grad_norm": 2.0363895893096924,
+      "learning_rate": 0.00014482629132391985,
+      "loss": 1.0555,
+      "step": 1931
+    },
+    {
+      "epoch": 0.37168141592920356,
+      "grad_norm": 1.873372197151184,
+      "learning_rate": 0.00014477058498198992,
+      "loss": 1.0285,
+      "step": 1932
+    },
+    {
+      "epoch": 0.3718737976144671,
+      "grad_norm": 1.5592864751815796,
+      "learning_rate": 0.00014471486125857742,
+      "loss": 0.8068,
+      "step": 1933
+    },
+    {
+      "epoch": 0.3720661792997307,
+      "grad_norm": 2.413073778152466,
+      "learning_rate": 0.0001446591201753162,
+      "loss": 0.9742,
+      "step": 1934
+    },
+    {
+      "epoch": 0.3722585609849942,
+      "grad_norm": 2.255188465118408,
+      "learning_rate": 0.0001446033617538469,
+      "loss": 1.0608,
+      "step": 1935
+    },
+    {
+      "epoch": 0.3724509426702578,
+      "grad_norm": 2.2581064701080322,
+      "learning_rate": 0.00014454758601581676,
+      "loss": 1.0092,
+      "step": 1936
+    },
+    {
+      "epoch": 0.37264332435552133,
+      "grad_norm": 2.1647725105285645,
+      "learning_rate": 0.00014449179298287997,
+      "loss": 1.0367,
+      "step": 1937
+    },
+    {
+      "epoch": 0.3728357060407849,
+      "grad_norm": 2.5473952293395996,
+      "learning_rate": 0.00014443598267669723,
+      "loss": 0.9582,
+      "step": 1938
+    },
+    {
+      "epoch": 0.3730280877260485,
+      "grad_norm": 1.9070287942886353,
+      "learning_rate": 0.000144380155118936,
+      "loss": 0.9454,
+      "step": 1939
+    },
+    {
+      "epoch": 0.37322046941131204,
+      "grad_norm": 2.819669008255005,
+      "learning_rate": 0.00014432431033127056,
+      "loss": 1.0735,
+      "step": 1940
+    },
+    {
+      "epoch": 0.3734128510965756,
+      "grad_norm": 2.362276792526245,
+      "learning_rate": 0.0001442684483353817,
+      "loss": 1.0025,
+      "step": 1941
+    },
+    {
+      "epoch": 0.37360523278183916,
+      "grad_norm": 2.6934146881103516,
+      "learning_rate": 0.00014421256915295695,
+      "loss": 0.9087,
+      "step": 1942
+    },
+    {
+      "epoch": 0.37379761446710275,
+      "grad_norm": 2.921886444091797,
+      "learning_rate": 0.00014415667280569063,
+      "loss": 0.9374,
+      "step": 1943
+    },
+    {
+      "epoch": 0.3739899961523663,
+      "grad_norm": 1.9552586078643799,
+      "learning_rate": 0.00014410075931528356,
+      "loss": 1.0571,
+      "step": 1944
+    },
+    {
+      "epoch": 0.37418237783762986,
+      "grad_norm": 2.3005006313323975,
+      "learning_rate": 0.00014404482870344323,
+      "loss": 1.0154,
+      "step": 1945
+    },
+    {
+      "epoch": 0.3743747595228934,
+      "grad_norm": 2.1670725345611572,
+      "learning_rate": 0.00014398888099188394,
+      "loss": 1.0794,
+      "step": 1946
+    },
+    {
+      "epoch": 0.374567141208157,
+      "grad_norm": 2.561225175857544,
+      "learning_rate": 0.00014393291620232645,
+      "loss": 0.9861,
+      "step": 1947
+    },
+    {
+      "epoch": 0.37475952289342057,
+      "grad_norm": 2.401137113571167,
+      "learning_rate": 0.00014387693435649826,
+      "loss": 1.0071,
+      "step": 1948
+    },
+    {
+      "epoch": 0.3749519045786841,
+      "grad_norm": 2.3451361656188965,
+      "learning_rate": 0.00014382093547613337,
+      "loss": 1.0408,
+      "step": 1949
+    },
+    {
+      "epoch": 0.3751442862639477,
+      "grad_norm": 1.980454683303833,
+      "learning_rate": 0.00014376491958297262,
+      "loss": 0.9504,
+      "step": 1950
+    },
+    {
+      "epoch": 0.3753366679492112,
+      "grad_norm": 2.5913100242614746,
+      "learning_rate": 0.00014370888669876317,
+      "loss": 1.0671,
+      "step": 1951
+    },
+    {
+      "epoch": 0.3755290496344748,
+      "grad_norm": 2.550048828125,
+      "learning_rate": 0.00014365283684525894,
+      "loss": 1.0378,
+      "step": 1952
+    },
+    {
+      "epoch": 0.37572143131973834,
+      "grad_norm": 2.186974048614502,
+      "learning_rate": 0.00014359677004422046,
+      "loss": 0.9688,
+      "step": 1953
+    },
+    {
+      "epoch": 0.37591381300500193,
+      "grad_norm": 2.0104832649230957,
+      "learning_rate": 0.00014354068631741476,
+      "loss": 1.0454,
+      "step": 1954
+    },
+    {
+      "epoch": 0.37610619469026546,
+      "grad_norm": 1.3720930814743042,
+      "learning_rate": 0.00014348458568661548,
+      "loss": 1.0969,
+      "step": 1955
+    },
+    {
+      "epoch": 0.37629857637552905,
+      "grad_norm": 2.211352586746216,
+      "learning_rate": 0.0001434284681736028,
+      "loss": 1.0622,
+      "step": 1956
+    },
+    {
+      "epoch": 0.37649095806079264,
+      "grad_norm": 2.0273633003234863,
+      "learning_rate": 0.00014337233380016352,
+      "loss": 1.0635,
+      "step": 1957
+    },
+    {
+      "epoch": 0.37668333974605617,
+      "grad_norm": 2.186833620071411,
+      "learning_rate": 0.0001433161825880909,
+      "loss": 1.0023,
+      "step": 1958
+    },
+    {
+      "epoch": 0.37687572143131975,
+      "grad_norm": 0.9019863605499268,
+      "learning_rate": 0.00014326001455918478,
+      "loss": 1.0043,
+      "step": 1959
+    },
+    {
+      "epoch": 0.3770681031165833,
+      "grad_norm": 1.386043906211853,
+      "learning_rate": 0.0001432038297352515,
+      "loss": 0.8034,
+      "step": 1960
+    },
+    {
+      "epoch": 0.3772604848018469,
+      "grad_norm": 2.330907106399536,
+      "learning_rate": 0.000143147628138104,
+      "loss": 1.0234,
+      "step": 1961
+    },
+    {
+      "epoch": 0.3774528664871104,
+      "grad_norm": 2.0674901008605957,
+      "learning_rate": 0.0001430914097895616,
+      "loss": 1.0631,
+      "step": 1962
+    },
+    {
+      "epoch": 0.377645248172374,
+      "grad_norm": 2.437772274017334,
+      "learning_rate": 0.0001430351747114503,
+      "loss": 1.0567,
+      "step": 1963
+    },
+    {
+      "epoch": 0.3778376298576376,
+      "grad_norm": 2.899590253829956,
+      "learning_rate": 0.0001429789229256024,
+      "loss": 0.9584,
+      "step": 1964
+    },
+    {
+      "epoch": 0.3780300115429011,
+      "grad_norm": 1.460975170135498,
+      "learning_rate": 0.0001429226544538568,
+      "loss": 0.7276,
+      "step": 1965
+    },
+    {
+      "epoch": 0.3782223932281647,
+      "grad_norm": 1.3714386224746704,
+      "learning_rate": 0.00014286636931805887,
+      "loss": 0.7318,
+      "step": 1966
+    },
+    {
+      "epoch": 0.37841477491342823,
+      "grad_norm": 2.116227865219116,
+      "learning_rate": 0.00014281006754006045,
+      "loss": 1.0173,
+      "step": 1967
+    },
+    {
+      "epoch": 0.3786071565986918,
+      "grad_norm": 1.5139482021331787,
+      "learning_rate": 0.0001427537491417198,
+      "loss": 1.0582,
+      "step": 1968
+    },
+    {
+      "epoch": 0.37879953828395535,
+      "grad_norm": 1.6767141819000244,
+      "learning_rate": 0.0001426974141449017,
+      "loss": 0.9268,
+      "step": 1969
+    },
+    {
+      "epoch": 0.37899191996921894,
+      "grad_norm": 2.2680904865264893,
+      "learning_rate": 0.00014264106257147732,
+      "loss": 1.0369,
+      "step": 1970
+    },
+    {
+      "epoch": 0.37918430165448247,
+      "grad_norm": 2.3876845836639404,
+      "learning_rate": 0.00014258469444332423,
+      "loss": 1.0214,
+      "step": 1971
+    },
+    {
+      "epoch": 0.37937668333974606,
+      "grad_norm": 2.296144962310791,
+      "learning_rate": 0.00014252830978232656,
+      "loss": 1.1011,
+      "step": 1972
+    },
+    {
+      "epoch": 0.37956906502500964,
+      "grad_norm": 2.267538547515869,
+      "learning_rate": 0.00014247190861037474,
+      "loss": 0.9966,
+      "step": 1973
+    },
+    {
+      "epoch": 0.3797614467102732,
+      "grad_norm": 2.4722414016723633,
+      "learning_rate": 0.00014241549094936567,
+      "loss": 1.0474,
+      "step": 1974
+    },
+    {
+      "epoch": 0.37995382839553676,
+      "grad_norm": 1.3094090223312378,
+      "learning_rate": 0.00014235905682120255,
+      "loss": 0.9889,
+      "step": 1975
+    },
+    {
+      "epoch": 0.3801462100808003,
+      "grad_norm": 2.2844350337982178,
+      "learning_rate": 0.00014230260624779512,
+      "loss": 1.0134,
+      "step": 1976
+    },
+    {
+      "epoch": 0.3803385917660639,
+      "grad_norm": 2.152385711669922,
+      "learning_rate": 0.00014224613925105947,
+      "loss": 0.9592,
+      "step": 1977
+    },
+    {
+      "epoch": 0.3805309734513274,
+      "grad_norm": 2.3622539043426514,
+      "learning_rate": 0.00014218965585291793,
+      "loss": 0.9938,
+      "step": 1978
+    },
+    {
+      "epoch": 0.380723355136591,
+      "grad_norm": 2.0622718334198,
+      "learning_rate": 0.00014213315607529938,
+      "loss": 1.0046,
+      "step": 1979
+    },
+    {
+      "epoch": 0.38091573682185453,
+      "grad_norm": 2.230142593383789,
+      "learning_rate": 0.00014207663994013895,
+      "loss": 0.9608,
+      "step": 1980
+    },
+    {
+      "epoch": 0.3811081185071181,
+      "grad_norm": 2.2188003063201904,
+      "learning_rate": 0.00014202010746937815,
+      "loss": 1.0565,
+      "step": 1981
+    },
+    {
+      "epoch": 0.3813005001923817,
+      "grad_norm": 2.774010181427002,
+      "learning_rate": 0.00014196355868496485,
+      "loss": 1.0518,
+      "step": 1982
+    },
+    {
+      "epoch": 0.38149288187764524,
+      "grad_norm": 2.278066873550415,
+      "learning_rate": 0.00014190699360885323,
+      "loss": 1.0413,
+      "step": 1983
+    },
+    {
+      "epoch": 0.3816852635629088,
+      "grad_norm": 1.5209215879440308,
+      "learning_rate": 0.00014185041226300374,
+      "loss": 0.9547,
+      "step": 1984
+    },
+    {
+      "epoch": 0.38187764524817236,
+      "grad_norm": 1.8342119455337524,
+      "learning_rate": 0.0001417938146693833,
+      "loss": 0.9736,
+      "step": 1985
+    },
+    {
+      "epoch": 0.38207002693343595,
+      "grad_norm": 2.217229127883911,
+      "learning_rate": 0.000141737200849965,
+      "loss": 0.9761,
+      "step": 1986
+    },
+    {
+      "epoch": 0.3822624086186995,
+      "grad_norm": 2.6626598834991455,
+      "learning_rate": 0.00014168057082672826,
+      "loss": 1.0937,
+      "step": 1987
+    },
+    {
+      "epoch": 0.38245479030396307,
+      "grad_norm": 2.4482226371765137,
+      "learning_rate": 0.00014162392462165884,
+      "loss": 1.0469,
+      "step": 1988
+    },
+    {
+      "epoch": 0.38264717198922665,
+      "grad_norm": 2.1072614192962646,
+      "learning_rate": 0.00014156726225674873,
+      "loss": 0.9671,
+      "step": 1989
+    },
+    {
+      "epoch": 0.3828395536744902,
+      "grad_norm": 2.718477487564087,
+      "learning_rate": 0.0001415105837539962,
+      "loss": 1.0256,
+      "step": 1990
+    },
+    {
+      "epoch": 0.3830319353597538,
+      "grad_norm": 2.484750747680664,
+      "learning_rate": 0.0001414538891354058,
+      "loss": 1.0062,
+      "step": 1991
+    },
+    {
+      "epoch": 0.3832243170450173,
+      "grad_norm": 2.20698881149292,
+      "learning_rate": 0.00014139717842298835,
+      "loss": 1.0205,
+      "step": 1992
+    },
+    {
+      "epoch": 0.3834166987302809,
+      "grad_norm": 2.4516382217407227,
+      "learning_rate": 0.00014134045163876087,
+      "loss": 1.0464,
+      "step": 1993
+    },
+    {
+      "epoch": 0.3836090804155444,
+      "grad_norm": 2.4734580516815186,
+      "learning_rate": 0.00014128370880474666,
+      "loss": 0.9903,
+      "step": 1994
+    },
+    {
+      "epoch": 0.383801462100808,
+      "grad_norm": 2.1748809814453125,
+      "learning_rate": 0.0001412269499429753,
+      "loss": 1.0238,
+      "step": 1995
+    },
+    {
+      "epoch": 0.38399384378607154,
+      "grad_norm": 2.6430609226226807,
+      "learning_rate": 0.00014117017507548244,
+      "loss": 1.0368,
+      "step": 1996
+    },
+    {
+      "epoch": 0.38418622547133513,
+      "grad_norm": 2.2624306678771973,
+      "learning_rate": 0.00014111338422431014,
+      "loss": 1.0422,
+      "step": 1997
+    },
+    {
+      "epoch": 0.3843786071565987,
+      "grad_norm": 2.248006582260132,
+      "learning_rate": 0.00014105657741150647,
+      "loss": 0.9068,
+      "step": 1998
+    },
+    {
+      "epoch": 0.38457098884186225,
+      "grad_norm": 2.191615343093872,
+      "learning_rate": 0.00014099975465912582,
+      "loss": 1.0592,
+      "step": 1999
+    },
+    {
+      "epoch": 0.38476337052712584,
+      "grad_norm": 2.2410755157470703,
+      "learning_rate": 0.00014094291598922877,
+      "loss": 1.0171,
+      "step": 2000
+    },
+    {
+      "epoch": 0.38495575221238937,
+      "grad_norm": 2.386687755584717,
+      "learning_rate": 0.00014088606142388202,
+      "loss": 1.0071,
+      "step": 2001
+    },
+    {
+      "epoch": 0.38514813389765296,
+      "grad_norm": 2.1394219398498535,
+      "learning_rate": 0.00014082919098515846,
+      "loss": 0.9921,
+      "step": 2002
+    },
+    {
+      "epoch": 0.3853405155829165,
+      "grad_norm": 2.080103874206543,
+      "learning_rate": 0.0001407723046951372,
+      "loss": 0.9578,
+      "step": 2003
+    },
+    {
+      "epoch": 0.3855328972681801,
+      "grad_norm": 2.5318148136138916,
+      "learning_rate": 0.0001407154025759034,
+      "loss": 0.9949,
+      "step": 2004
+    },
+    {
+      "epoch": 0.3857252789534436,
+      "grad_norm": 3.2196075916290283,
+      "learning_rate": 0.00014065848464954846,
+      "loss": 1.0664,
+      "step": 2005
+    },
+    {
+      "epoch": 0.3859176606387072,
+      "grad_norm": 2.2929482460021973,
+      "learning_rate": 0.00014060155093816986,
+      "loss": 0.9578,
+      "step": 2006
+    },
+    {
+      "epoch": 0.3861100423239708,
+      "grad_norm": 1.394026756286621,
+      "learning_rate": 0.00014054460146387124,
+      "loss": 0.7989,
+      "step": 2007
+    },
+    {
+      "epoch": 0.3863024240092343,
+      "grad_norm": 2.0632073879241943,
+      "learning_rate": 0.00014048763624876235,
+      "loss": 0.9274,
+      "step": 2008
+    },
+    {
+      "epoch": 0.3864948056944979,
+      "grad_norm": 2.864842653274536,
+      "learning_rate": 0.00014043065531495902,
+      "loss": 0.9462,
+      "step": 2009
+    },
+    {
+      "epoch": 0.38668718737976143,
+      "grad_norm": 2.595710039138794,
+      "learning_rate": 0.00014037365868458326,
+      "loss": 0.9567,
+      "step": 2010
+    },
+    {
+      "epoch": 0.386879569065025,
+      "grad_norm": 1.5200082063674927,
+      "learning_rate": 0.00014031664637976306,
+      "loss": 0.9839,
+      "step": 2011
+    },
+    {
+      "epoch": 0.38707195075028855,
+      "grad_norm": 1.2809231281280518,
+      "learning_rate": 0.0001402596184226326,
+      "loss": 0.7788,
+      "step": 2012
+    },
+    {
+      "epoch": 0.38726433243555214,
+      "grad_norm": 2.2823591232299805,
+      "learning_rate": 0.00014020257483533208,
+      "loss": 1.0329,
+      "step": 2013
+    },
+    {
+      "epoch": 0.38745671412081567,
+      "grad_norm": 2.1373653411865234,
+      "learning_rate": 0.0001401455156400078,
+      "loss": 1.0114,
+      "step": 2014
+    },
+    {
+      "epoch": 0.38764909580607926,
+      "grad_norm": 1.2135518789291382,
+      "learning_rate": 0.0001400884408588121,
+      "loss": 1.0154,
+      "step": 2015
+    },
+    {
+      "epoch": 0.38784147749134285,
+      "grad_norm": 1.9044760465621948,
+      "learning_rate": 0.0001400313505139034,
+      "loss": 1.0446,
+      "step": 2016
+    },
+    {
+      "epoch": 0.3880338591766064,
+      "grad_norm": 2.1260743141174316,
+      "learning_rate": 0.00013997424462744605,
+      "loss": 1.0142,
+      "step": 2017
+    },
+    {
+      "epoch": 0.38822624086186996,
+      "grad_norm": 2.378918170928955,
+      "learning_rate": 0.00013991712322161063,
+      "loss": 1.0333,
+      "step": 2018
+    },
+    {
+      "epoch": 0.3884186225471335,
+      "grad_norm": 2.5818562507629395,
+      "learning_rate": 0.00013985998631857357,
+      "loss": 1.0109,
+      "step": 2019
+    },
+    {
+      "epoch": 0.3886110042323971,
+      "grad_norm": 2.0968761444091797,
+      "learning_rate": 0.0001398028339405174,
+      "loss": 1.0374,
+      "step": 2020
+    },
+    {
+      "epoch": 0.3888033859176606,
+      "grad_norm": 1.329888105392456,
+      "learning_rate": 0.00013974566610963068,
+      "loss": 0.7736,
+      "step": 2021
+    },
+    {
+      "epoch": 0.3889957676029242,
+      "grad_norm": 2.6114044189453125,
+      "learning_rate": 0.00013968848284810786,
+      "loss": 0.9934,
+      "step": 2022
+    },
+    {
+      "epoch": 0.3891881492881878,
+      "grad_norm": 2.9008965492248535,
+      "learning_rate": 0.0001396312841781495,
+      "loss": 0.9687,
+      "step": 2023
+    },
+    {
+      "epoch": 0.3893805309734513,
+      "grad_norm": 2.812812328338623,
+      "learning_rate": 0.00013957407012196203,
+      "loss": 1.1018,
+      "step": 2024
+    },
+    {
+      "epoch": 0.3895729126587149,
+      "grad_norm": 1.5447956323623657,
+      "learning_rate": 0.000139516840701758,
+      "loss": 1.0667,
+      "step": 2025
+    },
+    {
+      "epoch": 0.38976529434397844,
+      "grad_norm": 2.4308981895446777,
+      "learning_rate": 0.0001394595959397558,
+      "loss": 1.0279,
+      "step": 2026
+    },
+    {
+      "epoch": 0.38995767602924203,
+      "grad_norm": 1.8810396194458008,
+      "learning_rate": 0.00013940233585817984,
+      "loss": 0.9934,
+      "step": 2027
+    },
+    {
+      "epoch": 0.39015005771450556,
+      "grad_norm": 2.7171528339385986,
+      "learning_rate": 0.0001393450604792604,
+      "loss": 0.9748,
+      "step": 2028
+    },
+    {
+      "epoch": 0.39034243939976915,
+      "grad_norm": 1.7177451848983765,
+      "learning_rate": 0.00013928776982523384,
+      "loss": 1.0314,
+      "step": 2029
+    },
+    {
+      "epoch": 0.3905348210850327,
+      "grad_norm": 2.2614712715148926,
+      "learning_rate": 0.00013923046391834228,
+      "loss": 1.0295,
+      "step": 2030
+    },
+    {
+      "epoch": 0.39072720277029627,
+      "grad_norm": 2.170705795288086,
+      "learning_rate": 0.0001391731427808339,
+      "loss": 0.9339,
+      "step": 2031
+    },
+    {
+      "epoch": 0.39091958445555985,
+      "grad_norm": 1.1974895000457764,
+      "learning_rate": 0.0001391158064349627,
+      "loss": 0.7818,
+      "step": 2032
+    },
+    {
+      "epoch": 0.3911119661408234,
+      "grad_norm": 2.2875094413757324,
+      "learning_rate": 0.00013905845490298866,
+      "loss": 0.9387,
+      "step": 2033
+    },
+    {
+      "epoch": 0.391304347826087,
+      "grad_norm": 2.4989264011383057,
+      "learning_rate": 0.0001390010882071776,
+      "loss": 1.0448,
+      "step": 2034
+    },
+    {
+      "epoch": 0.3914967295113505,
+      "grad_norm": 1.1905910968780518,
+      "learning_rate": 0.00013894370636980126,
+      "loss": 0.7691,
+      "step": 2035
+    },
+    {
+      "epoch": 0.3916891111966141,
+      "grad_norm": 2.1437323093414307,
+      "learning_rate": 0.00013888630941313727,
+      "loss": 0.9904,
+      "step": 2036
+    },
+    {
+      "epoch": 0.3918814928818776,
+      "grad_norm": 2.059305191040039,
+      "learning_rate": 0.000138828897359469,
+      "loss": 1.1072,
+      "step": 2037
+    },
+    {
+      "epoch": 0.3920738745671412,
+      "grad_norm": 1.524186372756958,
+      "learning_rate": 0.0001387714702310859,
+      "loss": 0.7894,
+      "step": 2038
+    },
+    {
+      "epoch": 0.39226625625240474,
+      "grad_norm": 1.2005305290222168,
+      "learning_rate": 0.00013871402805028314,
+      "loss": 1.0784,
+      "step": 2039
+    },
+    {
+      "epoch": 0.39245863793766833,
+      "grad_norm": 1.3715016841888428,
+      "learning_rate": 0.00013865657083936164,
+      "loss": 0.7531,
+      "step": 2040
+    },
+    {
+      "epoch": 0.3926510196229319,
+      "grad_norm": 1.1534337997436523,
+      "learning_rate": 0.00013859909862062842,
+      "loss": 1.0287,
+      "step": 2041
+    },
+    {
+      "epoch": 0.39284340130819545,
+      "grad_norm": 1.1162610054016113,
+      "learning_rate": 0.0001385416114163961,
+      "loss": 1.0575,
+      "step": 2042
+    },
+    {
+      "epoch": 0.39303578299345904,
+      "grad_norm": 2.200941324234009,
+      "learning_rate": 0.0001384841092489832,
+      "loss": 0.9481,
+      "step": 2043
+    },
+    {
+      "epoch": 0.39322816467872257,
+      "grad_norm": 2.064082145690918,
+      "learning_rate": 0.00013842659214071404,
+      "loss": 0.9922,
+      "step": 2044
+    },
+    {
+      "epoch": 0.39342054636398616,
+      "grad_norm": 2.0809133052825928,
+      "learning_rate": 0.00013836906011391878,
+      "loss": 1.0579,
+      "step": 2045
+    },
+    {
+      "epoch": 0.3936129280492497,
+      "grad_norm": 1.5401755571365356,
+      "learning_rate": 0.00013831151319093323,
+      "loss": 1.0343,
+      "step": 2046
+    },
+    {
+      "epoch": 0.3938053097345133,
+      "grad_norm": 1.3896150588989258,
+      "learning_rate": 0.0001382539513940992,
+      "loss": 0.9513,
+      "step": 2047
+    },
+    {
+      "epoch": 0.39399769141977686,
+      "grad_norm": 2.5641112327575684,
+      "learning_rate": 0.0001381963747457641,
+      "loss": 1.0449,
+      "step": 2048
+    },
+    {
+      "epoch": 0.3941900731050404,
+      "grad_norm": 2.221235513687134,
+      "learning_rate": 0.0001381387832682812,
+      "loss": 0.9658,
+      "step": 2049
+    },
+    {
+      "epoch": 0.394382454790304,
+      "grad_norm": 2.3585715293884277,
+      "learning_rate": 0.00013808117698400947,
+      "loss": 0.9864,
+      "step": 2050
+    },
+    {
+      "epoch": 0.3945748364755675,
+      "grad_norm": 1.0871765613555908,
+      "learning_rate": 0.00013802355591531365,
+      "loss": 1.0171,
+      "step": 2051
+    },
+    {
+      "epoch": 0.3947672181608311,
+      "grad_norm": 2.066192865371704,
+      "learning_rate": 0.00013796592008456427,
+      "loss": 0.956,
+      "step": 2052
+    },
+    {
+      "epoch": 0.39495959984609463,
+      "grad_norm": 1.2458277940750122,
+      "learning_rate": 0.00013790826951413747,
+      "loss": 0.7826,
+      "step": 2053
+    },
+    {
+      "epoch": 0.3951519815313582,
+      "grad_norm": 2.483954429626465,
+      "learning_rate": 0.00013785060422641526,
+      "loss": 0.9539,
+      "step": 2054
+    },
+    {
+      "epoch": 0.39534436321662175,
+      "grad_norm": 2.0900259017944336,
+      "learning_rate": 0.0001377929242437852,
+      "loss": 0.9775,
+      "step": 2055
+    },
+    {
+      "epoch": 0.39553674490188534,
+      "grad_norm": 2.2208642959594727,
+      "learning_rate": 0.00013773522958864076,
+      "loss": 0.955,
+      "step": 2056
+    },
+    {
+      "epoch": 0.3957291265871489,
+      "grad_norm": 0.8647241592407227,
+      "learning_rate": 0.0001376775202833809,
+      "loss": 1.0154,
+      "step": 2057
+    },
+    {
+      "epoch": 0.39592150827241246,
+      "grad_norm": 2.7264161109924316,
+      "learning_rate": 0.0001376197963504104,
+      "loss": 1.0846,
+      "step": 2058
+    },
+    {
+      "epoch": 0.39611388995767605,
+      "grad_norm": 2.2241272926330566,
+      "learning_rate": 0.00013756205781213963,
+      "loss": 0.9816,
+      "step": 2059
+    },
+    {
+      "epoch": 0.3963062716429396,
+      "grad_norm": 2.204205274581909,
+      "learning_rate": 0.00013750430469098478,
+      "loss": 0.98,
+      "step": 2060
+    },
+    {
+      "epoch": 0.39649865332820317,
+      "grad_norm": 2.9939072132110596,
+      "learning_rate": 0.00013744653700936753,
+      "loss": 1.0052,
+      "step": 2061
+    },
+    {
+      "epoch": 0.3966910350134667,
+      "grad_norm": 3.0493531227111816,
+      "learning_rate": 0.00013738875478971525,
+      "loss": 0.9532,
+      "step": 2062
+    },
+    {
+      "epoch": 0.3968834166987303,
+      "grad_norm": 1.8292204141616821,
+      "learning_rate": 0.00013733095805446106,
+      "loss": 1.1195,
+      "step": 2063
+    },
+    {
+      "epoch": 0.3970757983839938,
+      "grad_norm": 3.185791254043579,
+      "learning_rate": 0.0001372731468260436,
+      "loss": 0.9804,
+      "step": 2064
+    },
+    {
+      "epoch": 0.3972681800692574,
+      "grad_norm": 2.66125226020813,
+      "learning_rate": 0.0001372153211269072,
+      "loss": 1.0312,
+      "step": 2065
+    },
+    {
+      "epoch": 0.397460561754521,
+      "grad_norm": 2.4463164806365967,
+      "learning_rate": 0.00013715748097950175,
+      "loss": 0.9964,
+      "step": 2066
+    },
+    {
+      "epoch": 0.3976529434397845,
+      "grad_norm": 0.7782577276229858,
+      "learning_rate": 0.00013709962640628284,
+      "loss": 1.0444,
+      "step": 2067
+    },
+    {
+      "epoch": 0.3978453251250481,
+      "grad_norm": 1.7807798385620117,
+      "learning_rate": 0.00013704175742971158,
+      "loss": 1.0393,
+      "step": 2068
+    },
+    {
+      "epoch": 0.39803770681031164,
+      "grad_norm": 2.333333730697632,
+      "learning_rate": 0.0001369838740722547,
+      "loss": 0.9715,
+      "step": 2069
+    },
+    {
+      "epoch": 0.39823008849557523,
+      "grad_norm": 2.281967878341675,
+      "learning_rate": 0.0001369259763563845,
+      "loss": 0.9469,
+      "step": 2070
+    },
+    {
+      "epoch": 0.39842247018083876,
+      "grad_norm": 1.698356032371521,
+      "learning_rate": 0.0001368680643045789,
+      "loss": 0.9618,
+      "step": 2071
+    },
+    {
+      "epoch": 0.39861485186610235,
+      "grad_norm": 2.564318895339966,
+      "learning_rate": 0.0001368101379393213,
+      "loss": 0.9852,
+      "step": 2072
+    },
+    {
+      "epoch": 0.39880723355136594,
+      "grad_norm": 2.5909223556518555,
+      "learning_rate": 0.00013675219728310077,
+      "loss": 0.9746,
+      "step": 2073
+    },
+    {
+      "epoch": 0.39899961523662947,
+      "grad_norm": 2.314802646636963,
+      "learning_rate": 0.00013669424235841185,
+      "loss": 0.9875,
+      "step": 2074
+    },
+    {
+      "epoch": 0.39919199692189306,
+      "grad_norm": 0.959285318851471,
+      "learning_rate": 0.00013663627318775457,
+      "loss": 1.021,
+      "step": 2075
+    },
+    {
+      "epoch": 0.3993843786071566,
+      "grad_norm": 2.338844060897827,
+      "learning_rate": 0.00013657828979363467,
+      "loss": 1.0364,
+      "step": 2076
+    },
+    {
+      "epoch": 0.3995767602924202,
+      "grad_norm": 2.3757286071777344,
+      "learning_rate": 0.00013652029219856322,
+      "loss": 1.0536,
+      "step": 2077
+    },
+    {
+      "epoch": 0.3997691419776837,
+      "grad_norm": 1.4227973222732544,
+      "learning_rate": 0.00013646228042505694,
+      "loss": 0.8583,
+      "step": 2078
+    },
+    {
+      "epoch": 0.3999615236629473,
+      "grad_norm": 2.5857388973236084,
+      "learning_rate": 0.0001364042544956379,
+      "loss": 0.9643,
+      "step": 2079
+    },
+    {
+      "epoch": 0.4001539053482108,
+      "grad_norm": 2.3812410831451416,
+      "learning_rate": 0.00013634621443283387,
+      "loss": 1.011,
+      "step": 2080
+    },
+    {
+      "epoch": 0.4003462870334744,
+      "grad_norm": 1.3451483249664307,
+      "learning_rate": 0.00013628816025917798,
+      "loss": 1.0179,
+      "step": 2081
+    },
+    {
+      "epoch": 0.400538668718738,
+      "grad_norm": 2.1600303649902344,
+      "learning_rate": 0.00013623009199720883,
+      "loss": 1.0139,
+      "step": 2082
+    },
+    {
+      "epoch": 0.40073105040400153,
+      "grad_norm": 2.418637990951538,
+      "learning_rate": 0.00013617200966947052,
+      "loss": 1.066,
+      "step": 2083
+    },
+    {
+      "epoch": 0.4009234320892651,
+      "grad_norm": 2.5478858947753906,
+      "learning_rate": 0.00013611391329851263,
+      "loss": 1.0539,
+      "step": 2084
+    },
+    {
+      "epoch": 0.40111581377452865,
+      "grad_norm": 2.385664224624634,
+      "learning_rate": 0.00013605580290689014,
+      "loss": 1.0171,
+      "step": 2085
+    },
+    {
+      "epoch": 0.40130819545979224,
+      "grad_norm": 2.4962332248687744,
+      "learning_rate": 0.00013599767851716352,
+      "loss": 1.0486,
+      "step": 2086
+    },
+    {
+      "epoch": 0.40150057714505577,
+      "grad_norm": 2.8469555377960205,
+      "learning_rate": 0.00013593954015189867,
+      "loss": 1.0173,
+      "step": 2087
+    },
+    {
+      "epoch": 0.40169295883031936,
+      "grad_norm": 2.615797996520996,
+      "learning_rate": 0.0001358813878336669,
+      "loss": 1.024,
+      "step": 2088
+    },
+    {
+      "epoch": 0.4018853405155829,
+      "grad_norm": 3.0085480213165283,
+      "learning_rate": 0.00013582322158504494,
+      "loss": 0.9833,
+      "step": 2089
+    },
+    {
+      "epoch": 0.4020777222008465,
+      "grad_norm": 2.651336908340454,
+      "learning_rate": 0.00013576504142861495,
+      "loss": 1.0276,
+      "step": 2090
+    },
+    {
+      "epoch": 0.40227010388611006,
+      "grad_norm": 2.3057451248168945,
+      "learning_rate": 0.00013570684738696445,
+      "loss": 1.0765,
+      "step": 2091
+    },
+    {
+      "epoch": 0.4024624855713736,
+      "grad_norm": 2.5263404846191406,
+      "learning_rate": 0.0001356486394826863,
+      "loss": 0.9377,
+      "step": 2092
+    },
+    {
+      "epoch": 0.4026548672566372,
+      "grad_norm": 2.1794450283050537,
+      "learning_rate": 0.00013559041773837898,
+      "loss": 1.0872,
+      "step": 2093
+    },
+    {
+      "epoch": 0.4028472489419007,
+      "grad_norm": 2.184370517730713,
+      "learning_rate": 0.00013553218217664603,
+      "loss": 0.9952,
+      "step": 2094
+    },
+    {
+      "epoch": 0.4030396306271643,
+      "grad_norm": 2.3743715286254883,
+      "learning_rate": 0.00013547393282009655,
+      "loss": 1.0405,
+      "step": 2095
+    },
+    {
+      "epoch": 0.40323201231242783,
+      "grad_norm": 1.9796000719070435,
+      "learning_rate": 0.00013541566969134497,
+      "loss": 0.9262,
+      "step": 2096
+    },
+    {
+      "epoch": 0.4034243939976914,
+      "grad_norm": 2.4818122386932373,
+      "learning_rate": 0.000135357392813011,
+      "loss": 1.0504,
+      "step": 2097
+    },
+    {
+      "epoch": 0.403616775682955,
+      "grad_norm": 2.4162020683288574,
+      "learning_rate": 0.00013529910220771977,
+      "loss": 0.9513,
+      "step": 2098
+    },
+    {
+      "epoch": 0.40380915736821854,
+      "grad_norm": 2.458404779434204,
+      "learning_rate": 0.00013524079789810163,
+      "loss": 0.9506,
+      "step": 2099
+    },
+    {
+      "epoch": 0.40400153905348213,
+      "grad_norm": 2.3652219772338867,
+      "learning_rate": 0.00013518247990679241,
+      "loss": 0.9728,
+      "step": 2100
+    },
+    {
+      "epoch": 0.40419392073874566,
+      "grad_norm": 2.2657647132873535,
+      "learning_rate": 0.00013512414825643312,
+      "loss": 1.0329,
+      "step": 2101
+    },
+    {
+      "epoch": 0.40438630242400925,
+      "grad_norm": 2.310148000717163,
+      "learning_rate": 0.0001350658029696701,
+      "loss": 0.9677,
+      "step": 2102
+    },
+    {
+      "epoch": 0.4045786841092728,
+      "grad_norm": 1.6359156370162964,
+      "learning_rate": 0.00013500744406915504,
+      "loss": 0.9324,
+      "step": 2103
+    },
+    {
+      "epoch": 0.40477106579453637,
+      "grad_norm": 2.2912919521331787,
+      "learning_rate": 0.00013494907157754485,
+      "loss": 1.0359,
+      "step": 2104
+    },
+    {
+      "epoch": 0.4049634474797999,
+      "grad_norm": 2.1350061893463135,
+      "learning_rate": 0.0001348906855175017,
+      "loss": 0.9876,
+      "step": 2105
+    },
+    {
+      "epoch": 0.4051558291650635,
+      "grad_norm": 2.3606483936309814,
+      "learning_rate": 0.00013483228591169315,
+      "loss": 1.0008,
+      "step": 2106
+    },
+    {
+      "epoch": 0.4053482108503271,
+      "grad_norm": 2.2295539379119873,
+      "learning_rate": 0.0001347738727827919,
+      "loss": 0.9061,
+      "step": 2107
+    },
+    {
+      "epoch": 0.4055405925355906,
+      "grad_norm": 1.702531099319458,
+      "learning_rate": 0.0001347154461534759,
+      "loss": 0.9204,
+      "step": 2108
+    },
+    {
+      "epoch": 0.4057329742208542,
+      "grad_norm": 2.2621521949768066,
+      "learning_rate": 0.00013465700604642846,
+      "loss": 1.0034,
+      "step": 2109
+    },
+    {
+      "epoch": 0.4059253559061177,
+      "grad_norm": 1.948228120803833,
+      "learning_rate": 0.000134598552484338,
+      "loss": 1.0133,
+      "step": 2110
+    },
+    {
+      "epoch": 0.4061177375913813,
+      "grad_norm": 1.2004185914993286,
+      "learning_rate": 0.00013454008548989817,
+      "loss": 0.7433,
+      "step": 2111
+    },
+    {
+      "epoch": 0.40631011927664484,
+      "grad_norm": 2.4148664474487305,
+      "learning_rate": 0.0001344816050858079,
+      "loss": 0.9723,
+      "step": 2112
+    },
+    {
+      "epoch": 0.40650250096190843,
+      "grad_norm": 1.9818840026855469,
+      "learning_rate": 0.00013442311129477133,
+      "loss": 1.0762,
+      "step": 2113
+    },
+    {
+      "epoch": 0.40669488264717196,
+      "grad_norm": 1.270645022392273,
+      "learning_rate": 0.0001343646041394977,
+      "loss": 0.7615,
+      "step": 2114
+    },
+    {
+      "epoch": 0.40688726433243555,
+      "grad_norm": 1.341287612915039,
+      "learning_rate": 0.00013430608364270157,
+      "loss": 1.0328,
+      "step": 2115
+    },
+    {
+      "epoch": 0.40707964601769914,
+      "grad_norm": 2.0091190338134766,
+      "learning_rate": 0.00013424754982710254,
+      "loss": 1.0557,
+      "step": 2116
+    },
+    {
+      "epoch": 0.40727202770296267,
+      "grad_norm": 2.366304397583008,
+      "learning_rate": 0.0001341890027154255,
+      "loss": 0.9934,
+      "step": 2117
+    },
+    {
+      "epoch": 0.40746440938822626,
+      "grad_norm": 2.524237632751465,
+      "learning_rate": 0.00013413044233040043,
+      "loss": 1.0429,
+      "step": 2118
+    },
+    {
+      "epoch": 0.4076567910734898,
+      "grad_norm": 2.5381581783294678,
+      "learning_rate": 0.00013407186869476254,
+      "loss": 1.0181,
+      "step": 2119
+    },
+    {
+      "epoch": 0.4078491727587534,
+      "grad_norm": 1.0900909900665283,
+      "learning_rate": 0.00013401328183125207,
+      "loss": 0.9566,
+      "step": 2120
+    },
+    {
+      "epoch": 0.4080415544440169,
+      "grad_norm": 0.7273239493370056,
+      "learning_rate": 0.0001339546817626145,
+      "loss": 1.0129,
+      "step": 2121
+    },
+    {
+      "epoch": 0.4082339361292805,
+      "grad_norm": 2.276264190673828,
+      "learning_rate": 0.00013389606851160036,
+      "loss": 1.014,
+      "step": 2122
+    },
+    {
+      "epoch": 0.4084263178145441,
+      "grad_norm": 2.8681108951568604,
+      "learning_rate": 0.00013383744210096536,
+      "loss": 1.0117,
+      "step": 2123
+    },
+    {
+      "epoch": 0.4086186994998076,
+      "grad_norm": 1.9943767786026,
+      "learning_rate": 0.00013377880255347027,
+      "loss": 1.0236,
+      "step": 2124
+    },
+    {
+      "epoch": 0.4088110811850712,
+      "grad_norm": 1.822360634803772,
+      "learning_rate": 0.00013372014989188099,
+      "loss": 1.0399,
+      "step": 2125
+    },
+    {
+      "epoch": 0.40900346287033473,
+      "grad_norm": 1.7232239246368408,
+      "learning_rate": 0.0001336614841389685,
+      "loss": 0.9349,
+      "step": 2126
+    },
+    {
+      "epoch": 0.4091958445555983,
+      "grad_norm": 2.1910624504089355,
+      "learning_rate": 0.00013360280531750886,
+      "loss": 0.9805,
+      "step": 2127
+    },
+    {
+      "epoch": 0.40938822624086185,
+      "grad_norm": 1.5429714918136597,
+      "learning_rate": 0.00013354411345028325,
+      "loss": 1.1543,
+      "step": 2128
+    },
+    {
+      "epoch": 0.40958060792612544,
+      "grad_norm": 1.9436414241790771,
+      "learning_rate": 0.0001334854085600778,
+      "loss": 1.0781,
+      "step": 2129
+    },
+    {
+      "epoch": 0.40977298961138897,
+      "grad_norm": 1.7596856355667114,
+      "learning_rate": 0.00013342669066968385,
+      "loss": 0.9904,
+      "step": 2130
+    },
+    {
+      "epoch": 0.40996537129665256,
+      "grad_norm": 1.788402795791626,
+      "learning_rate": 0.0001333679598018976,
+      "loss": 1.0121,
+      "step": 2131
+    },
+    {
+      "epoch": 0.41015775298191615,
+      "grad_norm": 2.2617104053497314,
+      "learning_rate": 0.00013330921597952056,
+      "loss": 1.0212,
+      "step": 2132
+    },
+    {
+      "epoch": 0.4103501346671797,
+      "grad_norm": 2.3057329654693604,
+      "learning_rate": 0.00013325045922535895,
+      "loss": 0.9818,
+      "step": 2133
+    },
+    {
+      "epoch": 0.41054251635244327,
+      "grad_norm": 2.5402636528015137,
+      "learning_rate": 0.00013319168956222422,
+      "loss": 1.0474,
+      "step": 2134
+    },
+    {
+      "epoch": 0.4107348980377068,
+      "grad_norm": 2.221797466278076,
+      "learning_rate": 0.0001331329070129328,
+      "loss": 0.9698,
+      "step": 2135
+    },
+    {
+      "epoch": 0.4109272797229704,
+      "grad_norm": 2.6184704303741455,
+      "learning_rate": 0.00013307411160030607,
+      "loss": 1.0028,
+      "step": 2136
+    },
+    {
+      "epoch": 0.4111196614082339,
+      "grad_norm": 2.741410255432129,
+      "learning_rate": 0.00013301530334717046,
+      "loss": 1.0229,
+      "step": 2137
+    },
+    {
+      "epoch": 0.4113120430934975,
+      "grad_norm": 0.9947250485420227,
+      "learning_rate": 0.0001329564822763573,
+      "loss": 1.0709,
+      "step": 2138
+    },
+    {
+      "epoch": 0.41150442477876104,
+      "grad_norm": 1.4335472583770752,
+      "learning_rate": 0.000132897648410703,
+      "loss": 1.0094,
+      "step": 2139
+    },
+    {
+      "epoch": 0.4116968064640246,
+      "grad_norm": 2.256047487258911,
+      "learning_rate": 0.00013283880177304892,
+      "loss": 1.0335,
+      "step": 2140
+    },
+    {
+      "epoch": 0.4118891881492882,
+      "grad_norm": 2.4210164546966553,
+      "learning_rate": 0.0001327799423862413,
+      "loss": 0.9901,
+      "step": 2141
+    },
+    {
+      "epoch": 0.41208156983455174,
+      "grad_norm": 1.3404836654663086,
+      "learning_rate": 0.00013272107027313142,
+      "loss": 0.7715,
+      "step": 2142
+    },
+    {
+      "epoch": 0.41227395151981533,
+      "grad_norm": 2.4302797317504883,
+      "learning_rate": 0.00013266218545657541,
+      "loss": 1.0257,
+      "step": 2143
+    },
+    {
+      "epoch": 0.41246633320507886,
+      "grad_norm": 2.8346972465515137,
+      "learning_rate": 0.0001326032879594344,
+      "loss": 1.045,
+      "step": 2144
+    },
+    {
+      "epoch": 0.41265871489034245,
+      "grad_norm": 2.1370692253112793,
+      "learning_rate": 0.00013254437780457447,
+      "loss": 0.9626,
+      "step": 2145
+    },
+    {
+      "epoch": 0.412851096575606,
+      "grad_norm": 2.162801742553711,
+      "learning_rate": 0.00013248545501486653,
+      "loss": 0.9931,
+      "step": 2146
+    },
+    {
+      "epoch": 0.41304347826086957,
+      "grad_norm": 2.732614278793335,
+      "learning_rate": 0.00013242651961318646,
+      "loss": 1.0244,
+      "step": 2147
+    },
+    {
+      "epoch": 0.41323585994613315,
+      "grad_norm": 2.389601707458496,
+      "learning_rate": 0.000132367571622415,
+      "loss": 0.9872,
+      "step": 2148
+    },
+    {
+      "epoch": 0.4134282416313967,
+      "grad_norm": 2.2313787937164307,
+      "learning_rate": 0.00013230861106543776,
+      "loss": 1.0734,
+      "step": 2149
+    },
+    {
+      "epoch": 0.4136206233166603,
+      "grad_norm": 2.318730592727661,
+      "learning_rate": 0.0001322496379651453,
+      "loss": 1.0636,
+      "step": 2150
+    },
+    {
+      "epoch": 0.4138130050019238,
+      "grad_norm": 2.5190744400024414,
+      "learning_rate": 0.000132190652344433,
+      "loss": 1.0241,
+      "step": 2151
+    },
+    {
+      "epoch": 0.4140053866871874,
+      "grad_norm": 1.9321190118789673,
+      "learning_rate": 0.0001321316542262011,
+      "loss": 0.9716,
+      "step": 2152
+    },
+    {
+      "epoch": 0.4141977683724509,
+      "grad_norm": 2.268319606781006,
+      "learning_rate": 0.0001320726436333547,
+      "loss": 1.0096,
+      "step": 2153
+    },
+    {
+      "epoch": 0.4143901500577145,
+      "grad_norm": 1.315738320350647,
+      "learning_rate": 0.00013201362058880376,
+      "loss": 0.8163,
+      "step": 2154
+    },
+    {
+      "epoch": 0.41458253174297804,
+      "grad_norm": 2.7964563369750977,
+      "learning_rate": 0.00013195458511546307,
+      "loss": 0.9781,
+      "step": 2155
+    },
+    {
+      "epoch": 0.41477491342824163,
+      "grad_norm": 1.3313992023468018,
+      "learning_rate": 0.00013189553723625216,
+      "loss": 0.9971,
+      "step": 2156
+    },
+    {
+      "epoch": 0.4149672951135052,
+      "grad_norm": 2.656071186065674,
+      "learning_rate": 0.0001318364769740955,
+      "loss": 1.0612,
+      "step": 2157
+    },
+    {
+      "epoch": 0.41515967679876875,
+      "grad_norm": 1.278800129890442,
+      "learning_rate": 0.00013177740435192233,
+      "loss": 0.9933,
+      "step": 2158
+    },
+    {
+      "epoch": 0.41535205848403234,
+      "grad_norm": 2.1393814086914062,
+      "learning_rate": 0.00013171831939266668,
+      "loss": 1.0146,
+      "step": 2159
+    },
+    {
+      "epoch": 0.41554444016929587,
+      "grad_norm": 2.7138571739196777,
+      "learning_rate": 0.00013165922211926732,
+      "loss": 1.0826,
+      "step": 2160
+    },
+    {
+      "epoch": 0.41573682185455946,
+      "grad_norm": 2.179614782333374,
+      "learning_rate": 0.0001316001125546679,
+      "loss": 0.961,
+      "step": 2161
+    },
+    {
+      "epoch": 0.415929203539823,
+      "grad_norm": 2.2457833290100098,
+      "learning_rate": 0.00013154099072181676,
+      "loss": 1.0,
+      "step": 2162
+    },
+    {
+      "epoch": 0.4161215852250866,
+      "grad_norm": 2.460271120071411,
+      "learning_rate": 0.00013148185664366702,
+      "loss": 1.0207,
+      "step": 2163
+    },
+    {
+      "epoch": 0.4163139669103501,
+      "grad_norm": 2.5040926933288574,
+      "learning_rate": 0.0001314227103431766,
+      "loss": 0.9966,
+      "step": 2164
+    },
+    {
+      "epoch": 0.4165063485956137,
+      "grad_norm": 1.3953497409820557,
+      "learning_rate": 0.0001313635518433081,
+      "loss": 0.8041,
+      "step": 2165
+    },
+    {
+      "epoch": 0.4166987302808773,
+      "grad_norm": 2.4269542694091797,
+      "learning_rate": 0.00013130438116702888,
+      "loss": 1.0051,
+      "step": 2166
+    },
+    {
+      "epoch": 0.4168911119661408,
+      "grad_norm": 2.938689708709717,
+      "learning_rate": 0.00013124519833731106,
+      "loss": 1.0199,
+      "step": 2167
+    },
+    {
+      "epoch": 0.4170834936514044,
+      "grad_norm": 2.2541141510009766,
+      "learning_rate": 0.00013118600337713146,
+      "loss": 0.9614,
+      "step": 2168
+    },
+    {
+      "epoch": 0.41727587533666793,
+      "grad_norm": 1.653273105621338,
+      "learning_rate": 0.00013112679630947156,
+      "loss": 1.051,
+      "step": 2169
+    },
+    {
+      "epoch": 0.4174682570219315,
+      "grad_norm": 2.5672214031219482,
+      "learning_rate": 0.00013106757715731758,
+      "loss": 1.0731,
+      "step": 2170
+    },
+    {
+      "epoch": 0.41766063870719505,
+      "grad_norm": 2.8445069789886475,
+      "learning_rate": 0.0001310083459436605,
+      "loss": 0.9968,
+      "step": 2171
+    },
+    {
+      "epoch": 0.41785302039245864,
+      "grad_norm": 1.1662490367889404,
+      "learning_rate": 0.00013094910269149587,
+      "loss": 0.9708,
+      "step": 2172
+    },
+    {
+      "epoch": 0.41804540207772223,
+      "grad_norm": 2.0975537300109863,
+      "learning_rate": 0.00013088984742382396,
+      "loss": 0.9841,
+      "step": 2173
+    },
+    {
+      "epoch": 0.41823778376298576,
+      "grad_norm": 1.9598615169525146,
+      "learning_rate": 0.0001308305801636497,
+      "loss": 1.0014,
+      "step": 2174
+    },
+    {
+      "epoch": 0.41843016544824935,
+      "grad_norm": 2.709141254425049,
+      "learning_rate": 0.00013077130093398274,
+      "loss": 1.0353,
+      "step": 2175
+    },
+    {
+      "epoch": 0.4186225471335129,
+      "grad_norm": 2.3856029510498047,
+      "learning_rate": 0.00013071200975783723,
+      "loss": 0.9563,
+      "step": 2176
+    },
+    {
+      "epoch": 0.41881492881877647,
+      "grad_norm": 2.806976318359375,
+      "learning_rate": 0.00013065270665823205,
+      "loss": 1.0653,
+      "step": 2177
+    },
+    {
+      "epoch": 0.41900731050404,
+      "grad_norm": 2.2653510570526123,
+      "learning_rate": 0.00013059339165819083,
+      "loss": 1.0074,
+      "step": 2178
+    },
+    {
+      "epoch": 0.4191996921893036,
+      "grad_norm": 2.569941997528076,
+      "learning_rate": 0.00013053406478074156,
+      "loss": 1.0395,
+      "step": 2179
+    },
+    {
+      "epoch": 0.4193920738745671,
+      "grad_norm": 2.4211506843566895,
+      "learning_rate": 0.00013047472604891703,
+      "loss": 1.0396,
+      "step": 2180
+    },
+    {
+      "epoch": 0.4195844555598307,
+      "grad_norm": 1.0143215656280518,
+      "learning_rate": 0.00013041537548575457,
+      "loss": 0.9359,
+      "step": 2181
+    },
+    {
+      "epoch": 0.4197768372450943,
+      "grad_norm": 2.1877431869506836,
+      "learning_rate": 0.00013035601311429612,
+      "loss": 1.0495,
+      "step": 2182
+    },
+    {
+      "epoch": 0.4199692189303578,
+      "grad_norm": 2.5096518993377686,
+      "learning_rate": 0.00013029663895758813,
+      "loss": 1.0429,
+      "step": 2183
+    },
+    {
+      "epoch": 0.4201616006156214,
+      "grad_norm": 2.4985668659210205,
+      "learning_rate": 0.0001302372530386818,
+      "loss": 0.9521,
+      "step": 2184
+    },
+    {
+      "epoch": 0.42035398230088494,
+      "grad_norm": 1.221869945526123,
+      "learning_rate": 0.00013017785538063276,
+      "loss": 1.0107,
+      "step": 2185
+    },
+    {
+      "epoch": 0.42054636398614853,
+      "grad_norm": 2.363023281097412,
+      "learning_rate": 0.0001301184460065012,
+      "loss": 1.0572,
+      "step": 2186
+    },
+    {
+      "epoch": 0.42073874567141206,
+      "grad_norm": 2.9681644439697266,
+      "learning_rate": 0.0001300590249393519,
+      "loss": 1.0401,
+      "step": 2187
+    },
+    {
+      "epoch": 0.42093112735667565,
+      "grad_norm": 2.411137342453003,
+      "learning_rate": 0.00012999959220225414,
+      "loss": 1.0703,
+      "step": 2188
+    },
+    {
+      "epoch": 0.4211235090419392,
+      "grad_norm": 1.4854010343551636,
+      "learning_rate": 0.00012994014781828178,
+      "loss": 1.0594,
+      "step": 2189
+    },
+    {
+      "epoch": 0.42131589072720277,
+      "grad_norm": 1.3616106510162354,
+      "learning_rate": 0.00012988069181051314,
+      "loss": 0.7644,
+      "step": 2190
+    },
+    {
+      "epoch": 0.42150827241246636,
+      "grad_norm": 1.6271917819976807,
+      "learning_rate": 0.00012982122420203114,
+      "loss": 1.0078,
+      "step": 2191
+    },
+    {
+      "epoch": 0.4217006540977299,
+      "grad_norm": 2.4211161136627197,
+      "learning_rate": 0.00012976174501592313,
+      "loss": 1.0101,
+      "step": 2192
+    },
+    {
+      "epoch": 0.4218930357829935,
+      "grad_norm": 2.2215209007263184,
+      "learning_rate": 0.00012970225427528097,
+      "loss": 1.0136,
+      "step": 2193
+    },
+    {
+      "epoch": 0.422085417468257,
+      "grad_norm": 1.8052726984024048,
+      "learning_rate": 0.00012964275200320105,
+      "loss": 1.0262,
+      "step": 2194
+    },
+    {
+      "epoch": 0.4222777991535206,
+      "grad_norm": 0.7958564758300781,
+      "learning_rate": 0.00012958323822278412,
+      "loss": 1.0199,
+      "step": 2195
+    },
+    {
+      "epoch": 0.4224701808387841,
+      "grad_norm": 2.7244322299957275,
+      "learning_rate": 0.00012952371295713556,
+      "loss": 1.0394,
+      "step": 2196
+    },
+    {
+      "epoch": 0.4226625625240477,
+      "grad_norm": 2.440962076187134,
+      "learning_rate": 0.00012946417622936512,
+      "loss": 0.9651,
+      "step": 2197
+    },
+    {
+      "epoch": 0.4228549442093113,
+      "grad_norm": 3.0260767936706543,
+      "learning_rate": 0.00012940462806258695,
+      "loss": 1.0148,
+      "step": 2198
+    },
+    {
+      "epoch": 0.42304732589457483,
+      "grad_norm": 2.2724769115448,
+      "learning_rate": 0.00012934506847991976,
+      "loss": 1.0149,
+      "step": 2199
+    },
+    {
+      "epoch": 0.4232397075798384,
+      "grad_norm": 2.5502030849456787,
+      "learning_rate": 0.0001292854975044866,
+      "loss": 0.9433,
+      "step": 2200
+    },
+    {
+      "epoch": 0.42343208926510195,
+      "grad_norm": 2.3531219959259033,
+      "learning_rate": 0.00012922591515941498,
+      "loss": 1.0041,
+      "step": 2201
+    },
+    {
+      "epoch": 0.42362447095036554,
+      "grad_norm": 2.3544809818267822,
+      "learning_rate": 0.00012916632146783684,
+      "loss": 0.9623,
+      "step": 2202
+    },
+    {
+      "epoch": 0.42381685263562907,
+      "grad_norm": 2.2975785732269287,
+      "learning_rate": 0.0001291067164528884,
+      "loss": 1.078,
+      "step": 2203
+    },
+    {
+      "epoch": 0.42400923432089266,
+      "grad_norm": 2.602177619934082,
+      "learning_rate": 0.00012904710013771054,
+      "loss": 0.9282,
+      "step": 2204
+    },
+    {
+      "epoch": 0.4242016160061562,
+      "grad_norm": 2.551271677017212,
+      "learning_rate": 0.00012898747254544825,
+      "loss": 1.0536,
+      "step": 2205
+    },
+    {
+      "epoch": 0.4243939976914198,
+      "grad_norm": 1.9357308149337769,
+      "learning_rate": 0.00012892783369925105,
+      "loss": 0.9535,
+      "step": 2206
+    },
+    {
+      "epoch": 0.42458637937668336,
+      "grad_norm": 1.2973051071166992,
+      "learning_rate": 0.00012886818362227282,
+      "loss": 1.0126,
+      "step": 2207
+    },
+    {
+      "epoch": 0.4247787610619469,
+      "grad_norm": 1.82277512550354,
+      "learning_rate": 0.00012880852233767174,
+      "loss": 1.0176,
+      "step": 2208
+    },
+    {
+      "epoch": 0.4249711427472105,
+      "grad_norm": 2.873512029647827,
+      "learning_rate": 0.0001287488498686104,
+      "loss": 1.0428,
+      "step": 2209
+    },
+    {
+      "epoch": 0.425163524432474,
+      "grad_norm": 2.5375168323516846,
+      "learning_rate": 0.0001286891662382556,
+      "loss": 1.0265,
+      "step": 2210
+    },
+    {
+      "epoch": 0.4253559061177376,
+      "grad_norm": 2.4245636463165283,
+      "learning_rate": 0.00012862947146977875,
+      "loss": 0.9881,
+      "step": 2211
+    },
+    {
+      "epoch": 0.42554828780300114,
+      "grad_norm": 1.7671163082122803,
+      "learning_rate": 0.0001285697655863553,
+      "loss": 1.0185,
+      "step": 2212
+    },
+    {
+      "epoch": 0.4257406694882647,
+      "grad_norm": 1.1579701900482178,
+      "learning_rate": 0.00012851004861116517,
+      "loss": 1.059,
+      "step": 2213
+    },
+    {
+      "epoch": 0.42593305117352825,
+      "grad_norm": 2.415734052658081,
+      "learning_rate": 0.00012845032056739257,
+      "loss": 1.0306,
+      "step": 2214
+    },
+    {
+      "epoch": 0.42612543285879184,
+      "grad_norm": 2.1535348892211914,
+      "learning_rate": 0.00012839058147822593,
+      "loss": 1.003,
+      "step": 2215
+    },
+    {
+      "epoch": 0.42631781454405543,
+      "grad_norm": 2.569878101348877,
+      "learning_rate": 0.00012833083136685802,
+      "loss": 1.0137,
+      "step": 2216
+    },
+    {
+      "epoch": 0.42651019622931896,
+      "grad_norm": 2.2591562271118164,
+      "learning_rate": 0.00012827107025648594,
+      "loss": 1.0058,
+      "step": 2217
+    },
+    {
+      "epoch": 0.42670257791458255,
+      "grad_norm": 2.0434911251068115,
+      "learning_rate": 0.000128211298170311,
+      "loss": 1.0196,
+      "step": 2218
+    },
+    {
+      "epoch": 0.4268949595998461,
+      "grad_norm": 2.7149224281311035,
+      "learning_rate": 0.00012815151513153873,
+      "loss": 1.0154,
+      "step": 2219
+    },
+    {
+      "epoch": 0.42708734128510967,
+      "grad_norm": 2.461278200149536,
+      "learning_rate": 0.00012809172116337903,
+      "loss": 1.0627,
+      "step": 2220
+    },
+    {
+      "epoch": 0.4272797229703732,
+      "grad_norm": 2.693979024887085,
+      "learning_rate": 0.00012803191628904593,
+      "loss": 1.0065,
+      "step": 2221
+    },
+    {
+      "epoch": 0.4274721046556368,
+      "grad_norm": 2.3322811126708984,
+      "learning_rate": 0.0001279721005317578,
+      "loss": 0.9562,
+      "step": 2222
+    },
+    {
+      "epoch": 0.4276644863409004,
+      "grad_norm": 2.3082587718963623,
+      "learning_rate": 0.00012791227391473705,
+      "loss": 1.0273,
+      "step": 2223
+    },
+    {
+      "epoch": 0.4278568680261639,
+      "grad_norm": 2.42539381980896,
+      "learning_rate": 0.0001278524364612106,
+      "loss": 0.9662,
+      "step": 2224
+    },
+    {
+      "epoch": 0.4280492497114275,
+      "grad_norm": 2.318173408508301,
+      "learning_rate": 0.0001277925881944093,
+      "loss": 1.0186,
+      "step": 2225
+    },
+    {
+      "epoch": 0.428241631396691,
+      "grad_norm": 2.335205078125,
+      "learning_rate": 0.00012773272913756833,
+      "loss": 1.0425,
+      "step": 2226
+    },
+    {
+      "epoch": 0.4284340130819546,
+      "grad_norm": 2.3724725246429443,
+      "learning_rate": 0.00012767285931392704,
+      "loss": 1.0821,
+      "step": 2227
+    },
+    {
+      "epoch": 0.42862639476721814,
+      "grad_norm": 2.587892770767212,
+      "learning_rate": 0.000127612978746729,
+      "loss": 1.007,
+      "step": 2228
+    },
+    {
+      "epoch": 0.42881877645248173,
+      "grad_norm": 0.9510388374328613,
+      "learning_rate": 0.0001275530874592218,
+      "loss": 1.0668,
+      "step": 2229
+    },
+    {
+      "epoch": 0.42901115813774526,
+      "grad_norm": 1.8241522312164307,
+      "learning_rate": 0.0001274931854746574,
+      "loss": 1.0139,
+      "step": 2230
+    },
+    {
+      "epoch": 0.42920353982300885,
+      "grad_norm": 1.6252360343933105,
+      "learning_rate": 0.0001274332728162918,
+      "loss": 0.9455,
+      "step": 2231
+    },
+    {
+      "epoch": 0.42939592150827244,
+      "grad_norm": 2.556814670562744,
+      "learning_rate": 0.0001273733495073851,
+      "loss": 0.9981,
+      "step": 2232
+    },
+    {
+      "epoch": 0.42958830319353597,
+      "grad_norm": 1.3560316562652588,
+      "learning_rate": 0.00012731341557120168,
+      "loss": 0.8111,
+      "step": 2233
+    },
+    {
+      "epoch": 0.42978068487879956,
+      "grad_norm": 1.3201204538345337,
+      "learning_rate": 0.00012725347103100987,
+      "loss": 0.8191,
+      "step": 2234
+    },
+    {
+      "epoch": 0.4299730665640631,
+      "grad_norm": 1.909480094909668,
+      "learning_rate": 0.0001271935159100823,
+      "loss": 1.0591,
+      "step": 2235
+    },
+    {
+      "epoch": 0.4301654482493267,
+      "grad_norm": 1.6788324117660522,
+      "learning_rate": 0.00012713355023169547,
+      "loss": 0.8743,
+      "step": 2236
+    },
+    {
+      "epoch": 0.4303578299345902,
+      "grad_norm": 2.7671430110931396,
+      "learning_rate": 0.00012707357401913022,
+      "loss": 1.0478,
+      "step": 2237
+    },
+    {
+      "epoch": 0.4305502116198538,
+      "grad_norm": 1.6775882244110107,
+      "learning_rate": 0.0001270135872956714,
+      "loss": 1.005,
+      "step": 2238
+    },
+    {
+      "epoch": 0.4307425933051173,
+      "grad_norm": 1.1901050806045532,
+      "learning_rate": 0.00012695359008460785,
+      "loss": 1.0232,
+      "step": 2239
+    },
+    {
+      "epoch": 0.4309349749903809,
+      "grad_norm": 2.060504674911499,
+      "learning_rate": 0.00012689358240923262,
+      "loss": 0.8762,
+      "step": 2240
+    },
+    {
+      "epoch": 0.4311273566756445,
+      "grad_norm": 2.3213305473327637,
+      "learning_rate": 0.00012683356429284273,
+      "loss": 0.9536,
+      "step": 2241
+    },
+    {
+      "epoch": 0.43131973836090803,
+      "grad_norm": 2.123115301132202,
+      "learning_rate": 0.00012677353575873924,
+      "loss": 0.9745,
+      "step": 2242
+    },
+    {
+      "epoch": 0.4315121200461716,
+      "grad_norm": 2.1892831325531006,
+      "learning_rate": 0.00012671349683022735,
+      "loss": 1.0208,
+      "step": 2243
+    },
+    {
+      "epoch": 0.43170450173143515,
+      "grad_norm": 1.3210222721099854,
+      "learning_rate": 0.0001266534475306162,
+      "loss": 0.8344,
+      "step": 2244
+    },
+    {
+      "epoch": 0.43189688341669874,
+      "grad_norm": 1.7599616050720215,
+      "learning_rate": 0.00012659338788321905,
+      "loss": 0.8366,
+      "step": 2245
+    },
+    {
+      "epoch": 0.43208926510196227,
+      "grad_norm": 2.3361222743988037,
+      "learning_rate": 0.00012653331791135308,
+      "loss": 0.9956,
+      "step": 2246
+    },
+    {
+      "epoch": 0.43228164678722586,
+      "grad_norm": 2.534036636352539,
+      "learning_rate": 0.0001264732376383395,
+      "loss": 1.0273,
+      "step": 2247
+    },
+    {
+      "epoch": 0.4324740284724894,
+      "grad_norm": 2.143618583679199,
+      "learning_rate": 0.0001264131470875036,
+      "loss": 0.972,
+      "step": 2248
+    },
+    {
+      "epoch": 0.432666410157753,
+      "grad_norm": 0.7899037003517151,
+      "learning_rate": 0.00012635304628217452,
+      "loss": 1.045,
+      "step": 2249
+    },
+    {
+      "epoch": 0.43285879184301657,
+      "grad_norm": 2.934927225112915,
+      "learning_rate": 0.00012629293524568555,
+      "loss": 1.1252,
+      "step": 2250
+    },
+    {
+      "epoch": 0.4330511735282801,
+      "grad_norm": 1.337030053138733,
+      "learning_rate": 0.00012623281400137384,
+      "loss": 0.7436,
+      "step": 2251
+    },
+    {
+      "epoch": 0.4332435552135437,
+      "grad_norm": 2.440542221069336,
+      "learning_rate": 0.0001261726825725805,
+      "loss": 0.9517,
+      "step": 2252
+    },
+    {
+      "epoch": 0.4334359368988072,
+      "grad_norm": 2.3946423530578613,
+      "learning_rate": 0.0001261125409826506,
+      "loss": 1.0351,
+      "step": 2253
+    },
+    {
+      "epoch": 0.4336283185840708,
+      "grad_norm": 2.0878305435180664,
+      "learning_rate": 0.00012605238925493325,
+      "loss": 1.0425,
+      "step": 2254
+    },
+    {
+      "epoch": 0.43382070026933434,
+      "grad_norm": 2.2217800617218018,
+      "learning_rate": 0.00012599222741278135,
+      "loss": 1.0651,
+      "step": 2255
+    },
+    {
+      "epoch": 0.4340130819545979,
+      "grad_norm": 2.2886228561401367,
+      "learning_rate": 0.00012593205547955185,
+      "loss": 0.9434,
+      "step": 2256
+    },
+    {
+      "epoch": 0.4342054636398615,
+      "grad_norm": 1.2514402866363525,
+      "learning_rate": 0.00012587187347860553,
+      "loss": 1.0589,
+      "step": 2257
+    },
+    {
+      "epoch": 0.43439784532512504,
+      "grad_norm": 2.2382802963256836,
+      "learning_rate": 0.00012581168143330714,
+      "loss": 0.9493,
+      "step": 2258
+    },
+    {
+      "epoch": 0.43459022701038863,
+      "grad_norm": 2.2956554889678955,
+      "learning_rate": 0.0001257514793670253,
+      "loss": 0.9856,
+      "step": 2259
+    },
+    {
+      "epoch": 0.43478260869565216,
+      "grad_norm": 2.705209970474243,
+      "learning_rate": 0.00012569126730313255,
+      "loss": 1.0043,
+      "step": 2260
+    },
+    {
+      "epoch": 0.43497499038091575,
+      "grad_norm": 1.5324125289916992,
+      "learning_rate": 0.00012563104526500522,
+      "loss": 1.0249,
+      "step": 2261
+    },
+    {
+      "epoch": 0.4351673720661793,
+      "grad_norm": 1.4359239339828491,
+      "learning_rate": 0.0001255708132760236,
+      "loss": 0.8438,
+      "step": 2262
+    },
+    {
+      "epoch": 0.43535975375144287,
+      "grad_norm": 2.473400115966797,
+      "learning_rate": 0.00012551057135957186,
+      "loss": 0.9608,
+      "step": 2263
+    },
+    {
+      "epoch": 0.4355521354367064,
+      "grad_norm": 2.1563727855682373,
+      "learning_rate": 0.00012545031953903796,
+      "loss": 0.9669,
+      "step": 2264
+    },
+    {
+      "epoch": 0.43574451712197,
+      "grad_norm": 2.3325440883636475,
+      "learning_rate": 0.0001253900578378137,
+      "loss": 0.9957,
+      "step": 2265
+    },
+    {
+      "epoch": 0.4359368988072336,
+      "grad_norm": 2.073352813720703,
+      "learning_rate": 0.00012532978627929485,
+      "loss": 1.0739,
+      "step": 2266
+    },
+    {
+      "epoch": 0.4361292804924971,
+      "grad_norm": 2.5515716075897217,
+      "learning_rate": 0.00012526950488688082,
+      "loss": 1.0532,
+      "step": 2267
+    },
+    {
+      "epoch": 0.4363216621777607,
+      "grad_norm": 2.6769933700561523,
+      "learning_rate": 0.0001252092136839749,
+      "loss": 0.9956,
+      "step": 2268
+    },
+    {
+      "epoch": 0.4365140438630242,
+      "grad_norm": 1.9333584308624268,
+      "learning_rate": 0.00012514891269398427,
+      "loss": 0.9821,
+      "step": 2269
+    },
+    {
+      "epoch": 0.4367064255482878,
+      "grad_norm": 2.989109754562378,
+      "learning_rate": 0.00012508860194031986,
+      "loss": 0.973,
+      "step": 2270
+    },
+    {
+      "epoch": 0.43689880723355135,
+      "grad_norm": 1.4124975204467773,
+      "learning_rate": 0.00012502828144639628,
+      "loss": 0.8113,
+      "step": 2271
+    },
+    {
+      "epoch": 0.43709118891881493,
+      "grad_norm": 1.8631480932235718,
+      "learning_rate": 0.0001249679512356322,
+      "loss": 0.9977,
+      "step": 2272
+    },
+    {
+      "epoch": 0.43728357060407846,
+      "grad_norm": 2.495297431945801,
+      "learning_rate": 0.0001249076113314497,
+      "loss": 1.0066,
+      "step": 2273
+    },
+    {
+      "epoch": 0.43747595228934205,
+      "grad_norm": 2.1392993927001953,
+      "learning_rate": 0.0001248472617572749,
+      "loss": 0.896,
+      "step": 2274
+    },
+    {
+      "epoch": 0.43766833397460564,
+      "grad_norm": 2.467881917953491,
+      "learning_rate": 0.00012478690253653755,
+      "loss": 1.0415,
+      "step": 2275
+    },
+    {
+      "epoch": 0.43786071565986917,
+      "grad_norm": 2.410259246826172,
+      "learning_rate": 0.00012472653369267122,
+      "loss": 1.0021,
+      "step": 2276
+    },
+    {
+      "epoch": 0.43805309734513276,
+      "grad_norm": 1.7112387418746948,
+      "learning_rate": 0.00012466615524911315,
+      "loss": 0.9118,
+      "step": 2277
+    },
+    {
+      "epoch": 0.4382454790303963,
+      "grad_norm": 1.9393779039382935,
+      "learning_rate": 0.0001246057672293043,
+      "loss": 0.9992,
+      "step": 2278
+    },
+    {
+      "epoch": 0.4384378607156599,
+      "grad_norm": 2.287485361099243,
+      "learning_rate": 0.00012454536965668948,
+      "loss": 1.0142,
+      "step": 2279
+    },
+    {
+      "epoch": 0.4386302424009234,
+      "grad_norm": 2.8508615493774414,
+      "learning_rate": 0.000124484962554717,
+      "loss": 1.0957,
+      "step": 2280
+    },
+    {
+      "epoch": 0.438822624086187,
+      "grad_norm": 2.3042185306549072,
+      "learning_rate": 0.000124424545946839,
+      "loss": 0.9361,
+      "step": 2281
+    },
+    {
+      "epoch": 0.4390150057714506,
+      "grad_norm": 1.0008423328399658,
+      "learning_rate": 0.00012436411985651132,
+      "loss": 1.0729,
+      "step": 2282
+    },
+    {
+      "epoch": 0.4392073874567141,
+      "grad_norm": 2.2483553886413574,
+      "learning_rate": 0.0001243036843071934,
+      "loss": 0.9689,
+      "step": 2283
+    },
+    {
+      "epoch": 0.4393997691419777,
+      "grad_norm": 2.3147056102752686,
+      "learning_rate": 0.0001242432393223485,
+      "loss": 1.053,
+      "step": 2284
+    },
+    {
+      "epoch": 0.43959215082724123,
+      "grad_norm": 1.9770344495773315,
+      "learning_rate": 0.0001241827849254433,
+      "loss": 0.963,
+      "step": 2285
+    },
+    {
+      "epoch": 0.4397845325125048,
+      "grad_norm": 1.5617986917495728,
+      "learning_rate": 0.0001241223211399484,
+      "loss": 0.9852,
+      "step": 2286
+    },
+    {
+      "epoch": 0.43997691419776835,
+      "grad_norm": 0.873172402381897,
+      "learning_rate": 0.00012406184798933786,
+      "loss": 1.0038,
+      "step": 2287
+    },
+    {
+      "epoch": 0.44016929588303194,
+      "grad_norm": 0.9833274483680725,
+      "learning_rate": 0.00012400136549708946,
+      "loss": 1.0047,
+      "step": 2288
+    },
+    {
+      "epoch": 0.4403616775682955,
+      "grad_norm": 1.5796418190002441,
+      "learning_rate": 0.00012394087368668459,
+      "loss": 1.0282,
+      "step": 2289
+    },
+    {
+      "epoch": 0.44055405925355906,
+      "grad_norm": 2.388357162475586,
+      "learning_rate": 0.00012388037258160824,
+      "loss": 0.9392,
+      "step": 2290
+    },
+    {
+      "epoch": 0.44074644093882265,
+      "grad_norm": 2.7334156036376953,
+      "learning_rate": 0.000123819862205349,
+      "loss": 0.9739,
+      "step": 2291
+    },
+    {
+      "epoch": 0.4409388226240862,
+      "grad_norm": 1.4701759815216064,
+      "learning_rate": 0.00012375934258139915,
+      "loss": 1.0056,
+      "step": 2292
+    },
+    {
+      "epoch": 0.44113120430934977,
+      "grad_norm": 1.3947067260742188,
+      "learning_rate": 0.00012369881373325447,
+      "loss": 1.031,
+      "step": 2293
+    },
+    {
+      "epoch": 0.4413235859946133,
+      "grad_norm": 2.4266297817230225,
+      "learning_rate": 0.0001236382756844143,
+      "loss": 1.0746,
+      "step": 2294
+    },
+    {
+      "epoch": 0.4415159676798769,
+      "grad_norm": 1.3763823509216309,
+      "learning_rate": 0.00012357772845838157,
+      "loss": 0.8284,
+      "step": 2295
+    },
+    {
+      "epoch": 0.4417083493651404,
+      "grad_norm": 2.2267677783966064,
+      "learning_rate": 0.00012351717207866291,
+      "loss": 0.9672,
+      "step": 2296
+    },
+    {
+      "epoch": 0.441900731050404,
+      "grad_norm": 2.3321619033813477,
+      "learning_rate": 0.0001234566065687683,
+      "loss": 0.9629,
+      "step": 2297
+    },
+    {
+      "epoch": 0.44209311273566754,
+      "grad_norm": 1.128589153289795,
+      "learning_rate": 0.0001233960319522114,
+      "loss": 1.06,
+      "step": 2298
+    },
+    {
+      "epoch": 0.4422854944209311,
+      "grad_norm": 2.4580230712890625,
+      "learning_rate": 0.00012333544825250938,
+      "loss": 1.0464,
+      "step": 2299
+    },
+    {
+      "epoch": 0.4424778761061947,
+      "grad_norm": 2.380376100540161,
+      "learning_rate": 0.00012327485549318284,
+      "loss": 0.9747,
+      "step": 2300
+    },
+    {
+      "epoch": 0.44267025779145824,
+      "grad_norm": 2.238567590713501,
+      "learning_rate": 0.00012321425369775602,
+      "loss": 1.0335,
+      "step": 2301
+    },
+    {
+      "epoch": 0.44286263947672183,
+      "grad_norm": 2.392787456512451,
+      "learning_rate": 0.00012315364288975663,
+      "loss": 0.9809,
+      "step": 2302
+    },
+    {
+      "epoch": 0.44305502116198536,
+      "grad_norm": 2.3297054767608643,
+      "learning_rate": 0.00012309302309271587,
+      "loss": 0.9766,
+      "step": 2303
+    },
+    {
+      "epoch": 0.44324740284724895,
+      "grad_norm": 2.334482431411743,
+      "learning_rate": 0.00012303239433016842,
+      "loss": 1.0277,
+      "step": 2304
+    },
+    {
+      "epoch": 0.4434397845325125,
+      "grad_norm": 2.508192539215088,
+      "learning_rate": 0.00012297175662565247,
+      "loss": 1.0419,
+      "step": 2305
+    },
+    {
+      "epoch": 0.44363216621777607,
+      "grad_norm": 1.535554051399231,
+      "learning_rate": 0.0001229111100027097,
+      "loss": 0.8224,
+      "step": 2306
+    },
+    {
+      "epoch": 0.44382454790303966,
+      "grad_norm": 2.421172857284546,
+      "learning_rate": 0.00012285045448488511,
+      "loss": 1.0264,
+      "step": 2307
+    },
+    {
+      "epoch": 0.4440169295883032,
+      "grad_norm": 1.3813605308532715,
+      "learning_rate": 0.00012278979009572736,
+      "loss": 0.7629,
+      "step": 2308
+    },
+    {
+      "epoch": 0.4442093112735668,
+      "grad_norm": 2.4766886234283447,
+      "learning_rate": 0.0001227291168587884,
+      "loss": 0.9068,
+      "step": 2309
+    },
+    {
+      "epoch": 0.4444016929588303,
+      "grad_norm": 2.106126546859741,
+      "learning_rate": 0.0001226684347976237,
+      "loss": 1.0094,
+      "step": 2310
+    },
+    {
+      "epoch": 0.4445940746440939,
+      "grad_norm": 2.274082660675049,
+      "learning_rate": 0.0001226077439357921,
+      "loss": 0.9517,
+      "step": 2311
+    },
+    {
+      "epoch": 0.4447864563293574,
+      "grad_norm": 2.258535623550415,
+      "learning_rate": 0.00012254704429685593,
+      "loss": 1.0326,
+      "step": 2312
+    },
+    {
+      "epoch": 0.444978838014621,
+      "grad_norm": 2.3077077865600586,
+      "learning_rate": 0.00012248633590438083,
+      "loss": 1.0205,
+      "step": 2313
+    },
+    {
+      "epoch": 0.44517121969988455,
+      "grad_norm": 2.51712965965271,
+      "learning_rate": 0.00012242561878193588,
+      "loss": 0.9685,
+      "step": 2314
+    },
+    {
+      "epoch": 0.44536360138514813,
+      "grad_norm": 2.8264613151550293,
+      "learning_rate": 0.00012236489295309362,
+      "loss": 1.049,
+      "step": 2315
+    },
+    {
+      "epoch": 0.4455559830704117,
+      "grad_norm": 1.5576286315917969,
+      "learning_rate": 0.00012230415844142984,
+      "loss": 0.7817,
+      "step": 2316
+    },
+    {
+      "epoch": 0.44574836475567525,
+      "grad_norm": 2.6860666275024414,
+      "learning_rate": 0.0001222434152705238,
+      "loss": 1.0497,
+      "step": 2317
+    },
+    {
+      "epoch": 0.44594074644093884,
+      "grad_norm": 2.3129072189331055,
+      "learning_rate": 0.0001221826634639581,
+      "loss": 1.0103,
+      "step": 2318
+    },
+    {
+      "epoch": 0.44613312812620237,
+      "grad_norm": 2.5908756256103516,
+      "learning_rate": 0.0001221219030453187,
+      "loss": 1.0333,
+      "step": 2319
+    },
+    {
+      "epoch": 0.44632550981146596,
+      "grad_norm": 2.9254837036132812,
+      "learning_rate": 0.00012206113403819479,
+      "loss": 1.0208,
+      "step": 2320
+    },
+    {
+      "epoch": 0.4465178914967295,
+      "grad_norm": 1.9388304948806763,
+      "learning_rate": 0.00012200035646617911,
+      "loss": 1.0081,
+      "step": 2321
+    },
+    {
+      "epoch": 0.4467102731819931,
+      "grad_norm": 2.2436981201171875,
+      "learning_rate": 0.00012193957035286755,
+      "loss": 0.9266,
+      "step": 2322
+    },
+    {
+      "epoch": 0.4469026548672566,
+      "grad_norm": 2.8339574337005615,
+      "learning_rate": 0.00012187877572185937,
+      "loss": 0.9841,
+      "step": 2323
+    },
+    {
+      "epoch": 0.4470950365525202,
+      "grad_norm": 2.4214775562286377,
+      "learning_rate": 0.00012181797259675714,
+      "loss": 1.056,
+      "step": 2324
+    },
+    {
+      "epoch": 0.4472874182377838,
+      "grad_norm": 2.5033133029937744,
+      "learning_rate": 0.00012175716100116676,
+      "loss": 1.0441,
+      "step": 2325
+    },
+    {
+      "epoch": 0.4474797999230473,
+      "grad_norm": 2.018988609313965,
+      "learning_rate": 0.00012169634095869735,
+      "loss": 1.0002,
+      "step": 2326
+    },
+    {
+      "epoch": 0.4476721816083109,
+      "grad_norm": 3.1708977222442627,
+      "learning_rate": 0.00012163551249296133,
+      "loss": 1.0769,
+      "step": 2327
+    },
+    {
+      "epoch": 0.44786456329357444,
+      "grad_norm": 1.4504543542861938,
+      "learning_rate": 0.00012157467562757443,
+      "loss": 0.9674,
+      "step": 2328
+    },
+    {
+      "epoch": 0.448056944978838,
+      "grad_norm": 2.376091957092285,
+      "learning_rate": 0.00012151383038615563,
+      "loss": 0.9928,
+      "step": 2329
+    },
+    {
+      "epoch": 0.44824932666410156,
+      "grad_norm": 2.590000867843628,
+      "learning_rate": 0.0001214529767923271,
+      "loss": 0.9702,
+      "step": 2330
+    },
+    {
+      "epoch": 0.44844170834936514,
+      "grad_norm": 2.207052230834961,
+      "learning_rate": 0.00012139211486971436,
+      "loss": 0.9966,
+      "step": 2331
+    },
+    {
+      "epoch": 0.44863409003462873,
+      "grad_norm": 2.2230725288391113,
+      "learning_rate": 0.0001213312446419461,
+      "loss": 1.0058,
+      "step": 2332
+    },
+    {
+      "epoch": 0.44882647171989226,
+      "grad_norm": 2.1865758895874023,
+      "learning_rate": 0.00012127036613265417,
+      "loss": 1.0315,
+      "step": 2333
+    },
+    {
+      "epoch": 0.44901885340515585,
+      "grad_norm": 2.6034724712371826,
+      "learning_rate": 0.00012120947936547376,
+      "loss": 0.9812,
+      "step": 2334
+    },
+    {
+      "epoch": 0.4492112350904194,
+      "grad_norm": 1.3864617347717285,
+      "learning_rate": 0.00012114858436404321,
+      "loss": 0.762,
+      "step": 2335
+    },
+    {
+      "epoch": 0.44940361677568297,
+      "grad_norm": 1.1907415390014648,
+      "learning_rate": 0.00012108768115200404,
+      "loss": 0.8209,
+      "step": 2336
+    },
+    {
+      "epoch": 0.4495959984609465,
+      "grad_norm": 1.2113194465637207,
+      "learning_rate": 0.00012102676975300094,
+      "loss": 1.0386,
+      "step": 2337
+    },
+    {
+      "epoch": 0.4497883801462101,
+      "grad_norm": 2.400834798812866,
+      "learning_rate": 0.0001209658501906819,
+      "loss": 0.9834,
+      "step": 2338
+    },
+    {
+      "epoch": 0.4499807618314736,
+      "grad_norm": 2.324021100997925,
+      "learning_rate": 0.00012090492248869793,
+      "loss": 0.9243,
+      "step": 2339
+    },
+    {
+      "epoch": 0.4501731435167372,
+      "grad_norm": 2.144123077392578,
+      "learning_rate": 0.00012084398667070325,
+      "loss": 1.0394,
+      "step": 2340
+    },
+    {
+      "epoch": 0.4503655252020008,
+      "grad_norm": 1.9175926446914673,
+      "learning_rate": 0.00012078304276035526,
+      "loss": 0.9972,
+      "step": 2341
+    },
+    {
+      "epoch": 0.4505579068872643,
+      "grad_norm": 2.557673215866089,
+      "learning_rate": 0.00012072209078131451,
+      "loss": 1.0001,
+      "step": 2342
+    },
+    {
+      "epoch": 0.4507502885725279,
+      "grad_norm": 1.9061152935028076,
+      "learning_rate": 0.0001206611307572446,
+      "loss": 1.0141,
+      "step": 2343
+    },
+    {
+      "epoch": 0.45094267025779144,
+      "grad_norm": 2.5508508682250977,
+      "learning_rate": 0.0001206001627118124,
+      "loss": 1.0307,
+      "step": 2344
+    },
+    {
+      "epoch": 0.45113505194305503,
+      "grad_norm": 2.4778454303741455,
+      "learning_rate": 0.00012053918666868776,
+      "loss": 0.9614,
+      "step": 2345
+    },
+    {
+      "epoch": 0.45132743362831856,
+      "grad_norm": 0.979423463344574,
+      "learning_rate": 0.00012047820265154361,
+      "loss": 0.9656,
+      "step": 2346
+    },
+    {
+      "epoch": 0.45151981531358215,
+      "grad_norm": 2.206069231033325,
+      "learning_rate": 0.00012041721068405613,
+      "loss": 1.0032,
+      "step": 2347
+    },
+    {
+      "epoch": 0.4517121969988457,
+      "grad_norm": 2.3534610271453857,
+      "learning_rate": 0.00012035621078990449,
+      "loss": 0.9796,
+      "step": 2348
+    },
+    {
+      "epoch": 0.45190457868410927,
+      "grad_norm": 1.470333456993103,
+      "learning_rate": 0.00012029520299277095,
+      "loss": 0.9912,
+      "step": 2349
+    },
+    {
+      "epoch": 0.45209696036937286,
+      "grad_norm": 1.5125356912612915,
+      "learning_rate": 0.00012023418731634077,
+      "loss": 1.072,
+      "step": 2350
+    },
+    {
+      "epoch": 0.4522893420546364,
+      "grad_norm": 2.460198402404785,
+      "learning_rate": 0.00012017316378430243,
+      "loss": 0.9648,
+      "step": 2351
+    },
+    {
+      "epoch": 0.4524817237399,
+      "grad_norm": 2.3170177936553955,
+      "learning_rate": 0.00012011213242034733,
+      "loss": 1.0014,
+      "step": 2352
+    },
+    {
+      "epoch": 0.4526741054251635,
+      "grad_norm": 2.141688346862793,
+      "learning_rate": 0.00012005109324816992,
+      "loss": 0.942,
+      "step": 2353
+    },
+    {
+      "epoch": 0.4528664871104271,
+      "grad_norm": 2.6710097789764404,
+      "learning_rate": 0.00011999004629146775,
+      "loss": 1.0937,
+      "step": 2354
+    },
+    {
+      "epoch": 0.45305886879569063,
+      "grad_norm": 2.554018259048462,
+      "learning_rate": 0.00011992899157394133,
+      "loss": 1.0244,
+      "step": 2355
+    },
+    {
+      "epoch": 0.4532512504809542,
+      "grad_norm": 2.277078628540039,
+      "learning_rate": 0.00011986792911929419,
+      "loss": 0.9545,
+      "step": 2356
+    },
+    {
+      "epoch": 0.4534436321662178,
+      "grad_norm": 2.038804531097412,
+      "learning_rate": 0.00011980685895123289,
+      "loss": 0.9641,
+      "step": 2357
+    },
+    {
+      "epoch": 0.45363601385148133,
+      "grad_norm": 2.4609105587005615,
+      "learning_rate": 0.00011974578109346702,
+      "loss": 0.9566,
+      "step": 2358
+    },
+    {
+      "epoch": 0.4538283955367449,
+      "grad_norm": 1.4210039377212524,
+      "learning_rate": 0.00011968469556970905,
+      "loss": 0.7865,
+      "step": 2359
+    },
+    {
+      "epoch": 0.45402077722200845,
+      "grad_norm": 2.2436513900756836,
+      "learning_rate": 0.00011962360240367445,
+      "loss": 1.0973,
+      "step": 2360
+    },
+    {
+      "epoch": 0.45421315890727204,
+      "grad_norm": 1.8473749160766602,
+      "learning_rate": 0.00011956250161908178,
+      "loss": 1.0465,
+      "step": 2361
+    },
+    {
+      "epoch": 0.4544055405925356,
+      "grad_norm": 2.1628806591033936,
+      "learning_rate": 0.0001195013932396524,
+      "loss": 0.9225,
+      "step": 2362
+    },
+    {
+      "epoch": 0.45459792227779916,
+      "grad_norm": 2.323155403137207,
+      "learning_rate": 0.00011944027728911071,
+      "loss": 0.9885,
+      "step": 2363
+    },
+    {
+      "epoch": 0.4547903039630627,
+      "grad_norm": 1.6975098848342896,
+      "learning_rate": 0.00011937915379118405,
+      "loss": 1.0052,
+      "step": 2364
+    },
+    {
+      "epoch": 0.4549826856483263,
+      "grad_norm": 1.7289830446243286,
+      "learning_rate": 0.00011931802276960265,
+      "loss": 1.0747,
+      "step": 2365
+    },
+    {
+      "epoch": 0.45517506733358987,
+      "grad_norm": 2.6099469661712646,
+      "learning_rate": 0.00011925688424809964,
+      "loss": 1.0295,
+      "step": 2366
+    },
+    {
+      "epoch": 0.4553674490188534,
+      "grad_norm": 2.5253500938415527,
+      "learning_rate": 0.00011919573825041115,
+      "loss": 1.0413,
+      "step": 2367
+    },
+    {
+      "epoch": 0.455559830704117,
+      "grad_norm": 2.6535441875457764,
+      "learning_rate": 0.00011913458480027613,
+      "loss": 1.0188,
+      "step": 2368
+    },
+    {
+      "epoch": 0.4557522123893805,
+      "grad_norm": 1.415830373764038,
+      "learning_rate": 0.00011907342392143645,
+      "loss": 1.0754,
+      "step": 2369
+    },
+    {
+      "epoch": 0.4559445940746441,
+      "grad_norm": 2.238081932067871,
+      "learning_rate": 0.00011901225563763693,
+      "loss": 0.9695,
+      "step": 2370
+    },
+    {
+      "epoch": 0.45613697575990764,
+      "grad_norm": 2.4431114196777344,
+      "learning_rate": 0.00011895107997262515,
+      "loss": 0.9084,
+      "step": 2371
+    },
+    {
+      "epoch": 0.4563293574451712,
+      "grad_norm": 1.9115028381347656,
+      "learning_rate": 0.00011888989695015165,
+      "loss": 0.9486,
+      "step": 2372
+    },
+    {
+      "epoch": 0.45652173913043476,
+      "grad_norm": 2.5617661476135254,
+      "learning_rate": 0.00011882870659396968,
+      "loss": 0.9879,
+      "step": 2373
+    },
+    {
+      "epoch": 0.45671412081569834,
+      "grad_norm": 2.618346929550171,
+      "learning_rate": 0.00011876750892783557,
+      "loss": 0.9913,
+      "step": 2374
+    },
+    {
+      "epoch": 0.45690650250096193,
+      "grad_norm": 2.209566593170166,
+      "learning_rate": 0.00011870630397550831,
+      "loss": 1.0436,
+      "step": 2375
+    },
+    {
+      "epoch": 0.45709888418622546,
+      "grad_norm": 1.8555530309677124,
+      "learning_rate": 0.00011864509176074974,
+      "loss": 1.0394,
+      "step": 2376
+    },
+    {
+      "epoch": 0.45729126587148905,
+      "grad_norm": 2.9656896591186523,
+      "learning_rate": 0.0001185838723073246,
+      "loss": 1.0408,
+      "step": 2377
+    },
+    {
+      "epoch": 0.4574836475567526,
+      "grad_norm": 2.572023630142212,
+      "learning_rate": 0.00011852264563900038,
+      "loss": 1.0746,
+      "step": 2378
+    },
+    {
+      "epoch": 0.45767602924201617,
+      "grad_norm": 2.2676734924316406,
+      "learning_rate": 0.00011846141177954733,
+      "loss": 0.9895,
+      "step": 2379
+    },
+    {
+      "epoch": 0.4578684109272797,
+      "grad_norm": 2.1718060970306396,
+      "learning_rate": 0.00011840017075273861,
+      "loss": 1.0109,
+      "step": 2380
+    },
+    {
+      "epoch": 0.4580607926125433,
+      "grad_norm": 2.2853150367736816,
+      "learning_rate": 0.00011833892258235007,
+      "loss": 1.0261,
+      "step": 2381
+    },
+    {
+      "epoch": 0.4582531742978069,
+      "grad_norm": 2.063558578491211,
+      "learning_rate": 0.00011827766729216035,
+      "loss": 0.9694,
+      "step": 2382
+    },
+    {
+      "epoch": 0.4584455559830704,
+      "grad_norm": 2.2971231937408447,
+      "learning_rate": 0.00011821640490595084,
+      "loss": 1.055,
+      "step": 2383
+    },
+    {
+      "epoch": 0.458637937668334,
+      "grad_norm": 2.537545919418335,
+      "learning_rate": 0.00011815513544750579,
+      "loss": 0.9426,
+      "step": 2384
+    },
+    {
+      "epoch": 0.4588303193535975,
+      "grad_norm": 2.732342481613159,
+      "learning_rate": 0.00011809385894061205,
+      "loss": 0.9483,
+      "step": 2385
+    },
+    {
+      "epoch": 0.4590227010388611,
+      "grad_norm": 2.918212413787842,
+      "learning_rate": 0.00011803257540905925,
+      "loss": 1.0488,
+      "step": 2386
+    },
+    {
+      "epoch": 0.45921508272412465,
+      "grad_norm": 2.681605577468872,
+      "learning_rate": 0.00011797128487663981,
+      "loss": 1.0556,
+      "step": 2387
+    },
+    {
+      "epoch": 0.45940746440938823,
+      "grad_norm": 2.3075952529907227,
+      "learning_rate": 0.00011790998736714883,
+      "loss": 0.9791,
+      "step": 2388
+    },
+    {
+      "epoch": 0.45959984609465176,
+      "grad_norm": 2.8891680240631104,
+      "learning_rate": 0.00011784868290438404,
+      "loss": 0.9778,
+      "step": 2389
+    },
+    {
+      "epoch": 0.45979222777991535,
+      "grad_norm": 2.435753107070923,
+      "learning_rate": 0.00011778737151214605,
+      "loss": 0.9952,
+      "step": 2390
+    },
+    {
+      "epoch": 0.45998460946517894,
+      "grad_norm": 2.448780059814453,
+      "learning_rate": 0.00011772605321423799,
+      "loss": 0.9537,
+      "step": 2391
+    },
+    {
+      "epoch": 0.46017699115044247,
+      "grad_norm": 2.1232545375823975,
+      "learning_rate": 0.00011766472803446575,
+      "loss": 0.9786,
+      "step": 2392
+    },
+    {
+      "epoch": 0.46036937283570606,
+      "grad_norm": 1.6882396936416626,
+      "learning_rate": 0.00011760339599663787,
+      "loss": 1.0059,
+      "step": 2393
+    },
+    {
+      "epoch": 0.4605617545209696,
+      "grad_norm": 2.8120553493499756,
+      "learning_rate": 0.00011754205712456557,
+      "loss": 1.0026,
+      "step": 2394
+    },
+    {
+      "epoch": 0.4607541362062332,
+      "grad_norm": 2.4699742794036865,
+      "learning_rate": 0.00011748071144206266,
+      "loss": 1.0105,
+      "step": 2395
+    },
+    {
+      "epoch": 0.4609465178914967,
+      "grad_norm": 2.8250367641448975,
+      "learning_rate": 0.00011741935897294571,
+      "loss": 0.967,
+      "step": 2396
+    },
+    {
+      "epoch": 0.4611388995767603,
+      "grad_norm": 1.2501482963562012,
+      "learning_rate": 0.00011735799974103388,
+      "loss": 0.7519,
+      "step": 2397
+    },
+    {
+      "epoch": 0.46133128126202383,
+      "grad_norm": 2.3050153255462646,
+      "learning_rate": 0.00011729663377014888,
+      "loss": 1.0313,
+      "step": 2398
+    },
+    {
+      "epoch": 0.4615236629472874,
+      "grad_norm": 2.983377456665039,
+      "learning_rate": 0.00011723526108411509,
+      "loss": 0.9331,
+      "step": 2399
+    },
+    {
+      "epoch": 0.461716044632551,
+      "grad_norm": 2.433135509490967,
+      "learning_rate": 0.00011717388170675954,
+      "loss": 1.0196,
+      "step": 2400
+    },
+    {
+      "epoch": 0.46190842631781454,
+      "grad_norm": 1.2092583179473877,
+      "learning_rate": 0.00011711249566191178,
+      "loss": 1.007,
+      "step": 2401
+    },
+    {
+      "epoch": 0.4621008080030781,
+      "grad_norm": 1.0765597820281982,
+      "learning_rate": 0.00011705110297340397,
+      "loss": 1.0068,
+      "step": 2402
+    },
+    {
+      "epoch": 0.46229318968834165,
+      "grad_norm": 2.653658390045166,
+      "learning_rate": 0.00011698970366507096,
+      "loss": 0.9628,
+      "step": 2403
+    },
+    {
+      "epoch": 0.46248557137360524,
+      "grad_norm": 1.060054063796997,
+      "learning_rate": 0.00011692829776074998,
+      "loss": 1.0431,
+      "step": 2404
+    },
+    {
+      "epoch": 0.4626779530588688,
+      "grad_norm": 2.3312289714813232,
+      "learning_rate": 0.00011686688528428098,
+      "loss": 0.98,
+      "step": 2405
+    },
+    {
+      "epoch": 0.46287033474413236,
+      "grad_norm": 2.217858076095581,
+      "learning_rate": 0.00011680546625950635,
+      "loss": 1.0132,
+      "step": 2406
+    },
+    {
+      "epoch": 0.46306271642939595,
+      "grad_norm": 2.2629852294921875,
+      "learning_rate": 0.0001167440407102711,
+      "loss": 1.09,
+      "step": 2407
+    },
+    {
+      "epoch": 0.4632550981146595,
+      "grad_norm": 2.5339953899383545,
+      "learning_rate": 0.00011668260866042271,
+      "loss": 0.98,
+      "step": 2408
+    },
+    {
+      "epoch": 0.46344747979992307,
+      "grad_norm": 3.237168073654175,
+      "learning_rate": 0.00011662117013381126,
+      "loss": 0.9277,
+      "step": 2409
+    },
+    {
+      "epoch": 0.4636398614851866,
+      "grad_norm": 2.501052141189575,
+      "learning_rate": 0.00011655972515428928,
+      "loss": 1.0153,
+      "step": 2410
+    },
+    {
+      "epoch": 0.4638322431704502,
+      "grad_norm": 2.5871803760528564,
+      "learning_rate": 0.0001164982737457118,
+      "loss": 1.0454,
+      "step": 2411
+    },
+    {
+      "epoch": 0.4640246248557137,
+      "grad_norm": 2.4155197143554688,
+      "learning_rate": 0.00011643681593193643,
+      "loss": 0.9826,
+      "step": 2412
+    },
+    {
+      "epoch": 0.4642170065409773,
+      "grad_norm": 2.160922050476074,
+      "learning_rate": 0.00011637535173682316,
+      "loss": 0.9801,
+      "step": 2413
+    },
+    {
+      "epoch": 0.46440938822624084,
+      "grad_norm": 1.3618175983428955,
+      "learning_rate": 0.00011631388118423456,
+      "loss": 1.0459,
+      "step": 2414
+    },
+    {
+      "epoch": 0.4646017699115044,
+      "grad_norm": 1.5935145616531372,
+      "learning_rate": 0.00011625240429803552,
+      "loss": 0.8392,
+      "step": 2415
+    },
+    {
+      "epoch": 0.464794151596768,
+      "grad_norm": 1.2172508239746094,
+      "learning_rate": 0.0001161909211020936,
+      "loss": 0.7964,
+      "step": 2416
+    },
+    {
+      "epoch": 0.46498653328203154,
+      "grad_norm": 2.46657395362854,
+      "learning_rate": 0.00011612943162027863,
+      "loss": 1.0707,
+      "step": 2417
+    },
+    {
+      "epoch": 0.46517891496729513,
+      "grad_norm": 2.482252597808838,
+      "learning_rate": 0.00011606793587646294,
+      "loss": 0.9516,
+      "step": 2418
+    },
+    {
+      "epoch": 0.46537129665255866,
+      "grad_norm": 2.5039610862731934,
+      "learning_rate": 0.0001160064338945213,
+      "loss": 1.0196,
+      "step": 2419
+    },
+    {
+      "epoch": 0.46556367833782225,
+      "grad_norm": 2.2555506229400635,
+      "learning_rate": 0.00011594492569833094,
+      "loss": 1.0353,
+      "step": 2420
+    },
+    {
+      "epoch": 0.4657560600230858,
+      "grad_norm": 1.1694220304489136,
+      "learning_rate": 0.00011588341131177136,
+      "loss": 0.776,
+      "step": 2421
+    },
+    {
+      "epoch": 0.46594844170834937,
+      "grad_norm": 3.264285087585449,
+      "learning_rate": 0.00011582189075872468,
+      "loss": 1.049,
+      "step": 2422
+    },
+    {
+      "epoch": 0.4661408233936129,
+      "grad_norm": 1.3675886392593384,
+      "learning_rate": 0.00011576036406307522,
+      "loss": 0.9931,
+      "step": 2423
+    },
+    {
+      "epoch": 0.4663332050788765,
+      "grad_norm": 2.688699960708618,
+      "learning_rate": 0.00011569883124870979,
+      "loss": 1.0002,
+      "step": 2424
+    },
+    {
+      "epoch": 0.4665255867641401,
+      "grad_norm": 1.1168848276138306,
+      "learning_rate": 0.00011563729233951756,
+      "loss": 1.0676,
+      "step": 2425
+    },
+    {
+      "epoch": 0.4667179684494036,
+      "grad_norm": 2.256953716278076,
+      "learning_rate": 0.00011557574735939004,
+      "loss": 1.0077,
+      "step": 2426
+    },
+    {
+      "epoch": 0.4669103501346672,
+      "grad_norm": 2.325692892074585,
+      "learning_rate": 0.00011551419633222108,
+      "loss": 0.9854,
+      "step": 2427
+    },
+    {
+      "epoch": 0.4671027318199307,
+      "grad_norm": 2.8092095851898193,
+      "learning_rate": 0.00011545263928190693,
+      "loss": 1.018,
+      "step": 2428
+    },
+    {
+      "epoch": 0.4672951135051943,
+      "grad_norm": 2.126744508743286,
+      "learning_rate": 0.00011539107623234617,
+      "loss": 1.0257,
+      "step": 2429
+    },
+    {
+      "epoch": 0.46748749519045785,
+      "grad_norm": 2.105628728866577,
+      "learning_rate": 0.0001153295072074397,
+      "loss": 1.0342,
+      "step": 2430
+    },
+    {
+      "epoch": 0.46767987687572143,
+      "grad_norm": 2.2590365409851074,
+      "learning_rate": 0.00011526793223109071,
+      "loss": 1.0074,
+      "step": 2431
+    },
+    {
+      "epoch": 0.467872258560985,
+      "grad_norm": 2.3832221031188965,
+      "learning_rate": 0.00011520635132720474,
+      "loss": 0.9518,
+      "step": 2432
+    },
+    {
+      "epoch": 0.46806464024624855,
+      "grad_norm": 2.407407522201538,
+      "learning_rate": 0.0001151447645196896,
+      "loss": 1.0229,
+      "step": 2433
+    },
+    {
+      "epoch": 0.46825702193151214,
+      "grad_norm": 1.9372382164001465,
+      "learning_rate": 0.00011508317183245545,
+      "loss": 0.9456,
+      "step": 2434
+    },
+    {
+      "epoch": 0.4684494036167757,
+      "grad_norm": 2.6510555744171143,
+      "learning_rate": 0.00011502157328941465,
+      "loss": 0.965,
+      "step": 2435
+    },
+    {
+      "epoch": 0.46864178530203926,
+      "grad_norm": 1.8128416538238525,
+      "learning_rate": 0.00011495996891448189,
+      "loss": 1.0322,
+      "step": 2436
+    },
+    {
+      "epoch": 0.4688341669873028,
+      "grad_norm": 2.613481044769287,
+      "learning_rate": 0.00011489835873157413,
+      "loss": 1.0683,
+      "step": 2437
+    },
+    {
+      "epoch": 0.4690265486725664,
+      "grad_norm": 2.5035300254821777,
+      "learning_rate": 0.00011483674276461052,
+      "loss": 1.0062,
+      "step": 2438
+    },
+    {
+      "epoch": 0.4692189303578299,
+      "grad_norm": 2.2045578956604004,
+      "learning_rate": 0.00011477512103751253,
+      "loss": 1.0193,
+      "step": 2439
+    },
+    {
+      "epoch": 0.4694113120430935,
+      "grad_norm": 1.747169017791748,
+      "learning_rate": 0.00011471349357420384,
+      "loss": 0.9999,
+      "step": 2440
+    },
+    {
+      "epoch": 0.4696036937283571,
+      "grad_norm": 2.2418813705444336,
+      "learning_rate": 0.00011465186039861033,
+      "loss": 0.9549,
+      "step": 2441
+    },
+    {
+      "epoch": 0.4697960754136206,
+      "grad_norm": 2.5708892345428467,
+      "learning_rate": 0.00011459022153466015,
+      "loss": 1.0336,
+      "step": 2442
+    },
+    {
+      "epoch": 0.4699884570988842,
+      "grad_norm": 3.3080294132232666,
+      "learning_rate": 0.00011452857700628361,
+      "loss": 1.0063,
+      "step": 2443
+    },
+    {
+      "epoch": 0.47018083878414774,
+      "grad_norm": 2.735196828842163,
+      "learning_rate": 0.00011446692683741325,
+      "loss": 0.9771,
+      "step": 2444
+    },
+    {
+      "epoch": 0.4703732204694113,
+      "grad_norm": 1.794972538948059,
+      "learning_rate": 0.00011440527105198378,
+      "loss": 1.0254,
+      "step": 2445
+    },
+    {
+      "epoch": 0.47056560215467486,
+      "grad_norm": 2.404913902282715,
+      "learning_rate": 0.0001143436096739321,
+      "loss": 0.9287,
+      "step": 2446
+    },
+    {
+      "epoch": 0.47075798383993844,
+      "grad_norm": 2.554609775543213,
+      "learning_rate": 0.00011428194272719728,
+      "loss": 1.0287,
+      "step": 2447
+    },
+    {
+      "epoch": 0.470950365525202,
+      "grad_norm": 2.758056879043579,
+      "learning_rate": 0.00011422027023572051,
+      "loss": 1.0477,
+      "step": 2448
+    },
+    {
+      "epoch": 0.47114274721046556,
+      "grad_norm": 2.107764720916748,
+      "learning_rate": 0.00011415859222344525,
+      "loss": 1.0159,
+      "step": 2449
+    },
+    {
+      "epoch": 0.47133512889572915,
+      "grad_norm": 0.7311239838600159,
+      "learning_rate": 0.000114096908714317,
+      "loss": 1.0086,
+      "step": 2450
+    },
+    {
+      "epoch": 0.4715275105809927,
+      "grad_norm": 0.9692084789276123,
+      "learning_rate": 0.00011403521973228341,
+      "loss": 0.9861,
+      "step": 2451
+    },
+    {
+      "epoch": 0.47171989226625627,
+      "grad_norm": 0.9579837918281555,
+      "learning_rate": 0.00011397352530129429,
+      "loss": 1.0406,
+      "step": 2452
+    },
+    {
+      "epoch": 0.4719122739515198,
+      "grad_norm": 1.6997452974319458,
+      "learning_rate": 0.0001139118254453015,
+      "loss": 1.0351,
+      "step": 2453
+    },
+    {
+      "epoch": 0.4721046556367834,
+      "grad_norm": 1.0282096862792969,
+      "learning_rate": 0.00011385012018825907,
+      "loss": 0.9614,
+      "step": 2454
+    },
+    {
+      "epoch": 0.4722970373220469,
+      "grad_norm": 1.3151376247406006,
+      "learning_rate": 0.00011378840955412313,
+      "loss": 0.8042,
+      "step": 2455
+    },
+    {
+      "epoch": 0.4724894190073105,
+      "grad_norm": 2.2971372604370117,
+      "learning_rate": 0.00011372669356685183,
+      "loss": 1.0414,
+      "step": 2456
+    },
+    {
+      "epoch": 0.4726818006925741,
+      "grad_norm": 2.036301612854004,
+      "learning_rate": 0.00011366497225040548,
+      "loss": 0.9706,
+      "step": 2457
+    },
+    {
+      "epoch": 0.4728741823778376,
+      "grad_norm": 2.167152166366577,
+      "learning_rate": 0.00011360324562874643,
+      "loss": 0.9812,
+      "step": 2458
+    },
+    {
+      "epoch": 0.4730665640631012,
+      "grad_norm": 2.6004135608673096,
+      "learning_rate": 0.00011354151372583901,
+      "loss": 1.0567,
+      "step": 2459
+    },
+    {
+      "epoch": 0.47325894574836475,
+      "grad_norm": 2.068157911300659,
+      "learning_rate": 0.00011347977656564975,
+      "loss": 1.0032,
+      "step": 2460
+    },
+    {
+      "epoch": 0.47345132743362833,
+      "grad_norm": 2.3544118404388428,
+      "learning_rate": 0.00011341803417214705,
+      "loss": 0.9896,
+      "step": 2461
+    },
+    {
+      "epoch": 0.47364370911889186,
+      "grad_norm": 1.9077234268188477,
+      "learning_rate": 0.00011335628656930152,
+      "loss": 1.0009,
+      "step": 2462
+    },
+    {
+      "epoch": 0.47383609080415545,
+      "grad_norm": 2.038820266723633,
+      "learning_rate": 0.0001132945337810857,
+      "loss": 0.9629,
+      "step": 2463
+    },
+    {
+      "epoch": 0.474028472489419,
+      "grad_norm": 2.236497640609741,
+      "learning_rate": 0.00011323277583147409,
+      "loss": 0.9653,
+      "step": 2464
+    },
+    {
+      "epoch": 0.47422085417468257,
+      "grad_norm": 2.4429612159729004,
+      "learning_rate": 0.00011317101274444328,
+      "loss": 1.0484,
+      "step": 2465
+    },
+    {
+      "epoch": 0.47441323585994616,
+      "grad_norm": 2.1554958820343018,
+      "learning_rate": 0.00011310924454397187,
+      "loss": 0.9814,
+      "step": 2466
+    },
+    {
+      "epoch": 0.4746056175452097,
+      "grad_norm": 2.593327045440674,
+      "learning_rate": 0.00011304747125404031,
+      "loss": 0.9918,
+      "step": 2467
+    },
+    {
+      "epoch": 0.4747979992304733,
+      "grad_norm": 2.499300241470337,
+      "learning_rate": 0.00011298569289863119,
+      "loss": 1.0556,
+      "step": 2468
+    },
+    {
+      "epoch": 0.4749903809157368,
+      "grad_norm": 1.9360010623931885,
+      "learning_rate": 0.00011292390950172899,
+      "loss": 0.9142,
+      "step": 2469
+    },
+    {
+      "epoch": 0.4751827626010004,
+      "grad_norm": 1.7909685373306274,
+      "learning_rate": 0.00011286212108732015,
+      "loss": 0.997,
+      "step": 2470
+    },
+    {
+      "epoch": 0.47537514428626393,
+      "grad_norm": 2.404473066329956,
+      "learning_rate": 0.00011280032767939301,
+      "loss": 0.9353,
+      "step": 2471
+    },
+    {
+      "epoch": 0.4755675259715275,
+      "grad_norm": 2.1001715660095215,
+      "learning_rate": 0.00011273852930193798,
+      "loss": 0.9549,
+      "step": 2472
+    },
+    {
+      "epoch": 0.47575990765679105,
+      "grad_norm": 2.938225507736206,
+      "learning_rate": 0.00011267672597894725,
+      "loss": 1.0485,
+      "step": 2473
+    },
+    {
+      "epoch": 0.47595228934205464,
+      "grad_norm": 2.0799996852874756,
+      "learning_rate": 0.000112614917734415,
+      "loss": 0.8268,
+      "step": 2474
+    },
+    {
+      "epoch": 0.4761446710273182,
+      "grad_norm": 2.7278387546539307,
+      "learning_rate": 0.00011255310459233737,
+      "loss": 0.9967,
+      "step": 2475
+    },
+    {
+      "epoch": 0.47633705271258175,
+      "grad_norm": 2.4720346927642822,
+      "learning_rate": 0.00011249128657671233,
+      "loss": 1.106,
+      "step": 2476
+    },
+    {
+      "epoch": 0.47652943439784534,
+      "grad_norm": 2.0709640979766846,
+      "learning_rate": 0.00011242946371153972,
+      "loss": 1.0108,
+      "step": 2477
+    },
+    {
+      "epoch": 0.4767218160831089,
+      "grad_norm": 2.536436080932617,
+      "learning_rate": 0.00011236763602082137,
+      "loss": 1.0242,
+      "step": 2478
+    },
+    {
+      "epoch": 0.47691419776837246,
+      "grad_norm": 2.487779140472412,
+      "learning_rate": 0.00011230580352856088,
+      "loss": 1.0152,
+      "step": 2479
+    },
+    {
+      "epoch": 0.477106579453636,
+      "grad_norm": 2.5463221073150635,
+      "learning_rate": 0.00011224396625876375,
+      "loss": 1.0213,
+      "step": 2480
+    },
+    {
+      "epoch": 0.4772989611388996,
+      "grad_norm": 2.2382125854492188,
+      "learning_rate": 0.00011218212423543734,
+      "loss": 0.9789,
+      "step": 2481
+    },
+    {
+      "epoch": 0.4774913428241631,
+      "grad_norm": 3.203760862350464,
+      "learning_rate": 0.00011212027748259086,
+      "loss": 1.0859,
+      "step": 2482
+    },
+    {
+      "epoch": 0.4776837245094267,
+      "grad_norm": 2.587637186050415,
+      "learning_rate": 0.00011205842602423537,
+      "loss": 0.9661,
+      "step": 2483
+    },
+    {
+      "epoch": 0.4778761061946903,
+      "grad_norm": 1.5063904523849487,
+      "learning_rate": 0.00011199656988438372,
+      "loss": 0.9676,
+      "step": 2484
+    },
+    {
+      "epoch": 0.4780684878799538,
+      "grad_norm": 2.2403783798217773,
+      "learning_rate": 0.00011193470908705055,
+      "loss": 1.0579,
+      "step": 2485
+    },
+    {
+      "epoch": 0.4782608695652174,
+      "grad_norm": 2.6740736961364746,
+      "learning_rate": 0.00011187284365625242,
+      "loss": 1.0152,
+      "step": 2486
+    },
+    {
+      "epoch": 0.47845325125048094,
+      "grad_norm": 2.528129816055298,
+      "learning_rate": 0.00011181097361600755,
+      "loss": 0.9799,
+      "step": 2487
+    },
+    {
+      "epoch": 0.4786456329357445,
+      "grad_norm": 2.159771203994751,
+      "learning_rate": 0.00011174909899033607,
+      "loss": 0.8874,
+      "step": 2488
+    },
+    {
+      "epoch": 0.47883801462100806,
+      "grad_norm": 1.9930708408355713,
+      "learning_rate": 0.00011168721980325987,
+      "loss": 1.0569,
+      "step": 2489
+    },
+    {
+      "epoch": 0.47903039630627164,
+      "grad_norm": 1.0517325401306152,
+      "learning_rate": 0.0001116253360788025,
+      "loss": 1.0431,
+      "step": 2490
+    },
+    {
+      "epoch": 0.47922277799153523,
+      "grad_norm": 2.169846296310425,
+      "learning_rate": 0.00011156344784098942,
+      "loss": 1.0531,
+      "step": 2491
+    },
+    {
+      "epoch": 0.47941515967679876,
+      "grad_norm": 2.5158023834228516,
+      "learning_rate": 0.00011150155511384773,
+      "loss": 0.977,
+      "step": 2492
+    },
+    {
+      "epoch": 0.47960754136206235,
+      "grad_norm": 0.9373905658721924,
+      "learning_rate": 0.00011143965792140632,
+      "loss": 1.0398,
+      "step": 2493
+    },
+    {
+      "epoch": 0.4797999230473259,
+      "grad_norm": 2.2019424438476562,
+      "learning_rate": 0.00011137775628769584,
+      "loss": 1.0577,
+      "step": 2494
+    },
+    {
+      "epoch": 0.47999230473258947,
+      "grad_norm": 1.2052497863769531,
+      "learning_rate": 0.00011131585023674861,
+      "loss": 1.0461,
+      "step": 2495
+    },
+    {
+      "epoch": 0.480184686417853,
+      "grad_norm": 2.3645904064178467,
+      "learning_rate": 0.00011125393979259872,
+      "loss": 0.9421,
+      "step": 2496
+    },
+    {
+      "epoch": 0.4803770681031166,
+      "grad_norm": 1.8030023574829102,
+      "learning_rate": 0.00011119202497928192,
+      "loss": 0.9848,
+      "step": 2497
+    },
+    {
+      "epoch": 0.4805694497883801,
+      "grad_norm": 2.1907312870025635,
+      "learning_rate": 0.00011113010582083567,
+      "loss": 1.0399,
+      "step": 2498
+    },
+    {
+      "epoch": 0.4807618314736437,
+      "grad_norm": 1.9506596326828003,
+      "learning_rate": 0.00011106818234129912,
+      "loss": 1.0127,
+      "step": 2499
+    },
+    {
+      "epoch": 0.4809542131589073,
+      "grad_norm": 1.2162901163101196,
+      "learning_rate": 0.00011100625456471307,
+      "loss": 0.9968,
+      "step": 2500
+    },
+    {
+      "epoch": 0.4811465948441708,
+      "grad_norm": 2.2975902557373047,
+      "learning_rate": 0.00011094432251512005,
+      "loss": 0.983,
+      "step": 2501
+    },
+    {
+      "epoch": 0.4813389765294344,
+      "grad_norm": 2.4030232429504395,
+      "learning_rate": 0.00011088238621656421,
+      "loss": 0.9881,
+      "step": 2502
+    },
+    {
+      "epoch": 0.48153135821469795,
+      "grad_norm": 2.488842248916626,
+      "learning_rate": 0.00011082044569309137,
+      "loss": 1.0646,
+      "step": 2503
+    },
+    {
+      "epoch": 0.48172373989996153,
+      "grad_norm": 2.088616371154785,
+      "learning_rate": 0.00011075850096874893,
+      "loss": 1.0335,
+      "step": 2504
+    },
+    {
+      "epoch": 0.48191612158522507,
+      "grad_norm": 1.4234910011291504,
+      "learning_rate": 0.00011069655206758602,
+      "loss": 0.7888,
+      "step": 2505
+    },
+    {
+      "epoch": 0.48210850327048865,
+      "grad_norm": 2.2808971405029297,
+      "learning_rate": 0.00011063459901365325,
+      "loss": 1.0574,
+      "step": 2506
+    },
+    {
+      "epoch": 0.4823008849557522,
+      "grad_norm": 1.6067759990692139,
+      "learning_rate": 0.00011057264183100302,
+      "loss": 0.9692,
+      "step": 2507
+    },
+    {
+      "epoch": 0.48249326664101577,
+      "grad_norm": 2.318490743637085,
+      "learning_rate": 0.0001105106805436892,
+      "loss": 0.9742,
+      "step": 2508
+    },
+    {
+      "epoch": 0.48268564832627936,
+      "grad_norm": 1.537163496017456,
+      "learning_rate": 0.00011044871517576729,
+      "loss": 0.9903,
+      "step": 2509
+    },
+    {
+      "epoch": 0.4828780300115429,
+      "grad_norm": 3.2745325565338135,
+      "learning_rate": 0.00011038674575129441,
+      "loss": 0.9306,
+      "step": 2510
+    },
+    {
+      "epoch": 0.4830704116968065,
+      "grad_norm": 2.536001205444336,
+      "learning_rate": 0.0001103247722943292,
+      "loss": 0.9941,
+      "step": 2511
+    },
+    {
+      "epoch": 0.48326279338207,
+      "grad_norm": 1.5535314083099365,
+      "learning_rate": 0.00011026279482893187,
+      "loss": 0.8093,
+      "step": 2512
+    },
+    {
+      "epoch": 0.4834551750673336,
+      "grad_norm": 2.7451794147491455,
+      "learning_rate": 0.00011020081337916424,
+      "loss": 1.0149,
+      "step": 2513
+    },
+    {
+      "epoch": 0.48364755675259713,
+      "grad_norm": 1.1709246635437012,
+      "learning_rate": 0.00011013882796908963,
+      "loss": 0.7834,
+      "step": 2514
+    },
+    {
+      "epoch": 0.4838399384378607,
+      "grad_norm": 2.685180187225342,
+      "learning_rate": 0.00011007683862277292,
+      "loss": 0.9575,
+      "step": 2515
+    },
+    {
+      "epoch": 0.4840323201231243,
+      "grad_norm": 1.7856175899505615,
+      "learning_rate": 0.00011001484536428051,
+      "loss": 0.9654,
+      "step": 2516
+    },
+    {
+      "epoch": 0.48422470180838784,
+      "grad_norm": 2.3102705478668213,
+      "learning_rate": 0.00010995284821768028,
+      "loss": 1.0202,
+      "step": 2517
+    },
+    {
+      "epoch": 0.4844170834936514,
+      "grad_norm": 1.5802983045578003,
+      "learning_rate": 0.0001098908472070417,
+      "loss": 1.0176,
+      "step": 2518
+    },
+    {
+      "epoch": 0.48460946517891496,
+      "grad_norm": 2.7118899822235107,
+      "learning_rate": 0.00010982884235643567,
+      "loss": 1.033,
+      "step": 2519
+    },
+    {
+      "epoch": 0.48480184686417854,
+      "grad_norm": 1.689682960510254,
+      "learning_rate": 0.00010976683368993464,
+      "loss": 1.0285,
+      "step": 2520
+    },
+    {
+      "epoch": 0.4849942285494421,
+      "grad_norm": 2.464038848876953,
+      "learning_rate": 0.00010970482123161248,
+      "loss": 1.0977,
+      "step": 2521
+    },
+    {
+      "epoch": 0.48518661023470566,
+      "grad_norm": 2.3707492351531982,
+      "learning_rate": 0.00010964280500554458,
+      "loss": 1.0339,
+      "step": 2522
+    },
+    {
+      "epoch": 0.4853789919199692,
+      "grad_norm": 2.875131607055664,
+      "learning_rate": 0.00010958078503580775,
+      "loss": 1.0488,
+      "step": 2523
+    },
+    {
+      "epoch": 0.4855713736052328,
+      "grad_norm": 2.444225311279297,
+      "learning_rate": 0.00010951876134648032,
+      "loss": 0.988,
+      "step": 2524
+    },
+    {
+      "epoch": 0.48576375529049637,
+      "grad_norm": 2.237596273422241,
+      "learning_rate": 0.00010945673396164198,
+      "loss": 0.9985,
+      "step": 2525
+    },
+    {
+      "epoch": 0.4859561369757599,
+      "grad_norm": 2.9272358417510986,
+      "learning_rate": 0.00010939470290537388,
+      "loss": 1.0207,
+      "step": 2526
+    },
+    {
+      "epoch": 0.4861485186610235,
+      "grad_norm": 1.5333023071289062,
+      "learning_rate": 0.00010933266820175867,
+      "loss": 0.959,
+      "step": 2527
+    },
+    {
+      "epoch": 0.486340900346287,
+      "grad_norm": 1.3142082691192627,
+      "learning_rate": 0.00010927062987488035,
+      "loss": 0.7897,
+      "step": 2528
+    },
+    {
+      "epoch": 0.4865332820315506,
+      "grad_norm": 1.0694913864135742,
+      "learning_rate": 0.00010920858794882429,
+      "loss": 1.0433,
+      "step": 2529
+    },
+    {
+      "epoch": 0.48672566371681414,
+      "grad_norm": 1.8779606819152832,
+      "learning_rate": 0.00010914654244767734,
+      "loss": 1.0427,
+      "step": 2530
+    },
+    {
+      "epoch": 0.4869180454020777,
+      "grad_norm": 1.325904369354248,
+      "learning_rate": 0.00010908449339552768,
+      "loss": 0.7859,
+      "step": 2531
+    },
+    {
+      "epoch": 0.48711042708734126,
+      "grad_norm": 2.2861392498016357,
+      "learning_rate": 0.00010902244081646489,
+      "loss": 0.9234,
+      "step": 2532
+    },
+    {
+      "epoch": 0.48730280877260485,
+      "grad_norm": 2.3141419887542725,
+      "learning_rate": 0.00010896038473457993,
+      "loss": 0.9147,
+      "step": 2533
+    },
+    {
+      "epoch": 0.48749519045786843,
+      "grad_norm": 2.4254701137542725,
+      "learning_rate": 0.00010889832517396509,
+      "loss": 1.0006,
+      "step": 2534
+    },
+    {
+      "epoch": 0.48768757214313196,
+      "grad_norm": 2.355159044265747,
+      "learning_rate": 0.00010883626215871408,
+      "loss": 1.0358,
+      "step": 2535
+    },
+    {
+      "epoch": 0.48787995382839555,
+      "grad_norm": 2.852541446685791,
+      "learning_rate": 0.00010877419571292182,
+      "loss": 1.0261,
+      "step": 2536
+    },
+    {
+      "epoch": 0.4880723355136591,
+      "grad_norm": 2.526857376098633,
+      "learning_rate": 0.00010871212586068469,
+      "loss": 1.0614,
+      "step": 2537
+    },
+    {
+      "epoch": 0.48826471719892267,
+      "grad_norm": 2.3947198390960693,
+      "learning_rate": 0.00010865005262610033,
+      "loss": 0.982,
+      "step": 2538
+    },
+    {
+      "epoch": 0.4884570988841862,
+      "grad_norm": 2.008713722229004,
+      "learning_rate": 0.00010858797603326769,
+      "loss": 0.9819,
+      "step": 2539
+    },
+    {
+      "epoch": 0.4886494805694498,
+      "grad_norm": 1.1034409999847412,
+      "learning_rate": 0.0001085258961062871,
+      "loss": 0.9694,
+      "step": 2540
+    },
+    {
+      "epoch": 0.4888418622547134,
+      "grad_norm": 2.3556480407714844,
+      "learning_rate": 0.00010846381286926006,
+      "loss": 0.9751,
+      "step": 2541
+    },
+    {
+      "epoch": 0.4890342439399769,
+      "grad_norm": 2.723905086517334,
+      "learning_rate": 0.00010840172634628946,
+      "loss": 1.0029,
+      "step": 2542
+    },
+    {
+      "epoch": 0.4892266256252405,
+      "grad_norm": 2.14678692817688,
+      "learning_rate": 0.00010833963656147943,
+      "loss": 0.9741,
+      "step": 2543
+    },
+    {
+      "epoch": 0.48941900731050403,
+      "grad_norm": 1.2719849348068237,
+      "learning_rate": 0.00010827754353893532,
+      "loss": 0.7691,
+      "step": 2544
+    },
+    {
+      "epoch": 0.4896113889957676,
+      "grad_norm": 1.0171805620193481,
+      "learning_rate": 0.00010821544730276378,
+      "loss": 0.9909,
+      "step": 2545
+    },
+    {
+      "epoch": 0.48980377068103115,
+      "grad_norm": 2.36379075050354,
+      "learning_rate": 0.00010815334787707276,
+      "loss": 0.962,
+      "step": 2546
+    },
+    {
+      "epoch": 0.48999615236629473,
+      "grad_norm": 2.1662752628326416,
+      "learning_rate": 0.00010809124528597139,
+      "loss": 0.9919,
+      "step": 2547
+    },
+    {
+      "epoch": 0.49018853405155827,
+      "grad_norm": 1.154170274734497,
+      "learning_rate": 0.00010802913955356997,
+      "loss": 1.0145,
+      "step": 2548
+    },
+    {
+      "epoch": 0.49038091573682185,
+      "grad_norm": 2.6192257404327393,
+      "learning_rate": 0.00010796703070398015,
+      "loss": 0.9475,
+      "step": 2549
+    },
+    {
+      "epoch": 0.49057329742208544,
+      "grad_norm": 2.1254093647003174,
+      "learning_rate": 0.0001079049187613147,
+      "loss": 1.0087,
+      "step": 2550
+    },
+    {
+      "epoch": 0.490765679107349,
+      "grad_norm": 2.699007272720337,
+      "learning_rate": 0.0001078428037496876,
+      "loss": 0.9859,
+      "step": 2551
+    },
+    {
+      "epoch": 0.49095806079261256,
+      "grad_norm": 2.280832290649414,
+      "learning_rate": 0.00010778068569321403,
+      "loss": 1.0679,
+      "step": 2552
+    },
+    {
+      "epoch": 0.4911504424778761,
+      "grad_norm": 2.50931453704834,
+      "learning_rate": 0.0001077185646160104,
+      "loss": 0.9727,
+      "step": 2553
+    },
+    {
+      "epoch": 0.4913428241631397,
+      "grad_norm": 1.532430648803711,
+      "learning_rate": 0.00010765644054219421,
+      "loss": 1.0166,
+      "step": 2554
+    },
+    {
+      "epoch": 0.4915352058484032,
+      "grad_norm": 2.430523633956909,
+      "learning_rate": 0.0001075943134958842,
+      "loss": 0.9771,
+      "step": 2555
+    },
+    {
+      "epoch": 0.4917275875336668,
+      "grad_norm": 1.5213136672973633,
+      "learning_rate": 0.00010753218350120022,
+      "loss": 1.0545,
+      "step": 2556
+    },
+    {
+      "epoch": 0.49191996921893033,
+      "grad_norm": 2.2090723514556885,
+      "learning_rate": 0.00010747005058226324,
+      "loss": 0.9812,
+      "step": 2557
+    },
+    {
+      "epoch": 0.4921123509041939,
+      "grad_norm": 2.4196794033050537,
+      "learning_rate": 0.00010740791476319543,
+      "loss": 0.9696,
+      "step": 2558
+    },
+    {
+      "epoch": 0.4923047325894575,
+      "grad_norm": 1.4285922050476074,
+      "learning_rate": 0.00010734577606812007,
+      "loss": 0.9873,
+      "step": 2559
+    },
+    {
+      "epoch": 0.49249711427472104,
+      "grad_norm": 2.0300467014312744,
+      "learning_rate": 0.00010728363452116148,
+      "loss": 0.9747,
+      "step": 2560
+    },
+    {
+      "epoch": 0.4926894959599846,
+      "grad_norm": 1.5242557525634766,
+      "learning_rate": 0.00010722149014644522,
+      "loss": 0.9769,
+      "step": 2561
+    },
+    {
+      "epoch": 0.49288187764524816,
+      "grad_norm": 2.7415695190429688,
+      "learning_rate": 0.00010715934296809782,
+      "loss": 0.9754,
+      "step": 2562
+    },
+    {
+      "epoch": 0.49307425933051174,
+      "grad_norm": 2.2822928428649902,
+      "learning_rate": 0.000107097193010247,
+      "loss": 1.0044,
+      "step": 2563
+    },
+    {
+      "epoch": 0.4932666410157753,
+      "grad_norm": 3.1589107513427734,
+      "learning_rate": 0.00010703504029702148,
+      "loss": 1.0293,
+      "step": 2564
+    },
+    {
+      "epoch": 0.49345902270103886,
+      "grad_norm": 2.4638431072235107,
+      "learning_rate": 0.00010697288485255106,
+      "loss": 0.9938,
+      "step": 2565
+    },
+    {
+      "epoch": 0.49365140438630245,
+      "grad_norm": 1.453808069229126,
+      "learning_rate": 0.00010691072670096669,
+      "loss": 0.8019,
+      "step": 2566
+    },
+    {
+      "epoch": 0.493843786071566,
+      "grad_norm": 2.4564521312713623,
+      "learning_rate": 0.00010684856586640026,
+      "loss": 0.9836,
+      "step": 2567
+    },
+    {
+      "epoch": 0.49403616775682957,
+      "grad_norm": 2.4006683826446533,
+      "learning_rate": 0.00010678640237298476,
+      "loss": 0.9987,
+      "step": 2568
+    },
+    {
+      "epoch": 0.4942285494420931,
+      "grad_norm": 2.1085331439971924,
+      "learning_rate": 0.00010672423624485421,
+      "loss": 1.0381,
+      "step": 2569
+    },
+    {
+      "epoch": 0.4944209311273567,
+      "grad_norm": 2.6820693016052246,
+      "learning_rate": 0.00010666206750614362,
+      "loss": 0.973,
+      "step": 2570
+    },
+    {
+      "epoch": 0.4946133128126202,
+      "grad_norm": 2.1011390686035156,
+      "learning_rate": 0.00010659989618098905,
+      "loss": 1.019,
+      "step": 2571
+    },
+    {
+      "epoch": 0.4948056944978838,
+      "grad_norm": 1.9683669805526733,
+      "learning_rate": 0.00010653772229352749,
+      "loss": 1.043,
+      "step": 2572
+    },
+    {
+      "epoch": 0.49499807618314734,
+      "grad_norm": 2.202136754989624,
+      "learning_rate": 0.00010647554586789708,
+      "loss": 1.0523,
+      "step": 2573
+    },
+    {
+      "epoch": 0.4951904578684109,
+      "grad_norm": 2.279757499694824,
+      "learning_rate": 0.0001064133669282368,
+      "loss": 0.9723,
+      "step": 2574
+    },
+    {
+      "epoch": 0.4953828395536745,
+      "grad_norm": 2.7619516849517822,
+      "learning_rate": 0.00010635118549868668,
+      "loss": 1.0221,
+      "step": 2575
+    },
+    {
+      "epoch": 0.49557522123893805,
+      "grad_norm": 2.226632595062256,
+      "learning_rate": 0.00010628900160338764,
+      "loss": 0.9397,
+      "step": 2576
+    },
+    {
+      "epoch": 0.49576760292420163,
+      "grad_norm": 2.0882740020751953,
+      "learning_rate": 0.00010622681526648167,
+      "loss": 0.9515,
+      "step": 2577
+    },
+    {
+      "epoch": 0.49595998460946517,
+      "grad_norm": 1.6217576265335083,
+      "learning_rate": 0.00010616462651211157,
+      "loss": 0.984,
+      "step": 2578
+    },
+    {
+      "epoch": 0.49615236629472875,
+      "grad_norm": 2.153748035430908,
+      "learning_rate": 0.00010610243536442124,
+      "loss": 0.8871,
+      "step": 2579
+    },
+    {
+      "epoch": 0.4963447479799923,
+      "grad_norm": 1.5331519842147827,
+      "learning_rate": 0.00010604024184755539,
+      "loss": 0.7787,
+      "step": 2580
+    },
+    {
+      "epoch": 0.49653712966525587,
+      "grad_norm": 2.0030224323272705,
+      "learning_rate": 0.00010597804598565968,
+      "loss": 1.0864,
+      "step": 2581
+    },
+    {
+      "epoch": 0.4967295113505194,
+      "grad_norm": 2.3356902599334717,
+      "learning_rate": 0.00010591584780288068,
+      "loss": 0.9997,
+      "step": 2582
+    },
+    {
+      "epoch": 0.496921893035783,
+      "grad_norm": 2.4976017475128174,
+      "learning_rate": 0.00010585364732336587,
+      "loss": 0.9924,
+      "step": 2583
+    },
+    {
+      "epoch": 0.4971142747210466,
+      "grad_norm": 2.275448799133301,
+      "learning_rate": 0.00010579144457126364,
+      "loss": 0.9742,
+      "step": 2584
+    },
+    {
+      "epoch": 0.4973066564063101,
+      "grad_norm": 2.13071870803833,
+      "learning_rate": 0.00010572923957072321,
+      "loss": 0.9881,
+      "step": 2585
+    },
+    {
+      "epoch": 0.4974990380915737,
+      "grad_norm": 2.1338930130004883,
+      "learning_rate": 0.0001056670323458947,
+      "loss": 0.9958,
+      "step": 2586
+    },
+    {
+      "epoch": 0.49769141977683723,
+      "grad_norm": 1.4053316116333008,
+      "learning_rate": 0.00010560482292092913,
+      "loss": 0.7916,
+      "step": 2587
+    },
+    {
+      "epoch": 0.4978838014621008,
+      "grad_norm": 2.6473872661590576,
+      "learning_rate": 0.00010554261131997834,
+      "loss": 1.0453,
+      "step": 2588
+    },
+    {
+      "epoch": 0.49807618314736435,
+      "grad_norm": 2.1080636978149414,
+      "learning_rate": 0.00010548039756719497,
+      "loss": 1.0136,
+      "step": 2589
+    },
+    {
+      "epoch": 0.49826856483262794,
+      "grad_norm": 2.4503941535949707,
+      "learning_rate": 0.0001054181816867326,
+      "loss": 1.0337,
+      "step": 2590
+    },
+    {
+      "epoch": 0.4984609465178915,
+      "grad_norm": 2.420332908630371,
+      "learning_rate": 0.0001053559637027455,
+      "loss": 1.0495,
+      "step": 2591
+    },
+    {
+      "epoch": 0.49865332820315506,
+      "grad_norm": 2.464691400527954,
+      "learning_rate": 0.00010529374363938889,
+      "loss": 1.0589,
+      "step": 2592
+    },
+    {
+      "epoch": 0.49884570988841864,
+      "grad_norm": 1.9243513345718384,
+      "learning_rate": 0.00010523152152081874,
+      "loss": 0.8785,
+      "step": 2593
+    },
+    {
+      "epoch": 0.4990380915736822,
+      "grad_norm": 2.490107536315918,
+      "learning_rate": 0.00010516929737119181,
+      "loss": 1.0154,
+      "step": 2594
+    },
+    {
+      "epoch": 0.49923047325894576,
+      "grad_norm": 1.2543786764144897,
+      "learning_rate": 0.00010510707121466567,
+      "loss": 1.0403,
+      "step": 2595
+    },
+    {
+      "epoch": 0.4994228549442093,
+      "grad_norm": 1.6083873510360718,
+      "learning_rate": 0.00010504484307539863,
+      "loss": 0.8086,
+      "step": 2596
+    },
+    {
+      "epoch": 0.4996152366294729,
+      "grad_norm": 2.393200159072876,
+      "learning_rate": 0.00010498261297754983,
+      "loss": 1.0073,
+      "step": 2597
+    },
+    {
+      "epoch": 0.4998076183147364,
+      "grad_norm": 2.686574935913086,
+      "learning_rate": 0.00010492038094527907,
+      "loss": 1.0272,
+      "step": 2598
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9218977689743042,
+      "learning_rate": 0.00010485814700274706,
+      "loss": 0.9714,
+      "step": 2599
+    },
+    {
+      "epoch": 0.5001923816852636,
+      "grad_norm": 2.0254836082458496,
+      "learning_rate": 0.0001047959111741151,
+      "loss": 0.917,
+      "step": 2600
+    },
+    {
+      "epoch": 0.5003847633705272,
+      "grad_norm": 1.7803267240524292,
+      "learning_rate": 0.00010473367348354527,
+      "loss": 1.0715,
+      "step": 2601
+    },
+    {
+      "epoch": 0.5005771450557907,
+      "grad_norm": 2.1763429641723633,
+      "learning_rate": 0.00010467143395520043,
+      "loss": 0.9678,
+      "step": 2602
+    },
+    {
+      "epoch": 0.5007695267410542,
+      "grad_norm": 1.945468783378601,
+      "learning_rate": 0.0001046091926132441,
+      "loss": 0.9384,
+      "step": 2603
+    },
+    {
+      "epoch": 0.5009619084263178,
+      "grad_norm": 2.7939844131469727,
+      "learning_rate": 0.00010454694948184045,
+      "loss": 0.977,
+      "step": 2604
+    },
+    {
+      "epoch": 0.5011542901115814,
+      "grad_norm": 1.9334068298339844,
+      "learning_rate": 0.00010448470458515449,
+      "loss": 1.0241,
+      "step": 2605
+    },
+    {
+      "epoch": 0.5013466717968449,
+      "grad_norm": 2.459949254989624,
+      "learning_rate": 0.00010442245794735179,
+      "loss": 1.0196,
+      "step": 2606
+    },
+    {
+      "epoch": 0.5015390534821085,
+      "grad_norm": 2.322113037109375,
+      "learning_rate": 0.00010436020959259862,
+      "loss": 0.9448,
+      "step": 2607
+    },
+    {
+      "epoch": 0.5017314351673721,
+      "grad_norm": 2.845614433288574,
+      "learning_rate": 0.00010429795954506203,
+      "loss": 0.9862,
+      "step": 2608
+    },
+    {
+      "epoch": 0.5019238168526357,
+      "grad_norm": 1.714548110961914,
+      "learning_rate": 0.0001042357078289095,
+      "loss": 0.9976,
+      "step": 2609
+    },
+    {
+      "epoch": 0.5021161985378992,
+      "grad_norm": 2.505648136138916,
+      "learning_rate": 0.00010417345446830937,
+      "loss": 0.9185,
+      "step": 2610
+    },
+    {
+      "epoch": 0.5023085802231627,
+      "grad_norm": 2.1993138790130615,
+      "learning_rate": 0.00010411119948743051,
+      "loss": 0.9477,
+      "step": 2611
+    },
+    {
+      "epoch": 0.5025009619084263,
+      "grad_norm": 2.169628381729126,
+      "learning_rate": 0.00010404894291044246,
+      "loss": 0.9692,
+      "step": 2612
+    },
+    {
+      "epoch": 0.5026933435936899,
+      "grad_norm": 1.474808931350708,
+      "learning_rate": 0.00010398668476151538,
+      "loss": 0.9604,
+      "step": 2613
+    },
+    {
+      "epoch": 0.5028857252789535,
+      "grad_norm": 1.0392059087753296,
+      "learning_rate": 0.00010392442506481999,
+      "loss": 1.0735,
+      "step": 2614
+    },
+    {
+      "epoch": 0.503078106964217,
+      "grad_norm": 2.3223769664764404,
+      "learning_rate": 0.0001038621638445277,
+      "loss": 0.9745,
+      "step": 2615
+    },
+    {
+      "epoch": 0.5032704886494805,
+      "grad_norm": 2.177877187728882,
+      "learning_rate": 0.00010379990112481039,
+      "loss": 1.044,
+      "step": 2616
+    },
+    {
+      "epoch": 0.5034628703347441,
+      "grad_norm": 2.33948016166687,
+      "learning_rate": 0.00010373763692984062,
+      "loss": 1.0189,
+      "step": 2617
+    },
+    {
+      "epoch": 0.5036552520200077,
+      "grad_norm": 2.399751901626587,
+      "learning_rate": 0.00010367537128379154,
+      "loss": 1.0168,
+      "step": 2618
+    },
+    {
+      "epoch": 0.5038476337052713,
+      "grad_norm": 1.693411946296692,
+      "learning_rate": 0.00010361310421083676,
+      "loss": 0.9728,
+      "step": 2619
+    },
+    {
+      "epoch": 0.5040400153905348,
+      "grad_norm": 2.605984926223755,
+      "learning_rate": 0.00010355083573515051,
+      "loss": 0.9767,
+      "step": 2620
+    },
+    {
+      "epoch": 0.5042323970757984,
+      "grad_norm": 2.108011484146118,
+      "learning_rate": 0.00010348856588090763,
+      "loss": 0.9818,
+      "step": 2621
+    },
+    {
+      "epoch": 0.504424778761062,
+      "grad_norm": 2.36415696144104,
+      "learning_rate": 0.00010342629467228331,
+      "loss": 1.057,
+      "step": 2622
+    },
+    {
+      "epoch": 0.5046171604463255,
+      "grad_norm": 2.5718843936920166,
+      "learning_rate": 0.00010336402213345344,
+      "loss": 0.8936,
+      "step": 2623
+    },
+    {
+      "epoch": 0.504809542131589,
+      "grad_norm": 3.051175832748413,
+      "learning_rate": 0.00010330174828859433,
+      "loss": 0.9858,
+      "step": 2624
+    },
+    {
+      "epoch": 0.5050019238168526,
+      "grad_norm": 2.047734498977661,
+      "learning_rate": 0.00010323947316188288,
+      "loss": 1.0063,
+      "step": 2625
+    },
+    {
+      "epoch": 0.5051943055021162,
+      "grad_norm": 1.972113847732544,
+      "learning_rate": 0.0001031771967774964,
+      "loss": 0.9811,
+      "step": 2626
+    },
+    {
+      "epoch": 0.5053866871873798,
+      "grad_norm": 2.1569063663482666,
+      "learning_rate": 0.00010311491915961271,
+      "loss": 1.0151,
+      "step": 2627
+    },
+    {
+      "epoch": 0.5055790688726434,
+      "grad_norm": 2.4344377517700195,
+      "learning_rate": 0.0001030526403324102,
+      "loss": 1.0655,
+      "step": 2628
+    },
+    {
+      "epoch": 0.5057714505579068,
+      "grad_norm": 1.9591645002365112,
+      "learning_rate": 0.0001029903603200676,
+      "loss": 0.9919,
+      "step": 2629
+    },
+    {
+      "epoch": 0.5059638322431704,
+      "grad_norm": 1.8515256643295288,
+      "learning_rate": 0.00010292807914676412,
+      "loss": 0.9789,
+      "step": 2630
+    },
+    {
+      "epoch": 0.506156213928434,
+      "grad_norm": 2.4936742782592773,
+      "learning_rate": 0.00010286579683667952,
+      "loss": 0.9882,
+      "step": 2631
+    },
+    {
+      "epoch": 0.5063485956136976,
+      "grad_norm": 2.584876775741577,
+      "learning_rate": 0.00010280351341399392,
+      "loss": 0.9781,
+      "step": 2632
+    },
+    {
+      "epoch": 0.5065409772989612,
+      "grad_norm": 2.241424560546875,
+      "learning_rate": 0.0001027412289028879,
+      "loss": 0.9922,
+      "step": 2633
+    },
+    {
+      "epoch": 0.5067333589842247,
+      "grad_norm": 2.281438112258911,
+      "learning_rate": 0.00010267894332754242,
+      "loss": 1.0418,
+      "step": 2634
+    },
+    {
+      "epoch": 0.5069257406694883,
+      "grad_norm": 1.583991527557373,
+      "learning_rate": 0.00010261665671213891,
+      "loss": 0.8308,
+      "step": 2635
+    },
+    {
+      "epoch": 0.5071181223547518,
+      "grad_norm": 2.1300384998321533,
+      "learning_rate": 0.00010255436908085919,
+      "loss": 0.963,
+      "step": 2636
+    },
+    {
+      "epoch": 0.5073105040400154,
+      "grad_norm": 2.366366147994995,
+      "learning_rate": 0.00010249208045788538,
+      "loss": 1.0139,
+      "step": 2637
+    },
+    {
+      "epoch": 0.5075028857252789,
+      "grad_norm": 1.5810497999191284,
+      "learning_rate": 0.00010242979086740019,
+      "loss": 0.7794,
+      "step": 2638
+    },
+    {
+      "epoch": 0.5076952674105425,
+      "grad_norm": 1.8887430429458618,
+      "learning_rate": 0.00010236750033358648,
+      "loss": 0.9959,
+      "step": 2639
+    },
+    {
+      "epoch": 0.5078876490958061,
+      "grad_norm": 2.695683479309082,
+      "learning_rate": 0.00010230520888062764,
+      "loss": 0.9625,
+      "step": 2640
+    },
+    {
+      "epoch": 0.5080800307810697,
+      "grad_norm": 1.1898679733276367,
+      "learning_rate": 0.00010224291653270738,
+      "loss": 0.7904,
+      "step": 2641
+    },
+    {
+      "epoch": 0.5082724124663333,
+      "grad_norm": 1.689140796661377,
+      "learning_rate": 0.00010218062331400969,
+      "loss": 1.0282,
+      "step": 2642
+    },
+    {
+      "epoch": 0.5084647941515967,
+      "grad_norm": 2.5140857696533203,
+      "learning_rate": 0.00010211832924871889,
+      "loss": 0.9597,
+      "step": 2643
+    },
+    {
+      "epoch": 0.5086571758368603,
+      "grad_norm": 1.7013176679611206,
+      "learning_rate": 0.00010205603436101978,
+      "loss": 0.999,
+      "step": 2644
+    },
+    {
+      "epoch": 0.5088495575221239,
+      "grad_norm": 2.4796695709228516,
+      "learning_rate": 0.00010199373867509734,
+      "loss": 1.0592,
+      "step": 2645
+    },
+    {
+      "epoch": 0.5090419392073875,
+      "grad_norm": 1.3791605234146118,
+      "learning_rate": 0.0001019314422151369,
+      "loss": 0.9902,
+      "step": 2646
+    },
+    {
+      "epoch": 0.509234320892651,
+      "grad_norm": 2.2043683528900146,
+      "learning_rate": 0.00010186914500532407,
+      "loss": 1.0146,
+      "step": 2647
+    },
+    {
+      "epoch": 0.5094267025779146,
+      "grad_norm": 1.7607812881469727,
+      "learning_rate": 0.00010180684706984483,
+      "loss": 1.0103,
+      "step": 2648
+    },
+    {
+      "epoch": 0.5096190842631781,
+      "grad_norm": 2.5717039108276367,
+      "learning_rate": 0.00010174454843288532,
+      "loss": 0.9779,
+      "step": 2649
+    },
+    {
+      "epoch": 0.5098114659484417,
+      "grad_norm": 1.6791799068450928,
+      "learning_rate": 0.00010168224911863205,
+      "loss": 1.0277,
+      "step": 2650
+    },
+    {
+      "epoch": 0.5100038476337053,
+      "grad_norm": 1.7939454317092896,
+      "learning_rate": 0.00010161994915127173,
+      "loss": 1.0071,
+      "step": 2651
+    },
+    {
+      "epoch": 0.5101962293189688,
+      "grad_norm": 3.036653757095337,
+      "learning_rate": 0.00010155764855499139,
+      "loss": 0.901,
+      "step": 2652
+    },
+    {
+      "epoch": 0.5103886110042324,
+      "grad_norm": 2.229135513305664,
+      "learning_rate": 0.00010149534735397823,
+      "loss": 1.0084,
+      "step": 2653
+    },
+    {
+      "epoch": 0.510580992689496,
+      "grad_norm": 2.129707098007202,
+      "learning_rate": 0.0001014330455724198,
+      "loss": 0.9767,
+      "step": 2654
+    },
+    {
+      "epoch": 0.5107733743747596,
+      "grad_norm": 2.430269718170166,
+      "learning_rate": 0.00010137074323450371,
+      "loss": 1.0041,
+      "step": 2655
+    },
+    {
+      "epoch": 0.510965756060023,
+      "grad_norm": 2.831369161605835,
+      "learning_rate": 0.00010130844036441786,
+      "loss": 1.0033,
+      "step": 2656
+    },
+    {
+      "epoch": 0.5111581377452866,
+      "grad_norm": 2.733707904815674,
+      "learning_rate": 0.00010124613698635043,
+      "loss": 0.9307,
+      "step": 2657
+    },
+    {
+      "epoch": 0.5113505194305502,
+      "grad_norm": 2.5350611209869385,
+      "learning_rate": 0.00010118383312448974,
+      "loss": 0.9988,
+      "step": 2658
+    },
+    {
+      "epoch": 0.5115429011158138,
+      "grad_norm": 1.1077784299850464,
+      "learning_rate": 0.00010112152880302426,
+      "loss": 1.061,
+      "step": 2659
+    },
+    {
+      "epoch": 0.5117352828010774,
+      "grad_norm": 1.2093782424926758,
+      "learning_rate": 0.00010105922404614265,
+      "loss": 0.7936,
+      "step": 2660
+    },
+    {
+      "epoch": 0.5119276644863409,
+      "grad_norm": 2.5146234035491943,
+      "learning_rate": 0.00010099691887803384,
+      "loss": 1.0389,
+      "step": 2661
+    },
+    {
+      "epoch": 0.5121200461716044,
+      "grad_norm": 2.4163923263549805,
+      "learning_rate": 0.00010093461332288679,
+      "loss": 1.0289,
+      "step": 2662
+    },
+    {
+      "epoch": 0.512312427856868,
+      "grad_norm": 1.6616102457046509,
+      "learning_rate": 0.00010087230740489064,
+      "loss": 1.0631,
+      "step": 2663
+    },
+    {
+      "epoch": 0.5125048095421316,
+      "grad_norm": 1.6050125360488892,
+      "learning_rate": 0.00010081000114823474,
+      "loss": 0.8025,
+      "step": 2664
+    },
+    {
+      "epoch": 0.5126971912273951,
+      "grad_norm": 1.5980675220489502,
+      "learning_rate": 0.0001007476945771085,
+      "loss": 1.0434,
+      "step": 2665
+    },
+    {
+      "epoch": 0.5128895729126587,
+      "grad_norm": 2.5925650596618652,
+      "learning_rate": 0.00010068538771570148,
+      "loss": 0.9813,
+      "step": 2666
+    },
+    {
+      "epoch": 0.5130819545979223,
+      "grad_norm": 2.4062106609344482,
+      "learning_rate": 0.00010062308058820335,
+      "loss": 1.0038,
+      "step": 2667
+    },
+    {
+      "epoch": 0.5132743362831859,
+      "grad_norm": 1.5698583126068115,
+      "learning_rate": 0.00010056077321880393,
+      "loss": 1.0363,
+      "step": 2668
+    },
+    {
+      "epoch": 0.5134667179684494,
+      "grad_norm": 1.3952311277389526,
+      "learning_rate": 0.00010049846563169296,
+      "loss": 1.0029,
+      "step": 2669
+    },
+    {
+      "epoch": 0.5136590996537129,
+      "grad_norm": 2.728330612182617,
+      "learning_rate": 0.00010043615785106051,
+      "loss": 1.0052,
+      "step": 2670
+    },
+    {
+      "epoch": 0.5138514813389765,
+      "grad_norm": 2.3668015003204346,
+      "learning_rate": 0.00010037384990109657,
+      "loss": 1.0114,
+      "step": 2671
+    },
+    {
+      "epoch": 0.5140438630242401,
+      "grad_norm": 2.1383004188537598,
+      "learning_rate": 0.00010031154180599122,
+      "loss": 0.9675,
+      "step": 2672
+    },
+    {
+      "epoch": 0.5142362447095037,
+      "grad_norm": 2.7165982723236084,
+      "learning_rate": 0.0001002492335899346,
+      "loss": 1.0016,
+      "step": 2673
+    },
+    {
+      "epoch": 0.5144286263947672,
+      "grad_norm": 1.4290685653686523,
+      "learning_rate": 0.00010018692527711695,
+      "loss": 1.053,
+      "step": 2674
+    },
+    {
+      "epoch": 0.5146210080800308,
+      "grad_norm": 2.563387632369995,
+      "learning_rate": 0.00010012461689172845,
+      "loss": 0.9841,
+      "step": 2675
+    },
+    {
+      "epoch": 0.5148133897652943,
+      "grad_norm": 2.414367198944092,
+      "learning_rate": 0.00010006230845795936,
+      "loss": 0.9533,
+      "step": 2676
+    },
+    {
+      "epoch": 0.5150057714505579,
+      "grad_norm": 1.014197587966919,
+      "learning_rate": 0.0001,
+      "loss": 1.0381,
+      "step": 2677
+    },
+    {
+      "epoch": 0.5151981531358215,
+      "grad_norm": 1.9481650590896606,
+      "learning_rate": 9.993769154204063e-05,
+      "loss": 0.8975,
+      "step": 2678
+    },
+    {
+      "epoch": 0.515390534821085,
+      "grad_norm": 2.4129269123077393,
+      "learning_rate": 9.987538310827159e-05,
+      "loss": 1.027,
+      "step": 2679
+    },
+    {
+      "epoch": 0.5155829165063486,
+      "grad_norm": 2.4477248191833496,
+      "learning_rate": 9.981307472288309e-05,
+      "loss": 0.9625,
+      "step": 2680
+    },
+    {
+      "epoch": 0.5157752981916122,
+      "grad_norm": 2.829921007156372,
+      "learning_rate": 9.975076641006542e-05,
+      "loss": 1.0342,
+      "step": 2681
+    },
+    {
+      "epoch": 0.5159676798768758,
+      "grad_norm": 2.4257307052612305,
+      "learning_rate": 9.968845819400883e-05,
+      "loss": 0.9844,
+      "step": 2682
+    },
+    {
+      "epoch": 0.5161600615621393,
+      "grad_norm": 1.678660273551941,
+      "learning_rate": 9.962615009890346e-05,
+      "loss": 1.0132,
+      "step": 2683
+    },
+    {
+      "epoch": 0.5163524432474028,
+      "grad_norm": 2.053508758544922,
+      "learning_rate": 9.956384214893949e-05,
+      "loss": 1.0577,
+      "step": 2684
+    },
+    {
+      "epoch": 0.5165448249326664,
+      "grad_norm": 1.379069447517395,
+      "learning_rate": 9.950153436830706e-05,
+      "loss": 1.005,
+      "step": 2685
+    },
+    {
+      "epoch": 0.51673720661793,
+      "grad_norm": 1.009737491607666,
+      "learning_rate": 9.94392267811961e-05,
+      "loss": 1.0868,
+      "step": 2686
+    },
+    {
+      "epoch": 0.5169295883031936,
+      "grad_norm": 2.786947250366211,
+      "learning_rate": 9.937691941179665e-05,
+      "loss": 1.0113,
+      "step": 2687
+    },
+    {
+      "epoch": 0.517121969988457,
+      "grad_norm": 2.451378345489502,
+      "learning_rate": 9.931461228429854e-05,
+      "loss": 0.9892,
+      "step": 2688
+    },
+    {
+      "epoch": 0.5173143516737206,
+      "grad_norm": 2.3930094242095947,
+      "learning_rate": 9.925230542289152e-05,
+      "loss": 1.0153,
+      "step": 2689
+    },
+    {
+      "epoch": 0.5175067333589842,
+      "grad_norm": 2.7975008487701416,
+      "learning_rate": 9.91899988517653e-05,
+      "loss": 0.9642,
+      "step": 2690
+    },
+    {
+      "epoch": 0.5176991150442478,
+      "grad_norm": 1.5756298303604126,
+      "learning_rate": 9.912769259510938e-05,
+      "loss": 0.8186,
+      "step": 2691
+    },
+    {
+      "epoch": 0.5178914967295114,
+      "grad_norm": 1.7285665273666382,
+      "learning_rate": 9.906538667711322e-05,
+      "loss": 1.0031,
+      "step": 2692
+    },
+    {
+      "epoch": 0.5180838784147749,
+      "grad_norm": 2.662400245666504,
+      "learning_rate": 9.900308112196619e-05,
+      "loss": 0.9919,
+      "step": 2693
+    },
+    {
+      "epoch": 0.5182762601000385,
+      "grad_norm": 2.3194384574890137,
+      "learning_rate": 9.894077595385737e-05,
+      "loss": 1.0008,
+      "step": 2694
+    },
+    {
+      "epoch": 0.518468641785302,
+      "grad_norm": 2.7061543464660645,
+      "learning_rate": 9.887847119697578e-05,
+      "loss": 0.9596,
+      "step": 2695
+    },
+    {
+      "epoch": 0.5186610234705656,
+      "grad_norm": 2.2214133739471436,
+      "learning_rate": 9.88161668755103e-05,
+      "loss": 1.0896,
+      "step": 2696
+    },
+    {
+      "epoch": 0.5188534051558291,
+      "grad_norm": 2.7900686264038086,
+      "learning_rate": 9.875386301364958e-05,
+      "loss": 0.9588,
+      "step": 2697
+    },
+    {
+      "epoch": 0.5190457868410927,
+      "grad_norm": 2.620541572570801,
+      "learning_rate": 9.869155963558215e-05,
+      "loss": 0.9228,
+      "step": 2698
+    },
+    {
+      "epoch": 0.5192381685263563,
+      "grad_norm": 2.3041725158691406,
+      "learning_rate": 9.862925676549635e-05,
+      "loss": 0.9716,
+      "step": 2699
+    },
+    {
+      "epoch": 0.5194305502116199,
+      "grad_norm": 2.061610698699951,
+      "learning_rate": 9.856695442758023e-05,
+      "loss": 1.0019,
+      "step": 2700
+    },
+    {
+      "epoch": 0.5196229318968835,
+      "grad_norm": 1.288503885269165,
+      "learning_rate": 9.850465264602176e-05,
+      "loss": 0.9913,
+      "step": 2701
+    },
+    {
+      "epoch": 0.5198153135821469,
+      "grad_norm": 2.419347047805786,
+      "learning_rate": 9.844235144500864e-05,
+      "loss": 0.975,
+      "step": 2702
+    },
+    {
+      "epoch": 0.5200076952674105,
+      "grad_norm": 2.3524138927459717,
+      "learning_rate": 9.83800508487283e-05,
+      "loss": 0.9502,
+      "step": 2703
+    },
+    {
+      "epoch": 0.5202000769526741,
+      "grad_norm": 1.502383828163147,
+      "learning_rate": 9.831775088136796e-05,
+      "loss": 0.9471,
+      "step": 2704
+    },
+    {
+      "epoch": 0.5203924586379377,
+      "grad_norm": 2.352806806564331,
+      "learning_rate": 9.825545156711472e-05,
+      "loss": 0.9357,
+      "step": 2705
+    },
+    {
+      "epoch": 0.5205848403232012,
+      "grad_norm": 2.178239583969116,
+      "learning_rate": 9.819315293015518e-05,
+      "loss": 0.9682,
+      "step": 2706
+    },
+    {
+      "epoch": 0.5207772220084648,
+      "grad_norm": 1.8177083730697632,
+      "learning_rate": 9.813085499467594e-05,
+      "loss": 0.941,
+      "step": 2707
+    },
+    {
+      "epoch": 0.5209696036937284,
+      "grad_norm": 1.1784108877182007,
+      "learning_rate": 9.806855778486315e-05,
+      "loss": 1.0204,
+      "step": 2708
+    },
+    {
+      "epoch": 0.5211619853789919,
+      "grad_norm": 2.6698784828186035,
+      "learning_rate": 9.800626132490269e-05,
+      "loss": 1.015,
+      "step": 2709
+    },
+    {
+      "epoch": 0.5213543670642555,
+      "grad_norm": 2.1720590591430664,
+      "learning_rate": 9.794396563898022e-05,
+      "loss": 0.9417,
+      "step": 2710
+    },
+    {
+      "epoch": 0.521546748749519,
+      "grad_norm": 1.824765682220459,
+      "learning_rate": 9.788167075128113e-05,
+      "loss": 0.917,
+      "step": 2711
+    },
+    {
+      "epoch": 0.5217391304347826,
+      "grad_norm": 2.1400346755981445,
+      "learning_rate": 9.781937668599035e-05,
+      "loss": 0.9759,
+      "step": 2712
+    },
+    {
+      "epoch": 0.5219315121200462,
+      "grad_norm": 2.5830116271972656,
+      "learning_rate": 9.775708346729263e-05,
+      "loss": 0.9656,
+      "step": 2713
+    },
+    {
+      "epoch": 0.5221238938053098,
+      "grad_norm": 2.16532301902771,
+      "learning_rate": 9.769479111937238e-05,
+      "loss": 0.9745,
+      "step": 2714
+    },
+    {
+      "epoch": 0.5223162754905732,
+      "grad_norm": 1.4259300231933594,
+      "learning_rate": 9.763249966641353e-05,
+      "loss": 0.7845,
+      "step": 2715
+    },
+    {
+      "epoch": 0.5225086571758368,
+      "grad_norm": 2.2977073192596436,
+      "learning_rate": 9.757020913259987e-05,
+      "loss": 1.0497,
+      "step": 2716
+    },
+    {
+      "epoch": 0.5227010388611004,
+      "grad_norm": 2.091386556625366,
+      "learning_rate": 9.750791954211464e-05,
+      "loss": 1.0269,
+      "step": 2717
+    },
+    {
+      "epoch": 0.522893420546364,
+      "grad_norm": 2.4128551483154297,
+      "learning_rate": 9.744563091914085e-05,
+      "loss": 0.9631,
+      "step": 2718
+    },
+    {
+      "epoch": 0.5230858022316276,
+      "grad_norm": 2.4828128814697266,
+      "learning_rate": 9.738334328786112e-05,
+      "loss": 1.0192,
+      "step": 2719
+    },
+    {
+      "epoch": 0.5232781839168911,
+      "grad_norm": 1.367488145828247,
+      "learning_rate": 9.732105667245759e-05,
+      "loss": 1.0485,
+      "step": 2720
+    },
+    {
+      "epoch": 0.5234705656021547,
+      "grad_norm": 2.4676811695098877,
+      "learning_rate": 9.725877109711212e-05,
+      "loss": 1.0589,
+      "step": 2721
+    },
+    {
+      "epoch": 0.5236629472874182,
+      "grad_norm": 1.241827368736267,
+      "learning_rate": 9.71964865860061e-05,
+      "loss": 1.0082,
+      "step": 2722
+    },
+    {
+      "epoch": 0.5238553289726818,
+      "grad_norm": 1.8989801406860352,
+      "learning_rate": 9.713420316332049e-05,
+      "loss": 1.0009,
+      "step": 2723
+    },
+    {
+      "epoch": 0.5240477106579453,
+      "grad_norm": 2.160090446472168,
+      "learning_rate": 9.707192085323589e-05,
+      "loss": 1.045,
+      "step": 2724
+    },
+    {
+      "epoch": 0.5242400923432089,
+      "grad_norm": 1.283718228340149,
+      "learning_rate": 9.700963967993246e-05,
+      "loss": 0.9707,
+      "step": 2725
+    },
+    {
+      "epoch": 0.5244324740284725,
+      "grad_norm": 2.071525812149048,
+      "learning_rate": 9.694735966758982e-05,
+      "loss": 1.0217,
+      "step": 2726
+    },
+    {
+      "epoch": 0.5246248557137361,
+      "grad_norm": 2.2497403621673584,
+      "learning_rate": 9.688508084038729e-05,
+      "loss": 0.941,
+      "step": 2727
+    },
+    {
+      "epoch": 0.5248172373989997,
+      "grad_norm": 1.6205912828445435,
+      "learning_rate": 9.682280322250366e-05,
+      "loss": 0.9102,
+      "step": 2728
+    },
+    {
+      "epoch": 0.5250096190842631,
+      "grad_norm": 1.5740729570388794,
+      "learning_rate": 9.676052683811715e-05,
+      "loss": 1.0114,
+      "step": 2729
+    },
+    {
+      "epoch": 0.5252020007695267,
+      "grad_norm": 2.286900043487549,
+      "learning_rate": 9.669825171140568e-05,
+      "loss": 1.0138,
+      "step": 2730
+    },
+    {
+      "epoch": 0.5253943824547903,
+      "grad_norm": 2.2115304470062256,
+      "learning_rate": 9.66359778665466e-05,
+      "loss": 0.9527,
+      "step": 2731
+    },
+    {
+      "epoch": 0.5255867641400539,
+      "grad_norm": 2.207125663757324,
+      "learning_rate": 9.657370532771671e-05,
+      "loss": 0.9681,
+      "step": 2732
+    },
+    {
+      "epoch": 0.5257791458253175,
+      "grad_norm": 2.4019832611083984,
+      "learning_rate": 9.65114341190924e-05,
+      "loss": 0.9991,
+      "step": 2733
+    },
+    {
+      "epoch": 0.525971527510581,
+      "grad_norm": 2.0877134799957275,
+      "learning_rate": 9.64491642648495e-05,
+      "loss": 1.0477,
+      "step": 2734
+    },
+    {
+      "epoch": 0.5261639091958445,
+      "grad_norm": 2.366533041000366,
+      "learning_rate": 9.638689578916326e-05,
+      "loss": 1.0428,
+      "step": 2735
+    },
+    {
+      "epoch": 0.5263562908811081,
+      "grad_norm": 2.441722869873047,
+      "learning_rate": 9.632462871620847e-05,
+      "loss": 1.0538,
+      "step": 2736
+    },
+    {
+      "epoch": 0.5265486725663717,
+      "grad_norm": 2.6778390407562256,
+      "learning_rate": 9.62623630701594e-05,
+      "loss": 0.9884,
+      "step": 2737
+    },
+    {
+      "epoch": 0.5267410542516352,
+      "grad_norm": 1.3989328145980835,
+      "learning_rate": 9.620009887518963e-05,
+      "loss": 0.85,
+      "step": 2738
+    },
+    {
+      "epoch": 0.5269334359368988,
+      "grad_norm": 1.999101996421814,
+      "learning_rate": 9.613783615547233e-05,
+      "loss": 0.9724,
+      "step": 2739
+    },
+    {
+      "epoch": 0.5271258176221624,
+      "grad_norm": 2.5116889476776123,
+      "learning_rate": 9.607557493518005e-05,
+      "loss": 0.9618,
+      "step": 2740
+    },
+    {
+      "epoch": 0.527318199307426,
+      "grad_norm": 2.0821802616119385,
+      "learning_rate": 9.601331523848464e-05,
+      "loss": 0.9601,
+      "step": 2741
+    },
+    {
+      "epoch": 0.5275105809926895,
+      "grad_norm": 2.214961528778076,
+      "learning_rate": 9.595105708955757e-05,
+      "loss": 0.9144,
+      "step": 2742
+    },
+    {
+      "epoch": 0.527702962677953,
+      "grad_norm": 2.9128501415252686,
+      "learning_rate": 9.588880051256951e-05,
+      "loss": 1.0347,
+      "step": 2743
+    },
+    {
+      "epoch": 0.5278953443632166,
+      "grad_norm": 2.3447072505950928,
+      "learning_rate": 9.582654553169064e-05,
+      "loss": 0.9521,
+      "step": 2744
+    },
+    {
+      "epoch": 0.5280877260484802,
+      "grad_norm": 2.3202178478240967,
+      "learning_rate": 9.576429217109053e-05,
+      "loss": 0.9661,
+      "step": 2745
+    },
+    {
+      "epoch": 0.5282801077337438,
+      "grad_norm": 0.9210947155952454,
+      "learning_rate": 9.5702040454938e-05,
+      "loss": 0.7621,
+      "step": 2746
+    },
+    {
+      "epoch": 0.5284724894190073,
+      "grad_norm": 2.3480589389801025,
+      "learning_rate": 9.563979040740138e-05,
+      "loss": 0.9581,
+      "step": 2747
+    },
+    {
+      "epoch": 0.5286648711042709,
+      "grad_norm": 1.406495451927185,
+      "learning_rate": 9.557754205264826e-05,
+      "loss": 0.7737,
+      "step": 2748
+    },
+    {
+      "epoch": 0.5288572527895344,
+      "grad_norm": 2.341329336166382,
+      "learning_rate": 9.551529541484554e-05,
+      "loss": 1.0162,
+      "step": 2749
+    },
+    {
+      "epoch": 0.529049634474798,
+      "grad_norm": 2.8089146614074707,
+      "learning_rate": 9.545305051815956e-05,
+      "loss": 1.0145,
+      "step": 2750
+    },
+    {
+      "epoch": 0.5292420161600616,
+      "grad_norm": 2.510047197341919,
+      "learning_rate": 9.539080738675597e-05,
+      "loss": 1.0494,
+      "step": 2751
+    },
+    {
+      "epoch": 0.5294343978453251,
+      "grad_norm": 2.70169997215271,
+      "learning_rate": 9.53285660447996e-05,
+      "loss": 1.0155,
+      "step": 2752
+    },
+    {
+      "epoch": 0.5296267795305887,
+      "grad_norm": 2.35953426361084,
+      "learning_rate": 9.526632651645475e-05,
+      "loss": 0.9641,
+      "step": 2753
+    },
+    {
+      "epoch": 0.5298191612158523,
+      "grad_norm": 2.9310576915740967,
+      "learning_rate": 9.520408882588497e-05,
+      "loss": 1.0079,
+      "step": 2754
+    },
+    {
+      "epoch": 0.5300115429011159,
+      "grad_norm": 2.6701698303222656,
+      "learning_rate": 9.514185299725298e-05,
+      "loss": 1.024,
+      "step": 2755
+    },
+    {
+      "epoch": 0.5302039245863793,
+      "grad_norm": 2.0677106380462646,
+      "learning_rate": 9.507961905472093e-05,
+      "loss": 0.9576,
+      "step": 2756
+    },
+    {
+      "epoch": 0.5303963062716429,
+      "grad_norm": 2.25490140914917,
+      "learning_rate": 9.501738702245022e-05,
+      "loss": 0.99,
+      "step": 2757
+    },
+    {
+      "epoch": 0.5305886879569065,
+      "grad_norm": 2.222522735595703,
+      "learning_rate": 9.495515692460138e-05,
+      "loss": 0.9575,
+      "step": 2758
+    },
+    {
+      "epoch": 0.5307810696421701,
+      "grad_norm": 2.4623396396636963,
+      "learning_rate": 9.489292878533434e-05,
+      "loss": 0.9324,
+      "step": 2759
+    },
+    {
+      "epoch": 0.5309734513274337,
+      "grad_norm": 2.294175863265991,
+      "learning_rate": 9.483070262880822e-05,
+      "loss": 1.087,
+      "step": 2760
+    },
+    {
+      "epoch": 0.5311658330126972,
+      "grad_norm": 2.4922547340393066,
+      "learning_rate": 9.476847847918127e-05,
+      "loss": 1.0069,
+      "step": 2761
+    },
+    {
+      "epoch": 0.5313582146979607,
+      "grad_norm": 2.3559279441833496,
+      "learning_rate": 9.47062563606111e-05,
+      "loss": 1.0549,
+      "step": 2762
+    },
+    {
+      "epoch": 0.5315505963832243,
+      "grad_norm": 2.467170000076294,
+      "learning_rate": 9.464403629725453e-05,
+      "loss": 0.9681,
+      "step": 2763
+    },
+    {
+      "epoch": 0.5317429780684879,
+      "grad_norm": 2.303222894668579,
+      "learning_rate": 9.458181831326743e-05,
+      "loss": 0.9734,
+      "step": 2764
+    },
+    {
+      "epoch": 0.5319353597537514,
+      "grad_norm": 2.086104393005371,
+      "learning_rate": 9.451960243280506e-05,
+      "loss": 1.0235,
+      "step": 2765
+    },
+    {
+      "epoch": 0.532127741439015,
+      "grad_norm": 1.2941316366195679,
+      "learning_rate": 9.44573886800217e-05,
+      "loss": 1.0347,
+      "step": 2766
+    },
+    {
+      "epoch": 0.5323201231242786,
+      "grad_norm": 1.9823336601257324,
+      "learning_rate": 9.439517707907088e-05,
+      "loss": 1.0076,
+      "step": 2767
+    },
+    {
+      "epoch": 0.5325125048095422,
+      "grad_norm": 1.8195899724960327,
+      "learning_rate": 9.433296765410534e-05,
+      "loss": 0.9749,
+      "step": 2768
+    },
+    {
+      "epoch": 0.5327048864948057,
+      "grad_norm": 1.1837842464447021,
+      "learning_rate": 9.427076042927683e-05,
+      "loss": 0.7925,
+      "step": 2769
+    },
+    {
+      "epoch": 0.5328972681800692,
+      "grad_norm": 2.122567653656006,
+      "learning_rate": 9.420855542873637e-05,
+      "loss": 1.0023,
+      "step": 2770
+    },
+    {
+      "epoch": 0.5330896498653328,
+      "grad_norm": 2.481353998184204,
+      "learning_rate": 9.414635267663417e-05,
+      "loss": 0.9493,
+      "step": 2771
+    },
+    {
+      "epoch": 0.5332820315505964,
+      "grad_norm": 2.1729419231414795,
+      "learning_rate": 9.408415219711934e-05,
+      "loss": 0.9373,
+      "step": 2772
+    },
+    {
+      "epoch": 0.53347441323586,
+      "grad_norm": 2.255614757537842,
+      "learning_rate": 9.402195401434036e-05,
+      "loss": 0.8844,
+      "step": 2773
+    },
+    {
+      "epoch": 0.5336667949211235,
+      "grad_norm": 2.0349984169006348,
+      "learning_rate": 9.395975815244466e-05,
+      "loss": 0.9813,
+      "step": 2774
+    },
+    {
+      "epoch": 0.533859176606387,
+      "grad_norm": 1.848435640335083,
+      "learning_rate": 9.389756463557878e-05,
+      "loss": 0.9712,
+      "step": 2775
+    },
+    {
+      "epoch": 0.5340515582916506,
+      "grad_norm": 2.3729958534240723,
+      "learning_rate": 9.383537348788843e-05,
+      "loss": 1.0117,
+      "step": 2776
+    },
+    {
+      "epoch": 0.5342439399769142,
+      "grad_norm": 1.5366451740264893,
+      "learning_rate": 9.377318473351838e-05,
+      "loss": 0.8835,
+      "step": 2777
+    },
+    {
+      "epoch": 0.5344363216621778,
+      "grad_norm": 1.4311673641204834,
+      "learning_rate": 9.371099839661238e-05,
+      "loss": 0.814,
+      "step": 2778
+    },
+    {
+      "epoch": 0.5346287033474413,
+      "grad_norm": 2.306394100189209,
+      "learning_rate": 9.364881450131336e-05,
+      "loss": 0.9436,
+      "step": 2779
+    },
+    {
+      "epoch": 0.5348210850327049,
+      "grad_norm": 2.382890224456787,
+      "learning_rate": 9.358663307176323e-05,
+      "loss": 0.9758,
+      "step": 2780
+    },
+    {
+      "epoch": 0.5350134667179685,
+      "grad_norm": 2.8796603679656982,
+      "learning_rate": 9.352445413210293e-05,
+      "loss": 1.0068,
+      "step": 2781
+    },
+    {
+      "epoch": 0.535205848403232,
+      "grad_norm": 2.3932366371154785,
+      "learning_rate": 9.34622777064725e-05,
+      "loss": 1.0181,
+      "step": 2782
+    },
+    {
+      "epoch": 0.5353982300884956,
+      "grad_norm": 2.160850763320923,
+      "learning_rate": 9.340010381901101e-05,
+      "loss": 0.9547,
+      "step": 2783
+    },
+    {
+      "epoch": 0.5355906117737591,
+      "grad_norm": 1.2031151056289673,
+      "learning_rate": 9.33379324938564e-05,
+      "loss": 0.9904,
+      "step": 2784
+    },
+    {
+      "epoch": 0.5357829934590227,
+      "grad_norm": 2.400535821914673,
+      "learning_rate": 9.327576375514581e-05,
+      "loss": 0.9917,
+      "step": 2785
+    },
+    {
+      "epoch": 0.5359753751442863,
+      "grad_norm": 1.6929384469985962,
+      "learning_rate": 9.321359762701526e-05,
+      "loss": 1.0005,
+      "step": 2786
+    },
+    {
+      "epoch": 0.5361677568295499,
+      "grad_norm": 1.3004505634307861,
+      "learning_rate": 9.315143413359976e-05,
+      "loss": 0.7964,
+      "step": 2787
+    },
+    {
+      "epoch": 0.5363601385148133,
+      "grad_norm": 2.074916362762451,
+      "learning_rate": 9.308927329903332e-05,
+      "loss": 0.9094,
+      "step": 2788
+    },
+    {
+      "epoch": 0.5365525202000769,
+      "grad_norm": 2.224648952484131,
+      "learning_rate": 9.302711514744896e-05,
+      "loss": 1.0009,
+      "step": 2789
+    },
+    {
+      "epoch": 0.5367449018853405,
+      "grad_norm": 1.7067581415176392,
+      "learning_rate": 9.296495970297854e-05,
+      "loss": 1.0306,
+      "step": 2790
+    },
+    {
+      "epoch": 0.5369372835706041,
+      "grad_norm": 2.3670523166656494,
+      "learning_rate": 9.290280698975306e-05,
+      "loss": 0.935,
+      "step": 2791
+    },
+    {
+      "epoch": 0.5371296652558677,
+      "grad_norm": 2.7736823558807373,
+      "learning_rate": 9.28406570319022e-05,
+      "loss": 0.9894,
+      "step": 2792
+    },
+    {
+      "epoch": 0.5373220469411312,
+      "grad_norm": 2.3140783309936523,
+      "learning_rate": 9.27785098535548e-05,
+      "loss": 1.0144,
+      "step": 2793
+    },
+    {
+      "epoch": 0.5375144286263948,
+      "grad_norm": 2.1528942584991455,
+      "learning_rate": 9.271636547883856e-05,
+      "loss": 0.9231,
+      "step": 2794
+    },
+    {
+      "epoch": 0.5377068103116583,
+      "grad_norm": 2.5575783252716064,
+      "learning_rate": 9.265422393187997e-05,
+      "loss": 0.9533,
+      "step": 2795
+    },
+    {
+      "epoch": 0.5378991919969219,
+      "grad_norm": 2.3715262413024902,
+      "learning_rate": 9.259208523680458e-05,
+      "loss": 1.0202,
+      "step": 2796
+    },
+    {
+      "epoch": 0.5380915736821854,
+      "grad_norm": 2.2990000247955322,
+      "learning_rate": 9.252994941773678e-05,
+      "loss": 0.9347,
+      "step": 2797
+    },
+    {
+      "epoch": 0.538283955367449,
+      "grad_norm": 2.4314045906066895,
+      "learning_rate": 9.24678164987998e-05,
+      "loss": 0.9557,
+      "step": 2798
+    },
+    {
+      "epoch": 0.5384763370527126,
+      "grad_norm": 1.2306119203567505,
+      "learning_rate": 9.240568650411581e-05,
+      "loss": 0.9388,
+      "step": 2799
+    },
+    {
+      "epoch": 0.5386687187379762,
+      "grad_norm": 1.4322724342346191,
+      "learning_rate": 9.234355945780581e-05,
+      "loss": 0.9865,
+      "step": 2800
+    },
+    {
+      "epoch": 0.5388611004232398,
+      "grad_norm": 2.7461607456207275,
+      "learning_rate": 9.228143538398962e-05,
+      "loss": 1.0153,
+      "step": 2801
+    },
+    {
+      "epoch": 0.5390534821085032,
+      "grad_norm": 2.6075119972229004,
+      "learning_rate": 9.221931430678598e-05,
+      "loss": 1.0147,
+      "step": 2802
+    },
+    {
+      "epoch": 0.5392458637937668,
+      "grad_norm": 1.9562095403671265,
+      "learning_rate": 9.215719625031245e-05,
+      "loss": 0.983,
+      "step": 2803
+    },
+    {
+      "epoch": 0.5394382454790304,
+      "grad_norm": 2.8339924812316895,
+      "learning_rate": 9.209508123868533e-05,
+      "loss": 1.0422,
+      "step": 2804
+    },
+    {
+      "epoch": 0.539630627164294,
+      "grad_norm": 2.0118014812469482,
+      "learning_rate": 9.203296929601986e-05,
+      "loss": 0.9457,
+      "step": 2805
+    },
+    {
+      "epoch": 0.5398230088495575,
+      "grad_norm": 2.652696371078491,
+      "learning_rate": 9.197086044643004e-05,
+      "loss": 0.945,
+      "step": 2806
+    },
+    {
+      "epoch": 0.5400153905348211,
+      "grad_norm": 2.4902729988098145,
+      "learning_rate": 9.190875471402865e-05,
+      "loss": 0.9367,
+      "step": 2807
+    },
+    {
+      "epoch": 0.5402077722200846,
+      "grad_norm": 2.539745807647705,
+      "learning_rate": 9.184665212292722e-05,
+      "loss": 0.9748,
+      "step": 2808
+    },
+    {
+      "epoch": 0.5404001539053482,
+      "grad_norm": 2.2540934085845947,
+      "learning_rate": 9.178455269723623e-05,
+      "loss": 0.9908,
+      "step": 2809
+    },
+    {
+      "epoch": 0.5405925355906118,
+      "grad_norm": 1.4492368698120117,
+      "learning_rate": 9.172245646106469e-05,
+      "loss": 0.777,
+      "step": 2810
+    },
+    {
+      "epoch": 0.5407849172758753,
+      "grad_norm": 2.883185386657715,
+      "learning_rate": 9.16603634385206e-05,
+      "loss": 0.9784,
+      "step": 2811
+    },
+    {
+      "epoch": 0.5409772989611389,
+      "grad_norm": 1.4819533824920654,
+      "learning_rate": 9.159827365371056e-05,
+      "loss": 0.8315,
+      "step": 2812
+    },
+    {
+      "epoch": 0.5411696806464025,
+      "grad_norm": 2.507164239883423,
+      "learning_rate": 9.153618713073995e-05,
+      "loss": 1.0516,
+      "step": 2813
+    },
+    {
+      "epoch": 0.5413620623316661,
+      "grad_norm": 1.2241840362548828,
+      "learning_rate": 9.14741038937129e-05,
+      "loss": 0.9707,
+      "step": 2814
+    },
+    {
+      "epoch": 0.5415544440169295,
+      "grad_norm": 2.4441661834716797,
+      "learning_rate": 9.141202396673232e-05,
+      "loss": 1.0101,
+      "step": 2815
+    },
+    {
+      "epoch": 0.5417468257021931,
+      "grad_norm": 2.612522840499878,
+      "learning_rate": 9.134994737389969e-05,
+      "loss": 0.9965,
+      "step": 2816
+    },
+    {
+      "epoch": 0.5419392073874567,
+      "grad_norm": 2.2392992973327637,
+      "learning_rate": 9.128787413931536e-05,
+      "loss": 1.0224,
+      "step": 2817
+    },
+    {
+      "epoch": 0.5421315890727203,
+      "grad_norm": 1.791873812675476,
+      "learning_rate": 9.12258042870782e-05,
+      "loss": 0.9792,
+      "step": 2818
+    },
+    {
+      "epoch": 0.5423239707579839,
+      "grad_norm": 1.0507526397705078,
+      "learning_rate": 9.116373784128596e-05,
+      "loss": 0.9491,
+      "step": 2819
+    },
+    {
+      "epoch": 0.5425163524432474,
+      "grad_norm": 2.328798532485962,
+      "learning_rate": 9.110167482603494e-05,
+      "loss": 0.9367,
+      "step": 2820
+    },
+    {
+      "epoch": 0.542708734128511,
+      "grad_norm": 2.597554922103882,
+      "learning_rate": 9.10396152654201e-05,
+      "loss": 1.0161,
+      "step": 2821
+    },
+    {
+      "epoch": 0.5429011158137745,
+      "grad_norm": 2.138394832611084,
+      "learning_rate": 9.097755918353512e-05,
+      "loss": 1.0042,
+      "step": 2822
+    },
+    {
+      "epoch": 0.5430934974990381,
+      "grad_norm": 1.6371073722839355,
+      "learning_rate": 9.091550660447235e-05,
+      "loss": 1.0209,
+      "step": 2823
+    },
+    {
+      "epoch": 0.5432858791843016,
+      "grad_norm": 2.0671520233154297,
+      "learning_rate": 9.085345755232268e-05,
+      "loss": 1.0167,
+      "step": 2824
+    },
+    {
+      "epoch": 0.5434782608695652,
+      "grad_norm": 1.5008459091186523,
+      "learning_rate": 9.079141205117572e-05,
+      "loss": 0.9592,
+      "step": 2825
+    },
+    {
+      "epoch": 0.5436706425548288,
+      "grad_norm": 2.499250888824463,
+      "learning_rate": 9.072937012511969e-05,
+      "loss": 0.959,
+      "step": 2826
+    },
+    {
+      "epoch": 0.5438630242400924,
+      "grad_norm": 2.281681776046753,
+      "learning_rate": 9.066733179824134e-05,
+      "loss": 1.0727,
+      "step": 2827
+    },
+    {
+      "epoch": 0.544055405925356,
+      "grad_norm": 1.9935569763183594,
+      "learning_rate": 9.060529709462611e-05,
+      "loss": 0.8664,
+      "step": 2828
+    },
+    {
+      "epoch": 0.5442477876106194,
+      "grad_norm": 2.325981855392456,
+      "learning_rate": 9.054326603835807e-05,
+      "loss": 1.0232,
+      "step": 2829
+    },
+    {
+      "epoch": 0.544440169295883,
+      "grad_norm": 2.595695972442627,
+      "learning_rate": 9.04812386535197e-05,
+      "loss": 1.0373,
+      "step": 2830
+    },
+    {
+      "epoch": 0.5446325509811466,
+      "grad_norm": 2.1872215270996094,
+      "learning_rate": 9.041921496419226e-05,
+      "loss": 1.0144,
+      "step": 2831
+    },
+    {
+      "epoch": 0.5448249326664102,
+      "grad_norm": 2.8123013973236084,
+      "learning_rate": 9.035719499445544e-05,
+      "loss": 0.9885,
+      "step": 2832
+    },
+    {
+      "epoch": 0.5450173143516738,
+      "grad_norm": 2.5636675357818604,
+      "learning_rate": 9.029517876838755e-05,
+      "loss": 0.9536,
+      "step": 2833
+    },
+    {
+      "epoch": 0.5452096960369373,
+      "grad_norm": 1.4582874774932861,
+      "learning_rate": 9.023316631006537e-05,
+      "loss": 0.8056,
+      "step": 2834
+    },
+    {
+      "epoch": 0.5454020777222008,
+      "grad_norm": 1.4223617315292358,
+      "learning_rate": 9.017115764356435e-05,
+      "loss": 1.0136,
+      "step": 2835
+    },
+    {
+      "epoch": 0.5455944594074644,
+      "grad_norm": 1.3122024536132812,
+      "learning_rate": 9.010915279295831e-05,
+      "loss": 0.7654,
+      "step": 2836
+    },
+    {
+      "epoch": 0.545786841092728,
+      "grad_norm": 1.4013534784317017,
+      "learning_rate": 9.004715178231975e-05,
+      "loss": 0.9757,
+      "step": 2837
+    },
+    {
+      "epoch": 0.5459792227779915,
+      "grad_norm": 1.455893874168396,
+      "learning_rate": 8.998515463571953e-05,
+      "loss": 0.7751,
+      "step": 2838
+    },
+    {
+      "epoch": 0.5461716044632551,
+      "grad_norm": 2.0834717750549316,
+      "learning_rate": 8.99231613772271e-05,
+      "loss": 0.9407,
+      "step": 2839
+    },
+    {
+      "epoch": 0.5463639861485187,
+      "grad_norm": 2.5248594284057617,
+      "learning_rate": 8.986117203091042e-05,
+      "loss": 0.9384,
+      "step": 2840
+    },
+    {
+      "epoch": 0.5465563678337823,
+      "grad_norm": 2.283001661300659,
+      "learning_rate": 8.97991866208358e-05,
+      "loss": 1.0575,
+      "step": 2841
+    },
+    {
+      "epoch": 0.5467487495190458,
+      "grad_norm": 2.602569103240967,
+      "learning_rate": 8.973720517106814e-05,
+      "loss": 0.9297,
+      "step": 2842
+    },
+    {
+      "epoch": 0.5469411312043093,
+      "grad_norm": 1.1233973503112793,
+      "learning_rate": 8.967522770567084e-05,
+      "loss": 1.0289,
+      "step": 2843
+    },
+    {
+      "epoch": 0.5471335128895729,
+      "grad_norm": 2.6593034267425537,
+      "learning_rate": 8.961325424870561e-05,
+      "loss": 0.9561,
+      "step": 2844
+    },
+    {
+      "epoch": 0.5473258945748365,
+      "grad_norm": 2.6924736499786377,
+      "learning_rate": 8.955128482423272e-05,
+      "loss": 1.0583,
+      "step": 2845
+    },
+    {
+      "epoch": 0.5475182762601001,
+      "grad_norm": 1.276232361793518,
+      "learning_rate": 8.948931945631082e-05,
+      "loss": 1.031,
+      "step": 2846
+    },
+    {
+      "epoch": 0.5477106579453636,
+      "grad_norm": 2.3787026405334473,
+      "learning_rate": 8.9427358168997e-05,
+      "loss": 1.0484,
+      "step": 2847
+    },
+    {
+      "epoch": 0.5479030396306271,
+      "grad_norm": 2.4146783351898193,
+      "learning_rate": 8.936540098634674e-05,
+      "loss": 0.9669,
+      "step": 2848
+    },
+    {
+      "epoch": 0.5480954213158907,
+      "grad_norm": 1.959571123123169,
+      "learning_rate": 8.930344793241403e-05,
+      "loss": 0.9303,
+      "step": 2849
+    },
+    {
+      "epoch": 0.5482878030011543,
+      "grad_norm": 2.351572275161743,
+      "learning_rate": 8.924149903125108e-05,
+      "loss": 0.9436,
+      "step": 2850
+    },
+    {
+      "epoch": 0.5484801846864179,
+      "grad_norm": 2.6218981742858887,
+      "learning_rate": 8.917955430690865e-05,
+      "loss": 0.9971,
+      "step": 2851
+    },
+    {
+      "epoch": 0.5486725663716814,
+      "grad_norm": 1.8094024658203125,
+      "learning_rate": 8.91176137834358e-05,
+      "loss": 0.962,
+      "step": 2852
+    },
+    {
+      "epoch": 0.548864948056945,
+      "grad_norm": 2.3851394653320312,
+      "learning_rate": 8.905567748487996e-05,
+      "loss": 0.9588,
+      "step": 2853
+    },
+    {
+      "epoch": 0.5490573297422086,
+      "grad_norm": 2.660266160964966,
+      "learning_rate": 8.899374543528694e-05,
+      "loss": 1.008,
+      "step": 2854
+    },
+    {
+      "epoch": 0.5492497114274721,
+      "grad_norm": 1.3268409967422485,
+      "learning_rate": 8.893181765870094e-05,
+      "loss": 0.7903,
+      "step": 2855
+    },
+    {
+      "epoch": 0.5494420931127356,
+      "grad_norm": 2.263685941696167,
+      "learning_rate": 8.886989417916435e-05,
+      "loss": 0.9539,
+      "step": 2856
+    },
+    {
+      "epoch": 0.5496344747979992,
+      "grad_norm": 2.4439759254455566,
+      "learning_rate": 8.880797502071809e-05,
+      "loss": 0.9508,
+      "step": 2857
+    },
+    {
+      "epoch": 0.5498268564832628,
+      "grad_norm": 3.0336687564849854,
+      "learning_rate": 8.87460602074013e-05,
+      "loss": 0.9629,
+      "step": 2858
+    },
+    {
+      "epoch": 0.5500192381685264,
+      "grad_norm": 2.191901922225952,
+      "learning_rate": 8.86841497632514e-05,
+      "loss": 0.9363,
+      "step": 2859
+    },
+    {
+      "epoch": 0.55021161985379,
+      "grad_norm": 2.8833389282226562,
+      "learning_rate": 8.862224371230417e-05,
+      "loss": 0.908,
+      "step": 2860
+    },
+    {
+      "epoch": 0.5504040015390534,
+      "grad_norm": 1.6624855995178223,
+      "learning_rate": 8.85603420785937e-05,
+      "loss": 0.9253,
+      "step": 2861
+    },
+    {
+      "epoch": 0.550596383224317,
+      "grad_norm": 2.296391725540161,
+      "learning_rate": 8.84984448861523e-05,
+      "loss": 1.0645,
+      "step": 2862
+    },
+    {
+      "epoch": 0.5507887649095806,
+      "grad_norm": 2.2961931228637695,
+      "learning_rate": 8.843655215901059e-05,
+      "loss": 0.9248,
+      "step": 2863
+    },
+    {
+      "epoch": 0.5509811465948442,
+      "grad_norm": 2.086000442504883,
+      "learning_rate": 8.837466392119752e-05,
+      "loss": 0.8737,
+      "step": 2864
+    },
+    {
+      "epoch": 0.5511735282801077,
+      "grad_norm": 2.5884346961975098,
+      "learning_rate": 8.831278019674016e-05,
+      "loss": 0.9616,
+      "step": 2865
+    },
+    {
+      "epoch": 0.5513659099653713,
+      "grad_norm": 2.645202875137329,
+      "learning_rate": 8.825090100966395e-05,
+      "loss": 0.9543,
+      "step": 2866
+    },
+    {
+      "epoch": 0.5515582916506349,
+      "grad_norm": 2.4129624366760254,
+      "learning_rate": 8.818902638399246e-05,
+      "loss": 0.9634,
+      "step": 2867
+    },
+    {
+      "epoch": 0.5517506733358984,
+      "grad_norm": 2.9746413230895996,
+      "learning_rate": 8.81271563437476e-05,
+      "loss": 1.0068,
+      "step": 2868
+    },
+    {
+      "epoch": 0.551943055021162,
+      "grad_norm": 1.8207952976226807,
+      "learning_rate": 8.806529091294948e-05,
+      "loss": 0.9495,
+      "step": 2869
+    },
+    {
+      "epoch": 0.5521354367064255,
+      "grad_norm": 2.3081514835357666,
+      "learning_rate": 8.800343011561632e-05,
+      "loss": 1.0178,
+      "step": 2870
+    },
+    {
+      "epoch": 0.5523278183916891,
+      "grad_norm": 2.0466349124908447,
+      "learning_rate": 8.794157397576464e-05,
+      "loss": 0.97,
+      "step": 2871
+    },
+    {
+      "epoch": 0.5525202000769527,
+      "grad_norm": 1.487589955329895,
+      "learning_rate": 8.787972251740915e-05,
+      "loss": 0.8284,
+      "step": 2872
+    },
+    {
+      "epoch": 0.5527125817622163,
+      "grad_norm": 2.176469564437866,
+      "learning_rate": 8.781787576456268e-05,
+      "loss": 0.9246,
+      "step": 2873
+    },
+    {
+      "epoch": 0.5529049634474797,
+      "grad_norm": 2.570037603378296,
+      "learning_rate": 8.775603374123626e-05,
+      "loss": 1.0011,
+      "step": 2874
+    },
+    {
+      "epoch": 0.5530973451327433,
+      "grad_norm": 2.529344320297241,
+      "learning_rate": 8.769419647143917e-05,
+      "loss": 0.9753,
+      "step": 2875
+    },
+    {
+      "epoch": 0.5532897268180069,
+      "grad_norm": 1.9869474172592163,
+      "learning_rate": 8.763236397917864e-05,
+      "loss": 0.9586,
+      "step": 2876
+    },
+    {
+      "epoch": 0.5534821085032705,
+      "grad_norm": 2.4638688564300537,
+      "learning_rate": 8.757053628846029e-05,
+      "loss": 0.9651,
+      "step": 2877
+    },
+    {
+      "epoch": 0.5536744901885341,
+      "grad_norm": 1.51243257522583,
+      "learning_rate": 8.750871342328771e-05,
+      "loss": 0.978,
+      "step": 2878
+    },
+    {
+      "epoch": 0.5538668718737976,
+      "grad_norm": 2.544590711593628,
+      "learning_rate": 8.744689540766264e-05,
+      "loss": 0.9876,
+      "step": 2879
+    },
+    {
+      "epoch": 0.5540592535590612,
+      "grad_norm": 1.0219151973724365,
+      "learning_rate": 8.7385082265585e-05,
+      "loss": 1.0045,
+      "step": 2880
+    },
+    {
+      "epoch": 0.5542516352443247,
+      "grad_norm": 2.1152772903442383,
+      "learning_rate": 8.732327402105279e-05,
+      "loss": 0.9501,
+      "step": 2881
+    },
+    {
+      "epoch": 0.5544440169295883,
+      "grad_norm": 2.21590518951416,
+      "learning_rate": 8.726147069806205e-05,
+      "loss": 0.9953,
+      "step": 2882
+    },
+    {
+      "epoch": 0.5546363986148518,
+      "grad_norm": 2.654555320739746,
+      "learning_rate": 8.719967232060697e-05,
+      "loss": 0.9717,
+      "step": 2883
+    },
+    {
+      "epoch": 0.5548287803001154,
+      "grad_norm": 2.308943271636963,
+      "learning_rate": 8.713787891267989e-05,
+      "loss": 0.9142,
+      "step": 2884
+    },
+    {
+      "epoch": 0.555021161985379,
+      "grad_norm": 2.5446982383728027,
+      "learning_rate": 8.707609049827102e-05,
+      "loss": 0.9822,
+      "step": 2885
+    },
+    {
+      "epoch": 0.5552135436706426,
+      "grad_norm": 2.3957126140594482,
+      "learning_rate": 8.70143071013688e-05,
+      "loss": 0.9537,
+      "step": 2886
+    },
+    {
+      "epoch": 0.5554059253559062,
+      "grad_norm": 1.5333623886108398,
+      "learning_rate": 8.695252874595971e-05,
+      "loss": 0.9298,
+      "step": 2887
+    },
+    {
+      "epoch": 0.5555983070411696,
+      "grad_norm": 1.265673041343689,
+      "learning_rate": 8.689075545602815e-05,
+      "loss": 0.9432,
+      "step": 2888
+    },
+    {
+      "epoch": 0.5557906887264332,
+      "grad_norm": 0.9394447207450867,
+      "learning_rate": 8.68289872555567e-05,
+      "loss": 0.7703,
+      "step": 2889
+    },
+    {
+      "epoch": 0.5559830704116968,
+      "grad_norm": 2.6259982585906982,
+      "learning_rate": 8.676722416852594e-05,
+      "loss": 0.9792,
+      "step": 2890
+    },
+    {
+      "epoch": 0.5561754520969604,
+      "grad_norm": 2.4074759483337402,
+      "learning_rate": 8.670546621891434e-05,
+      "loss": 0.9836,
+      "step": 2891
+    },
+    {
+      "epoch": 0.556367833782224,
+      "grad_norm": 2.5017449855804443,
+      "learning_rate": 8.664371343069849e-05,
+      "loss": 1.0112,
+      "step": 2892
+    },
+    {
+      "epoch": 0.5565602154674875,
+      "grad_norm": 2.1577064990997314,
+      "learning_rate": 8.658196582785296e-05,
+      "loss": 0.9434,
+      "step": 2893
+    },
+    {
+      "epoch": 0.556752597152751,
+      "grad_norm": 2.331444501876831,
+      "learning_rate": 8.652022343435028e-05,
+      "loss": 0.9822,
+      "step": 2894
+    },
+    {
+      "epoch": 0.5569449788380146,
+      "grad_norm": 2.4514801502227783,
+      "learning_rate": 8.645848627416103e-05,
+      "loss": 0.9576,
+      "step": 2895
+    },
+    {
+      "epoch": 0.5571373605232782,
+      "grad_norm": 2.56319522857666,
+      "learning_rate": 8.63967543712536e-05,
+      "loss": 0.9865,
+      "step": 2896
+    },
+    {
+      "epoch": 0.5573297422085417,
+      "grad_norm": 2.035451889038086,
+      "learning_rate": 8.633502774959453e-05,
+      "loss": 0.9929,
+      "step": 2897
+    },
+    {
+      "epoch": 0.5575221238938053,
+      "grad_norm": 1.1011499166488647,
+      "learning_rate": 8.627330643314818e-05,
+      "loss": 0.9579,
+      "step": 2898
+    },
+    {
+      "epoch": 0.5577145055790689,
+      "grad_norm": 2.560652494430542,
+      "learning_rate": 8.62115904458769e-05,
+      "loss": 0.954,
+      "step": 2899
+    },
+    {
+      "epoch": 0.5579068872643325,
+      "grad_norm": 2.000943422317505,
+      "learning_rate": 8.614987981174094e-05,
+      "loss": 0.9476,
+      "step": 2900
+    },
+    {
+      "epoch": 0.558099268949596,
+      "grad_norm": 2.8439745903015137,
+      "learning_rate": 8.608817455469854e-05,
+      "loss": 0.9556,
+      "step": 2901
+    },
+    {
+      "epoch": 0.5582916506348595,
+      "grad_norm": 1.6371135711669922,
+      "learning_rate": 8.602647469870574e-05,
+      "loss": 1.0922,
+      "step": 2902
+    },
+    {
+      "epoch": 0.5584840323201231,
+      "grad_norm": 1.6030280590057373,
+      "learning_rate": 8.596478026771659e-05,
+      "loss": 1.0366,
+      "step": 2903
+    },
+    {
+      "epoch": 0.5586764140053867,
+      "grad_norm": 2.6914680004119873,
+      "learning_rate": 8.590309128568303e-05,
+      "loss": 1.0183,
+      "step": 2904
+    },
+    {
+      "epoch": 0.5588687956906503,
+      "grad_norm": 1.0260518789291382,
+      "learning_rate": 8.584140777655476e-05,
+      "loss": 1.0274,
+      "step": 2905
+    },
+    {
+      "epoch": 0.5590611773759138,
+      "grad_norm": 0.8957763910293579,
+      "learning_rate": 8.577972976427949e-05,
+      "loss": 0.9675,
+      "step": 2906
+    },
+    {
+      "epoch": 0.5592535590611774,
+      "grad_norm": 1.0725457668304443,
+      "learning_rate": 8.571805727280277e-05,
+      "loss": 0.9618,
+      "step": 2907
+    },
+    {
+      "epoch": 0.5594459407464409,
+      "grad_norm": 1.8119853734970093,
+      "learning_rate": 8.565639032606794e-05,
+      "loss": 1.0119,
+      "step": 2908
+    },
+    {
+      "epoch": 0.5596383224317045,
+      "grad_norm": 1.4730665683746338,
+      "learning_rate": 8.559472894801623e-05,
+      "loss": 0.9393,
+      "step": 2909
+    },
+    {
+      "epoch": 0.5598307041169681,
+      "grad_norm": 2.6006827354431152,
+      "learning_rate": 8.553307316258677e-05,
+      "loss": 0.9336,
+      "step": 2910
+    },
+    {
+      "epoch": 0.5600230858022316,
+      "grad_norm": 2.8509881496429443,
+      "learning_rate": 8.547142299371641e-05,
+      "loss": 0.9346,
+      "step": 2911
+    },
+    {
+      "epoch": 0.5602154674874952,
+      "grad_norm": 1.9600105285644531,
+      "learning_rate": 8.540977846533985e-05,
+      "loss": 0.9874,
+      "step": 2912
+    },
+    {
+      "epoch": 0.5604078491727588,
+      "grad_norm": 2.1424479484558105,
+      "learning_rate": 8.534813960138969e-05,
+      "loss": 0.9757,
+      "step": 2913
+    },
+    {
+      "epoch": 0.5606002308580224,
+      "grad_norm": 2.426703453063965,
+      "learning_rate": 8.528650642579617e-05,
+      "loss": 0.9652,
+      "step": 2914
+    },
+    {
+      "epoch": 0.5607926125432858,
+      "grad_norm": 2.3616063594818115,
+      "learning_rate": 8.522487896248749e-05,
+      "loss": 1.0282,
+      "step": 2915
+    },
+    {
+      "epoch": 0.5609849942285494,
+      "grad_norm": 2.363126516342163,
+      "learning_rate": 8.51632572353895e-05,
+      "loss": 1.0741,
+      "step": 2916
+    },
+    {
+      "epoch": 0.561177375913813,
+      "grad_norm": 2.5406930446624756,
+      "learning_rate": 8.51016412684259e-05,
+      "loss": 0.968,
+      "step": 2917
+    },
+    {
+      "epoch": 0.5613697575990766,
+      "grad_norm": 2.3184618949890137,
+      "learning_rate": 8.504003108551814e-05,
+      "loss": 1.0208,
+      "step": 2918
+    },
+    {
+      "epoch": 0.5615621392843402,
+      "grad_norm": 2.451568126678467,
+      "learning_rate": 8.497842671058538e-05,
+      "loss": 1.0465,
+      "step": 2919
+    },
+    {
+      "epoch": 0.5617545209696037,
+      "grad_norm": 2.056405782699585,
+      "learning_rate": 8.491682816754456e-05,
+      "loss": 0.9726,
+      "step": 2920
+    },
+    {
+      "epoch": 0.5619469026548672,
+      "grad_norm": 2.2048871517181396,
+      "learning_rate": 8.485523548031042e-05,
+      "loss": 1.0244,
+      "step": 2921
+    },
+    {
+      "epoch": 0.5621392843401308,
+      "grad_norm": 2.683910846710205,
+      "learning_rate": 8.479364867279529e-05,
+      "loss": 1.0041,
+      "step": 2922
+    },
+    {
+      "epoch": 0.5623316660253944,
+      "grad_norm": 2.8482115268707275,
+      "learning_rate": 8.47320677689093e-05,
+      "loss": 0.9992,
+      "step": 2923
+    },
+    {
+      "epoch": 0.5625240477106579,
+      "grad_norm": 2.202807664871216,
+      "learning_rate": 8.467049279256033e-05,
+      "loss": 0.961,
+      "step": 2924
+    },
+    {
+      "epoch": 0.5627164293959215,
+      "grad_norm": 1.1087948083877563,
+      "learning_rate": 8.460892376765386e-05,
+      "loss": 1.0654,
+      "step": 2925
+    },
+    {
+      "epoch": 0.5629088110811851,
+      "grad_norm": 2.195103645324707,
+      "learning_rate": 8.454736071809308e-05,
+      "loss": 0.9213,
+      "step": 2926
+    },
+    {
+      "epoch": 0.5631011927664487,
+      "grad_norm": 2.3391573429107666,
+      "learning_rate": 8.448580366777897e-05,
+      "loss": 0.9904,
+      "step": 2927
+    },
+    {
+      "epoch": 0.5632935744517122,
+      "grad_norm": 2.106934070587158,
+      "learning_rate": 8.442425264061e-05,
+      "loss": 0.9934,
+      "step": 2928
+    },
+    {
+      "epoch": 0.5634859561369757,
+      "grad_norm": 2.68582820892334,
+      "learning_rate": 8.436270766048245e-05,
+      "loss": 0.9473,
+      "step": 2929
+    },
+    {
+      "epoch": 0.5636783378222393,
+      "grad_norm": 2.6358673572540283,
+      "learning_rate": 8.430116875129022e-05,
+      "loss": 0.9903,
+      "step": 2930
+    },
+    {
+      "epoch": 0.5638707195075029,
+      "grad_norm": 1.4265855550765991,
+      "learning_rate": 8.42396359369248e-05,
+      "loss": 0.9769,
+      "step": 2931
+    },
+    {
+      "epoch": 0.5640631011927665,
+      "grad_norm": 2.2840688228607178,
+      "learning_rate": 8.417810924127533e-05,
+      "loss": 1.0277,
+      "step": 2932
+    },
+    {
+      "epoch": 0.56425548287803,
+      "grad_norm": 1.8403658866882324,
+      "learning_rate": 8.411658868822866e-05,
+      "loss": 0.9983,
+      "step": 2933
+    },
+    {
+      "epoch": 0.5644478645632935,
+      "grad_norm": 2.0796451568603516,
+      "learning_rate": 8.40550743016691e-05,
+      "loss": 0.9272,
+      "step": 2934
+    },
+    {
+      "epoch": 0.5646402462485571,
+      "grad_norm": 2.3246867656707764,
+      "learning_rate": 8.399356610547869e-05,
+      "loss": 0.9654,
+      "step": 2935
+    },
+    {
+      "epoch": 0.5648326279338207,
+      "grad_norm": 2.299088954925537,
+      "learning_rate": 8.39320641235371e-05,
+      "loss": 0.958,
+      "step": 2936
+    },
+    {
+      "epoch": 0.5650250096190843,
+      "grad_norm": 1.8961584568023682,
+      "learning_rate": 8.38705683797214e-05,
+      "loss": 0.977,
+      "step": 2937
+    },
+    {
+      "epoch": 0.5652173913043478,
+      "grad_norm": 1.9893218278884888,
+      "learning_rate": 8.380907889790641e-05,
+      "loss": 0.9201,
+      "step": 2938
+    },
+    {
+      "epoch": 0.5654097729896114,
+      "grad_norm": 2.3838741779327393,
+      "learning_rate": 8.374759570196449e-05,
+      "loss": 0.9051,
+      "step": 2939
+    },
+    {
+      "epoch": 0.565602154674875,
+      "grad_norm": 2.0770463943481445,
+      "learning_rate": 8.368611881576547e-05,
+      "loss": 0.9598,
+      "step": 2940
+    },
+    {
+      "epoch": 0.5657945363601385,
+      "grad_norm": 2.3089005947113037,
+      "learning_rate": 8.362464826317686e-05,
+      "loss": 0.9633,
+      "step": 2941
+    },
+    {
+      "epoch": 0.5659869180454021,
+      "grad_norm": 2.3526206016540527,
+      "learning_rate": 8.356318406806358e-05,
+      "loss": 0.9979,
+      "step": 2942
+    },
+    {
+      "epoch": 0.5661792997306656,
+      "grad_norm": 1.4229483604431152,
+      "learning_rate": 8.35017262542882e-05,
+      "loss": 0.9795,
+      "step": 2943
+    },
+    {
+      "epoch": 0.5663716814159292,
+      "grad_norm": 1.6258875131607056,
+      "learning_rate": 8.344027484571075e-05,
+      "loss": 1.0124,
+      "step": 2944
+    },
+    {
+      "epoch": 0.5665640631011928,
+      "grad_norm": 2.4749748706817627,
+      "learning_rate": 8.337882986618876e-05,
+      "loss": 0.9225,
+      "step": 2945
+    },
+    {
+      "epoch": 0.5667564447864564,
+      "grad_norm": 2.285562038421631,
+      "learning_rate": 8.331739133957728e-05,
+      "loss": 0.9395,
+      "step": 2946
+    },
+    {
+      "epoch": 0.5669488264717198,
+      "grad_norm": 2.4392495155334473,
+      "learning_rate": 8.325595928972894e-05,
+      "loss": 0.9609,
+      "step": 2947
+    },
+    {
+      "epoch": 0.5671412081569834,
+      "grad_norm": 1.0204542875289917,
+      "learning_rate": 8.319453374049366e-05,
+      "loss": 0.976,
+      "step": 2948
+    },
+    {
+      "epoch": 0.567333589842247,
+      "grad_norm": 2.1571195125579834,
+      "learning_rate": 8.313311471571901e-05,
+      "loss": 0.9284,
+      "step": 2949
+    },
+    {
+      "epoch": 0.5675259715275106,
+      "grad_norm": 2.040574073791504,
+      "learning_rate": 8.307170223925003e-05,
+      "loss": 0.9245,
+      "step": 2950
+    },
+    {
+      "epoch": 0.5677183532127742,
+      "grad_norm": 2.282928228378296,
+      "learning_rate": 8.301029633492905e-05,
+      "loss": 0.984,
+      "step": 2951
+    },
+    {
+      "epoch": 0.5679107348980377,
+      "grad_norm": 1.3967009782791138,
+      "learning_rate": 8.294889702659602e-05,
+      "loss": 1.0168,
+      "step": 2952
+    },
+    {
+      "epoch": 0.5681031165833013,
+      "grad_norm": 2.1526477336883545,
+      "learning_rate": 8.288750433808827e-05,
+      "loss": 1.0019,
+      "step": 2953
+    },
+    {
+      "epoch": 0.5682954982685648,
+      "grad_norm": 2.3585240840911865,
+      "learning_rate": 8.28261182932405e-05,
+      "loss": 1.0105,
+      "step": 2954
+    },
+    {
+      "epoch": 0.5684878799538284,
+      "grad_norm": 2.3312392234802246,
+      "learning_rate": 8.276473891588492e-05,
+      "loss": 0.983,
+      "step": 2955
+    },
+    {
+      "epoch": 0.5686802616390919,
+      "grad_norm": 2.490865707397461,
+      "learning_rate": 8.270336622985116e-05,
+      "loss": 0.9857,
+      "step": 2956
+    },
+    {
+      "epoch": 0.5688726433243555,
+      "grad_norm": 2.279103994369507,
+      "learning_rate": 8.264200025896616e-05,
+      "loss": 1.0192,
+      "step": 2957
+    },
+    {
+      "epoch": 0.5690650250096191,
+      "grad_norm": 1.6234809160232544,
+      "learning_rate": 8.258064102705427e-05,
+      "loss": 0.9808,
+      "step": 2958
+    },
+    {
+      "epoch": 0.5692574066948827,
+      "grad_norm": 1.8026167154312134,
+      "learning_rate": 8.251928855793735e-05,
+      "loss": 1.0261,
+      "step": 2959
+    },
+    {
+      "epoch": 0.5694497883801463,
+      "grad_norm": 2.430746078491211,
+      "learning_rate": 8.245794287543447e-05,
+      "loss": 0.9406,
+      "step": 2960
+    },
+    {
+      "epoch": 0.5696421700654097,
+      "grad_norm": 1.5460247993469238,
+      "learning_rate": 8.239660400336214e-05,
+      "loss": 0.9706,
+      "step": 2961
+    },
+    {
+      "epoch": 0.5698345517506733,
+      "grad_norm": 3.213765859603882,
+      "learning_rate": 8.233527196553427e-05,
+      "loss": 1.0311,
+      "step": 2962
+    },
+    {
+      "epoch": 0.5700269334359369,
+      "grad_norm": 2.613698959350586,
+      "learning_rate": 8.227394678576203e-05,
+      "loss": 0.9543,
+      "step": 2963
+    },
+    {
+      "epoch": 0.5702193151212005,
+      "grad_norm": 2.6336817741394043,
+      "learning_rate": 8.221262848785395e-05,
+      "loss": 0.9609,
+      "step": 2964
+    },
+    {
+      "epoch": 0.570411696806464,
+      "grad_norm": 1.7937211990356445,
+      "learning_rate": 8.215131709561598e-05,
+      "loss": 1.0531,
+      "step": 2965
+    },
+    {
+      "epoch": 0.5706040784917276,
+      "grad_norm": 2.3503024578094482,
+      "learning_rate": 8.20900126328512e-05,
+      "loss": 0.9354,
+      "step": 2966
+    },
+    {
+      "epoch": 0.5707964601769911,
+      "grad_norm": 1.243074893951416,
+      "learning_rate": 8.202871512336024e-05,
+      "loss": 0.982,
+      "step": 2967
+    },
+    {
+      "epoch": 0.5709888418622547,
+      "grad_norm": 1.8825892210006714,
+      "learning_rate": 8.196742459094077e-05,
+      "loss": 1.0392,
+      "step": 2968
+    },
+    {
+      "epoch": 0.5711812235475183,
+      "grad_norm": 1.4940333366394043,
+      "learning_rate": 8.190614105938796e-05,
+      "loss": 0.8106,
+      "step": 2969
+    },
+    {
+      "epoch": 0.5713736052327818,
+      "grad_norm": 2.2627973556518555,
+      "learning_rate": 8.184486455249424e-05,
+      "loss": 0.957,
+      "step": 2970
+    },
+    {
+      "epoch": 0.5715659869180454,
+      "grad_norm": 2.0414226055145264,
+      "learning_rate": 8.178359509404917e-05,
+      "loss": 0.9927,
+      "step": 2971
+    },
+    {
+      "epoch": 0.571758368603309,
+      "grad_norm": 2.164883613586426,
+      "learning_rate": 8.172233270783966e-05,
+      "loss": 0.9333,
+      "step": 2972
+    },
+    {
+      "epoch": 0.5719507502885726,
+      "grad_norm": 2.2567594051361084,
+      "learning_rate": 8.166107741764997e-05,
+      "loss": 0.9376,
+      "step": 2973
+    },
+    {
+      "epoch": 0.572143131973836,
+      "grad_norm": 1.6009259223937988,
+      "learning_rate": 8.159982924726141e-05,
+      "loss": 0.9618,
+      "step": 2974
+    },
+    {
+      "epoch": 0.5723355136590996,
+      "grad_norm": 2.723482608795166,
+      "learning_rate": 8.153858822045266e-05,
+      "loss": 0.9819,
+      "step": 2975
+    },
+    {
+      "epoch": 0.5725278953443632,
+      "grad_norm": 2.230592966079712,
+      "learning_rate": 8.147735436099967e-05,
+      "loss": 0.9437,
+      "step": 2976
+    },
+    {
+      "epoch": 0.5727202770296268,
+      "grad_norm": 2.71713924407959,
+      "learning_rate": 8.141612769267544e-05,
+      "loss": 0.9818,
+      "step": 2977
+    },
+    {
+      "epoch": 0.5729126587148904,
+      "grad_norm": 2.0010218620300293,
+      "learning_rate": 8.135490823925027e-05,
+      "loss": 0.957,
+      "step": 2978
+    },
+    {
+      "epoch": 0.5731050404001539,
+      "grad_norm": 2.4697353839874268,
+      "learning_rate": 8.129369602449175e-05,
+      "loss": 0.9866,
+      "step": 2979
+    },
+    {
+      "epoch": 0.5732974220854175,
+      "grad_norm": 1.5031039714813232,
+      "learning_rate": 8.123249107216446e-05,
+      "loss": 0.9336,
+      "step": 2980
+    },
+    {
+      "epoch": 0.573489803770681,
+      "grad_norm": 2.3132028579711914,
+      "learning_rate": 8.117129340603032e-05,
+      "loss": 0.9555,
+      "step": 2981
+    },
+    {
+      "epoch": 0.5736821854559446,
+      "grad_norm": 1.6299386024475098,
+      "learning_rate": 8.111010304984841e-05,
+      "loss": 0.9244,
+      "step": 2982
+    },
+    {
+      "epoch": 0.5738745671412081,
+      "grad_norm": 2.4035587310791016,
+      "learning_rate": 8.104892002737487e-05,
+      "loss": 0.8893,
+      "step": 2983
+    },
+    {
+      "epoch": 0.5740669488264717,
+      "grad_norm": 1.9704598188400269,
+      "learning_rate": 8.098774436236308e-05,
+      "loss": 0.9561,
+      "step": 2984
+    },
+    {
+      "epoch": 0.5742593305117353,
+      "grad_norm": 2.6687705516815186,
+      "learning_rate": 8.092657607856356e-05,
+      "loss": 0.9963,
+      "step": 2985
+    },
+    {
+      "epoch": 0.5744517121969989,
+      "grad_norm": 1.5619436502456665,
+      "learning_rate": 8.086541519972388e-05,
+      "loss": 0.7819,
+      "step": 2986
+    },
+    {
+      "epoch": 0.5746440938822625,
+      "grad_norm": 2.3015499114990234,
+      "learning_rate": 8.080426174958886e-05,
+      "loss": 0.9674,
+      "step": 2987
+    },
+    {
+      "epoch": 0.5748364755675259,
+      "grad_norm": 1.821818470954895,
+      "learning_rate": 8.074311575190038e-05,
+      "loss": 0.9729,
+      "step": 2988
+    },
+    {
+      "epoch": 0.5750288572527895,
+      "grad_norm": 2.4858620166778564,
+      "learning_rate": 8.068197723039737e-05,
+      "loss": 0.9406,
+      "step": 2989
+    },
+    {
+      "epoch": 0.5752212389380531,
+      "grad_norm": 1.4262418746948242,
+      "learning_rate": 8.062084620881597e-05,
+      "loss": 0.8228,
+      "step": 2990
+    },
+    {
+      "epoch": 0.5754136206233167,
+      "grad_norm": 1.4863728284835815,
+      "learning_rate": 8.055972271088932e-05,
+      "loss": 0.8028,
+      "step": 2991
+    },
+    {
+      "epoch": 0.5756060023085803,
+      "grad_norm": 1.9669840335845947,
+      "learning_rate": 8.049860676034761e-05,
+      "loss": 0.9028,
+      "step": 2992
+    },
+    {
+      "epoch": 0.5757983839938438,
+      "grad_norm": 2.4914638996124268,
+      "learning_rate": 8.043749838091827e-05,
+      "loss": 0.9408,
+      "step": 2993
+    },
+    {
+      "epoch": 0.5759907656791073,
+      "grad_norm": 2.6181435585021973,
+      "learning_rate": 8.037639759632556e-05,
+      "loss": 0.972,
+      "step": 2994
+    },
+    {
+      "epoch": 0.5761831473643709,
+      "grad_norm": 2.2281534671783447,
+      "learning_rate": 8.031530443029099e-05,
+      "loss": 1.0115,
+      "step": 2995
+    },
+    {
+      "epoch": 0.5763755290496345,
+      "grad_norm": 2.042670488357544,
+      "learning_rate": 8.025421890653302e-05,
+      "loss": 0.9713,
+      "step": 2996
+    },
+    {
+      "epoch": 0.576567910734898,
+      "grad_norm": 2.2801671028137207,
+      "learning_rate": 8.019314104876712e-05,
+      "loss": 0.954,
+      "step": 2997
+    },
+    {
+      "epoch": 0.5767602924201616,
+      "grad_norm": 2.695814371109009,
+      "learning_rate": 8.013207088070582e-05,
+      "loss": 0.9783,
+      "step": 2998
+    },
+    {
+      "epoch": 0.5769526741054252,
+      "grad_norm": 1.2960927486419678,
+      "learning_rate": 8.007100842605872e-05,
+      "loss": 0.812,
+      "step": 2999
+    },
+    {
+      "epoch": 0.5771450557906888,
+      "grad_norm": 2.5093579292297363,
+      "learning_rate": 8.000995370853227e-05,
+      "loss": 0.9004,
+      "step": 3000
+    },
+    {
+      "epoch": 0.5773374374759523,
+      "grad_norm": 1.8676396608352661,
+      "learning_rate": 7.994890675183008e-05,
+      "loss": 0.9653,
+      "step": 3001
+    },
+    {
+      "epoch": 0.5775298191612158,
+      "grad_norm": 2.4046225547790527,
+      "learning_rate": 7.98878675796527e-05,
+      "loss": 0.8986,
+      "step": 3002
+    },
+    {
+      "epoch": 0.5777222008464794,
+      "grad_norm": 1.5028818845748901,
+      "learning_rate": 7.98268362156976e-05,
+      "loss": 0.8157,
+      "step": 3003
+    },
+    {
+      "epoch": 0.577914582531743,
+      "grad_norm": 2.5474462509155273,
+      "learning_rate": 7.976581268365923e-05,
+      "loss": 1.046,
+      "step": 3004
+    },
+    {
+      "epoch": 0.5781069642170066,
+      "grad_norm": 2.501190185546875,
+      "learning_rate": 7.97047970072291e-05,
+      "loss": 0.9549,
+      "step": 3005
+    },
+    {
+      "epoch": 0.5782993459022701,
+      "grad_norm": 2.625722885131836,
+      "learning_rate": 7.964378921009552e-05,
+      "loss": 0.9708,
+      "step": 3006
+    },
+    {
+      "epoch": 0.5784917275875336,
+      "grad_norm": 2.572222948074341,
+      "learning_rate": 7.958278931594385e-05,
+      "loss": 0.9617,
+      "step": 3007
+    },
+    {
+      "epoch": 0.5786841092727972,
+      "grad_norm": 2.2147834300994873,
+      "learning_rate": 7.952179734845641e-05,
+      "loss": 1.0029,
+      "step": 3008
+    },
+    {
+      "epoch": 0.5788764909580608,
+      "grad_norm": 2.1288554668426514,
+      "learning_rate": 7.946081333131227e-05,
+      "loss": 0.9557,
+      "step": 3009
+    },
+    {
+      "epoch": 0.5790688726433244,
+      "grad_norm": 2.2991089820861816,
+      "learning_rate": 7.93998372881876e-05,
+      "loss": 0.9844,
+      "step": 3010
+    },
+    {
+      "epoch": 0.5792612543285879,
+      "grad_norm": 2.672959089279175,
+      "learning_rate": 7.93388692427554e-05,
+      "loss": 0.9941,
+      "step": 3011
+    },
+    {
+      "epoch": 0.5794536360138515,
+      "grad_norm": 1.8224947452545166,
+      "learning_rate": 7.92779092186855e-05,
+      "loss": 0.9769,
+      "step": 3012
+    },
+    {
+      "epoch": 0.5796460176991151,
+      "grad_norm": 1.728642225265503,
+      "learning_rate": 7.921695723964472e-05,
+      "loss": 0.9738,
+      "step": 3013
+    },
+    {
+      "epoch": 0.5798383993843786,
+      "grad_norm": 2.2238874435424805,
+      "learning_rate": 7.915601332929679e-05,
+      "loss": 0.9933,
+      "step": 3014
+    },
+    {
+      "epoch": 0.5800307810696421,
+      "grad_norm": 2.2212283611297607,
+      "learning_rate": 7.909507751130209e-05,
+      "loss": 0.9498,
+      "step": 3015
+    },
+    {
+      "epoch": 0.5802231627549057,
+      "grad_norm": 2.1713554859161377,
+      "learning_rate": 7.903414980931814e-05,
+      "loss": 0.9748,
+      "step": 3016
+    },
+    {
+      "epoch": 0.5804155444401693,
+      "grad_norm": 2.5553112030029297,
+      "learning_rate": 7.897323024699907e-05,
+      "loss": 1.0114,
+      "step": 3017
+    },
+    {
+      "epoch": 0.5806079261254329,
+      "grad_norm": 2.4614005088806152,
+      "learning_rate": 7.891231884799599e-05,
+      "loss": 1.0087,
+      "step": 3018
+    },
+    {
+      "epoch": 0.5808003078106965,
+      "grad_norm": 2.560838222503662,
+      "learning_rate": 7.885141563595684e-05,
+      "loss": 1.063,
+      "step": 3019
+    },
+    {
+      "epoch": 0.58099268949596,
+      "grad_norm": 1.53875732421875,
+      "learning_rate": 7.879052063452625e-05,
+      "loss": 0.9844,
+      "step": 3020
+    },
+    {
+      "epoch": 0.5811850711812235,
+      "grad_norm": 1.5490409135818481,
+      "learning_rate": 7.872963386734583e-05,
+      "loss": 0.9965,
+      "step": 3021
+    },
+    {
+      "epoch": 0.5813774528664871,
+      "grad_norm": 2.518900156021118,
+      "learning_rate": 7.866875535805394e-05,
+      "loss": 1.0642,
+      "step": 3022
+    },
+    {
+      "epoch": 0.5815698345517507,
+      "grad_norm": 2.467639923095703,
+      "learning_rate": 7.860788513028565e-05,
+      "loss": 0.9052,
+      "step": 3023
+    },
+    {
+      "epoch": 0.5817622162370142,
+      "grad_norm": 2.5492095947265625,
+      "learning_rate": 7.854702320767289e-05,
+      "loss": 0.9101,
+      "step": 3024
+    },
+    {
+      "epoch": 0.5819545979222778,
+      "grad_norm": 1.3729774951934814,
+      "learning_rate": 7.848616961384442e-05,
+      "loss": 0.8988,
+      "step": 3025
+    },
+    {
+      "epoch": 0.5821469796075414,
+      "grad_norm": 2.572052001953125,
+      "learning_rate": 7.842532437242559e-05,
+      "loss": 0.9684,
+      "step": 3026
+    },
+    {
+      "epoch": 0.582339361292805,
+      "grad_norm": 2.3798367977142334,
+      "learning_rate": 7.836448750703868e-05,
+      "loss": 0.949,
+      "step": 3027
+    },
+    {
+      "epoch": 0.5825317429780685,
+      "grad_norm": 1.423946738243103,
+      "learning_rate": 7.83036590413027e-05,
+      "loss": 0.927,
+      "step": 3028
+    },
+    {
+      "epoch": 0.582724124663332,
+      "grad_norm": 1.969128966331482,
+      "learning_rate": 7.824283899883326e-05,
+      "loss": 1.0183,
+      "step": 3029
+    },
+    {
+      "epoch": 0.5829165063485956,
+      "grad_norm": 2.645221471786499,
+      "learning_rate": 7.818202740324286e-05,
+      "loss": 0.9573,
+      "step": 3030
+    },
+    {
+      "epoch": 0.5831088880338592,
+      "grad_norm": 1.4257628917694092,
+      "learning_rate": 7.812122427814068e-05,
+      "loss": 0.7558,
+      "step": 3031
+    },
+    {
+      "epoch": 0.5833012697191228,
+      "grad_norm": 2.388965129852295,
+      "learning_rate": 7.806042964713247e-05,
+      "loss": 0.9609,
+      "step": 3032
+    },
+    {
+      "epoch": 0.5834936514043862,
+      "grad_norm": 1.826205849647522,
+      "learning_rate": 7.799964353382089e-05,
+      "loss": 0.9791,
+      "step": 3033
+    },
+    {
+      "epoch": 0.5836860330896498,
+      "grad_norm": 1.7370930910110474,
+      "learning_rate": 7.793886596180522e-05,
+      "loss": 1.0258,
+      "step": 3034
+    },
+    {
+      "epoch": 0.5838784147749134,
+      "grad_norm": 1.410161018371582,
+      "learning_rate": 7.787809695468133e-05,
+      "loss": 0.8115,
+      "step": 3035
+    },
+    {
+      "epoch": 0.584070796460177,
+      "grad_norm": 2.2277512550354004,
+      "learning_rate": 7.78173365360419e-05,
+      "loss": 0.9629,
+      "step": 3036
+    },
+    {
+      "epoch": 0.5842631781454406,
+      "grad_norm": 2.4949684143066406,
+      "learning_rate": 7.775658472947623e-05,
+      "loss": 0.9992,
+      "step": 3037
+    },
+    {
+      "epoch": 0.5844555598307041,
+      "grad_norm": 1.5214868783950806,
+      "learning_rate": 7.769584155857018e-05,
+      "loss": 1.024,
+      "step": 3038
+    },
+    {
+      "epoch": 0.5846479415159677,
+      "grad_norm": 2.5681357383728027,
+      "learning_rate": 7.763510704690644e-05,
+      "loss": 0.8721,
+      "step": 3039
+    },
+    {
+      "epoch": 0.5848403232012312,
+      "grad_norm": 1.3334184885025024,
+      "learning_rate": 7.757438121806414e-05,
+      "loss": 1.0494,
+      "step": 3040
+    },
+    {
+      "epoch": 0.5850327048864948,
+      "grad_norm": 2.535418748855591,
+      "learning_rate": 7.75136640956192e-05,
+      "loss": 0.9559,
+      "step": 3041
+    },
+    {
+      "epoch": 0.5852250865717584,
+      "grad_norm": 2.9983770847320557,
+      "learning_rate": 7.745295570314411e-05,
+      "loss": 1.0172,
+      "step": 3042
+    },
+    {
+      "epoch": 0.5854174682570219,
+      "grad_norm": 2.900477170944214,
+      "learning_rate": 7.739225606420793e-05,
+      "loss": 0.9151,
+      "step": 3043
+    },
+    {
+      "epoch": 0.5856098499422855,
+      "grad_norm": 2.698265552520752,
+      "learning_rate": 7.733156520237632e-05,
+      "loss": 0.984,
+      "step": 3044
+    },
+    {
+      "epoch": 0.5858022316275491,
+      "grad_norm": 2.7079031467437744,
+      "learning_rate": 7.727088314121165e-05,
+      "loss": 0.9882,
+      "step": 3045
+    },
+    {
+      "epoch": 0.5859946133128127,
+      "grad_norm": 2.7068569660186768,
+      "learning_rate": 7.721020990427268e-05,
+      "loss": 0.9838,
+      "step": 3046
+    },
+    {
+      "epoch": 0.5861869949980761,
+      "grad_norm": 2.6314587593078613,
+      "learning_rate": 7.71495455151149e-05,
+      "loss": 0.8976,
+      "step": 3047
+    },
+    {
+      "epoch": 0.5863793766833397,
+      "grad_norm": 2.2872323989868164,
+      "learning_rate": 7.708888999729036e-05,
+      "loss": 0.9999,
+      "step": 3048
+    },
+    {
+      "epoch": 0.5865717583686033,
+      "grad_norm": 2.741396427154541,
+      "learning_rate": 7.702824337434755e-05,
+      "loss": 0.9691,
+      "step": 3049
+    },
+    {
+      "epoch": 0.5867641400538669,
+      "grad_norm": 2.029571771621704,
+      "learning_rate": 7.696760566983159e-05,
+      "loss": 0.9614,
+      "step": 3050
+    },
+    {
+      "epoch": 0.5869565217391305,
+      "grad_norm": 2.515380382537842,
+      "learning_rate": 7.690697690728416e-05,
+      "loss": 0.9355,
+      "step": 3051
+    },
+    {
+      "epoch": 0.587148903424394,
+      "grad_norm": 2.0886178016662598,
+      "learning_rate": 7.68463571102434e-05,
+      "loss": 1.05,
+      "step": 3052
+    },
+    {
+      "epoch": 0.5873412851096576,
+      "grad_norm": 2.3460121154785156,
+      "learning_rate": 7.6785746302244e-05,
+      "loss": 1.0088,
+      "step": 3053
+    },
+    {
+      "epoch": 0.5875336667949211,
+      "grad_norm": 2.839282274246216,
+      "learning_rate": 7.67251445068172e-05,
+      "loss": 0.9569,
+      "step": 3054
+    },
+    {
+      "epoch": 0.5877260484801847,
+      "grad_norm": 2.1096842288970947,
+      "learning_rate": 7.666455174749066e-05,
+      "loss": 0.9933,
+      "step": 3055
+    },
+    {
+      "epoch": 0.5879184301654482,
+      "grad_norm": 2.0542962551116943,
+      "learning_rate": 7.66039680477886e-05,
+      "loss": 0.9737,
+      "step": 3056
+    },
+    {
+      "epoch": 0.5881108118507118,
+      "grad_norm": 1.2538275718688965,
+      "learning_rate": 7.654339343123173e-05,
+      "loss": 0.8009,
+      "step": 3057
+    },
+    {
+      "epoch": 0.5883031935359754,
+      "grad_norm": 2.226579427719116,
+      "learning_rate": 7.648282792133711e-05,
+      "loss": 0.9637,
+      "step": 3058
+    },
+    {
+      "epoch": 0.588495575221239,
+      "grad_norm": 2.4232301712036133,
+      "learning_rate": 7.642227154161841e-05,
+      "loss": 0.9812,
+      "step": 3059
+    },
+    {
+      "epoch": 0.5886879569065026,
+      "grad_norm": 2.3625378608703613,
+      "learning_rate": 7.636172431558575e-05,
+      "loss": 0.9637,
+      "step": 3060
+    },
+    {
+      "epoch": 0.588880338591766,
+      "grad_norm": 1.4711060523986816,
+      "learning_rate": 7.630118626674556e-05,
+      "loss": 0.9622,
+      "step": 3061
+    },
+    {
+      "epoch": 0.5890727202770296,
+      "grad_norm": 2.390500545501709,
+      "learning_rate": 7.624065741860084e-05,
+      "loss": 0.9374,
+      "step": 3062
+    },
+    {
+      "epoch": 0.5892651019622932,
+      "grad_norm": 2.4689321517944336,
+      "learning_rate": 7.618013779465101e-05,
+      "loss": 0.9909,
+      "step": 3063
+    },
+    {
+      "epoch": 0.5894574836475568,
+      "grad_norm": 1.929559350013733,
+      "learning_rate": 7.611962741839178e-05,
+      "loss": 1.0137,
+      "step": 3064
+    },
+    {
+      "epoch": 0.5896498653328203,
+      "grad_norm": 2.7735602855682373,
+      "learning_rate": 7.605912631331546e-05,
+      "loss": 0.9813,
+      "step": 3065
+    },
+    {
+      "epoch": 0.5898422470180839,
+      "grad_norm": 2.3141210079193115,
+      "learning_rate": 7.599863450291055e-05,
+      "loss": 0.9299,
+      "step": 3066
+    },
+    {
+      "epoch": 0.5900346287033474,
+      "grad_norm": 2.815855026245117,
+      "learning_rate": 7.593815201066215e-05,
+      "loss": 1.0162,
+      "step": 3067
+    },
+    {
+      "epoch": 0.590227010388611,
+      "grad_norm": 1.4739654064178467,
+      "learning_rate": 7.587767886005163e-05,
+      "loss": 1.042,
+      "step": 3068
+    },
+    {
+      "epoch": 0.5904193920738746,
+      "grad_norm": 2.298842668533325,
+      "learning_rate": 7.581721507455673e-05,
+      "loss": 1.0086,
+      "step": 3069
+    },
+    {
+      "epoch": 0.5906117737591381,
+      "grad_norm": 2.562221050262451,
+      "learning_rate": 7.575676067765154e-05,
+      "loss": 1.0031,
+      "step": 3070
+    },
+    {
+      "epoch": 0.5908041554444017,
+      "grad_norm": 2.5621631145477295,
+      "learning_rate": 7.569631569280661e-05,
+      "loss": 0.9536,
+      "step": 3071
+    },
+    {
+      "epoch": 0.5909965371296653,
+      "grad_norm": 2.1468679904937744,
+      "learning_rate": 7.56358801434887e-05,
+      "loss": 0.9683,
+      "step": 3072
+    },
+    {
+      "epoch": 0.5911889188149289,
+      "grad_norm": 1.3177669048309326,
+      "learning_rate": 7.5575454053161e-05,
+      "loss": 0.7794,
+      "step": 3073
+    },
+    {
+      "epoch": 0.5913813005001923,
+      "grad_norm": 2.428596258163452,
+      "learning_rate": 7.551503744528304e-05,
+      "loss": 0.9541,
+      "step": 3074
+    },
+    {
+      "epoch": 0.5915736821854559,
+      "grad_norm": 1.1514543294906616,
+      "learning_rate": 7.545463034331053e-05,
+      "loss": 1.029,
+      "step": 3075
+    },
+    {
+      "epoch": 0.5917660638707195,
+      "grad_norm": 2.5084643363952637,
+      "learning_rate": 7.539423277069568e-05,
+      "loss": 0.9751,
+      "step": 3076
+    },
+    {
+      "epoch": 0.5919584455559831,
+      "grad_norm": 3.4945223331451416,
+      "learning_rate": 7.53338447508869e-05,
+      "loss": 0.9265,
+      "step": 3077
+    },
+    {
+      "epoch": 0.5921508272412467,
+      "grad_norm": 2.4107792377471924,
+      "learning_rate": 7.52734663073288e-05,
+      "loss": 0.9313,
+      "step": 3078
+    },
+    {
+      "epoch": 0.5923432089265102,
+      "grad_norm": 2.5077195167541504,
+      "learning_rate": 7.521309746346246e-05,
+      "loss": 0.9326,
+      "step": 3079
+    },
+    {
+      "epoch": 0.5925355906117737,
+      "grad_norm": 1.2014130353927612,
+      "learning_rate": 7.515273824272514e-05,
+      "loss": 0.9979,
+      "step": 3080
+    },
+    {
+      "epoch": 0.5927279722970373,
+      "grad_norm": 2.3815224170684814,
+      "learning_rate": 7.509238866855033e-05,
+      "loss": 0.9726,
+      "step": 3081
+    },
+    {
+      "epoch": 0.5929203539823009,
+      "grad_norm": 2.638853073120117,
+      "learning_rate": 7.503204876436785e-05,
+      "loss": 0.9293,
+      "step": 3082
+    },
+    {
+      "epoch": 0.5931127356675644,
+      "grad_norm": 2.628793954849243,
+      "learning_rate": 7.497171855360371e-05,
+      "loss": 0.9702,
+      "step": 3083
+    },
+    {
+      "epoch": 0.593305117352828,
+      "grad_norm": 2.560638427734375,
+      "learning_rate": 7.491139805968018e-05,
+      "loss": 0.9793,
+      "step": 3084
+    },
+    {
+      "epoch": 0.5934974990380916,
+      "grad_norm": 2.3393752574920654,
+      "learning_rate": 7.485108730601571e-05,
+      "loss": 0.985,
+      "step": 3085
+    },
+    {
+      "epoch": 0.5936898807233552,
+      "grad_norm": 2.1935019493103027,
+      "learning_rate": 7.479078631602512e-05,
+      "loss": 0.9487,
+      "step": 3086
+    },
+    {
+      "epoch": 0.5938822624086187,
+      "grad_norm": 2.2085673809051514,
+      "learning_rate": 7.47304951131192e-05,
+      "loss": 1.077,
+      "step": 3087
+    },
+    {
+      "epoch": 0.5940746440938822,
+      "grad_norm": 2.5997302532196045,
+      "learning_rate": 7.467021372070516e-05,
+      "loss": 0.9627,
+      "step": 3088
+    },
+    {
+      "epoch": 0.5942670257791458,
+      "grad_norm": 2.4768340587615967,
+      "learning_rate": 7.46099421621863e-05,
+      "loss": 0.997,
+      "step": 3089
+    },
+    {
+      "epoch": 0.5944594074644094,
+      "grad_norm": 2.9473509788513184,
+      "learning_rate": 7.454968046096204e-05,
+      "loss": 0.9553,
+      "step": 3090
+    },
+    {
+      "epoch": 0.594651789149673,
+      "grad_norm": 2.140681505203247,
+      "learning_rate": 7.448942864042818e-05,
+      "loss": 0.9045,
+      "step": 3091
+    },
+    {
+      "epoch": 0.5948441708349365,
+      "grad_norm": 2.396846055984497,
+      "learning_rate": 7.442918672397642e-05,
+      "loss": 0.9871,
+      "step": 3092
+    },
+    {
+      "epoch": 0.5950365525202,
+      "grad_norm": 2.440427780151367,
+      "learning_rate": 7.43689547349948e-05,
+      "loss": 0.9581,
+      "step": 3093
+    },
+    {
+      "epoch": 0.5952289342054636,
+      "grad_norm": 2.301631450653076,
+      "learning_rate": 7.43087326968675e-05,
+      "loss": 0.9468,
+      "step": 3094
+    },
+    {
+      "epoch": 0.5954213158907272,
+      "grad_norm": 2.5089406967163086,
+      "learning_rate": 7.424852063297469e-05,
+      "loss": 0.9684,
+      "step": 3095
+    },
+    {
+      "epoch": 0.5956136975759908,
+      "grad_norm": 1.2552037239074707,
+      "learning_rate": 7.418831856669286e-05,
+      "loss": 1.0728,
+      "step": 3096
+    },
+    {
+      "epoch": 0.5958060792612543,
+      "grad_norm": 2.5644888877868652,
+      "learning_rate": 7.41281265213945e-05,
+      "loss": 0.9461,
+      "step": 3097
+    },
+    {
+      "epoch": 0.5959984609465179,
+      "grad_norm": 2.3302698135375977,
+      "learning_rate": 7.406794452044816e-05,
+      "loss": 0.9366,
+      "step": 3098
+    },
+    {
+      "epoch": 0.5961908426317815,
+      "grad_norm": 1.46123206615448,
+      "learning_rate": 7.400777258721865e-05,
+      "loss": 0.7997,
+      "step": 3099
+    },
+    {
+      "epoch": 0.596383224317045,
+      "grad_norm": 1.945576786994934,
+      "learning_rate": 7.394761074506679e-05,
+      "loss": 0.9657,
+      "step": 3100
+    },
+    {
+      "epoch": 0.5965756060023086,
+      "grad_norm": 1.5564539432525635,
+      "learning_rate": 7.38874590173494e-05,
+      "loss": 0.9666,
+      "step": 3101
+    },
+    {
+      "epoch": 0.5967679876875721,
+      "grad_norm": 1.8886839151382446,
+      "learning_rate": 7.382731742741952e-05,
+      "loss": 0.8965,
+      "step": 3102
+    },
+    {
+      "epoch": 0.5969603693728357,
+      "grad_norm": 2.6267921924591064,
+      "learning_rate": 7.376718599862621e-05,
+      "loss": 0.9992,
+      "step": 3103
+    },
+    {
+      "epoch": 0.5971527510580993,
+      "grad_norm": 2.20382022857666,
+      "learning_rate": 7.370706475431446e-05,
+      "loss": 0.9069,
+      "step": 3104
+    },
+    {
+      "epoch": 0.5973451327433629,
+      "grad_norm": 2.4351985454559326,
+      "learning_rate": 7.364695371782546e-05,
+      "loss": 0.9418,
+      "step": 3105
+    },
+    {
+      "epoch": 0.5975375144286263,
+      "grad_norm": 2.6241419315338135,
+      "learning_rate": 7.358685291249644e-05,
+      "loss": 0.9831,
+      "step": 3106
+    },
+    {
+      "epoch": 0.5977298961138899,
+      "grad_norm": 2.0473713874816895,
+      "learning_rate": 7.352676236166051e-05,
+      "loss": 0.9751,
+      "step": 3107
+    },
+    {
+      "epoch": 0.5979222777991535,
+      "grad_norm": 1.4935952425003052,
+      "learning_rate": 7.346668208864695e-05,
+      "loss": 0.9263,
+      "step": 3108
+    },
+    {
+      "epoch": 0.5981146594844171,
+      "grad_norm": 2.3961925506591797,
+      "learning_rate": 7.340661211678099e-05,
+      "loss": 0.9502,
+      "step": 3109
+    },
+    {
+      "epoch": 0.5983070411696807,
+      "grad_norm": 2.561790943145752,
+      "learning_rate": 7.33465524693838e-05,
+      "loss": 0.9022,
+      "step": 3110
+    },
+    {
+      "epoch": 0.5984994228549442,
+      "grad_norm": 2.215423822402954,
+      "learning_rate": 7.328650316977264e-05,
+      "loss": 0.9253,
+      "step": 3111
+    },
+    {
+      "epoch": 0.5986918045402078,
+      "grad_norm": 1.8758195638656616,
+      "learning_rate": 7.322646424126078e-05,
+      "loss": 0.965,
+      "step": 3112
+    },
+    {
+      "epoch": 0.5988841862254713,
+      "grad_norm": 2.802783489227295,
+      "learning_rate": 7.31664357071573e-05,
+      "loss": 0.9354,
+      "step": 3113
+    },
+    {
+      "epoch": 0.5990765679107349,
+      "grad_norm": 2.156964063644409,
+      "learning_rate": 7.310641759076741e-05,
+      "loss": 0.862,
+      "step": 3114
+    },
+    {
+      "epoch": 0.5992689495959984,
+      "grad_norm": 2.8961167335510254,
+      "learning_rate": 7.304640991539215e-05,
+      "loss": 1.0559,
+      "step": 3115
+    },
+    {
+      "epoch": 0.599461331281262,
+      "grad_norm": 2.402493715286255,
+      "learning_rate": 7.298641270432862e-05,
+      "loss": 0.9355,
+      "step": 3116
+    },
+    {
+      "epoch": 0.5996537129665256,
+      "grad_norm": 1.8459093570709229,
+      "learning_rate": 7.292642598086981e-05,
+      "loss": 0.9906,
+      "step": 3117
+    },
+    {
+      "epoch": 0.5998460946517892,
+      "grad_norm": 2.8664441108703613,
+      "learning_rate": 7.286644976830457e-05,
+      "loss": 0.9324,
+      "step": 3118
+    },
+    {
+      "epoch": 0.6000384763370528,
+      "grad_norm": 1.702199101448059,
+      "learning_rate": 7.280648408991775e-05,
+      "loss": 0.9152,
+      "step": 3119
+    },
+    {
+      "epoch": 0.6002308580223162,
+      "grad_norm": 2.7765071392059326,
+      "learning_rate": 7.274652896899015e-05,
+      "loss": 1.0771,
+      "step": 3120
+    },
+    {
+      "epoch": 0.6004232397075798,
+      "grad_norm": 2.218201160430908,
+      "learning_rate": 7.268658442879833e-05,
+      "loss": 1.0197,
+      "step": 3121
+    },
+    {
+      "epoch": 0.6006156213928434,
+      "grad_norm": 2.617560863494873,
+      "learning_rate": 7.262665049261488e-05,
+      "loss": 0.9279,
+      "step": 3122
+    },
+    {
+      "epoch": 0.600808003078107,
+      "grad_norm": 2.3066165447235107,
+      "learning_rate": 7.256672718370824e-05,
+      "loss": 1.0493,
+      "step": 3123
+    },
+    {
+      "epoch": 0.6010003847633705,
+      "grad_norm": 2.242368459701538,
+      "learning_rate": 7.25068145253426e-05,
+      "loss": 0.8933,
+      "step": 3124
+    },
+    {
+      "epoch": 0.6011927664486341,
+      "grad_norm": 2.5176830291748047,
+      "learning_rate": 7.244691254077819e-05,
+      "loss": 0.9668,
+      "step": 3125
+    },
+    {
+      "epoch": 0.6013851481338977,
+      "grad_norm": 2.340456247329712,
+      "learning_rate": 7.238702125327106e-05,
+      "loss": 0.9678,
+      "step": 3126
+    },
+    {
+      "epoch": 0.6015775298191612,
+      "grad_norm": 2.0646274089813232,
+      "learning_rate": 7.232714068607295e-05,
+      "loss": 0.9538,
+      "step": 3127
+    },
+    {
+      "epoch": 0.6017699115044248,
+      "grad_norm": 2.167071580886841,
+      "learning_rate": 7.226727086243168e-05,
+      "loss": 0.9384,
+      "step": 3128
+    },
+    {
+      "epoch": 0.6019622931896883,
+      "grad_norm": 2.5887038707733154,
+      "learning_rate": 7.220741180559073e-05,
+      "loss": 1.0312,
+      "step": 3129
+    },
+    {
+      "epoch": 0.6021546748749519,
+      "grad_norm": 3.132523775100708,
+      "learning_rate": 7.214756353878942e-05,
+      "loss": 0.9374,
+      "step": 3130
+    },
+    {
+      "epoch": 0.6023470565602155,
+      "grad_norm": 2.3079707622528076,
+      "learning_rate": 7.208772608526293e-05,
+      "loss": 0.9479,
+      "step": 3131
+    },
+    {
+      "epoch": 0.6025394382454791,
+      "grad_norm": 2.2616333961486816,
+      "learning_rate": 7.202789946824227e-05,
+      "loss": 0.8848,
+      "step": 3132
+    },
+    {
+      "epoch": 0.6027318199307425,
+      "grad_norm": 2.6574676036834717,
+      "learning_rate": 7.196808371095408e-05,
+      "loss": 0.9782,
+      "step": 3133
+    },
+    {
+      "epoch": 0.6029242016160061,
+      "grad_norm": 1.3155982494354248,
+      "learning_rate": 7.190827883662099e-05,
+      "loss": 0.7821,
+      "step": 3134
+    },
+    {
+      "epoch": 0.6031165833012697,
+      "grad_norm": 2.541020154953003,
+      "learning_rate": 7.184848486846128e-05,
+      "loss": 0.9515,
+      "step": 3135
+    },
+    {
+      "epoch": 0.6033089649865333,
+      "grad_norm": 2.840898036956787,
+      "learning_rate": 7.178870182968903e-05,
+      "loss": 0.9212,
+      "step": 3136
+    },
+    {
+      "epoch": 0.6035013466717969,
+      "grad_norm": 2.4104793071746826,
+      "learning_rate": 7.172892974351406e-05,
+      "loss": 0.9822,
+      "step": 3137
+    },
+    {
+      "epoch": 0.6036937283570604,
+      "grad_norm": 2.257800817489624,
+      "learning_rate": 7.166916863314199e-05,
+      "loss": 1.0212,
+      "step": 3138
+    },
+    {
+      "epoch": 0.603886110042324,
+      "grad_norm": 2.5174777507781982,
+      "learning_rate": 7.160941852177409e-05,
+      "loss": 0.9684,
+      "step": 3139
+    },
+    {
+      "epoch": 0.6040784917275875,
+      "grad_norm": 1.4015836715698242,
+      "learning_rate": 7.154967943260748e-05,
+      "loss": 1.0603,
+      "step": 3140
+    },
+    {
+      "epoch": 0.6042708734128511,
+      "grad_norm": 2.307697057723999,
+      "learning_rate": 7.148995138883483e-05,
+      "loss": 1.0321,
+      "step": 3141
+    },
+    {
+      "epoch": 0.6044632550981146,
+      "grad_norm": 2.929783821105957,
+      "learning_rate": 7.143023441364472e-05,
+      "loss": 0.9666,
+      "step": 3142
+    },
+    {
+      "epoch": 0.6046556367833782,
+      "grad_norm": 1.706547737121582,
+      "learning_rate": 7.13705285302213e-05,
+      "loss": 0.9484,
+      "step": 3143
+    },
+    {
+      "epoch": 0.6048480184686418,
+      "grad_norm": 2.401860475540161,
+      "learning_rate": 7.131083376174441e-05,
+      "loss": 0.9585,
+      "step": 3144
+    },
+    {
+      "epoch": 0.6050404001539054,
+      "grad_norm": 2.349437713623047,
+      "learning_rate": 7.125115013138966e-05,
+      "loss": 0.9953,
+      "step": 3145
+    },
+    {
+      "epoch": 0.605232781839169,
+      "grad_norm": 2.37510085105896,
+      "learning_rate": 7.119147766232833e-05,
+      "loss": 0.9245,
+      "step": 3146
+    },
+    {
+      "epoch": 0.6054251635244324,
+      "grad_norm": 2.205674886703491,
+      "learning_rate": 7.113181637772721e-05,
+      "loss": 1.0031,
+      "step": 3147
+    },
+    {
+      "epoch": 0.605617545209696,
+      "grad_norm": 2.1904592514038086,
+      "learning_rate": 7.107216630074896e-05,
+      "loss": 0.9133,
+      "step": 3148
+    },
+    {
+      "epoch": 0.6058099268949596,
+      "grad_norm": 1.0645915269851685,
+      "learning_rate": 7.101252745455179e-05,
+      "loss": 0.7421,
+      "step": 3149
+    },
+    {
+      "epoch": 0.6060023085802232,
+      "grad_norm": 2.4612069129943848,
+      "learning_rate": 7.09528998622895e-05,
+      "loss": 1.0204,
+      "step": 3150
+    },
+    {
+      "epoch": 0.6061946902654868,
+      "grad_norm": 2.2636115550994873,
+      "learning_rate": 7.089328354711159e-05,
+      "loss": 0.9259,
+      "step": 3151
+    },
+    {
+      "epoch": 0.6063870719507503,
+      "grad_norm": 1.2048448324203491,
+      "learning_rate": 7.083367853216323e-05,
+      "loss": 1.0299,
+      "step": 3152
+    },
+    {
+      "epoch": 0.6065794536360138,
+      "grad_norm": 3.0441040992736816,
+      "learning_rate": 7.077408484058505e-05,
+      "loss": 1.0039,
+      "step": 3153
+    },
+    {
+      "epoch": 0.6067718353212774,
+      "grad_norm": 2.280167579650879,
+      "learning_rate": 7.071450249551341e-05,
+      "loss": 0.8773,
+      "step": 3154
+    },
+    {
+      "epoch": 0.606964217006541,
+      "grad_norm": 2.282015323638916,
+      "learning_rate": 7.065493152008026e-05,
+      "loss": 1.0308,
+      "step": 3155
+    },
+    {
+      "epoch": 0.6071565986918045,
+      "grad_norm": 2.621990203857422,
+      "learning_rate": 7.059537193741306e-05,
+      "loss": 1.0043,
+      "step": 3156
+    },
+    {
+      "epoch": 0.6073489803770681,
+      "grad_norm": 2.3815219402313232,
+      "learning_rate": 7.053582377063489e-05,
+      "loss": 0.9811,
+      "step": 3157
+    },
+    {
+      "epoch": 0.6075413620623317,
+      "grad_norm": 2.3689916133880615,
+      "learning_rate": 7.047628704286446e-05,
+      "loss": 0.9452,
+      "step": 3158
+    },
+    {
+      "epoch": 0.6077337437475953,
+      "grad_norm": 2.398207426071167,
+      "learning_rate": 7.041676177721588e-05,
+      "loss": 1.0586,
+      "step": 3159
+    },
+    {
+      "epoch": 0.6079261254328588,
+      "grad_norm": 1.1286805868148804,
+      "learning_rate": 7.035724799679898e-05,
+      "loss": 0.8176,
+      "step": 3160
+    },
+    {
+      "epoch": 0.6081185071181223,
+      "grad_norm": 1.6200053691864014,
+      "learning_rate": 7.029774572471903e-05,
+      "loss": 0.7879,
+      "step": 3161
+    },
+    {
+      "epoch": 0.6083108888033859,
+      "grad_norm": 2.2450897693634033,
+      "learning_rate": 7.023825498407689e-05,
+      "loss": 0.9736,
+      "step": 3162
+    },
+    {
+      "epoch": 0.6085032704886495,
+      "grad_norm": 2.1610565185546875,
+      "learning_rate": 7.017877579796885e-05,
+      "loss": 0.9772,
+      "step": 3163
+    },
+    {
+      "epoch": 0.6086956521739131,
+      "grad_norm": 2.483011245727539,
+      "learning_rate": 7.011930818948688e-05,
+      "loss": 0.9559,
+      "step": 3164
+    },
+    {
+      "epoch": 0.6088880338591766,
+      "grad_norm": 2.654447555541992,
+      "learning_rate": 7.005985218171825e-05,
+      "loss": 1.025,
+      "step": 3165
+    },
+    {
+      "epoch": 0.6090804155444401,
+      "grad_norm": 2.5617029666900635,
+      "learning_rate": 7.000040779774591e-05,
+      "loss": 0.9938,
+      "step": 3166
+    },
+    {
+      "epoch": 0.6092727972297037,
+      "grad_norm": 2.4265642166137695,
+      "learning_rate": 6.994097506064812e-05,
+      "loss": 0.9948,
+      "step": 3167
+    },
+    {
+      "epoch": 0.6094651789149673,
+      "grad_norm": 1.8030345439910889,
+      "learning_rate": 6.988155399349881e-05,
+      "loss": 0.915,
+      "step": 3168
+    },
+    {
+      "epoch": 0.6096575606002309,
+      "grad_norm": 2.799443483352661,
+      "learning_rate": 6.982214461936727e-05,
+      "loss": 1.0157,
+      "step": 3169
+    },
+    {
+      "epoch": 0.6098499422854944,
+      "grad_norm": 2.2371671199798584,
+      "learning_rate": 6.97627469613182e-05,
+      "loss": 0.9707,
+      "step": 3170
+    },
+    {
+      "epoch": 0.610042323970758,
+      "grad_norm": 2.3917031288146973,
+      "learning_rate": 6.970336104241186e-05,
+      "loss": 0.9217,
+      "step": 3171
+    },
+    {
+      "epoch": 0.6102347056560216,
+      "grad_norm": 1.2773996591567993,
+      "learning_rate": 6.964398688570395e-05,
+      "loss": 0.9433,
+      "step": 3172
+    },
+    {
+      "epoch": 0.6104270873412851,
+      "grad_norm": 2.6609926223754883,
+      "learning_rate": 6.958462451424547e-05,
+      "loss": 0.9548,
+      "step": 3173
+    },
+    {
+      "epoch": 0.6106194690265486,
+      "grad_norm": 3.143282890319824,
+      "learning_rate": 6.952527395108302e-05,
+      "loss": 0.9694,
+      "step": 3174
+    },
+    {
+      "epoch": 0.6108118507118122,
+      "grad_norm": 2.3275909423828125,
+      "learning_rate": 6.946593521925851e-05,
+      "loss": 0.9807,
+      "step": 3175
+    },
+    {
+      "epoch": 0.6110042323970758,
+      "grad_norm": 1.525238275527954,
+      "learning_rate": 6.940660834180921e-05,
+      "loss": 0.8108,
+      "step": 3176
+    },
+    {
+      "epoch": 0.6111966140823394,
+      "grad_norm": 2.1447947025299072,
+      "learning_rate": 6.934729334176792e-05,
+      "loss": 0.9355,
+      "step": 3177
+    },
+    {
+      "epoch": 0.611388995767603,
+      "grad_norm": 0.9312242865562439,
+      "learning_rate": 6.928799024216282e-05,
+      "loss": 0.7654,
+      "step": 3178
+    },
+    {
+      "epoch": 0.6115813774528664,
+      "grad_norm": 2.3158962726593018,
+      "learning_rate": 6.92286990660173e-05,
+      "loss": 0.937,
+      "step": 3179
+    },
+    {
+      "epoch": 0.61177375913813,
+      "grad_norm": 2.115764856338501,
+      "learning_rate": 6.91694198363503e-05,
+      "loss": 0.9723,
+      "step": 3180
+    },
+    {
+      "epoch": 0.6119661408233936,
+      "grad_norm": 2.4408371448516846,
+      "learning_rate": 6.911015257617605e-05,
+      "loss": 0.9941,
+      "step": 3181
+    },
+    {
+      "epoch": 0.6121585225086572,
+      "grad_norm": 1.7151280641555786,
+      "learning_rate": 6.905089730850416e-05,
+      "loss": 1.0041,
+      "step": 3182
+    },
+    {
+      "epoch": 0.6123509041939207,
+      "grad_norm": 2.3649587631225586,
+      "learning_rate": 6.89916540563395e-05,
+      "loss": 0.9682,
+      "step": 3183
+    },
+    {
+      "epoch": 0.6125432858791843,
+      "grad_norm": 2.7657108306884766,
+      "learning_rate": 6.893242284268244e-05,
+      "loss": 0.9705,
+      "step": 3184
+    },
+    {
+      "epoch": 0.6127356675644479,
+      "grad_norm": 1.6930965185165405,
+      "learning_rate": 6.887320369052848e-05,
+      "loss": 0.9765,
+      "step": 3185
+    },
+    {
+      "epoch": 0.6129280492497114,
+      "grad_norm": 2.098160743713379,
+      "learning_rate": 6.881399662286858e-05,
+      "loss": 1.0353,
+      "step": 3186
+    },
+    {
+      "epoch": 0.613120430934975,
+      "grad_norm": 2.2025485038757324,
+      "learning_rate": 6.875480166268896e-05,
+      "loss": 0.9884,
+      "step": 3187
+    },
+    {
+      "epoch": 0.6133128126202385,
+      "grad_norm": 2.717766761779785,
+      "learning_rate": 6.869561883297116e-05,
+      "loss": 0.9583,
+      "step": 3188
+    },
+    {
+      "epoch": 0.6135051943055021,
+      "grad_norm": 2.280907392501831,
+      "learning_rate": 6.863644815669197e-05,
+      "loss": 0.9396,
+      "step": 3189
+    },
+    {
+      "epoch": 0.6136975759907657,
+      "grad_norm": 1.7802282571792603,
+      "learning_rate": 6.857728965682345e-05,
+      "loss": 0.9894,
+      "step": 3190
+    },
+    {
+      "epoch": 0.6138899576760293,
+      "grad_norm": 2.541487216949463,
+      "learning_rate": 6.851814335633297e-05,
+      "loss": 0.8963,
+      "step": 3191
+    },
+    {
+      "epoch": 0.6140823393612928,
+      "grad_norm": 2.8396036624908447,
+      "learning_rate": 6.845900927818327e-05,
+      "loss": 0.9797,
+      "step": 3192
+    },
+    {
+      "epoch": 0.6142747210465563,
+      "grad_norm": 2.5017940998077393,
+      "learning_rate": 6.839988744533211e-05,
+      "loss": 0.9416,
+      "step": 3193
+    },
+    {
+      "epoch": 0.6144671027318199,
+      "grad_norm": 1.6978691816329956,
+      "learning_rate": 6.834077788073267e-05,
+      "loss": 0.9666,
+      "step": 3194
+    },
+    {
+      "epoch": 0.6146594844170835,
+      "grad_norm": 2.256253957748413,
+      "learning_rate": 6.828168060733337e-05,
+      "loss": 0.9284,
+      "step": 3195
+    },
+    {
+      "epoch": 0.6148518661023471,
+      "grad_norm": 2.2945687770843506,
+      "learning_rate": 6.822259564807768e-05,
+      "loss": 0.9788,
+      "step": 3196
+    },
+    {
+      "epoch": 0.6150442477876106,
+      "grad_norm": 2.5562586784362793,
+      "learning_rate": 6.816352302590449e-05,
+      "loss": 0.9986,
+      "step": 3197
+    },
+    {
+      "epoch": 0.6152366294728742,
+      "grad_norm": 1.531530737876892,
+      "learning_rate": 6.810446276374789e-05,
+      "loss": 0.9222,
+      "step": 3198
+    },
+    {
+      "epoch": 0.6154290111581378,
+      "grad_norm": 1.063672661781311,
+      "learning_rate": 6.804541488453697e-05,
+      "loss": 1.0123,
+      "step": 3199
+    },
+    {
+      "epoch": 0.6156213928434013,
+      "grad_norm": 2.2200071811676025,
+      "learning_rate": 6.798637941119625e-05,
+      "loss": 0.9621,
+      "step": 3200
+    },
+    {
+      "epoch": 0.6158137745286649,
+      "grad_norm": 2.1942708492279053,
+      "learning_rate": 6.792735636664532e-05,
+      "loss": 0.9852,
+      "step": 3201
+    },
+    {
+      "epoch": 0.6160061562139284,
+      "grad_norm": 1.0851913690567017,
+      "learning_rate": 6.786834577379892e-05,
+      "loss": 0.9634,
+      "step": 3202
+    },
+    {
+      "epoch": 0.616198537899192,
+      "grad_norm": 2.4275035858154297,
+      "learning_rate": 6.780934765556701e-05,
+      "loss": 0.8744,
+      "step": 3203
+    },
+    {
+      "epoch": 0.6163909195844556,
+      "grad_norm": 1.2528818845748901,
+      "learning_rate": 6.775036203485473e-05,
+      "loss": 0.9525,
+      "step": 3204
+    },
+    {
+      "epoch": 0.6165833012697192,
+      "grad_norm": 2.3385517597198486,
+      "learning_rate": 6.769138893456225e-05,
+      "loss": 0.9504,
+      "step": 3205
+    },
+    {
+      "epoch": 0.6167756829549826,
+      "grad_norm": 2.396711587905884,
+      "learning_rate": 6.763242837758504e-05,
+      "loss": 0.9623,
+      "step": 3206
+    },
+    {
+      "epoch": 0.6169680646402462,
+      "grad_norm": 2.399165391921997,
+      "learning_rate": 6.757348038681357e-05,
+      "loss": 0.9803,
+      "step": 3207
+    },
+    {
+      "epoch": 0.6171604463255098,
+      "grad_norm": 2.489154100418091,
+      "learning_rate": 6.751454498513348e-05,
+      "loss": 0.9168,
+      "step": 3208
+    },
+    {
+      "epoch": 0.6173528280107734,
+      "grad_norm": 1.9397058486938477,
+      "learning_rate": 6.745562219542554e-05,
+      "loss": 1.0257,
+      "step": 3209
+    },
+    {
+      "epoch": 0.617545209696037,
+      "grad_norm": 2.0896153450012207,
+      "learning_rate": 6.739671204056562e-05,
+      "loss": 0.8978,
+      "step": 3210
+    },
+    {
+      "epoch": 0.6177375913813005,
+      "grad_norm": 2.513519525527954,
+      "learning_rate": 6.733781454342462e-05,
+      "loss": 0.9185,
+      "step": 3211
+    },
+    {
+      "epoch": 0.617929973066564,
+      "grad_norm": 2.6459460258483887,
+      "learning_rate": 6.72789297268686e-05,
+      "loss": 0.895,
+      "step": 3212
+    },
+    {
+      "epoch": 0.6181223547518276,
+      "grad_norm": 1.4594855308532715,
+      "learning_rate": 6.722005761375873e-05,
+      "loss": 0.8139,
+      "step": 3213
+    },
+    {
+      "epoch": 0.6183147364370912,
+      "grad_norm": 2.354123592376709,
+      "learning_rate": 6.71611982269511e-05,
+      "loss": 0.9093,
+      "step": 3214
+    },
+    {
+      "epoch": 0.6185071181223547,
+      "grad_norm": 1.0529342889785767,
+      "learning_rate": 6.710235158929703e-05,
+      "loss": 1.0065,
+      "step": 3215
+    },
+    {
+      "epoch": 0.6186994998076183,
+      "grad_norm": 1.3223927021026611,
+      "learning_rate": 6.704351772364274e-05,
+      "loss": 0.9653,
+      "step": 3216
+    },
+    {
+      "epoch": 0.6188918814928819,
+      "grad_norm": 1.290169358253479,
+      "learning_rate": 6.698469665282957e-05,
+      "loss": 1.0328,
+      "step": 3217
+    },
+    {
+      "epoch": 0.6190842631781455,
+      "grad_norm": 2.56853985786438,
+      "learning_rate": 6.692588839969397e-05,
+      "loss": 0.9384,
+      "step": 3218
+    },
+    {
+      "epoch": 0.619276644863409,
+      "grad_norm": 2.4183781147003174,
+      "learning_rate": 6.686709298706722e-05,
+      "loss": 1.0208,
+      "step": 3219
+    },
+    {
+      "epoch": 0.6194690265486725,
+      "grad_norm": 2.7658841609954834,
+      "learning_rate": 6.680831043777579e-05,
+      "loss": 0.9905,
+      "step": 3220
+    },
+    {
+      "epoch": 0.6196614082339361,
+      "grad_norm": 1.3934415578842163,
+      "learning_rate": 6.674954077464107e-05,
+      "loss": 1.0665,
+      "step": 3221
+    },
+    {
+      "epoch": 0.6198537899191997,
+      "grad_norm": 2.7599377632141113,
+      "learning_rate": 6.669078402047948e-05,
+      "loss": 0.9967,
+      "step": 3222
+    },
+    {
+      "epoch": 0.6200461716044633,
+      "grad_norm": 2.55301833152771,
+      "learning_rate": 6.663204019810237e-05,
+      "loss": 0.9594,
+      "step": 3223
+    },
+    {
+      "epoch": 0.6202385532897268,
+      "grad_norm": 2.2115235328674316,
+      "learning_rate": 6.657330933031619e-05,
+      "loss": 0.9561,
+      "step": 3224
+    },
+    {
+      "epoch": 0.6204309349749904,
+      "grad_norm": 2.405978202819824,
+      "learning_rate": 6.65145914399222e-05,
+      "loss": 1.0412,
+      "step": 3225
+    },
+    {
+      "epoch": 0.6206233166602539,
+      "grad_norm": 2.3598663806915283,
+      "learning_rate": 6.645588654971678e-05,
+      "loss": 0.9765,
+      "step": 3226
+    },
+    {
+      "epoch": 0.6208156983455175,
+      "grad_norm": 2.7505757808685303,
+      "learning_rate": 6.639719468249115e-05,
+      "loss": 1.0096,
+      "step": 3227
+    },
+    {
+      "epoch": 0.6210080800307811,
+      "grad_norm": 2.226233959197998,
+      "learning_rate": 6.633851586103153e-05,
+      "loss": 0.9611,
+      "step": 3228
+    },
+    {
+      "epoch": 0.6212004617160446,
+      "grad_norm": 2.188546657562256,
+      "learning_rate": 6.627985010811902e-05,
+      "loss": 0.9512,
+      "step": 3229
+    },
+    {
+      "epoch": 0.6213928434013082,
+      "grad_norm": 2.7669975757598877,
+      "learning_rate": 6.622119744652978e-05,
+      "loss": 0.9542,
+      "step": 3230
+    },
+    {
+      "epoch": 0.6215852250865718,
+      "grad_norm": 1.3693945407867432,
+      "learning_rate": 6.616255789903466e-05,
+      "loss": 0.9421,
+      "step": 3231
+    },
+    {
+      "epoch": 0.6217776067718354,
+      "grad_norm": 2.7345938682556152,
+      "learning_rate": 6.610393148839964e-05,
+      "loss": 1.0017,
+      "step": 3232
+    },
+    {
+      "epoch": 0.6219699884570988,
+      "grad_norm": 2.576396942138672,
+      "learning_rate": 6.604531823738552e-05,
+      "loss": 1.0787,
+      "step": 3233
+    },
+    {
+      "epoch": 0.6221623701423624,
+      "grad_norm": 2.471769094467163,
+      "learning_rate": 6.598671816874794e-05,
+      "loss": 0.9841,
+      "step": 3234
+    },
+    {
+      "epoch": 0.622354751827626,
+      "grad_norm": 1.8817650079727173,
+      "learning_rate": 6.592813130523746e-05,
+      "loss": 0.9454,
+      "step": 3235
+    },
+    {
+      "epoch": 0.6225471335128896,
+      "grad_norm": 2.462751865386963,
+      "learning_rate": 6.586955766959957e-05,
+      "loss": 1.0002,
+      "step": 3236
+    },
+    {
+      "epoch": 0.6227395151981532,
+      "grad_norm": 2.4656996726989746,
+      "learning_rate": 6.581099728457451e-05,
+      "loss": 0.8967,
+      "step": 3237
+    },
+    {
+      "epoch": 0.6229318968834167,
+      "grad_norm": 2.394299268722534,
+      "learning_rate": 6.575245017289746e-05,
+      "loss": 1.0212,
+      "step": 3238
+    },
+    {
+      "epoch": 0.6231242785686802,
+      "grad_norm": 1.8570151329040527,
+      "learning_rate": 6.569391635729847e-05,
+      "loss": 0.9902,
+      "step": 3239
+    },
+    {
+      "epoch": 0.6233166602539438,
+      "grad_norm": 2.347092866897583,
+      "learning_rate": 6.563539586050233e-05,
+      "loss": 0.98,
+      "step": 3240
+    },
+    {
+      "epoch": 0.6235090419392074,
+      "grad_norm": 1.1084256172180176,
+      "learning_rate": 6.557688870522871e-05,
+      "loss": 1.0425,
+      "step": 3241
+    },
+    {
+      "epoch": 0.6237014236244709,
+      "grad_norm": 2.4646644592285156,
+      "learning_rate": 6.551839491419213e-05,
+      "loss": 0.9505,
+      "step": 3242
+    },
+    {
+      "epoch": 0.6238938053097345,
+      "grad_norm": 2.462282180786133,
+      "learning_rate": 6.545991451010185e-05,
+      "loss": 0.9568,
+      "step": 3243
+    },
+    {
+      "epoch": 0.6240861869949981,
+      "grad_norm": 2.489811897277832,
+      "learning_rate": 6.540144751566207e-05,
+      "loss": 0.9342,
+      "step": 3244
+    },
+    {
+      "epoch": 0.6242785686802617,
+      "grad_norm": 2.158465623855591,
+      "learning_rate": 6.534299395357155e-05,
+      "loss": 0.918,
+      "step": 3245
+    },
+    {
+      "epoch": 0.6244709503655252,
+      "grad_norm": 2.171138286590576,
+      "learning_rate": 6.52845538465241e-05,
+      "loss": 0.9391,
+      "step": 3246
+    },
+    {
+      "epoch": 0.6246633320507887,
+      "grad_norm": 2.773089647293091,
+      "learning_rate": 6.522612721720814e-05,
+      "loss": 1.0388,
+      "step": 3247
+    },
+    {
+      "epoch": 0.6248557137360523,
+      "grad_norm": 1.7613635063171387,
+      "learning_rate": 6.516771408830686e-05,
+      "loss": 0.993,
+      "step": 3248
+    },
+    {
+      "epoch": 0.6250480954213159,
+      "grad_norm": 1.1622836589813232,
+      "learning_rate": 6.51093144824983e-05,
+      "loss": 0.9704,
+      "step": 3249
+    },
+    {
+      "epoch": 0.6252404771065795,
+      "grad_norm": 1.713948130607605,
+      "learning_rate": 6.505092842245519e-05,
+      "loss": 0.9732,
+      "step": 3250
+    },
+    {
+      "epoch": 0.6254328587918431,
+      "grad_norm": 2.3551204204559326,
+      "learning_rate": 6.499255593084498e-05,
+      "loss": 0.9994,
+      "step": 3251
+    },
+    {
+      "epoch": 0.6256252404771065,
+      "grad_norm": 2.8866779804229736,
+      "learning_rate": 6.49341970303299e-05,
+      "loss": 0.9974,
+      "step": 3252
+    },
+    {
+      "epoch": 0.6258176221623701,
+      "grad_norm": 1.7916817665100098,
+      "learning_rate": 6.48758517435669e-05,
+      "loss": 1.0064,
+      "step": 3253
+    },
+    {
+      "epoch": 0.6260100038476337,
+      "grad_norm": 3.320791482925415,
+      "learning_rate": 6.481752009320761e-05,
+      "loss": 0.9955,
+      "step": 3254
+    },
+    {
+      "epoch": 0.6262023855328973,
+      "grad_norm": 2.0784599781036377,
+      "learning_rate": 6.475920210189836e-05,
+      "loss": 0.9207,
+      "step": 3255
+    },
+    {
+      "epoch": 0.6263947672181608,
+      "grad_norm": 3.0699403285980225,
+      "learning_rate": 6.470089779228028e-05,
+      "loss": 1.0156,
+      "step": 3256
+    },
+    {
+      "epoch": 0.6265871489034244,
+      "grad_norm": 2.288912057876587,
+      "learning_rate": 6.464260718698902e-05,
+      "loss": 0.9755,
+      "step": 3257
+    },
+    {
+      "epoch": 0.626779530588688,
+      "grad_norm": 2.695952892303467,
+      "learning_rate": 6.458433030865502e-05,
+      "loss": 0.9769,
+      "step": 3258
+    },
+    {
+      "epoch": 0.6269719122739515,
+      "grad_norm": 3.0405049324035645,
+      "learning_rate": 6.452606717990346e-05,
+      "loss": 0.9394,
+      "step": 3259
+    },
+    {
+      "epoch": 0.6271642939592151,
+      "grad_norm": 2.4883100986480713,
+      "learning_rate": 6.4467817823354e-05,
+      "loss": 1.0511,
+      "step": 3260
+    },
+    {
+      "epoch": 0.6273566756444786,
+      "grad_norm": 2.613882303237915,
+      "learning_rate": 6.440958226162104e-05,
+      "loss": 0.966,
+      "step": 3261
+    },
+    {
+      "epoch": 0.6275490573297422,
+      "grad_norm": 2.09665846824646,
+      "learning_rate": 6.43513605173137e-05,
+      "loss": 0.9158,
+      "step": 3262
+    },
+    {
+      "epoch": 0.6277414390150058,
+      "grad_norm": 1.390740156173706,
+      "learning_rate": 6.429315261303559e-05,
+      "loss": 0.9431,
+      "step": 3263
+    },
+    {
+      "epoch": 0.6279338207002694,
+      "grad_norm": 2.4367618560791016,
+      "learning_rate": 6.423495857138509e-05,
+      "loss": 0.9605,
+      "step": 3264
+    },
+    {
+      "epoch": 0.6281262023855329,
+      "grad_norm": 1.2645679712295532,
+      "learning_rate": 6.417677841495507e-05,
+      "loss": 0.9834,
+      "step": 3265
+    },
+    {
+      "epoch": 0.6283185840707964,
+      "grad_norm": 1.7312002182006836,
+      "learning_rate": 6.411861216633313e-05,
+      "loss": 0.8848,
+      "step": 3266
+    },
+    {
+      "epoch": 0.62851096575606,
+      "grad_norm": 1.4525166749954224,
+      "learning_rate": 6.406045984810135e-05,
+      "loss": 0.9669,
+      "step": 3267
+    },
+    {
+      "epoch": 0.6287033474413236,
+      "grad_norm": 2.8594558238983154,
+      "learning_rate": 6.400232148283651e-05,
+      "loss": 0.9468,
+      "step": 3268
+    },
+    {
+      "epoch": 0.6288957291265872,
+      "grad_norm": 2.722553014755249,
+      "learning_rate": 6.394419709310988e-05,
+      "loss": 0.9525,
+      "step": 3269
+    },
+    {
+      "epoch": 0.6290881108118507,
+      "grad_norm": 2.522448778152466,
+      "learning_rate": 6.388608670148741e-05,
+      "loss": 0.9001,
+      "step": 3270
+    },
+    {
+      "epoch": 0.6292804924971143,
+      "grad_norm": 1.4077800512313843,
+      "learning_rate": 6.38279903305295e-05,
+      "loss": 1.0392,
+      "step": 3271
+    },
+    {
+      "epoch": 0.6294728741823779,
+      "grad_norm": 2.317842721939087,
+      "learning_rate": 6.376990800279119e-05,
+      "loss": 0.9367,
+      "step": 3272
+    },
+    {
+      "epoch": 0.6296652558676414,
+      "grad_norm": 2.38257098197937,
+      "learning_rate": 6.371183974082203e-05,
+      "loss": 0.9804,
+      "step": 3273
+    },
+    {
+      "epoch": 0.6298576375529049,
+      "grad_norm": 2.6125688552856445,
+      "learning_rate": 6.365378556716612e-05,
+      "loss": 0.9964,
+      "step": 3274
+    },
+    {
+      "epoch": 0.6300500192381685,
+      "grad_norm": 2.6158595085144043,
+      "learning_rate": 6.359574550436208e-05,
+      "loss": 0.9081,
+      "step": 3275
+    },
+    {
+      "epoch": 0.6302424009234321,
+      "grad_norm": 1.637289047241211,
+      "learning_rate": 6.35377195749431e-05,
+      "loss": 0.9961,
+      "step": 3276
+    },
+    {
+      "epoch": 0.6304347826086957,
+      "grad_norm": 2.423309564590454,
+      "learning_rate": 6.347970780143678e-05,
+      "loss": 0.9656,
+      "step": 3277
+    },
+    {
+      "epoch": 0.6306271642939593,
+      "grad_norm": 1.2076420783996582,
+      "learning_rate": 6.342171020636533e-05,
+      "loss": 0.7881,
+      "step": 3278
+    },
+    {
+      "epoch": 0.6308195459792227,
+      "grad_norm": 1.9615356922149658,
+      "learning_rate": 6.336372681224542e-05,
+      "loss": 0.9267,
+      "step": 3279
+    },
+    {
+      "epoch": 0.6310119276644863,
+      "grad_norm": 2.2101938724517822,
+      "learning_rate": 6.33057576415882e-05,
+      "loss": 1.0197,
+      "step": 3280
+    },
+    {
+      "epoch": 0.6312043093497499,
+      "grad_norm": 2.811171531677246,
+      "learning_rate": 6.324780271689923e-05,
+      "loss": 0.9651,
+      "step": 3281
+    },
+    {
+      "epoch": 0.6313966910350135,
+      "grad_norm": 1.4341286420822144,
+      "learning_rate": 6.318986206067872e-05,
+      "loss": 1.0616,
+      "step": 3282
+    },
+    {
+      "epoch": 0.631589072720277,
+      "grad_norm": 1.5778615474700928,
+      "learning_rate": 6.313193569542113e-05,
+      "loss": 0.8168,
+      "step": 3283
+    },
+    {
+      "epoch": 0.6317814544055406,
+      "grad_norm": 2.369802951812744,
+      "learning_rate": 6.30740236436155e-05,
+      "loss": 0.966,
+      "step": 3284
+    },
+    {
+      "epoch": 0.6319738360908042,
+      "grad_norm": 1.6139453649520874,
+      "learning_rate": 6.301612592774533e-05,
+      "loss": 0.953,
+      "step": 3285
+    },
+    {
+      "epoch": 0.6321662177760677,
+      "grad_norm": 3.0832958221435547,
+      "learning_rate": 6.295824257028844e-05,
+      "loss": 0.9573,
+      "step": 3286
+    },
+    {
+      "epoch": 0.6323585994613313,
+      "grad_norm": 2.311791181564331,
+      "learning_rate": 6.290037359371717e-05,
+      "loss": 1.0243,
+      "step": 3287
+    },
+    {
+      "epoch": 0.6325509811465948,
+      "grad_norm": 1.2486827373504639,
+      "learning_rate": 6.284251902049827e-05,
+      "loss": 1.0367,
+      "step": 3288
+    },
+    {
+      "epoch": 0.6327433628318584,
+      "grad_norm": 2.388258695602417,
+      "learning_rate": 6.278467887309283e-05,
+      "loss": 0.9486,
+      "step": 3289
+    },
+    {
+      "epoch": 0.632935744517122,
+      "grad_norm": 2.985443353652954,
+      "learning_rate": 6.272685317395643e-05,
+      "loss": 0.9944,
+      "step": 3290
+    },
+    {
+      "epoch": 0.6331281262023856,
+      "grad_norm": 2.241227388381958,
+      "learning_rate": 6.266904194553895e-05,
+      "loss": 0.9844,
+      "step": 3291
+    },
+    {
+      "epoch": 0.633320507887649,
+      "grad_norm": 2.506291389465332,
+      "learning_rate": 6.261124521028476e-05,
+      "loss": 0.9868,
+      "step": 3292
+    },
+    {
+      "epoch": 0.6335128895729126,
+      "grad_norm": 1.6716350317001343,
+      "learning_rate": 6.255346299063252e-05,
+      "loss": 0.9422,
+      "step": 3293
+    },
+    {
+      "epoch": 0.6337052712581762,
+      "grad_norm": 1.1551481485366821,
+      "learning_rate": 6.249569530901524e-05,
+      "loss": 0.9681,
+      "step": 3294
+    },
+    {
+      "epoch": 0.6338976529434398,
+      "grad_norm": 2.6215145587921143,
+      "learning_rate": 6.243794218786035e-05,
+      "loss": 0.9443,
+      "step": 3295
+    },
+    {
+      "epoch": 0.6340900346287034,
+      "grad_norm": 1.7708081007003784,
+      "learning_rate": 6.238020364958965e-05,
+      "loss": 1.0269,
+      "step": 3296
+    },
+    {
+      "epoch": 0.6342824163139669,
+      "grad_norm": 1.4964066743850708,
+      "learning_rate": 6.232247971661912e-05,
+      "loss": 0.9659,
+      "step": 3297
+    },
+    {
+      "epoch": 0.6344747979992305,
+      "grad_norm": 2.3863258361816406,
+      "learning_rate": 6.226477041135926e-05,
+      "loss": 0.9752,
+      "step": 3298
+    },
+    {
+      "epoch": 0.634667179684494,
+      "grad_norm": 2.9236197471618652,
+      "learning_rate": 6.22070757562148e-05,
+      "loss": 0.9552,
+      "step": 3299
+    },
+    {
+      "epoch": 0.6348595613697576,
+      "grad_norm": 2.37355375289917,
+      "learning_rate": 6.214939577358478e-05,
+      "loss": 0.9338,
+      "step": 3300
+    },
+    {
+      "epoch": 0.6350519430550212,
+      "grad_norm": 2.9562172889709473,
+      "learning_rate": 6.209173048586253e-05,
+      "loss": 1.0194,
+      "step": 3301
+    },
+    {
+      "epoch": 0.6352443247402847,
+      "grad_norm": 2.306720733642578,
+      "learning_rate": 6.203407991543577e-05,
+      "loss": 0.9763,
+      "step": 3302
+    },
+    {
+      "epoch": 0.6354367064255483,
+      "grad_norm": 2.999331474304199,
+      "learning_rate": 6.197644408468635e-05,
+      "loss": 1.0344,
+      "step": 3303
+    },
+    {
+      "epoch": 0.6356290881108119,
+      "grad_norm": 1.8216288089752197,
+      "learning_rate": 6.191882301599052e-05,
+      "loss": 1.0099,
+      "step": 3304
+    },
+    {
+      "epoch": 0.6358214697960755,
+      "grad_norm": 2.4252545833587646,
+      "learning_rate": 6.186121673171882e-05,
+      "loss": 0.9402,
+      "step": 3305
+    },
+    {
+      "epoch": 0.6360138514813389,
+      "grad_norm": 2.7971341609954834,
+      "learning_rate": 6.180362525423591e-05,
+      "loss": 1.0342,
+      "step": 3306
+    },
+    {
+      "epoch": 0.6362062331666025,
+      "grad_norm": 2.666170358657837,
+      "learning_rate": 6.174604860590081e-05,
+      "loss": 0.9931,
+      "step": 3307
+    },
+    {
+      "epoch": 0.6363986148518661,
+      "grad_norm": 2.4208927154541016,
+      "learning_rate": 6.168848680906678e-05,
+      "loss": 0.9335,
+      "step": 3308
+    },
+    {
+      "epoch": 0.6365909965371297,
+      "grad_norm": 2.184971332550049,
+      "learning_rate": 6.163093988608127e-05,
+      "loss": 0.9565,
+      "step": 3309
+    },
+    {
+      "epoch": 0.6367833782223933,
+      "grad_norm": 2.4999923706054688,
+      "learning_rate": 6.157340785928595e-05,
+      "loss": 0.9545,
+      "step": 3310
+    },
+    {
+      "epoch": 0.6369757599076568,
+      "grad_norm": 1.984115481376648,
+      "learning_rate": 6.151589075101682e-05,
+      "loss": 0.9854,
+      "step": 3311
+    },
+    {
+      "epoch": 0.6371681415929203,
+      "grad_norm": 2.8163092136383057,
+      "learning_rate": 6.145838858360391e-05,
+      "loss": 0.9304,
+      "step": 3312
+    },
+    {
+      "epoch": 0.6373605232781839,
+      "grad_norm": 1.798768401145935,
+      "learning_rate": 6.140090137937158e-05,
+      "loss": 0.8882,
+      "step": 3313
+    },
+    {
+      "epoch": 0.6375529049634475,
+      "grad_norm": 2.184311866760254,
+      "learning_rate": 6.134342916063838e-05,
+      "loss": 1.0323,
+      "step": 3314
+    },
+    {
+      "epoch": 0.637745286648711,
+      "grad_norm": 2.5491795539855957,
+      "learning_rate": 6.12859719497169e-05,
+      "loss": 0.987,
+      "step": 3315
+    },
+    {
+      "epoch": 0.6379376683339746,
+      "grad_norm": 1.7168686389923096,
+      "learning_rate": 6.122852976891414e-05,
+      "loss": 0.9594,
+      "step": 3316
+    },
+    {
+      "epoch": 0.6381300500192382,
+      "grad_norm": 2.354623317718506,
+      "learning_rate": 6.1171102640531e-05,
+      "loss": 0.9126,
+      "step": 3317
+    },
+    {
+      "epoch": 0.6383224317045018,
+      "grad_norm": 1.8594534397125244,
+      "learning_rate": 6.111369058686276e-05,
+      "loss": 0.9166,
+      "step": 3318
+    },
+    {
+      "epoch": 0.6385148133897653,
+      "grad_norm": 2.0176923274993896,
+      "learning_rate": 6.105629363019875e-05,
+      "loss": 0.9123,
+      "step": 3319
+    },
+    {
+      "epoch": 0.6387071950750288,
+      "grad_norm": 2.1912670135498047,
+      "learning_rate": 6.099891179282241e-05,
+      "loss": 0.8345,
+      "step": 3320
+    },
+    {
+      "epoch": 0.6388995767602924,
+      "grad_norm": 1.477917194366455,
+      "learning_rate": 6.094154509701133e-05,
+      "loss": 0.7671,
+      "step": 3321
+    },
+    {
+      "epoch": 0.639091958445556,
+      "grad_norm": 2.056518793106079,
+      "learning_rate": 6.088419356503732e-05,
+      "loss": 0.8934,
+      "step": 3322
+    },
+    {
+      "epoch": 0.6392843401308196,
+      "grad_norm": 2.0055034160614014,
+      "learning_rate": 6.0826857219166124e-05,
+      "loss": 0.9812,
+      "step": 3323
+    },
+    {
+      "epoch": 0.6394767218160831,
+      "grad_norm": 2.6301820278167725,
+      "learning_rate": 6.0769536081657717e-05,
+      "loss": 1.0516,
+      "step": 3324
+    },
+    {
+      "epoch": 0.6396691035013466,
+      "grad_norm": 1.6487125158309937,
+      "learning_rate": 6.0712230174766194e-05,
+      "loss": 0.9902,
+      "step": 3325
+    },
+    {
+      "epoch": 0.6398614851866102,
+      "grad_norm": 1.464025616645813,
+      "learning_rate": 6.065493952073961e-05,
+      "loss": 0.9829,
+      "step": 3326
+    },
+    {
+      "epoch": 0.6400538668718738,
+      "grad_norm": 2.321436643600464,
+      "learning_rate": 6.0597664141820176e-05,
+      "loss": 0.9882,
+      "step": 3327
+    },
+    {
+      "epoch": 0.6402462485571374,
+      "grad_norm": 2.738729476928711,
+      "learning_rate": 6.0540404060244214e-05,
+      "loss": 0.8892,
+      "step": 3328
+    },
+    {
+      "epoch": 0.6404386302424009,
+      "grad_norm": 2.2472007274627686,
+      "learning_rate": 6.0483159298242e-05,
+      "loss": 0.9938,
+      "step": 3329
+    },
+    {
+      "epoch": 0.6406310119276645,
+      "grad_norm": 1.488610863685608,
+      "learning_rate": 6.042592987803796e-05,
+      "loss": 0.9415,
+      "step": 3330
+    },
+    {
+      "epoch": 0.6408233936129281,
+      "grad_norm": 1.156364917755127,
+      "learning_rate": 6.036871582185054e-05,
+      "loss": 0.9983,
+      "step": 3331
+    },
+    {
+      "epoch": 0.6410157752981916,
+      "grad_norm": 2.385824203491211,
+      "learning_rate": 6.031151715189217e-05,
+      "loss": 0.9569,
+      "step": 3332
+    },
+    {
+      "epoch": 0.6412081569834551,
+      "grad_norm": 2.6182119846343994,
+      "learning_rate": 6.025433389036934e-05,
+      "loss": 0.989,
+      "step": 3333
+    },
+    {
+      "epoch": 0.6414005386687187,
+      "grad_norm": 2.096961259841919,
+      "learning_rate": 6.019716605948261e-05,
+      "loss": 0.9891,
+      "step": 3334
+    },
+    {
+      "epoch": 0.6415929203539823,
+      "grad_norm": 2.2330899238586426,
+      "learning_rate": 6.014001368142643e-05,
+      "loss": 1.0208,
+      "step": 3335
+    },
+    {
+      "epoch": 0.6417853020392459,
+      "grad_norm": 2.548557758331299,
+      "learning_rate": 6.0082876778389365e-05,
+      "loss": 0.8795,
+      "step": 3336
+    },
+    {
+      "epoch": 0.6419776837245095,
+      "grad_norm": 2.5374605655670166,
+      "learning_rate": 6.002575537255395e-05,
+      "loss": 1.0552,
+      "step": 3337
+    },
+    {
+      "epoch": 0.642170065409773,
+      "grad_norm": 2.713286876678467,
+      "learning_rate": 5.996864948609662e-05,
+      "loss": 1.0047,
+      "step": 3338
+    },
+    {
+      "epoch": 0.6423624470950365,
+      "grad_norm": 1.355481505393982,
+      "learning_rate": 5.991155914118792e-05,
+      "loss": 0.8292,
+      "step": 3339
+    },
+    {
+      "epoch": 0.6425548287803001,
+      "grad_norm": 1.5099447965621948,
+      "learning_rate": 5.985448435999224e-05,
+      "loss": 1.0401,
+      "step": 3340
+    },
+    {
+      "epoch": 0.6427472104655637,
+      "grad_norm": 2.2036921977996826,
+      "learning_rate": 5.979742516466793e-05,
+      "loss": 0.9706,
+      "step": 3341
+    },
+    {
+      "epoch": 0.6429395921508272,
+      "grad_norm": 1.5249569416046143,
+      "learning_rate": 5.9740381577367455e-05,
+      "loss": 0.9791,
+      "step": 3342
+    },
+    {
+      "epoch": 0.6431319738360908,
+      "grad_norm": 2.2304975986480713,
+      "learning_rate": 5.968335362023697e-05,
+      "loss": 0.9566,
+      "step": 3343
+    },
+    {
+      "epoch": 0.6433243555213544,
+      "grad_norm": 2.03932523727417,
+      "learning_rate": 5.9626341315416754e-05,
+      "loss": 1.0176,
+      "step": 3344
+    },
+    {
+      "epoch": 0.643516737206618,
+      "grad_norm": 2.4192965030670166,
+      "learning_rate": 5.9569344685041006e-05,
+      "loss": 0.9435,
+      "step": 3345
+    },
+    {
+      "epoch": 0.6437091188918815,
+      "grad_norm": 1.940470576286316,
+      "learning_rate": 5.9512363751237677e-05,
+      "loss": 0.9883,
+      "step": 3346
+    },
+    {
+      "epoch": 0.643901500577145,
+      "grad_norm": 1.4149129390716553,
+      "learning_rate": 5.945539853612876e-05,
+      "loss": 0.9735,
+      "step": 3347
+    },
+    {
+      "epoch": 0.6440938822624086,
+      "grad_norm": 2.5840630531311035,
+      "learning_rate": 5.9398449061830164e-05,
+      "loss": 1.0567,
+      "step": 3348
+    },
+    {
+      "epoch": 0.6442862639476722,
+      "grad_norm": 2.7011470794677734,
+      "learning_rate": 5.934151535045156e-05,
+      "loss": 0.9848,
+      "step": 3349
+    },
+    {
+      "epoch": 0.6444786456329358,
+      "grad_norm": 2.413450002670288,
+      "learning_rate": 5.92845974240966e-05,
+      "loss": 0.9113,
+      "step": 3350
+    },
+    {
+      "epoch": 0.6446710273181993,
+      "grad_norm": 2.3527302742004395,
+      "learning_rate": 5.9227695304862826e-05,
+      "loss": 0.9755,
+      "step": 3351
+    },
+    {
+      "epoch": 0.6448634090034628,
+      "grad_norm": 2.1975204944610596,
+      "learning_rate": 5.917080901484155e-05,
+      "loss": 0.9344,
+      "step": 3352
+    },
+    {
+      "epoch": 0.6450557906887264,
+      "grad_norm": 2.531773328781128,
+      "learning_rate": 5.9113938576118e-05,
+      "loss": 0.9311,
+      "step": 3353
+    },
+    {
+      "epoch": 0.64524817237399,
+      "grad_norm": 2.3448033332824707,
+      "learning_rate": 5.905708401077128e-05,
+      "loss": 1.0325,
+      "step": 3354
+    },
+    {
+      "epoch": 0.6454405540592536,
+      "grad_norm": 2.0003268718719482,
+      "learning_rate": 5.900024534087421e-05,
+      "loss": 0.9619,
+      "step": 3355
+    },
+    {
+      "epoch": 0.6456329357445171,
+      "grad_norm": 2.217491626739502,
+      "learning_rate": 5.8943422588493544e-05,
+      "loss": 0.9638,
+      "step": 3356
+    },
+    {
+      "epoch": 0.6458253174297807,
+      "grad_norm": 2.112985372543335,
+      "learning_rate": 5.8886615775689904e-05,
+      "loss": 0.9345,
+      "step": 3357
+    },
+    {
+      "epoch": 0.6460176991150443,
+      "grad_norm": 2.476762294769287,
+      "learning_rate": 5.882982492451757e-05,
+      "loss": 0.9714,
+      "step": 3358
+    },
+    {
+      "epoch": 0.6462100808003078,
+      "grad_norm": 2.2994234561920166,
+      "learning_rate": 5.8773050057024716e-05,
+      "loss": 0.9452,
+      "step": 3359
+    },
+    {
+      "epoch": 0.6464024624855714,
+      "grad_norm": 2.9972784519195557,
+      "learning_rate": 5.871629119525335e-05,
+      "loss": 1.0215,
+      "step": 3360
+    },
+    {
+      "epoch": 0.6465948441708349,
+      "grad_norm": 2.3470401763916016,
+      "learning_rate": 5.865954836123915e-05,
+      "loss": 0.9405,
+      "step": 3361
+    },
+    {
+      "epoch": 0.6467872258560985,
+      "grad_norm": 2.6758532524108887,
+      "learning_rate": 5.860282157701167e-05,
+      "loss": 0.9317,
+      "step": 3362
+    },
+    {
+      "epoch": 0.6469796075413621,
+      "grad_norm": 2.3979790210723877,
+      "learning_rate": 5.854611086459423e-05,
+      "loss": 0.9563,
+      "step": 3363
+    },
+    {
+      "epoch": 0.6471719892266257,
+      "grad_norm": 2.334205389022827,
+      "learning_rate": 5.848941624600381e-05,
+      "loss": 0.955,
+      "step": 3364
+    },
+    {
+      "epoch": 0.6473643709118891,
+      "grad_norm": 2.4089534282684326,
+      "learning_rate": 5.843273774325131e-05,
+      "loss": 0.9836,
+      "step": 3365
+    },
+    {
+      "epoch": 0.6475567525971527,
+      "grad_norm": 2.5862209796905518,
+      "learning_rate": 5.837607537834119e-05,
+      "loss": 0.9699,
+      "step": 3366
+    },
+    {
+      "epoch": 0.6477491342824163,
+      "grad_norm": 2.156008243560791,
+      "learning_rate": 5.8319429173271725e-05,
+      "loss": 0.9447,
+      "step": 3367
+    },
+    {
+      "epoch": 0.6479415159676799,
+      "grad_norm": 2.0802550315856934,
+      "learning_rate": 5.826279915003503e-05,
+      "loss": 0.9533,
+      "step": 3368
+    },
+    {
+      "epoch": 0.6481338976529435,
+      "grad_norm": 2.2902519702911377,
+      "learning_rate": 5.820618533061672e-05,
+      "loss": 1.0012,
+      "step": 3369
+    },
+    {
+      "epoch": 0.648326279338207,
+      "grad_norm": 1.2091543674468994,
+      "learning_rate": 5.8149587736996246e-05,
+      "loss": 1.001,
+      "step": 3370
+    },
+    {
+      "epoch": 0.6485186610234706,
+      "grad_norm": 2.287061929702759,
+      "learning_rate": 5.809300639114683e-05,
+      "loss": 0.9219,
+      "step": 3371
+    },
+    {
+      "epoch": 0.6487110427087341,
+      "grad_norm": 2.458564043045044,
+      "learning_rate": 5.803644131503516e-05,
+      "loss": 0.9551,
+      "step": 3372
+    },
+    {
+      "epoch": 0.6489034243939977,
+      "grad_norm": 2.4707717895507812,
+      "learning_rate": 5.797989253062186e-05,
+      "loss": 1.0143,
+      "step": 3373
+    },
+    {
+      "epoch": 0.6490958060792612,
+      "grad_norm": 1.0362117290496826,
+      "learning_rate": 5.792336005986105e-05,
+      "loss": 1.0301,
+      "step": 3374
+    },
+    {
+      "epoch": 0.6492881877645248,
+      "grad_norm": 2.4774742126464844,
+      "learning_rate": 5.7866843924700634e-05,
+      "loss": 0.903,
+      "step": 3375
+    },
+    {
+      "epoch": 0.6494805694497884,
+      "grad_norm": 1.6794586181640625,
+      "learning_rate": 5.7810344147082086e-05,
+      "loss": 0.965,
+      "step": 3376
+    },
+    {
+      "epoch": 0.649672951135052,
+      "grad_norm": 2.404803991317749,
+      "learning_rate": 5.775386074894058e-05,
+      "loss": 0.9601,
+      "step": 3377
+    },
+    {
+      "epoch": 0.6498653328203156,
+      "grad_norm": 2.2669475078582764,
+      "learning_rate": 5.7697393752204887e-05,
+      "loss": 0.9623,
+      "step": 3378
+    },
+    {
+      "epoch": 0.650057714505579,
+      "grad_norm": 1.5213100910186768,
+      "learning_rate": 5.7640943178797444e-05,
+      "loss": 0.9702,
+      "step": 3379
+    },
+    {
+      "epoch": 0.6502500961908426,
+      "grad_norm": 2.5395090579986572,
+      "learning_rate": 5.758450905063439e-05,
+      "loss": 0.9375,
+      "step": 3380
+    },
+    {
+      "epoch": 0.6504424778761062,
+      "grad_norm": 2.181779623031616,
+      "learning_rate": 5.7528091389625247e-05,
+      "loss": 0.9566,
+      "step": 3381
+    },
+    {
+      "epoch": 0.6506348595613698,
+      "grad_norm": 2.6626124382019043,
+      "learning_rate": 5.747169021767343e-05,
+      "loss": 0.8937,
+      "step": 3382
+    },
+    {
+      "epoch": 0.6508272412466333,
+      "grad_norm": 2.7504382133483887,
+      "learning_rate": 5.7415305556675805e-05,
+      "loss": 0.9937,
+      "step": 3383
+    },
+    {
+      "epoch": 0.6510196229318969,
+      "grad_norm": 2.330202341079712,
+      "learning_rate": 5.7358937428522694e-05,
+      "loss": 0.9203,
+      "step": 3384
+    },
+    {
+      "epoch": 0.6512120046171604,
+      "grad_norm": 2.629986047744751,
+      "learning_rate": 5.7302585855098315e-05,
+      "loss": 0.8958,
+      "step": 3385
+    },
+    {
+      "epoch": 0.651404386302424,
+      "grad_norm": 2.2752363681793213,
+      "learning_rate": 5.7246250858280214e-05,
+      "loss": 0.983,
+      "step": 3386
+    },
+    {
+      "epoch": 0.6515967679876876,
+      "grad_norm": 2.118659257888794,
+      "learning_rate": 5.7189932459939575e-05,
+      "loss": 1.0063,
+      "step": 3387
+    },
+    {
+      "epoch": 0.6517891496729511,
+      "grad_norm": 2.650852680206299,
+      "learning_rate": 5.713363068194115e-05,
+      "loss": 0.9793,
+      "step": 3388
+    },
+    {
+      "epoch": 0.6519815313582147,
+      "grad_norm": 2.4462292194366455,
+      "learning_rate": 5.7077345546143235e-05,
+      "loss": 0.9404,
+      "step": 3389
+    },
+    {
+      "epoch": 0.6521739130434783,
+      "grad_norm": 2.7696518898010254,
+      "learning_rate": 5.7021077074397654e-05,
+      "loss": 0.9548,
+      "step": 3390
+    },
+    {
+      "epoch": 0.6523662947287419,
+      "grad_norm": 2.346057176589966,
+      "learning_rate": 5.696482528854974e-05,
+      "loss": 1.0296,
+      "step": 3391
+    },
+    {
+      "epoch": 0.6525586764140053,
+      "grad_norm": 2.4601633548736572,
+      "learning_rate": 5.6908590210438416e-05,
+      "loss": 0.8591,
+      "step": 3392
+    },
+    {
+      "epoch": 0.6527510580992689,
+      "grad_norm": 2.1330342292785645,
+      "learning_rate": 5.6852371861896e-05,
+      "loss": 0.9576,
+      "step": 3393
+    },
+    {
+      "epoch": 0.6529434397845325,
+      "grad_norm": 2.295461654663086,
+      "learning_rate": 5.6796170264748525e-05,
+      "loss": 0.8974,
+      "step": 3394
+    },
+    {
+      "epoch": 0.6531358214697961,
+      "grad_norm": 1.931517481803894,
+      "learning_rate": 5.6739985440815266e-05,
+      "loss": 0.9718,
+      "step": 3395
+    },
+    {
+      "epoch": 0.6533282031550597,
+      "grad_norm": 2.6259524822235107,
+      "learning_rate": 5.6683817411909114e-05,
+      "loss": 0.9794,
+      "step": 3396
+    },
+    {
+      "epoch": 0.6535205848403232,
+      "grad_norm": 1.4613306522369385,
+      "learning_rate": 5.6627666199836526e-05,
+      "loss": 0.8017,
+      "step": 3397
+    },
+    {
+      "epoch": 0.6537129665255867,
+      "grad_norm": 2.4664857387542725,
+      "learning_rate": 5.657153182639721e-05,
+      "loss": 0.944,
+      "step": 3398
+    },
+    {
+      "epoch": 0.6539053482108503,
+      "grad_norm": 2.8625173568725586,
+      "learning_rate": 5.651541431338454e-05,
+      "loss": 0.9727,
+      "step": 3399
+    },
+    {
+      "epoch": 0.6540977298961139,
+      "grad_norm": 2.406597137451172,
+      "learning_rate": 5.645931368258527e-05,
+      "loss": 0.9765,
+      "step": 3400
+    },
+    {
+      "epoch": 0.6542901115813774,
+      "grad_norm": 2.344656467437744,
+      "learning_rate": 5.640322995577958e-05,
+      "loss": 0.9035,
+      "step": 3401
+    },
+    {
+      "epoch": 0.654482493266641,
+      "grad_norm": 2.522895097732544,
+      "learning_rate": 5.634716315474109e-05,
+      "loss": 1.0525,
+      "step": 3402
+    },
+    {
+      "epoch": 0.6546748749519046,
+      "grad_norm": 2.656095504760742,
+      "learning_rate": 5.629111330123689e-05,
+      "loss": 0.9321,
+      "step": 3403
+    },
+    {
+      "epoch": 0.6548672566371682,
+      "grad_norm": 2.3086700439453125,
+      "learning_rate": 5.623508041702743e-05,
+      "loss": 0.9037,
+      "step": 3404
+    },
+    {
+      "epoch": 0.6550596383224317,
+      "grad_norm": 2.1504595279693604,
+      "learning_rate": 5.617906452386659e-05,
+      "loss": 0.9881,
+      "step": 3405
+    },
+    {
+      "epoch": 0.6552520200076952,
+      "grad_norm": 2.7710070610046387,
+      "learning_rate": 5.612306564350178e-05,
+      "loss": 0.9853,
+      "step": 3406
+    },
+    {
+      "epoch": 0.6554444016929588,
+      "grad_norm": 2.4275968074798584,
+      "learning_rate": 5.6067083797673536e-05,
+      "loss": 0.953,
+      "step": 3407
+    },
+    {
+      "epoch": 0.6556367833782224,
+      "grad_norm": 1.3648707866668701,
+      "learning_rate": 5.601111900811606e-05,
+      "loss": 1.0146,
+      "step": 3408
+    },
+    {
+      "epoch": 0.655829165063486,
+      "grad_norm": 2.2468395233154297,
+      "learning_rate": 5.595517129655681e-05,
+      "loss": 0.8993,
+      "step": 3409
+    },
+    {
+      "epoch": 0.6560215467487496,
+      "grad_norm": 2.8260912895202637,
+      "learning_rate": 5.589924068471648e-05,
+      "loss": 0.9604,
+      "step": 3410
+    },
+    {
+      "epoch": 0.656213928434013,
+      "grad_norm": 2.52848482131958,
+      "learning_rate": 5.5843327194309394e-05,
+      "loss": 0.8776,
+      "step": 3411
+    },
+    {
+      "epoch": 0.6564063101192766,
+      "grad_norm": 1.8191566467285156,
+      "learning_rate": 5.578743084704305e-05,
+      "loss": 0.9547,
+      "step": 3412
+    },
+    {
+      "epoch": 0.6565986918045402,
+      "grad_norm": 2.5644404888153076,
+      "learning_rate": 5.573155166461833e-05,
+      "loss": 0.9948,
+      "step": 3413
+    },
+    {
+      "epoch": 0.6567910734898038,
+      "grad_norm": 2.5447404384613037,
+      "learning_rate": 5.567568966872947e-05,
+      "loss": 0.9986,
+      "step": 3414
+    },
+    {
+      "epoch": 0.6569834551750673,
+      "grad_norm": 2.436457395553589,
+      "learning_rate": 5.5619844881064e-05,
+      "loss": 1.0123,
+      "step": 3415
+    },
+    {
+      "epoch": 0.6571758368603309,
+      "grad_norm": 2.2638046741485596,
+      "learning_rate": 5.556401732330281e-05,
+      "loss": 0.9515,
+      "step": 3416
+    },
+    {
+      "epoch": 0.6573682185455945,
+      "grad_norm": 2.925525426864624,
+      "learning_rate": 5.550820701712007e-05,
+      "loss": 1.0409,
+      "step": 3417
+    },
+    {
+      "epoch": 0.657560600230858,
+      "grad_norm": 2.9931693077087402,
+      "learning_rate": 5.545241398418326e-05,
+      "loss": 0.9622,
+      "step": 3418
+    },
+    {
+      "epoch": 0.6577529819161216,
+      "grad_norm": 1.9204435348510742,
+      "learning_rate": 5.539663824615312e-05,
+      "loss": 0.9204,
+      "step": 3419
+    },
+    {
+      "epoch": 0.6579453636013851,
+      "grad_norm": 3.173076868057251,
+      "learning_rate": 5.5340879824683834e-05,
+      "loss": 0.9191,
+      "step": 3420
+    },
+    {
+      "epoch": 0.6581377452866487,
+      "grad_norm": 2.401522397994995,
+      "learning_rate": 5.528513874142262e-05,
+      "loss": 0.971,
+      "step": 3421
+    },
+    {
+      "epoch": 0.6583301269719123,
+      "grad_norm": 2.752213954925537,
+      "learning_rate": 5.522941501801008e-05,
+      "loss": 1.0012,
+      "step": 3422
+    },
+    {
+      "epoch": 0.6585225086571759,
+      "grad_norm": 2.4237020015716553,
+      "learning_rate": 5.51737086760802e-05,
+      "loss": 0.9104,
+      "step": 3423
+    },
+    {
+      "epoch": 0.6587148903424394,
+      "grad_norm": 2.1651384830474854,
+      "learning_rate": 5.5118019737259966e-05,
+      "loss": 0.8926,
+      "step": 3424
+    },
+    {
+      "epoch": 0.6589072720277029,
+      "grad_norm": 2.3554060459136963,
+      "learning_rate": 5.5062348223169825e-05,
+      "loss": 0.9106,
+      "step": 3425
+    },
+    {
+      "epoch": 0.6590996537129665,
+      "grad_norm": 2.0181286334991455,
+      "learning_rate": 5.5006694155423356e-05,
+      "loss": 1.0592,
+      "step": 3426
+    },
+    {
+      "epoch": 0.6592920353982301,
+      "grad_norm": 1.8012245893478394,
+      "learning_rate": 5.495105755562738e-05,
+      "loss": 0.9629,
+      "step": 3427
+    },
+    {
+      "epoch": 0.6594844170834937,
+      "grad_norm": 2.442751407623291,
+      "learning_rate": 5.489543844538194e-05,
+      "loss": 0.941,
+      "step": 3428
+    },
+    {
+      "epoch": 0.6596767987687572,
+      "grad_norm": 1.5500468015670776,
+      "learning_rate": 5.48398368462803e-05,
+      "loss": 0.9464,
+      "step": 3429
+    },
+    {
+      "epoch": 0.6598691804540208,
+      "grad_norm": 1.6931859254837036,
+      "learning_rate": 5.4784252779908905e-05,
+      "loss": 1.027,
+      "step": 3430
+    },
+    {
+      "epoch": 0.6600615621392844,
+      "grad_norm": 2.9665722846984863,
+      "learning_rate": 5.4728686267847354e-05,
+      "loss": 0.9695,
+      "step": 3431
+    },
+    {
+      "epoch": 0.6602539438245479,
+      "grad_norm": 2.7147412300109863,
+      "learning_rate": 5.467313733166862e-05,
+      "loss": 0.9747,
+      "step": 3432
+    },
+    {
+      "epoch": 0.6604463255098114,
+      "grad_norm": 2.1497859954833984,
+      "learning_rate": 5.461760599293855e-05,
+      "loss": 0.9222,
+      "step": 3433
+    },
+    {
+      "epoch": 0.660638707195075,
+      "grad_norm": 2.2817792892456055,
+      "learning_rate": 5.456209227321642e-05,
+      "loss": 0.9891,
+      "step": 3434
+    },
+    {
+      "epoch": 0.6608310888803386,
+      "grad_norm": 2.331648588180542,
+      "learning_rate": 5.4506596194054584e-05,
+      "loss": 0.96,
+      "step": 3435
+    },
+    {
+      "epoch": 0.6610234705656022,
+      "grad_norm": 1.935443639755249,
+      "learning_rate": 5.445111777699842e-05,
+      "loss": 0.8497,
+      "step": 3436
+    },
+    {
+      "epoch": 0.6612158522508658,
+      "grad_norm": 1.27147376537323,
+      "learning_rate": 5.439565704358667e-05,
+      "loss": 0.9314,
+      "step": 3437
+    },
+    {
+      "epoch": 0.6614082339361292,
+      "grad_norm": 1.6918551921844482,
+      "learning_rate": 5.4340214015351046e-05,
+      "loss": 0.9941,
+      "step": 3438
+    },
+    {
+      "epoch": 0.6616006156213928,
+      "grad_norm": 2.4095218181610107,
+      "learning_rate": 5.428478871381646e-05,
+      "loss": 0.8822,
+      "step": 3439
+    },
+    {
+      "epoch": 0.6617929973066564,
+      "grad_norm": 2.3597452640533447,
+      "learning_rate": 5.422938116050091e-05,
+      "loss": 0.9968,
+      "step": 3440
+    },
+    {
+      "epoch": 0.66198537899192,
+      "grad_norm": 2.7621591091156006,
+      "learning_rate": 5.417399137691552e-05,
+      "loss": 1.0083,
+      "step": 3441
+    },
+    {
+      "epoch": 0.6621777606771835,
+      "grad_norm": 1.2740557193756104,
+      "learning_rate": 5.411861938456453e-05,
+      "loss": 0.8406,
+      "step": 3442
+    },
+    {
+      "epoch": 0.6623701423624471,
+      "grad_norm": 1.1876320838928223,
+      "learning_rate": 5.4063265204945224e-05,
+      "loss": 0.9703,
+      "step": 3443
+    },
+    {
+      "epoch": 0.6625625240477107,
+      "grad_norm": 2.9521658420562744,
+      "learning_rate": 5.400792885954802e-05,
+      "loss": 0.9773,
+      "step": 3444
+    },
+    {
+      "epoch": 0.6627549057329742,
+      "grad_norm": 2.5784530639648438,
+      "learning_rate": 5.395261036985635e-05,
+      "loss": 0.9976,
+      "step": 3445
+    },
+    {
+      "epoch": 0.6629472874182378,
+      "grad_norm": 2.591533660888672,
+      "learning_rate": 5.389730975734686e-05,
+      "loss": 0.9771,
+      "step": 3446
+    },
+    {
+      "epoch": 0.6631396691035013,
+      "grad_norm": 2.653059244155884,
+      "learning_rate": 5.384202704348902e-05,
+      "loss": 0.9835,
+      "step": 3447
+    },
+    {
+      "epoch": 0.6633320507887649,
+      "grad_norm": 2.376098871231079,
+      "learning_rate": 5.378676224974557e-05,
+      "loss": 0.9866,
+      "step": 3448
+    },
+    {
+      "epoch": 0.6635244324740285,
+      "grad_norm": 2.349334478378296,
+      "learning_rate": 5.373151539757224e-05,
+      "loss": 0.9358,
+      "step": 3449
+    },
+    {
+      "epoch": 0.6637168141592921,
+      "grad_norm": 2.45475172996521,
+      "learning_rate": 5.36762865084176e-05,
+      "loss": 0.9527,
+      "step": 3450
+    },
+    {
+      "epoch": 0.6639091958445555,
+      "grad_norm": 2.293672561645508,
+      "learning_rate": 5.362107560372358e-05,
+      "loss": 0.9605,
+      "step": 3451
+    },
+    {
+      "epoch": 0.6641015775298191,
+      "grad_norm": 2.198302984237671,
+      "learning_rate": 5.356588270492487e-05,
+      "loss": 0.9498,
+      "step": 3452
+    },
+    {
+      "epoch": 0.6642939592150827,
+      "grad_norm": 1.9639376401901245,
+      "learning_rate": 5.351070783344926e-05,
+      "loss": 0.9855,
+      "step": 3453
+    },
+    {
+      "epoch": 0.6644863409003463,
+      "grad_norm": 2.606722831726074,
+      "learning_rate": 5.3455551010717545e-05,
+      "loss": 1.0317,
+      "step": 3454
+    },
+    {
+      "epoch": 0.6646787225856099,
+      "grad_norm": 1.051749348640442,
+      "learning_rate": 5.340041225814349e-05,
+      "loss": 1.0094,
+      "step": 3455
+    },
+    {
+      "epoch": 0.6648711042708734,
+      "grad_norm": 2.4465694427490234,
+      "learning_rate": 5.334529159713388e-05,
+      "loss": 1.0431,
+      "step": 3456
+    },
+    {
+      "epoch": 0.665063485956137,
+      "grad_norm": 2.465240955352783,
+      "learning_rate": 5.3290189049088416e-05,
+      "loss": 0.9081,
+      "step": 3457
+    },
+    {
+      "epoch": 0.6652558676414005,
+      "grad_norm": 2.356602430343628,
+      "learning_rate": 5.323510463539989e-05,
+      "loss": 0.8994,
+      "step": 3458
+    },
+    {
+      "epoch": 0.6654482493266641,
+      "grad_norm": 1.4018046855926514,
+      "learning_rate": 5.3180038377453826e-05,
+      "loss": 0.9942,
+      "step": 3459
+    },
+    {
+      "epoch": 0.6656406310119277,
+      "grad_norm": 1.5344510078430176,
+      "learning_rate": 5.3124990296628976e-05,
+      "loss": 0.9628,
+      "step": 3460
+    },
+    {
+      "epoch": 0.6658330126971912,
+      "grad_norm": 1.874816656112671,
+      "learning_rate": 5.306996041429687e-05,
+      "loss": 1.0049,
+      "step": 3461
+    },
+    {
+      "epoch": 0.6660253943824548,
+      "grad_norm": 2.162172555923462,
+      "learning_rate": 5.3014948751821915e-05,
+      "loss": 0.8412,
+      "step": 3462
+    },
+    {
+      "epoch": 0.6662177760677184,
+      "grad_norm": 1.2799543142318726,
+      "learning_rate": 5.295995533056162e-05,
+      "loss": 0.8198,
+      "step": 3463
+    },
+    {
+      "epoch": 0.666410157752982,
+      "grad_norm": 1.9108601808547974,
+      "learning_rate": 5.290498017186631e-05,
+      "loss": 0.9952,
+      "step": 3464
+    },
+    {
+      "epoch": 0.6666025394382454,
+      "grad_norm": 1.1828819513320923,
+      "learning_rate": 5.285002329707923e-05,
+      "loss": 1.024,
+      "step": 3465
+    },
+    {
+      "epoch": 0.666794921123509,
+      "grad_norm": 2.6735105514526367,
+      "learning_rate": 5.2795084727536536e-05,
+      "loss": 0.9711,
+      "step": 3466
+    },
+    {
+      "epoch": 0.6669873028087726,
+      "grad_norm": 2.0140464305877686,
+      "learning_rate": 5.274016448456724e-05,
+      "loss": 0.9778,
+      "step": 3467
+    },
+    {
+      "epoch": 0.6671796844940362,
+      "grad_norm": 1.004148244857788,
+      "learning_rate": 5.268526258949331e-05,
+      "loss": 0.9922,
+      "step": 3468
+    },
+    {
+      "epoch": 0.6673720661792998,
+      "grad_norm": 1.9275201559066772,
+      "learning_rate": 5.2630379063629534e-05,
+      "loss": 0.9263,
+      "step": 3469
+    },
+    {
+      "epoch": 0.6675644478645633,
+      "grad_norm": 1.0002166032791138,
+      "learning_rate": 5.2575513928283594e-05,
+      "loss": 0.765,
+      "step": 3470
+    },
+    {
+      "epoch": 0.6677568295498268,
+      "grad_norm": 2.5798444747924805,
+      "learning_rate": 5.252066720475597e-05,
+      "loss": 0.9703,
+      "step": 3471
+    },
+    {
+      "epoch": 0.6679492112350904,
+      "grad_norm": 2.1841065883636475,
+      "learning_rate": 5.246583891434018e-05,
+      "loss": 1.0044,
+      "step": 3472
+    },
+    {
+      "epoch": 0.668141592920354,
+      "grad_norm": 1.6043471097946167,
+      "learning_rate": 5.2411029078322314e-05,
+      "loss": 0.9515,
+      "step": 3473
+    },
+    {
+      "epoch": 0.6683339746056175,
+      "grad_norm": 2.6248514652252197,
+      "learning_rate": 5.235623771798151e-05,
+      "loss": 0.9707,
+      "step": 3474
+    },
+    {
+      "epoch": 0.6685263562908811,
+      "grad_norm": 2.751303195953369,
+      "learning_rate": 5.2301464854589696e-05,
+      "loss": 0.9856,
+      "step": 3475
+    },
+    {
+      "epoch": 0.6687187379761447,
+      "grad_norm": 2.6814770698547363,
+      "learning_rate": 5.224671050941146e-05,
+      "loss": 1.0016,
+      "step": 3476
+    },
+    {
+      "epoch": 0.6689111196614083,
+      "grad_norm": 2.291055679321289,
+      "learning_rate": 5.219197470370443e-05,
+      "loss": 0.9204,
+      "step": 3477
+    },
+    {
+      "epoch": 0.6691035013466718,
+      "grad_norm": 2.519230365753174,
+      "learning_rate": 5.2137257458718894e-05,
+      "loss": 0.9755,
+      "step": 3478
+    },
+    {
+      "epoch": 0.6692958830319353,
+      "grad_norm": 2.108391046524048,
+      "learning_rate": 5.208255879569799e-05,
+      "loss": 0.9095,
+      "step": 3479
+    },
+    {
+      "epoch": 0.6694882647171989,
+      "grad_norm": 2.480151653289795,
+      "learning_rate": 5.20278787358776e-05,
+      "loss": 0.9638,
+      "step": 3480
+    },
+    {
+      "epoch": 0.6696806464024625,
+      "grad_norm": 2.5209107398986816,
+      "learning_rate": 5.197321730048641e-05,
+      "loss": 0.9437,
+      "step": 3481
+    },
+    {
+      "epoch": 0.6698730280877261,
+      "grad_norm": 2.1742115020751953,
+      "learning_rate": 5.191857451074587e-05,
+      "loss": 0.9522,
+      "step": 3482
+    },
+    {
+      "epoch": 0.6700654097729896,
+      "grad_norm": 2.9360151290893555,
+      "learning_rate": 5.1863950387870176e-05,
+      "loss": 1.0224,
+      "step": 3483
+    },
+    {
+      "epoch": 0.6702577914582531,
+      "grad_norm": 2.4164886474609375,
+      "learning_rate": 5.1809344953066374e-05,
+      "loss": 0.9477,
+      "step": 3484
+    },
+    {
+      "epoch": 0.6704501731435167,
+      "grad_norm": 2.471514940261841,
+      "learning_rate": 5.175475822753404e-05,
+      "loss": 0.9759,
+      "step": 3485
+    },
+    {
+      "epoch": 0.6706425548287803,
+      "grad_norm": 2.3363516330718994,
+      "learning_rate": 5.170019023246574e-05,
+      "loss": 0.9794,
+      "step": 3486
+    },
+    {
+      "epoch": 0.6708349365140439,
+      "grad_norm": 2.3177313804626465,
+      "learning_rate": 5.1645640989046595e-05,
+      "loss": 0.912,
+      "step": 3487
+    },
+    {
+      "epoch": 0.6710273181993074,
+      "grad_norm": 2.2096424102783203,
+      "learning_rate": 5.159111051845451e-05,
+      "loss": 1.0019,
+      "step": 3488
+    },
+    {
+      "epoch": 0.671219699884571,
+      "grad_norm": 2.1326913833618164,
+      "learning_rate": 5.153659884186013e-05,
+      "loss": 0.9583,
+      "step": 3489
+    },
+    {
+      "epoch": 0.6714120815698346,
+      "grad_norm": 2.460239887237549,
+      "learning_rate": 5.1482105980426643e-05,
+      "loss": 0.9667,
+      "step": 3490
+    },
+    {
+      "epoch": 0.6716044632550981,
+      "grad_norm": 2.417346715927124,
+      "learning_rate": 5.142763195531017e-05,
+      "loss": 0.9979,
+      "step": 3491
+    },
+    {
+      "epoch": 0.6717968449403616,
+      "grad_norm": 2.4863669872283936,
+      "learning_rate": 5.1373176787659386e-05,
+      "loss": 0.9576,
+      "step": 3492
+    },
+    {
+      "epoch": 0.6719892266256252,
+      "grad_norm": 2.4866552352905273,
+      "learning_rate": 5.1318740498615624e-05,
+      "loss": 0.9595,
+      "step": 3493
+    },
+    {
+      "epoch": 0.6721816083108888,
+      "grad_norm": 2.551398754119873,
+      "learning_rate": 5.126432310931295e-05,
+      "loss": 0.99,
+      "step": 3494
+    },
+    {
+      "epoch": 0.6723739899961524,
+      "grad_norm": 2.2957143783569336,
+      "learning_rate": 5.120992464087807e-05,
+      "loss": 0.9465,
+      "step": 3495
+    },
+    {
+      "epoch": 0.672566371681416,
+      "grad_norm": 2.3917911052703857,
+      "learning_rate": 5.115554511443033e-05,
+      "loss": 0.9876,
+      "step": 3496
+    },
+    {
+      "epoch": 0.6727587533666795,
+      "grad_norm": 2.3765242099761963,
+      "learning_rate": 5.11011845510817e-05,
+      "loss": 0.951,
+      "step": 3497
+    },
+    {
+      "epoch": 0.672951135051943,
+      "grad_norm": 2.339386224746704,
+      "learning_rate": 5.1046842971936936e-05,
+      "loss": 0.9181,
+      "step": 3498
+    },
+    {
+      "epoch": 0.6731435167372066,
+      "grad_norm": 1.9488368034362793,
+      "learning_rate": 5.099252039809317e-05,
+      "loss": 0.9814,
+      "step": 3499
+    },
+    {
+      "epoch": 0.6733358984224702,
+      "grad_norm": 2.2047853469848633,
+      "learning_rate": 5.0938216850640396e-05,
+      "loss": 0.9996,
+      "step": 3500
+    },
+    {
+      "epoch": 0.6735282801077337,
+      "grad_norm": 3.0381877422332764,
+      "learning_rate": 5.0883932350661134e-05,
+      "loss": 0.9239,
+      "step": 3501
+    },
+    {
+      "epoch": 0.6737206617929973,
+      "grad_norm": 2.2990128993988037,
+      "learning_rate": 5.082966691923037e-05,
+      "loss": 0.9487,
+      "step": 3502
+    },
+    {
+      "epoch": 0.6739130434782609,
+      "grad_norm": 2.6305630207061768,
+      "learning_rate": 5.077542057741592e-05,
+      "loss": 0.9385,
+      "step": 3503
+    },
+    {
+      "epoch": 0.6741054251635245,
+      "grad_norm": 2.482391119003296,
+      "learning_rate": 5.072119334627806e-05,
+      "loss": 0.9432,
+      "step": 3504
+    },
+    {
+      "epoch": 0.674297806848788,
+      "grad_norm": 2.305507183074951,
+      "learning_rate": 5.066698524686966e-05,
+      "loss": 1.0401,
+      "step": 3505
+    },
+    {
+      "epoch": 0.6744901885340515,
+      "grad_norm": 2.824415922164917,
+      "learning_rate": 5.0612796300236175e-05,
+      "loss": 0.9538,
+      "step": 3506
+    },
+    {
+      "epoch": 0.6746825702193151,
+      "grad_norm": 2.4567654132843018,
+      "learning_rate": 5.055862652741562e-05,
+      "loss": 0.9884,
+      "step": 3507
+    },
+    {
+      "epoch": 0.6748749519045787,
+      "grad_norm": 2.610133171081543,
+      "learning_rate": 5.0504475949438555e-05,
+      "loss": 0.9692,
+      "step": 3508
+    },
+    {
+      "epoch": 0.6750673335898423,
+      "grad_norm": 2.3146893978118896,
+      "learning_rate": 5.045034458732808e-05,
+      "loss": 0.9298,
+      "step": 3509
+    },
+    {
+      "epoch": 0.6752597152751059,
+      "grad_norm": 1.2273074388504028,
+      "learning_rate": 5.039623246209995e-05,
+      "loss": 0.7821,
+      "step": 3510
+    },
+    {
+      "epoch": 0.6754520969603693,
+      "grad_norm": 2.50974702835083,
+      "learning_rate": 5.0342139594762216e-05,
+      "loss": 0.9605,
+      "step": 3511
+    },
+    {
+      "epoch": 0.6756444786456329,
+      "grad_norm": 2.877890110015869,
+      "learning_rate": 5.028806600631569e-05,
+      "loss": 0.9542,
+      "step": 3512
+    },
+    {
+      "epoch": 0.6758368603308965,
+      "grad_norm": 2.665099859237671,
+      "learning_rate": 5.023401171775357e-05,
+      "loss": 0.9219,
+      "step": 3513
+    },
+    {
+      "epoch": 0.6760292420161601,
+      "grad_norm": 2.4421656131744385,
+      "learning_rate": 5.0179976750061605e-05,
+      "loss": 0.9921,
+      "step": 3514
+    },
+    {
+      "epoch": 0.6762216237014236,
+      "grad_norm": 2.06158709526062,
+      "learning_rate": 5.012596112421806e-05,
+      "loss": 0.9402,
+      "step": 3515
+    },
+    {
+      "epoch": 0.6764140053866872,
+      "grad_norm": 2.511909246444702,
+      "learning_rate": 5.007196486119354e-05,
+      "loss": 0.9736,
+      "step": 3516
+    },
+    {
+      "epoch": 0.6766063870719508,
+      "grad_norm": 2.421481132507324,
+      "learning_rate": 5.001798798195136e-05,
+      "loss": 0.9284,
+      "step": 3517
+    },
+    {
+      "epoch": 0.6767987687572143,
+      "grad_norm": 2.970036506652832,
+      "learning_rate": 4.996403050744719e-05,
+      "loss": 0.9817,
+      "step": 3518
+    },
+    {
+      "epoch": 0.6769911504424779,
+      "grad_norm": 2.5078461170196533,
+      "learning_rate": 4.991009245862917e-05,
+      "loss": 0.9771,
+      "step": 3519
+    },
+    {
+      "epoch": 0.6771835321277414,
+      "grad_norm": 1.168923258781433,
+      "learning_rate": 4.985617385643789e-05,
+      "loss": 1.0287,
+      "step": 3520
+    },
+    {
+      "epoch": 0.677375913813005,
+      "grad_norm": 2.4029483795166016,
+      "learning_rate": 4.980227472180643e-05,
+      "loss": 0.9102,
+      "step": 3521
+    },
+    {
+      "epoch": 0.6775682954982686,
+      "grad_norm": 2.893893241882324,
+      "learning_rate": 4.974839507566027e-05,
+      "loss": 0.9362,
+      "step": 3522
+    },
+    {
+      "epoch": 0.6777606771835322,
+      "grad_norm": 1.6903738975524902,
+      "learning_rate": 4.969453493891732e-05,
+      "loss": 0.9684,
+      "step": 3523
+    },
+    {
+      "epoch": 0.6779530588687956,
+      "grad_norm": 1.957715630531311,
+      "learning_rate": 4.964069433248807e-05,
+      "loss": 0.8818,
+      "step": 3524
+    },
+    {
+      "epoch": 0.6781454405540592,
+      "grad_norm": 2.620218276977539,
+      "learning_rate": 4.95868732772751e-05,
+      "loss": 1.0102,
+      "step": 3525
+    },
+    {
+      "epoch": 0.6783378222393228,
+      "grad_norm": 2.460089921951294,
+      "learning_rate": 4.953307179417376e-05,
+      "loss": 0.8991,
+      "step": 3526
+    },
+    {
+      "epoch": 0.6785302039245864,
+      "grad_norm": 2.23116397857666,
+      "learning_rate": 4.947928990407156e-05,
+      "loss": 0.9098,
+      "step": 3527
+    },
+    {
+      "epoch": 0.67872258560985,
+      "grad_norm": 1.7849981784820557,
+      "learning_rate": 4.94255276278485e-05,
+      "loss": 1.0094,
+      "step": 3528
+    },
+    {
+      "epoch": 0.6789149672951135,
+      "grad_norm": 1.401058554649353,
+      "learning_rate": 4.937178498637696e-05,
+      "loss": 0.7599,
+      "step": 3529
+    },
+    {
+      "epoch": 0.6791073489803771,
+      "grad_norm": 2.2015373706817627,
+      "learning_rate": 4.9318062000521644e-05,
+      "loss": 0.9828,
+      "step": 3530
+    },
+    {
+      "epoch": 0.6792997306656406,
+      "grad_norm": 2.4133269786834717,
+      "learning_rate": 4.926435869113971e-05,
+      "loss": 0.9424,
+      "step": 3531
+    },
+    {
+      "epoch": 0.6794921123509042,
+      "grad_norm": 1.448472023010254,
+      "learning_rate": 4.9210675079080607e-05,
+      "loss": 0.7809,
+      "step": 3532
+    },
+    {
+      "epoch": 0.6796844940361677,
+      "grad_norm": 2.5161828994750977,
+      "learning_rate": 4.915701118518615e-05,
+      "loss": 0.9669,
+      "step": 3533
+    },
+    {
+      "epoch": 0.6798768757214313,
+      "grad_norm": 1.617413878440857,
+      "learning_rate": 4.9103367030290547e-05,
+      "loss": 0.9573,
+      "step": 3534
+    },
+    {
+      "epoch": 0.6800692574066949,
+      "grad_norm": 2.932157278060913,
+      "learning_rate": 4.9049742635220243e-05,
+      "loss": 1.0133,
+      "step": 3535
+    },
+    {
+      "epoch": 0.6802616390919585,
+      "grad_norm": 2.4612739086151123,
+      "learning_rate": 4.899613802079419e-05,
+      "loss": 0.926,
+      "step": 3536
+    },
+    {
+      "epoch": 0.6804540207772221,
+      "grad_norm": 1.9807894229888916,
+      "learning_rate": 4.8942553207823395e-05,
+      "loss": 1.02,
+      "step": 3537
+    },
+    {
+      "epoch": 0.6806464024624855,
+      "grad_norm": 2.1038522720336914,
+      "learning_rate": 4.8888988217111444e-05,
+      "loss": 0.9198,
+      "step": 3538
+    },
+    {
+      "epoch": 0.6808387841477491,
+      "grad_norm": 1.2644178867340088,
+      "learning_rate": 4.883544306945407e-05,
+      "loss": 1.0052,
+      "step": 3539
+    },
+    {
+      "epoch": 0.6810311658330127,
+      "grad_norm": 1.9556598663330078,
+      "learning_rate": 4.878191778563934e-05,
+      "loss": 0.9495,
+      "step": 3540
+    },
+    {
+      "epoch": 0.6812235475182763,
+      "grad_norm": 2.6190366744995117,
+      "learning_rate": 4.8728412386447654e-05,
+      "loss": 0.9246,
+      "step": 3541
+    },
+    {
+      "epoch": 0.6814159292035398,
+      "grad_norm": 2.2594964504241943,
+      "learning_rate": 4.867492689265154e-05,
+      "loss": 0.996,
+      "step": 3542
+    },
+    {
+      "epoch": 0.6816083108888034,
+      "grad_norm": 2.1641013622283936,
+      "learning_rate": 4.862146132501602e-05,
+      "loss": 0.9345,
+      "step": 3543
+    },
+    {
+      "epoch": 0.681800692574067,
+      "grad_norm": 1.3573044538497925,
+      "learning_rate": 4.856801570429822e-05,
+      "loss": 1.0193,
+      "step": 3544
+    },
+    {
+      "epoch": 0.6819930742593305,
+      "grad_norm": 2.2086081504821777,
+      "learning_rate": 4.851459005124759e-05,
+      "loss": 0.9384,
+      "step": 3545
+    },
+    {
+      "epoch": 0.6821854559445941,
+      "grad_norm": 2.1364166736602783,
+      "learning_rate": 4.8461184386605785e-05,
+      "loss": 0.9542,
+      "step": 3546
+    },
+    {
+      "epoch": 0.6823778376298576,
+      "grad_norm": 1.8380147218704224,
+      "learning_rate": 4.8407798731106746e-05,
+      "loss": 0.9636,
+      "step": 3547
+    },
+    {
+      "epoch": 0.6825702193151212,
+      "grad_norm": 1.6382418870925903,
+      "learning_rate": 4.8354433105476604e-05,
+      "loss": 0.7798,
+      "step": 3548
+    },
+    {
+      "epoch": 0.6827626010003848,
+      "grad_norm": 2.5110251903533936,
+      "learning_rate": 4.83010875304337e-05,
+      "loss": 0.9594,
+      "step": 3549
+    },
+    {
+      "epoch": 0.6829549826856484,
+      "grad_norm": 1.287500023841858,
+      "learning_rate": 4.8247762026688746e-05,
+      "loss": 0.9791,
+      "step": 3550
+    },
+    {
+      "epoch": 0.6831473643709118,
+      "grad_norm": 2.3565914630889893,
+      "learning_rate": 4.8194456614944375e-05,
+      "loss": 0.9175,
+      "step": 3551
+    },
+    {
+      "epoch": 0.6833397460561754,
+      "grad_norm": 2.350785970687866,
+      "learning_rate": 4.814117131589569e-05,
+      "loss": 0.9372,
+      "step": 3552
+    },
+    {
+      "epoch": 0.683532127741439,
+      "grad_norm": 1.1176879405975342,
+      "learning_rate": 4.8087906150229866e-05,
+      "loss": 1.0182,
+      "step": 3553
+    },
+    {
+      "epoch": 0.6837245094267026,
+      "grad_norm": 2.0610530376434326,
+      "learning_rate": 4.8034661138626255e-05,
+      "loss": 1.0231,
+      "step": 3554
+    },
+    {
+      "epoch": 0.6839168911119662,
+      "grad_norm": 2.345567226409912,
+      "learning_rate": 4.798143630175642e-05,
+      "loss": 0.9048,
+      "step": 3555
+    },
+    {
+      "epoch": 0.6841092727972297,
+      "grad_norm": 1.8238990306854248,
+      "learning_rate": 4.7928231660284054e-05,
+      "loss": 0.9515,
+      "step": 3556
+    },
+    {
+      "epoch": 0.6843016544824932,
+      "grad_norm": 1.9509066343307495,
+      "learning_rate": 4.7875047234865045e-05,
+      "loss": 1.0006,
+      "step": 3557
+    },
+    {
+      "epoch": 0.6844940361677568,
+      "grad_norm": 1.6572667360305786,
+      "learning_rate": 4.782188304614741e-05,
+      "loss": 0.9988,
+      "step": 3558
+    },
+    {
+      "epoch": 0.6846864178530204,
+      "grad_norm": 1.1591516733169556,
+      "learning_rate": 4.776873911477132e-05,
+      "loss": 1.0081,
+      "step": 3559
+    },
+    {
+      "epoch": 0.6848787995382839,
+      "grad_norm": 2.429577112197876,
+      "learning_rate": 4.771561546136908e-05,
+      "loss": 0.9707,
+      "step": 3560
+    },
+    {
+      "epoch": 0.6850711812235475,
+      "grad_norm": 2.7014219760894775,
+      "learning_rate": 4.766251210656509e-05,
+      "loss": 0.9557,
+      "step": 3561
+    },
+    {
+      "epoch": 0.6852635629088111,
+      "grad_norm": 2.560462713241577,
+      "learning_rate": 4.7609429070976006e-05,
+      "loss": 0.9182,
+      "step": 3562
+    },
+    {
+      "epoch": 0.6854559445940747,
+      "grad_norm": 1.690266728401184,
+      "learning_rate": 4.7556366375210345e-05,
+      "loss": 0.8429,
+      "step": 3563
+    },
+    {
+      "epoch": 0.6856483262793382,
+      "grad_norm": 1.717684030532837,
+      "learning_rate": 4.750332403986901e-05,
+      "loss": 0.9565,
+      "step": 3564
+    },
+    {
+      "epoch": 0.6858407079646017,
+      "grad_norm": 2.6270174980163574,
+      "learning_rate": 4.745030208554473e-05,
+      "loss": 0.9471,
+      "step": 3565
+    },
+    {
+      "epoch": 0.6860330896498653,
+      "grad_norm": 2.0913078784942627,
+      "learning_rate": 4.739730053282255e-05,
+      "loss": 0.9411,
+      "step": 3566
+    },
+    {
+      "epoch": 0.6862254713351289,
+      "grad_norm": 2.039673089981079,
+      "learning_rate": 4.734431940227951e-05,
+      "loss": 0.9711,
+      "step": 3567
+    },
+    {
+      "epoch": 0.6864178530203925,
+      "grad_norm": 1.0807304382324219,
+      "learning_rate": 4.729135871448459e-05,
+      "loss": 1.0082,
+      "step": 3568
+    },
+    {
+      "epoch": 0.6866102347056561,
+      "grad_norm": 2.829622745513916,
+      "learning_rate": 4.723841848999907e-05,
+      "loss": 0.9705,
+      "step": 3569
+    },
+    {
+      "epoch": 0.6868026163909196,
+      "grad_norm": 1.9166561365127563,
+      "learning_rate": 4.7185498749376125e-05,
+      "loss": 0.9937,
+      "step": 3570
+    },
+    {
+      "epoch": 0.6869949980761831,
+      "grad_norm": 2.7111732959747314,
+      "learning_rate": 4.713259951316103e-05,
+      "loss": 0.9098,
+      "step": 3571
+    },
+    {
+      "epoch": 0.6871873797614467,
+      "grad_norm": 1.1792412996292114,
+      "learning_rate": 4.7079720801891055e-05,
+      "loss": 0.9128,
+      "step": 3572
+    },
+    {
+      "epoch": 0.6873797614467103,
+      "grad_norm": 2.7696125507354736,
+      "learning_rate": 4.702686263609559e-05,
+      "loss": 0.9312,
+      "step": 3573
+    },
+    {
+      "epoch": 0.6875721431319738,
+      "grad_norm": 1.9944347143173218,
+      "learning_rate": 4.6974025036295954e-05,
+      "loss": 0.9015,
+      "step": 3574
+    },
+    {
+      "epoch": 0.6877645248172374,
+      "grad_norm": 1.5115556716918945,
+      "learning_rate": 4.6921208023005505e-05,
+      "loss": 0.9476,
+      "step": 3575
+    },
+    {
+      "epoch": 0.687956906502501,
+      "grad_norm": 2.1956491470336914,
+      "learning_rate": 4.6868411616729734e-05,
+      "loss": 1.0065,
+      "step": 3576
+    },
+    {
+      "epoch": 0.6881492881877646,
+      "grad_norm": 2.14644718170166,
+      "learning_rate": 4.681563583796587e-05,
+      "loss": 0.9809,
+      "step": 3577
+    },
+    {
+      "epoch": 0.6883416698730281,
+      "grad_norm": 1.9469869136810303,
+      "learning_rate": 4.6762880707203404e-05,
+      "loss": 0.9004,
+      "step": 3578
+    },
+    {
+      "epoch": 0.6885340515582916,
+      "grad_norm": 2.4530997276306152,
+      "learning_rate": 4.671014624492365e-05,
+      "loss": 0.9853,
+      "step": 3579
+    },
+    {
+      "epoch": 0.6887264332435552,
+      "grad_norm": 1.8070074319839478,
+      "learning_rate": 4.665743247159995e-05,
+      "loss": 0.9114,
+      "step": 3580
+    },
+    {
+      "epoch": 0.6889188149288188,
+      "grad_norm": 2.3112411499023438,
+      "learning_rate": 4.660473940769761e-05,
+      "loss": 0.9589,
+      "step": 3581
+    },
+    {
+      "epoch": 0.6891111966140824,
+      "grad_norm": 2.495697498321533,
+      "learning_rate": 4.6552067073673876e-05,
+      "loss": 0.9425,
+      "step": 3582
+    },
+    {
+      "epoch": 0.6893035782993459,
+      "grad_norm": 0.9760496020317078,
+      "learning_rate": 4.649941548997797e-05,
+      "loss": 1.011,
+      "step": 3583
+    },
+    {
+      "epoch": 0.6894959599846094,
+      "grad_norm": 1.2422596216201782,
+      "learning_rate": 4.644678467705101e-05,
+      "loss": 0.9455,
+      "step": 3584
+    },
+    {
+      "epoch": 0.689688341669873,
+      "grad_norm": 2.604342460632324,
+      "learning_rate": 4.6394174655326216e-05,
+      "loss": 0.9179,
+      "step": 3585
+    },
+    {
+      "epoch": 0.6898807233551366,
+      "grad_norm": 1.9366099834442139,
+      "learning_rate": 4.6341585445228485e-05,
+      "loss": 0.9561,
+      "step": 3586
+    },
+    {
+      "epoch": 0.6900731050404002,
+      "grad_norm": 2.48215651512146,
+      "learning_rate": 4.628901706717476e-05,
+      "loss": 0.9596,
+      "step": 3587
+    },
+    {
+      "epoch": 0.6902654867256637,
+      "grad_norm": 2.610804557800293,
+      "learning_rate": 4.623646954157399e-05,
+      "loss": 0.9404,
+      "step": 3588
+    },
+    {
+      "epoch": 0.6904578684109273,
+      "grad_norm": 2.15207839012146,
+      "learning_rate": 4.618394288882681e-05,
+      "loss": 0.9821,
+      "step": 3589
+    },
+    {
+      "epoch": 0.6906502500961909,
+      "grad_norm": 1.8594437837600708,
+      "learning_rate": 4.6131437129326025e-05,
+      "loss": 0.9493,
+      "step": 3590
+    },
+    {
+      "epoch": 0.6908426317814544,
+      "grad_norm": 1.261702537536621,
+      "learning_rate": 4.6078952283456024e-05,
+      "loss": 0.7936,
+      "step": 3591
+    },
+    {
+      "epoch": 0.6910350134667179,
+      "grad_norm": 0.7717821002006531,
+      "learning_rate": 4.6026488371593325e-05,
+      "loss": 0.9431,
+      "step": 3592
+    },
+    {
+      "epoch": 0.6912273951519815,
+      "grad_norm": 1.5624136924743652,
+      "learning_rate": 4.5974045414106216e-05,
+      "loss": 0.8201,
+      "step": 3593
+    },
+    {
+      "epoch": 0.6914197768372451,
+      "grad_norm": 2.3129239082336426,
+      "learning_rate": 4.592162343135483e-05,
+      "loss": 0.9437,
+      "step": 3594
+    },
+    {
+      "epoch": 0.6916121585225087,
+      "grad_norm": 2.1169686317443848,
+      "learning_rate": 4.586922244369122e-05,
+      "loss": 0.9508,
+      "step": 3595
+    },
+    {
+      "epoch": 0.6918045402077723,
+      "grad_norm": 2.4472224712371826,
+      "learning_rate": 4.581684247145922e-05,
+      "loss": 0.9238,
+      "step": 3596
+    },
+    {
+      "epoch": 0.6919969218930357,
+      "grad_norm": 1.437218189239502,
+      "learning_rate": 4.576448353499456e-05,
+      "loss": 1.0074,
+      "step": 3597
+    },
+    {
+      "epoch": 0.6921893035782993,
+      "grad_norm": 2.2439820766448975,
+      "learning_rate": 4.5712145654624774e-05,
+      "loss": 0.8902,
+      "step": 3598
+    },
+    {
+      "epoch": 0.6923816852635629,
+      "grad_norm": 1.603169322013855,
+      "learning_rate": 4.565982885066923e-05,
+      "loss": 0.8825,
+      "step": 3599
+    },
+    {
+      "epoch": 0.6925740669488265,
+      "grad_norm": 2.5067992210388184,
+      "learning_rate": 4.560753314343912e-05,
+      "loss": 0.9341,
+      "step": 3600
+    },
+    {
+      "epoch": 0.69276644863409,
+      "grad_norm": 2.0341849327087402,
+      "learning_rate": 4.555525855323738e-05,
+      "loss": 0.898,
+      "step": 3601
+    },
+    {
+      "epoch": 0.6929588303193536,
+      "grad_norm": 2.3477623462677,
+      "learning_rate": 4.5503005100358945e-05,
+      "loss": 0.9034,
+      "step": 3602
+    },
+    {
+      "epoch": 0.6931512120046172,
+      "grad_norm": 1.0688493251800537,
+      "learning_rate": 4.5450772805090226e-05,
+      "loss": 0.8164,
+      "step": 3603
+    },
+    {
+      "epoch": 0.6933435936898807,
+      "grad_norm": 1.1931219100952148,
+      "learning_rate": 4.539856168770974e-05,
+      "loss": 0.7595,
+      "step": 3604
+    },
+    {
+      "epoch": 0.6935359753751443,
+      "grad_norm": 2.2727198600769043,
+      "learning_rate": 4.5346371768487586e-05,
+      "loss": 0.9811,
+      "step": 3605
+    },
+    {
+      "epoch": 0.6937283570604078,
+      "grad_norm": 2.3567593097686768,
+      "learning_rate": 4.529420306768569e-05,
+      "loss": 1.0107,
+      "step": 3606
+    },
+    {
+      "epoch": 0.6939207387456714,
+      "grad_norm": 1.6884970664978027,
+      "learning_rate": 4.524205560555774e-05,
+      "loss": 0.8852,
+      "step": 3607
+    },
+    {
+      "epoch": 0.694113120430935,
+      "grad_norm": 1.9373750686645508,
+      "learning_rate": 4.518992940234917e-05,
+      "loss": 0.9437,
+      "step": 3608
+    },
+    {
+      "epoch": 0.6943055021161986,
+      "grad_norm": 2.4543495178222656,
+      "learning_rate": 4.5137824478297177e-05,
+      "loss": 0.9735,
+      "step": 3609
+    },
+    {
+      "epoch": 0.694497883801462,
+      "grad_norm": 2.7026314735412598,
+      "learning_rate": 4.508574085363064e-05,
+      "loss": 0.9457,
+      "step": 3610
+    },
+    {
+      "epoch": 0.6946902654867256,
+      "grad_norm": 1.3363816738128662,
+      "learning_rate": 4.5033678548570346e-05,
+      "loss": 0.9909,
+      "step": 3611
+    },
+    {
+      "epoch": 0.6948826471719892,
+      "grad_norm": 1.0404704809188843,
+      "learning_rate": 4.498163758332853e-05,
+      "loss": 0.8571,
+      "step": 3612
+    },
+    {
+      "epoch": 0.6950750288572528,
+      "grad_norm": 2.620303153991699,
+      "learning_rate": 4.4929617978109315e-05,
+      "loss": 1.0139,
+      "step": 3613
+    },
+    {
+      "epoch": 0.6952674105425164,
+      "grad_norm": 2.651364803314209,
+      "learning_rate": 4.48776197531086e-05,
+      "loss": 0.8944,
+      "step": 3614
+    },
+    {
+      "epoch": 0.6954597922277799,
+      "grad_norm": 2.0717689990997314,
+      "learning_rate": 4.4825642928513744e-05,
+      "loss": 0.925,
+      "step": 3615
+    },
+    {
+      "epoch": 0.6956521739130435,
+      "grad_norm": 2.075225591659546,
+      "learning_rate": 4.4773687524504084e-05,
+      "loss": 0.9593,
+      "step": 3616
+    },
+    {
+      "epoch": 0.695844555598307,
+      "grad_norm": 2.3270859718322754,
+      "learning_rate": 4.4721753561250354e-05,
+      "loss": 0.8998,
+      "step": 3617
+    },
+    {
+      "epoch": 0.6960369372835706,
+      "grad_norm": 2.3785927295684814,
+      "learning_rate": 4.466984105891521e-05,
+      "loss": 0.9396,
+      "step": 3618
+    },
+    {
+      "epoch": 0.6962293189688342,
+      "grad_norm": 2.2723803520202637,
+      "learning_rate": 4.4617950037652846e-05,
+      "loss": 1.0149,
+      "step": 3619
+    },
+    {
+      "epoch": 0.6964217006540977,
+      "grad_norm": 2.123861789703369,
+      "learning_rate": 4.4566080517609134e-05,
+      "loss": 0.9911,
+      "step": 3620
+    },
+    {
+      "epoch": 0.6966140823393613,
+      "grad_norm": 2.4288551807403564,
+      "learning_rate": 4.4514232518921605e-05,
+      "loss": 0.9675,
+      "step": 3621
+    },
+    {
+      "epoch": 0.6968064640246249,
+      "grad_norm": 2.548661947250366,
+      "learning_rate": 4.4462406061719455e-05,
+      "loss": 0.9576,
+      "step": 3622
+    },
+    {
+      "epoch": 0.6969988457098885,
+      "grad_norm": 1.9252896308898926,
+      "learning_rate": 4.441060116612348e-05,
+      "loss": 0.9701,
+      "step": 3623
+    },
+    {
+      "epoch": 0.6971912273951519,
+      "grad_norm": 1.4989999532699585,
+      "learning_rate": 4.435881785224613e-05,
+      "loss": 0.8042,
+      "step": 3624
+    },
+    {
+      "epoch": 0.6973836090804155,
+      "grad_norm": 2.239698886871338,
+      "learning_rate": 4.430705614019147e-05,
+      "loss": 0.9124,
+      "step": 3625
+    },
+    {
+      "epoch": 0.6975759907656791,
+      "grad_norm": 2.2189810276031494,
+      "learning_rate": 4.425531605005518e-05,
+      "loss": 0.9972,
+      "step": 3626
+    },
+    {
+      "epoch": 0.6977683724509427,
+      "grad_norm": 0.986457884311676,
+      "learning_rate": 4.420359760192452e-05,
+      "loss": 1.0018,
+      "step": 3627
+    },
+    {
+      "epoch": 0.6979607541362063,
+      "grad_norm": 3.0673553943634033,
+      "learning_rate": 4.415190081587845e-05,
+      "loss": 0.8899,
+      "step": 3628
+    },
+    {
+      "epoch": 0.6981531358214698,
+      "grad_norm": 2.8216331005096436,
+      "learning_rate": 4.4100225711987344e-05,
+      "loss": 0.9428,
+      "step": 3629
+    },
+    {
+      "epoch": 0.6983455175067333,
+      "grad_norm": 2.3743374347686768,
+      "learning_rate": 4.404857231031332e-05,
+      "loss": 0.9424,
+      "step": 3630
+    },
+    {
+      "epoch": 0.6985378991919969,
+      "grad_norm": 1.0655142068862915,
+      "learning_rate": 4.3996940630909985e-05,
+      "loss": 1.0136,
+      "step": 3631
+    },
+    {
+      "epoch": 0.6987302808772605,
+      "grad_norm": 1.8283427953720093,
+      "learning_rate": 4.394533069382255e-05,
+      "loss": 1.0119,
+      "step": 3632
+    },
+    {
+      "epoch": 0.698922662562524,
+      "grad_norm": 0.9408661127090454,
+      "learning_rate": 4.3893742519087754e-05,
+      "loss": 0.9205,
+      "step": 3633
+    },
+    {
+      "epoch": 0.6991150442477876,
+      "grad_norm": 2.5338339805603027,
+      "learning_rate": 4.3842176126733916e-05,
+      "loss": 0.9665,
+      "step": 3634
+    },
+    {
+      "epoch": 0.6993074259330512,
+      "grad_norm": 1.604905128479004,
+      "learning_rate": 4.379063153678087e-05,
+      "loss": 0.9515,
+      "step": 3635
+    },
+    {
+      "epoch": 0.6994998076183148,
+      "grad_norm": 2.1519346237182617,
+      "learning_rate": 4.373910876923997e-05,
+      "loss": 1.0121,
+      "step": 3636
+    },
+    {
+      "epoch": 0.6996921893035783,
+      "grad_norm": 2.298600673675537,
+      "learning_rate": 4.368760784411423e-05,
+      "loss": 0.9014,
+      "step": 3637
+    },
+    {
+      "epoch": 0.6998845709888418,
+      "grad_norm": 1.1726566553115845,
+      "learning_rate": 4.363612878139799e-05,
+      "loss": 0.9468,
+      "step": 3638
+    },
+    {
+      "epoch": 0.7000769526741054,
+      "grad_norm": 2.4414029121398926,
+      "learning_rate": 4.358467160107722e-05,
+      "loss": 0.9996,
+      "step": 3639
+    },
+    {
+      "epoch": 0.700269334359369,
+      "grad_norm": 2.382627010345459,
+      "learning_rate": 4.353323632312938e-05,
+      "loss": 0.9976,
+      "step": 3640
+    },
+    {
+      "epoch": 0.7004617160446326,
+      "grad_norm": 0.9833115339279175,
+      "learning_rate": 4.348182296752336e-05,
+      "loss": 1.0063,
+      "step": 3641
+    },
+    {
+      "epoch": 0.7006540977298961,
+      "grad_norm": 2.1566519737243652,
+      "learning_rate": 4.343043155421971e-05,
+      "loss": 1.072,
+      "step": 3642
+    },
+    {
+      "epoch": 0.7008464794151597,
+      "grad_norm": 2.141711950302124,
+      "learning_rate": 4.337906210317021e-05,
+      "loss": 0.9844,
+      "step": 3643
+    },
+    {
+      "epoch": 0.7010388611004232,
+      "grad_norm": 1.704615592956543,
+      "learning_rate": 4.332771463431836e-05,
+      "loss": 0.9526,
+      "step": 3644
+    },
+    {
+      "epoch": 0.7012312427856868,
+      "grad_norm": 2.4329190254211426,
+      "learning_rate": 4.3276389167598975e-05,
+      "loss": 0.9405,
+      "step": 3645
+    },
+    {
+      "epoch": 0.7014236244709504,
+      "grad_norm": 2.2455036640167236,
+      "learning_rate": 4.322508572293835e-05,
+      "loss": 0.9567,
+      "step": 3646
+    },
+    {
+      "epoch": 0.7016160061562139,
+      "grad_norm": 1.7846271991729736,
+      "learning_rate": 4.3173804320254275e-05,
+      "loss": 0.9724,
+      "step": 3647
+    },
+    {
+      "epoch": 0.7018083878414775,
+      "grad_norm": 1.809287428855896,
+      "learning_rate": 4.312254497945595e-05,
+      "loss": 0.8133,
+      "step": 3648
+    },
+    {
+      "epoch": 0.7020007695267411,
+      "grad_norm": 2.547714948654175,
+      "learning_rate": 4.307130772044401e-05,
+      "loss": 0.961,
+      "step": 3649
+    },
+    {
+      "epoch": 0.7021931512120047,
+      "grad_norm": 2.277801036834717,
+      "learning_rate": 4.302009256311048e-05,
+      "loss": 0.942,
+      "step": 3650
+    },
+    {
+      "epoch": 0.7023855328972681,
+      "grad_norm": 2.5043978691101074,
+      "learning_rate": 4.296889952733898e-05,
+      "loss": 0.9095,
+      "step": 3651
+    },
+    {
+      "epoch": 0.7025779145825317,
+      "grad_norm": 2.3441593647003174,
+      "learning_rate": 4.291772863300427e-05,
+      "loss": 0.9546,
+      "step": 3652
+    },
+    {
+      "epoch": 0.7027702962677953,
+      "grad_norm": 1.1003873348236084,
+      "learning_rate": 4.286657989997268e-05,
+      "loss": 1.0127,
+      "step": 3653
+    },
+    {
+      "epoch": 0.7029626779530589,
+      "grad_norm": 2.5243847370147705,
+      "learning_rate": 4.281545334810201e-05,
+      "loss": 0.8881,
+      "step": 3654
+    },
+    {
+      "epoch": 0.7031550596383225,
+      "grad_norm": 2.2407445907592773,
+      "learning_rate": 4.276434899724119e-05,
+      "loss": 1.014,
+      "step": 3655
+    },
+    {
+      "epoch": 0.703347441323586,
+      "grad_norm": 1.2455822229385376,
+      "learning_rate": 4.2713266867230804e-05,
+      "loss": 0.9153,
+      "step": 3656
+    },
+    {
+      "epoch": 0.7035398230088495,
+      "grad_norm": 2.071686029434204,
+      "learning_rate": 4.2662206977902664e-05,
+      "loss": 0.98,
+      "step": 3657
+    },
+    {
+      "epoch": 0.7037322046941131,
+      "grad_norm": 2.0418829917907715,
+      "learning_rate": 4.261116934907998e-05,
+      "loss": 0.9914,
+      "step": 3658
+    },
+    {
+      "epoch": 0.7039245863793767,
+      "grad_norm": 2.7027108669281006,
+      "learning_rate": 4.2560154000577304e-05,
+      "loss": 0.9707,
+      "step": 3659
+    },
+    {
+      "epoch": 0.7041169680646402,
+      "grad_norm": 2.4418880939483643,
+      "learning_rate": 4.250916095220056e-05,
+      "loss": 0.9856,
+      "step": 3660
+    },
+    {
+      "epoch": 0.7043093497499038,
+      "grad_norm": 2.656592845916748,
+      "learning_rate": 4.2458190223747e-05,
+      "loss": 0.9285,
+      "step": 3661
+    },
+    {
+      "epoch": 0.7045017314351674,
+      "grad_norm": 2.684037923812866,
+      "learning_rate": 4.2407241835005176e-05,
+      "loss": 1.0203,
+      "step": 3662
+    },
+    {
+      "epoch": 0.704694113120431,
+      "grad_norm": 1.7849632501602173,
+      "learning_rate": 4.235631580575513e-05,
+      "loss": 1.0129,
+      "step": 3663
+    },
+    {
+      "epoch": 0.7048864948056945,
+      "grad_norm": 2.1810660362243652,
+      "learning_rate": 4.230541215576798e-05,
+      "loss": 0.9894,
+      "step": 3664
+    },
+    {
+      "epoch": 0.705078876490958,
+      "grad_norm": 3.011782169342041,
+      "learning_rate": 4.225453090480631e-05,
+      "loss": 1.0057,
+      "step": 3665
+    },
+    {
+      "epoch": 0.7052712581762216,
+      "grad_norm": 1.5482877492904663,
+      "learning_rate": 4.220367207262398e-05,
+      "loss": 0.8451,
+      "step": 3666
+    },
+    {
+      "epoch": 0.7054636398614852,
+      "grad_norm": 2.5375537872314453,
+      "learning_rate": 4.2152835678966096e-05,
+      "loss": 0.9832,
+      "step": 3667
+    },
+    {
+      "epoch": 0.7056560215467488,
+      "grad_norm": 1.683024525642395,
+      "learning_rate": 4.210202174356922e-05,
+      "loss": 1.0139,
+      "step": 3668
+    },
+    {
+      "epoch": 0.7058484032320124,
+      "grad_norm": 2.4783387184143066,
+      "learning_rate": 4.20512302861609e-05,
+      "loss": 0.9025,
+      "step": 3669
+    },
+    {
+      "epoch": 0.7060407849172758,
+      "grad_norm": 2.5154237747192383,
+      "learning_rate": 4.200046132646027e-05,
+      "loss": 0.9591,
+      "step": 3670
+    },
+    {
+      "epoch": 0.7062331666025394,
+      "grad_norm": 1.5583683252334595,
+      "learning_rate": 4.1949714884177525e-05,
+      "loss": 0.9673,
+      "step": 3671
+    },
+    {
+      "epoch": 0.706425548287803,
+      "grad_norm": 2.8709700107574463,
+      "learning_rate": 4.1898990979014205e-05,
+      "loss": 1.0138,
+      "step": 3672
+    },
+    {
+      "epoch": 0.7066179299730666,
+      "grad_norm": 1.6406999826431274,
+      "learning_rate": 4.184828963066305e-05,
+      "loss": 0.9875,
+      "step": 3673
+    },
+    {
+      "epoch": 0.7068103116583301,
+      "grad_norm": 3.075387954711914,
+      "learning_rate": 4.179761085880809e-05,
+      "loss": 0.959,
+      "step": 3674
+    },
+    {
+      "epoch": 0.7070026933435937,
+      "grad_norm": 2.369678497314453,
+      "learning_rate": 4.174695468312456e-05,
+      "loss": 0.9108,
+      "step": 3675
+    },
+    {
+      "epoch": 0.7071950750288573,
+      "grad_norm": 2.7644827365875244,
+      "learning_rate": 4.16963211232789e-05,
+      "loss": 0.9532,
+      "step": 3676
+    },
+    {
+      "epoch": 0.7073874567141208,
+      "grad_norm": 0.9318640828132629,
+      "learning_rate": 4.16457101989289e-05,
+      "loss": 0.9965,
+      "step": 3677
+    },
+    {
+      "epoch": 0.7075798383993844,
+      "grad_norm": 2.464282751083374,
+      "learning_rate": 4.159512192972337e-05,
+      "loss": 0.9273,
+      "step": 3678
+    },
+    {
+      "epoch": 0.7077722200846479,
+      "grad_norm": 2.4922072887420654,
+      "learning_rate": 4.15445563353024e-05,
+      "loss": 0.9484,
+      "step": 3679
+    },
+    {
+      "epoch": 0.7079646017699115,
+      "grad_norm": 1.3588006496429443,
+      "learning_rate": 4.1494013435297416e-05,
+      "loss": 0.754,
+      "step": 3680
+    },
+    {
+      "epoch": 0.7081569834551751,
+      "grad_norm": 1.8317455053329468,
+      "learning_rate": 4.1443493249330766e-05,
+      "loss": 0.912,
+      "step": 3681
+    },
+    {
+      "epoch": 0.7083493651404387,
+      "grad_norm": 2.0321757793426514,
+      "learning_rate": 4.139299579701623e-05,
+      "loss": 0.9104,
+      "step": 3682
+    },
+    {
+      "epoch": 0.7085417468257021,
+      "grad_norm": 2.601956367492676,
+      "learning_rate": 4.134252109795863e-05,
+      "loss": 0.8875,
+      "step": 3683
+    },
+    {
+      "epoch": 0.7087341285109657,
+      "grad_norm": 2.4826266765594482,
+      "learning_rate": 4.129206917175397e-05,
+      "loss": 1.0018,
+      "step": 3684
+    },
+    {
+      "epoch": 0.7089265101962293,
+      "grad_norm": 2.3331143856048584,
+      "learning_rate": 4.124164003798944e-05,
+      "loss": 0.9903,
+      "step": 3685
+    },
+    {
+      "epoch": 0.7091188918814929,
+      "grad_norm": 2.640472412109375,
+      "learning_rate": 4.119123371624335e-05,
+      "loss": 0.9876,
+      "step": 3686
+    },
+    {
+      "epoch": 0.7093112735667565,
+      "grad_norm": 2.421396255493164,
+      "learning_rate": 4.114085022608517e-05,
+      "loss": 0.9446,
+      "step": 3687
+    },
+    {
+      "epoch": 0.70950365525202,
+      "grad_norm": 3.07413649559021,
+      "learning_rate": 4.109048958707552e-05,
+      "loss": 1.0277,
+      "step": 3688
+    },
+    {
+      "epoch": 0.7096960369372836,
+      "grad_norm": 2.7209784984588623,
+      "learning_rate": 4.104015181876612e-05,
+      "loss": 0.9358,
+      "step": 3689
+    },
+    {
+      "epoch": 0.7098884186225471,
+      "grad_norm": 2.216418981552124,
+      "learning_rate": 4.09898369406998e-05,
+      "loss": 0.9514,
+      "step": 3690
+    },
+    {
+      "epoch": 0.7100808003078107,
+      "grad_norm": 2.6160953044891357,
+      "learning_rate": 4.093954497241064e-05,
+      "loss": 1.011,
+      "step": 3691
+    },
+    {
+      "epoch": 0.7102731819930742,
+      "grad_norm": 1.1653450727462769,
+      "learning_rate": 4.088927593342358e-05,
+      "loss": 0.9435,
+      "step": 3692
+    },
+    {
+      "epoch": 0.7104655636783378,
+      "grad_norm": 2.4288201332092285,
+      "learning_rate": 4.0839029843254814e-05,
+      "loss": 0.9063,
+      "step": 3693
+    },
+    {
+      "epoch": 0.7106579453636014,
+      "grad_norm": 2.0413899421691895,
+      "learning_rate": 4.078880672141171e-05,
+      "loss": 0.9118,
+      "step": 3694
+    },
+    {
+      "epoch": 0.710850327048865,
+      "grad_norm": 1.6144226789474487,
+      "learning_rate": 4.073860658739246e-05,
+      "loss": 1.0394,
+      "step": 3695
+    },
+    {
+      "epoch": 0.7110427087341286,
+      "grad_norm": 2.8568384647369385,
+      "learning_rate": 4.0688429460686605e-05,
+      "loss": 0.9202,
+      "step": 3696
+    },
+    {
+      "epoch": 0.711235090419392,
+      "grad_norm": 2.254446506500244,
+      "learning_rate": 4.0638275360774594e-05,
+      "loss": 0.8934,
+      "step": 3697
+    },
+    {
+      "epoch": 0.7114274721046556,
+      "grad_norm": 2.5451204776763916,
+      "learning_rate": 4.058814430712796e-05,
+      "loss": 0.9186,
+      "step": 3698
+    },
+    {
+      "epoch": 0.7116198537899192,
+      "grad_norm": 2.6671204566955566,
+      "learning_rate": 4.053803631920933e-05,
+      "loss": 0.9838,
+      "step": 3699
+    },
+    {
+      "epoch": 0.7118122354751828,
+      "grad_norm": 1.3629897832870483,
+      "learning_rate": 4.0487951416472326e-05,
+      "loss": 0.9807,
+      "step": 3700
+    },
+    {
+      "epoch": 0.7120046171604463,
+      "grad_norm": 1.8729207515716553,
+      "learning_rate": 4.0437889618361635e-05,
+      "loss": 0.9943,
+      "step": 3701
+    },
+    {
+      "epoch": 0.7121969988457099,
+      "grad_norm": 1.544274091720581,
+      "learning_rate": 4.0387850944312957e-05,
+      "loss": 0.9699,
+      "step": 3702
+    },
+    {
+      "epoch": 0.7123893805309734,
+      "grad_norm": 1.290138840675354,
+      "learning_rate": 4.033783541375311e-05,
+      "loss": 0.7289,
+      "step": 3703
+    },
+    {
+      "epoch": 0.712581762216237,
+      "grad_norm": 2.213052272796631,
+      "learning_rate": 4.028784304609976e-05,
+      "loss": 0.9826,
+      "step": 3704
+    },
+    {
+      "epoch": 0.7127741439015006,
+      "grad_norm": 2.5312366485595703,
+      "learning_rate": 4.0237873860761646e-05,
+      "loss": 0.969,
+      "step": 3705
+    },
+    {
+      "epoch": 0.7129665255867641,
+      "grad_norm": 1.5929797887802124,
+      "learning_rate": 4.018792787713864e-05,
+      "loss": 0.9436,
+      "step": 3706
+    },
+    {
+      "epoch": 0.7131589072720277,
+      "grad_norm": 1.397552728652954,
+      "learning_rate": 4.013800511462135e-05,
+      "loss": 0.9293,
+      "step": 3707
+    },
+    {
+      "epoch": 0.7133512889572913,
+      "grad_norm": 2.3914384841918945,
+      "learning_rate": 4.0088105592591616e-05,
+      "loss": 0.9271,
+      "step": 3708
+    },
+    {
+      "epoch": 0.7135436706425549,
+      "grad_norm": 1.2146204710006714,
+      "learning_rate": 4.003822933042213e-05,
+      "loss": 0.9445,
+      "step": 3709
+    },
+    {
+      "epoch": 0.7137360523278183,
+      "grad_norm": 1.7355114221572876,
+      "learning_rate": 3.998837634747655e-05,
+      "loss": 0.9193,
+      "step": 3710
+    },
+    {
+      "epoch": 0.7139284340130819,
+      "grad_norm": 2.355496883392334,
+      "learning_rate": 3.9938546663109544e-05,
+      "loss": 0.9445,
+      "step": 3711
+    },
+    {
+      "epoch": 0.7141208156983455,
+      "grad_norm": 2.1330971717834473,
+      "learning_rate": 3.9888740296666696e-05,
+      "loss": 1.0309,
+      "step": 3712
+    },
+    {
+      "epoch": 0.7143131973836091,
+      "grad_norm": 1.844709038734436,
+      "learning_rate": 3.983895726748455e-05,
+      "loss": 0.9349,
+      "step": 3713
+    },
+    {
+      "epoch": 0.7145055790688727,
+      "grad_norm": 1.243660807609558,
+      "learning_rate": 3.9789197594890606e-05,
+      "loss": 0.9866,
+      "step": 3714
+    },
+    {
+      "epoch": 0.7146979607541362,
+      "grad_norm": 2.304955005645752,
+      "learning_rate": 3.973946129820326e-05,
+      "loss": 0.9492,
+      "step": 3715
+    },
+    {
+      "epoch": 0.7148903424393998,
+      "grad_norm": 2.3657190799713135,
+      "learning_rate": 3.9689748396731854e-05,
+      "loss": 0.8558,
+      "step": 3716
+    },
+    {
+      "epoch": 0.7150827241246633,
+      "grad_norm": 2.519242286682129,
+      "learning_rate": 3.964005890977672e-05,
+      "loss": 0.9831,
+      "step": 3717
+    },
+    {
+      "epoch": 0.7152751058099269,
+      "grad_norm": 1.226691484451294,
+      "learning_rate": 3.959039285662894e-05,
+      "loss": 0.7459,
+      "step": 3718
+    },
+    {
+      "epoch": 0.7154674874951905,
+      "grad_norm": 1.8221051692962646,
+      "learning_rate": 3.954075025657058e-05,
+      "loss": 0.9174,
+      "step": 3719
+    },
+    {
+      "epoch": 0.715659869180454,
+      "grad_norm": 2.012173891067505,
+      "learning_rate": 3.949113112887471e-05,
+      "loss": 0.992,
+      "step": 3720
+    },
+    {
+      "epoch": 0.7158522508657176,
+      "grad_norm": 2.483128070831299,
+      "learning_rate": 3.944153549280506e-05,
+      "loss": 1.0392,
+      "step": 3721
+    },
+    {
+      "epoch": 0.7160446325509812,
+      "grad_norm": 2.482053279876709,
+      "learning_rate": 3.939196336761645e-05,
+      "loss": 0.9212,
+      "step": 3722
+    },
+    {
+      "epoch": 0.7162370142362448,
+      "grad_norm": 2.6003963947296143,
+      "learning_rate": 3.934241477255445e-05,
+      "loss": 1.0042,
+      "step": 3723
+    },
+    {
+      "epoch": 0.7164293959215082,
+      "grad_norm": 2.1753733158111572,
+      "learning_rate": 3.929288972685554e-05,
+      "loss": 0.8954,
+      "step": 3724
+    },
+    {
+      "epoch": 0.7166217776067718,
+      "grad_norm": 2.2712011337280273,
+      "learning_rate": 3.9243388249747046e-05,
+      "loss": 1.0131,
+      "step": 3725
+    },
+    {
+      "epoch": 0.7168141592920354,
+      "grad_norm": 1.4051201343536377,
+      "learning_rate": 3.919391036044715e-05,
+      "loss": 0.769,
+      "step": 3726
+    },
+    {
+      "epoch": 0.717006540977299,
+      "grad_norm": 2.004920482635498,
+      "learning_rate": 3.914445607816486e-05,
+      "loss": 0.8901,
+      "step": 3727
+    },
+    {
+      "epoch": 0.7171989226625626,
+      "grad_norm": 1.0040428638458252,
+      "learning_rate": 3.9095025422100003e-05,
+      "loss": 0.98,
+      "step": 3728
+    },
+    {
+      "epoch": 0.717391304347826,
+      "grad_norm": 2.555549144744873,
+      "learning_rate": 3.9045618411443374e-05,
+      "loss": 0.9259,
+      "step": 3729
+    },
+    {
+      "epoch": 0.7175836860330896,
+      "grad_norm": 2.810460090637207,
+      "learning_rate": 3.8996235065376354e-05,
+      "loss": 0.9697,
+      "step": 3730
+    },
+    {
+      "epoch": 0.7177760677183532,
+      "grad_norm": 1.7384542226791382,
+      "learning_rate": 3.894687540307127e-05,
+      "loss": 0.9455,
+      "step": 3731
+    },
+    {
+      "epoch": 0.7179684494036168,
+      "grad_norm": 2.581728458404541,
+      "learning_rate": 3.889753944369135e-05,
+      "loss": 1.0543,
+      "step": 3732
+    },
+    {
+      "epoch": 0.7181608310888803,
+      "grad_norm": 1.200973629951477,
+      "learning_rate": 3.8848227206390364e-05,
+      "loss": 0.9882,
+      "step": 3733
+    },
+    {
+      "epoch": 0.7183532127741439,
+      "grad_norm": 2.3249688148498535,
+      "learning_rate": 3.879893871031314e-05,
+      "loss": 0.9005,
+      "step": 3734
+    },
+    {
+      "epoch": 0.7185455944594075,
+      "grad_norm": 2.855922222137451,
+      "learning_rate": 3.8749673974595105e-05,
+      "loss": 0.926,
+      "step": 3735
+    },
+    {
+      "epoch": 0.718737976144671,
+      "grad_norm": 2.3155975341796875,
+      "learning_rate": 3.870043301836256e-05,
+      "loss": 0.9468,
+      "step": 3736
+    },
+    {
+      "epoch": 0.7189303578299346,
+      "grad_norm": 2.1903791427612305,
+      "learning_rate": 3.865121586073251e-05,
+      "loss": 1.0032,
+      "step": 3737
+    },
+    {
+      "epoch": 0.7191227395151981,
+      "grad_norm": 3.194258689880371,
+      "learning_rate": 3.8602022520812764e-05,
+      "loss": 0.9954,
+      "step": 3738
+    },
+    {
+      "epoch": 0.7193151212004617,
+      "grad_norm": 2.4303500652313232,
+      "learning_rate": 3.855285301770187e-05,
+      "loss": 0.9409,
+      "step": 3739
+    },
+    {
+      "epoch": 0.7195075028857253,
+      "grad_norm": 2.3586716651916504,
+      "learning_rate": 3.850370737048913e-05,
+      "loss": 0.9147,
+      "step": 3740
+    },
+    {
+      "epoch": 0.7196998845709889,
+      "grad_norm": 2.2538273334503174,
+      "learning_rate": 3.845458559825457e-05,
+      "loss": 0.9169,
+      "step": 3741
+    },
+    {
+      "epoch": 0.7198922662562524,
+      "grad_norm": 2.4147613048553467,
+      "learning_rate": 3.840548772006891e-05,
+      "loss": 0.9179,
+      "step": 3742
+    },
+    {
+      "epoch": 0.7200846479415159,
+      "grad_norm": 2.120490550994873,
+      "learning_rate": 3.835641375499375e-05,
+      "loss": 0.9385,
+      "step": 3743
+    },
+    {
+      "epoch": 0.7202770296267795,
+      "grad_norm": 1.640967607498169,
+      "learning_rate": 3.830736372208118e-05,
+      "loss": 0.9821,
+      "step": 3744
+    },
+    {
+      "epoch": 0.7204694113120431,
+      "grad_norm": 2.556706666946411,
+      "learning_rate": 3.825833764037412e-05,
+      "loss": 0.9478,
+      "step": 3745
+    },
+    {
+      "epoch": 0.7206617929973067,
+      "grad_norm": 2.5956296920776367,
+      "learning_rate": 3.820933552890629e-05,
+      "loss": 0.8982,
+      "step": 3746
+    },
+    {
+      "epoch": 0.7208541746825702,
+      "grad_norm": 2.274062395095825,
+      "learning_rate": 3.816035740670185e-05,
+      "loss": 0.9374,
+      "step": 3747
+    },
+    {
+      "epoch": 0.7210465563678338,
+      "grad_norm": 1.1468229293823242,
+      "learning_rate": 3.8111403292775905e-05,
+      "loss": 0.9608,
+      "step": 3748
+    },
+    {
+      "epoch": 0.7212389380530974,
+      "grad_norm": 1.7839531898498535,
+      "learning_rate": 3.8062473206134083e-05,
+      "loss": 0.8171,
+      "step": 3749
+    },
+    {
+      "epoch": 0.7214313197383609,
+      "grad_norm": 0.9098085165023804,
+      "learning_rate": 3.801356716577273e-05,
+      "loss": 0.9837,
+      "step": 3750
+    },
+    {
+      "epoch": 0.7216237014236244,
+      "grad_norm": 1.98843514919281,
+      "learning_rate": 3.796468519067887e-05,
+      "loss": 0.9187,
+      "step": 3751
+    },
+    {
+      "epoch": 0.721816083108888,
+      "grad_norm": 2.2435121536254883,
+      "learning_rate": 3.7915827299830155e-05,
+      "loss": 0.9549,
+      "step": 3752
+    },
+    {
+      "epoch": 0.7220084647941516,
+      "grad_norm": 2.4917397499084473,
+      "learning_rate": 3.786699351219489e-05,
+      "loss": 0.897,
+      "step": 3753
+    },
+    {
+      "epoch": 0.7222008464794152,
+      "grad_norm": 2.3775811195373535,
+      "learning_rate": 3.7818183846732024e-05,
+      "loss": 0.9469,
+      "step": 3754
+    },
+    {
+      "epoch": 0.7223932281646788,
+      "grad_norm": 2.142181873321533,
+      "learning_rate": 3.776939832239125e-05,
+      "loss": 0.8182,
+      "step": 3755
+    },
+    {
+      "epoch": 0.7225856098499422,
+      "grad_norm": 2.293530225753784,
+      "learning_rate": 3.772063695811262e-05,
+      "loss": 0.8869,
+      "step": 3756
+    },
+    {
+      "epoch": 0.7227779915352058,
+      "grad_norm": 1.0942679643630981,
+      "learning_rate": 3.767189977282711e-05,
+      "loss": 0.9749,
+      "step": 3757
+    },
+    {
+      "epoch": 0.7229703732204694,
+      "grad_norm": 1.856986165046692,
+      "learning_rate": 3.762318678545616e-05,
+      "loss": 0.9174,
+      "step": 3758
+    },
+    {
+      "epoch": 0.723162754905733,
+      "grad_norm": 2.2710602283477783,
+      "learning_rate": 3.7574498014911716e-05,
+      "loss": 0.8988,
+      "step": 3759
+    },
+    {
+      "epoch": 0.7233551365909965,
+      "grad_norm": 2.355398178100586,
+      "learning_rate": 3.752583348009657e-05,
+      "loss": 0.8225,
+      "step": 3760
+    },
+    {
+      "epoch": 0.7235475182762601,
+      "grad_norm": 2.676571846008301,
+      "learning_rate": 3.7477193199903904e-05,
+      "loss": 0.9941,
+      "step": 3761
+    },
+    {
+      "epoch": 0.7237398999615237,
+      "grad_norm": 2.358551263809204,
+      "learning_rate": 3.742857719321756e-05,
+      "loss": 0.9786,
+      "step": 3762
+    },
+    {
+      "epoch": 0.7239322816467872,
+      "grad_norm": 2.1370701789855957,
+      "learning_rate": 3.737998547891195e-05,
+      "loss": 0.9708,
+      "step": 3763
+    },
+    {
+      "epoch": 0.7241246633320508,
+      "grad_norm": 1.8180936574935913,
+      "learning_rate": 3.7331418075852053e-05,
+      "loss": 0.9875,
+      "step": 3764
+    },
+    {
+      "epoch": 0.7243170450173143,
+      "grad_norm": 2.383790969848633,
+      "learning_rate": 3.728287500289339e-05,
+      "loss": 0.987,
+      "step": 3765
+    },
+    {
+      "epoch": 0.7245094267025779,
+      "grad_norm": 2.477219820022583,
+      "learning_rate": 3.723435627888208e-05,
+      "loss": 0.9498,
+      "step": 3766
+    },
+    {
+      "epoch": 0.7247018083878415,
+      "grad_norm": 2.159367561340332,
+      "learning_rate": 3.718586192265473e-05,
+      "loss": 0.917,
+      "step": 3767
+    },
+    {
+      "epoch": 0.7248941900731051,
+      "grad_norm": 2.7767930030822754,
+      "learning_rate": 3.7137391953038514e-05,
+      "loss": 1.0415,
+      "step": 3768
+    },
+    {
+      "epoch": 0.7250865717583687,
+      "grad_norm": 1.822189450263977,
+      "learning_rate": 3.708894638885122e-05,
+      "loss": 0.9262,
+      "step": 3769
+    },
+    {
+      "epoch": 0.7252789534436321,
+      "grad_norm": 2.5770721435546875,
+      "learning_rate": 3.7040525248901005e-05,
+      "loss": 0.9251,
+      "step": 3770
+    },
+    {
+      "epoch": 0.7254713351288957,
+      "grad_norm": 3.0554754734039307,
+      "learning_rate": 3.6992128551986616e-05,
+      "loss": 1.0333,
+      "step": 3771
+    },
+    {
+      "epoch": 0.7256637168141593,
+      "grad_norm": 2.098115921020508,
+      "learning_rate": 3.6943756316897404e-05,
+      "loss": 0.972,
+      "step": 3772
+    },
+    {
+      "epoch": 0.7258560984994229,
+      "grad_norm": 1.0479977130889893,
+      "learning_rate": 3.6895408562413025e-05,
+      "loss": 0.974,
+      "step": 3773
+    },
+    {
+      "epoch": 0.7260484801846864,
+      "grad_norm": 2.7691922187805176,
+      "learning_rate": 3.684708530730382e-05,
+      "loss": 0.9513,
+      "step": 3774
+    },
+    {
+      "epoch": 0.72624086186995,
+      "grad_norm": 2.5486621856689453,
+      "learning_rate": 3.6798786570330525e-05,
+      "loss": 0.9606,
+      "step": 3775
+    },
+    {
+      "epoch": 0.7264332435552135,
+      "grad_norm": 2.4007132053375244,
+      "learning_rate": 3.675051237024436e-05,
+      "loss": 0.9324,
+      "step": 3776
+    },
+    {
+      "epoch": 0.7266256252404771,
+      "grad_norm": 2.337639331817627,
+      "learning_rate": 3.670226272578704e-05,
+      "loss": 0.9446,
+      "step": 3777
+    },
+    {
+      "epoch": 0.7268180069257407,
+      "grad_norm": 2.9339699745178223,
+      "learning_rate": 3.665403765569073e-05,
+      "loss": 0.9375,
+      "step": 3778
+    },
+    {
+      "epoch": 0.7270103886110042,
+      "grad_norm": 1.181991457939148,
+      "learning_rate": 3.660583717867807e-05,
+      "loss": 1.0057,
+      "step": 3779
+    },
+    {
+      "epoch": 0.7272027702962678,
+      "grad_norm": 2.413912773132324,
+      "learning_rate": 3.655766131346211e-05,
+      "loss": 1.0353,
+      "step": 3780
+    },
+    {
+      "epoch": 0.7273951519815314,
+      "grad_norm": 2.3493611812591553,
+      "learning_rate": 3.650951007874648e-05,
+      "loss": 1.0435,
+      "step": 3781
+    },
+    {
+      "epoch": 0.727587533666795,
+      "grad_norm": 2.342694044113159,
+      "learning_rate": 3.646138349322501e-05,
+      "loss": 0.951,
+      "step": 3782
+    },
+    {
+      "epoch": 0.7277799153520584,
+      "grad_norm": 1.9056546688079834,
+      "learning_rate": 3.641328157558219e-05,
+      "loss": 0.9171,
+      "step": 3783
+    },
+    {
+      "epoch": 0.727972297037322,
+      "grad_norm": 2.5930206775665283,
+      "learning_rate": 3.6365204344492867e-05,
+      "loss": 0.9893,
+      "step": 3784
+    },
+    {
+      "epoch": 0.7281646787225856,
+      "grad_norm": 1.5241378545761108,
+      "learning_rate": 3.631715181862215e-05,
+      "loss": 0.893,
+      "step": 3785
+    },
+    {
+      "epoch": 0.7283570604078492,
+      "grad_norm": 2.2311649322509766,
+      "learning_rate": 3.62691240166258e-05,
+      "loss": 0.9362,
+      "step": 3786
+    },
+    {
+      "epoch": 0.7285494420931128,
+      "grad_norm": 2.6636040210723877,
+      "learning_rate": 3.6221120957149825e-05,
+      "loss": 1.0083,
+      "step": 3787
+    },
+    {
+      "epoch": 0.7287418237783763,
+      "grad_norm": 2.258820056915283,
+      "learning_rate": 3.617314265883066e-05,
+      "loss": 0.9177,
+      "step": 3788
+    },
+    {
+      "epoch": 0.7289342054636399,
+      "grad_norm": 3.1157655715942383,
+      "learning_rate": 3.6125189140295145e-05,
+      "loss": 0.9643,
+      "step": 3789
+    },
+    {
+      "epoch": 0.7291265871489034,
+      "grad_norm": 2.6368567943573,
+      "learning_rate": 3.607726042016049e-05,
+      "loss": 0.98,
+      "step": 3790
+    },
+    {
+      "epoch": 0.729318968834167,
+      "grad_norm": 2.3443024158477783,
+      "learning_rate": 3.602935651703424e-05,
+      "loss": 0.9925,
+      "step": 3791
+    },
+    {
+      "epoch": 0.7295113505194305,
+      "grad_norm": 2.6623315811157227,
+      "learning_rate": 3.598147744951438e-05,
+      "loss": 0.9004,
+      "step": 3792
+    },
+    {
+      "epoch": 0.7297037322046941,
+      "grad_norm": 2.6795613765716553,
+      "learning_rate": 3.5933623236189195e-05,
+      "loss": 0.9837,
+      "step": 3793
+    },
+    {
+      "epoch": 0.7298961138899577,
+      "grad_norm": 2.2525198459625244,
+      "learning_rate": 3.58857938956373e-05,
+      "loss": 0.966,
+      "step": 3794
+    },
+    {
+      "epoch": 0.7300884955752213,
+      "grad_norm": 1.2505733966827393,
+      "learning_rate": 3.58379894464278e-05,
+      "loss": 0.7571,
+      "step": 3795
+    },
+    {
+      "epoch": 0.7302808772604849,
+      "grad_norm": 2.853820323944092,
+      "learning_rate": 3.57902099071199e-05,
+      "loss": 0.984,
+      "step": 3796
+    },
+    {
+      "epoch": 0.7304732589457483,
+      "grad_norm": 2.3997600078582764,
+      "learning_rate": 3.574245529626334e-05,
+      "loss": 1.0107,
+      "step": 3797
+    },
+    {
+      "epoch": 0.7306656406310119,
+      "grad_norm": 2.4308624267578125,
+      "learning_rate": 3.569472563239814e-05,
+      "loss": 0.9695,
+      "step": 3798
+    },
+    {
+      "epoch": 0.7308580223162755,
+      "grad_norm": 2.400533676147461,
+      "learning_rate": 3.5647020934054464e-05,
+      "loss": 0.8468,
+      "step": 3799
+    },
+    {
+      "epoch": 0.7310504040015391,
+      "grad_norm": 2.514747381210327,
+      "learning_rate": 3.559934121975304e-05,
+      "loss": 0.8764,
+      "step": 3800
+    },
+    {
+      "epoch": 0.7312427856868026,
+      "grad_norm": 1.9845913648605347,
+      "learning_rate": 3.555168650800473e-05,
+      "loss": 1.0414,
+      "step": 3801
+    },
+    {
+      "epoch": 0.7314351673720662,
+      "grad_norm": 2.351522922515869,
+      "learning_rate": 3.5504056817310735e-05,
+      "loss": 0.9677,
+      "step": 3802
+    },
+    {
+      "epoch": 0.7316275490573297,
+      "grad_norm": 1.805031180381775,
+      "learning_rate": 3.5456452166162545e-05,
+      "loss": 0.9595,
+      "step": 3803
+    },
+    {
+      "epoch": 0.7318199307425933,
+      "grad_norm": 2.0242161750793457,
+      "learning_rate": 3.540887257304193e-05,
+      "loss": 0.9413,
+      "step": 3804
+    },
+    {
+      "epoch": 0.7320123124278569,
+      "grad_norm": 2.0787012577056885,
+      "learning_rate": 3.5361318056420925e-05,
+      "loss": 0.9789,
+      "step": 3805
+    },
+    {
+      "epoch": 0.7322046941131204,
+      "grad_norm": 2.264993906021118,
+      "learning_rate": 3.531378863476178e-05,
+      "loss": 0.9548,
+      "step": 3806
+    },
+    {
+      "epoch": 0.732397075798384,
+      "grad_norm": 1.5229896306991577,
+      "learning_rate": 3.526628432651716e-05,
+      "loss": 0.7561,
+      "step": 3807
+    },
+    {
+      "epoch": 0.7325894574836476,
+      "grad_norm": 2.7864739894866943,
+      "learning_rate": 3.5218805150129755e-05,
+      "loss": 0.9656,
+      "step": 3808
+    },
+    {
+      "epoch": 0.7327818391689112,
+      "grad_norm": 2.24433970451355,
+      "learning_rate": 3.51713511240327e-05,
+      "loss": 0.9278,
+      "step": 3809
+    },
+    {
+      "epoch": 0.7329742208541746,
+      "grad_norm": 1.5711196660995483,
+      "learning_rate": 3.5123922266649303e-05,
+      "loss": 0.7897,
+      "step": 3810
+    },
+    {
+      "epoch": 0.7331666025394382,
+      "grad_norm": 2.140418767929077,
+      "learning_rate": 3.507651859639295e-05,
+      "loss": 0.9661,
+      "step": 3811
+    },
+    {
+      "epoch": 0.7333589842247018,
+      "grad_norm": 2.8539063930511475,
+      "learning_rate": 3.502914013166749e-05,
+      "loss": 0.9462,
+      "step": 3812
+    },
+    {
+      "epoch": 0.7335513659099654,
+      "grad_norm": 1.2900595664978027,
+      "learning_rate": 3.4981786890866854e-05,
+      "loss": 0.9333,
+      "step": 3813
+    },
+    {
+      "epoch": 0.733743747595229,
+      "grad_norm": 2.5251893997192383,
+      "learning_rate": 3.493445889237518e-05,
+      "loss": 0.9487,
+      "step": 3814
+    },
+    {
+      "epoch": 0.7339361292804925,
+      "grad_norm": 1.3003565073013306,
+      "learning_rate": 3.488715615456685e-05,
+      "loss": 0.9611,
+      "step": 3815
+    },
+    {
+      "epoch": 0.734128510965756,
+      "grad_norm": 2.3944621086120605,
+      "learning_rate": 3.483987869580638e-05,
+      "loss": 0.9876,
+      "step": 3816
+    },
+    {
+      "epoch": 0.7343208926510196,
+      "grad_norm": 1.5785565376281738,
+      "learning_rate": 3.4792626534448544e-05,
+      "loss": 0.9712,
+      "step": 3817
+    },
+    {
+      "epoch": 0.7345132743362832,
+      "grad_norm": 1.5962023735046387,
+      "learning_rate": 3.474539968883824e-05,
+      "loss": 1.0143,
+      "step": 3818
+    },
+    {
+      "epoch": 0.7347056560215467,
+      "grad_norm": 2.721494197845459,
+      "learning_rate": 3.469819817731056e-05,
+      "loss": 0.9591,
+      "step": 3819
+    },
+    {
+      "epoch": 0.7348980377068103,
+      "grad_norm": 1.3078265190124512,
+      "learning_rate": 3.4651022018190715e-05,
+      "loss": 0.9458,
+      "step": 3820
+    },
+    {
+      "epoch": 0.7350904193920739,
+      "grad_norm": 3.150883197784424,
+      "learning_rate": 3.460387122979423e-05,
+      "loss": 0.9627,
+      "step": 3821
+    },
+    {
+      "epoch": 0.7352828010773375,
+      "grad_norm": 0.9423494338989258,
+      "learning_rate": 3.455674583042652e-05,
+      "loss": 0.9532,
+      "step": 3822
+    },
+    {
+      "epoch": 0.735475182762601,
+      "grad_norm": 1.5937261581420898,
+      "learning_rate": 3.450964583838339e-05,
+      "loss": 0.8057,
+      "step": 3823
+    },
+    {
+      "epoch": 0.7356675644478645,
+      "grad_norm": 2.1796798706054688,
+      "learning_rate": 3.446257127195067e-05,
+      "loss": 0.9238,
+      "step": 3824
+    },
+    {
+      "epoch": 0.7358599461331281,
+      "grad_norm": 2.729487895965576,
+      "learning_rate": 3.4415522149404235e-05,
+      "loss": 0.9393,
+      "step": 3825
+    },
+    {
+      "epoch": 0.7360523278183917,
+      "grad_norm": 2.56174373626709,
+      "learning_rate": 3.4368498489010283e-05,
+      "loss": 0.9934,
+      "step": 3826
+    },
+    {
+      "epoch": 0.7362447095036553,
+      "grad_norm": 2.7279481887817383,
+      "learning_rate": 3.432150030902497e-05,
+      "loss": 1.0092,
+      "step": 3827
+    },
+    {
+      "epoch": 0.7364370911889189,
+      "grad_norm": 2.3831698894500732,
+      "learning_rate": 3.427452762769462e-05,
+      "loss": 0.9702,
+      "step": 3828
+    },
+    {
+      "epoch": 0.7366294728741823,
+      "grad_norm": 1.543399453163147,
+      "learning_rate": 3.422758046325562e-05,
+      "loss": 0.9483,
+      "step": 3829
+    },
+    {
+      "epoch": 0.7368218545594459,
+      "grad_norm": 1.4908251762390137,
+      "learning_rate": 3.418065883393452e-05,
+      "loss": 0.8275,
+      "step": 3830
+    },
+    {
+      "epoch": 0.7370142362447095,
+      "grad_norm": 2.344865560531616,
+      "learning_rate": 3.4133762757947874e-05,
+      "loss": 0.9633,
+      "step": 3831
+    },
+    {
+      "epoch": 0.7372066179299731,
+      "grad_norm": 2.2170355319976807,
+      "learning_rate": 3.408689225350234e-05,
+      "loss": 0.9605,
+      "step": 3832
+    },
+    {
+      "epoch": 0.7373989996152366,
+      "grad_norm": 2.5441906452178955,
+      "learning_rate": 3.4040047338794756e-05,
+      "loss": 0.8876,
+      "step": 3833
+    },
+    {
+      "epoch": 0.7375913813005002,
+      "grad_norm": 1.6234301328659058,
+      "learning_rate": 3.399322803201178e-05,
+      "loss": 0.9535,
+      "step": 3834
+    },
+    {
+      "epoch": 0.7377837629857638,
+      "grad_norm": 1.4903086423873901,
+      "learning_rate": 3.394643435133041e-05,
+      "loss": 0.9272,
+      "step": 3835
+    },
+    {
+      "epoch": 0.7379761446710273,
+      "grad_norm": 1.4795399904251099,
+      "learning_rate": 3.389966631491751e-05,
+      "loss": 0.7944,
+      "step": 3836
+    },
+    {
+      "epoch": 0.7381685263562909,
+      "grad_norm": 2.4129555225372314,
+      "learning_rate": 3.385292394093006e-05,
+      "loss": 0.9651,
+      "step": 3837
+    },
+    {
+      "epoch": 0.7383609080415544,
+      "grad_norm": 2.228811025619507,
+      "learning_rate": 3.380620724751506e-05,
+      "loss": 0.9697,
+      "step": 3838
+    },
+    {
+      "epoch": 0.738553289726818,
+      "grad_norm": 2.53407883644104,
+      "learning_rate": 3.3759516252809477e-05,
+      "loss": 0.9188,
+      "step": 3839
+    },
+    {
+      "epoch": 0.7387456714120816,
+      "grad_norm": 2.1077351570129395,
+      "learning_rate": 3.3712850974940435e-05,
+      "loss": 0.9196,
+      "step": 3840
+    },
+    {
+      "epoch": 0.7389380530973452,
+      "grad_norm": 2.229576349258423,
+      "learning_rate": 3.366621143202498e-05,
+      "loss": 0.9075,
+      "step": 3841
+    },
+    {
+      "epoch": 0.7391304347826086,
+      "grad_norm": 2.2330482006073,
+      "learning_rate": 3.361959764217017e-05,
+      "loss": 0.9796,
+      "step": 3842
+    },
+    {
+      "epoch": 0.7393228164678722,
+      "grad_norm": 1.7443095445632935,
+      "learning_rate": 3.3573009623473126e-05,
+      "loss": 1.0472,
+      "step": 3843
+    },
+    {
+      "epoch": 0.7395151981531358,
+      "grad_norm": 1.967947244644165,
+      "learning_rate": 3.352644739402089e-05,
+      "loss": 0.966,
+      "step": 3844
+    },
+    {
+      "epoch": 0.7397075798383994,
+      "grad_norm": 1.1639596223831177,
+      "learning_rate": 3.3479910971890515e-05,
+      "loss": 0.9824,
+      "step": 3845
+    },
+    {
+      "epoch": 0.739899961523663,
+      "grad_norm": 1.5797845125198364,
+      "learning_rate": 3.3433400375149026e-05,
+      "loss": 0.9828,
+      "step": 3846
+    },
+    {
+      "epoch": 0.7400923432089265,
+      "grad_norm": 2.3101768493652344,
+      "learning_rate": 3.338691562185353e-05,
+      "loss": 0.9831,
+      "step": 3847
+    },
+    {
+      "epoch": 0.7402847248941901,
+      "grad_norm": 2.6812949180603027,
+      "learning_rate": 3.3340456730050884e-05,
+      "loss": 0.9942,
+      "step": 3848
+    },
+    {
+      "epoch": 0.7404771065794536,
+      "grad_norm": 1.486631155014038,
+      "learning_rate": 3.3294023717778125e-05,
+      "loss": 0.9628,
+      "step": 3849
+    },
+    {
+      "epoch": 0.7406694882647172,
+      "grad_norm": 2.254702568054199,
+      "learning_rate": 3.324761660306215e-05,
+      "loss": 0.9897,
+      "step": 3850
+    },
+    {
+      "epoch": 0.7408618699499807,
+      "grad_norm": 2.5484023094177246,
+      "learning_rate": 3.320123540391968e-05,
+      "loss": 0.903,
+      "step": 3851
+    },
+    {
+      "epoch": 0.7410542516352443,
+      "grad_norm": 1.488093614578247,
+      "learning_rate": 3.315488013835762e-05,
+      "loss": 0.813,
+      "step": 3852
+    },
+    {
+      "epoch": 0.7412466333205079,
+      "grad_norm": 2.129958152770996,
+      "learning_rate": 3.310855082437263e-05,
+      "loss": 0.9452,
+      "step": 3853
+    },
+    {
+      "epoch": 0.7414390150057715,
+      "grad_norm": 3.0452218055725098,
+      "learning_rate": 3.306224747995136e-05,
+      "loss": 0.926,
+      "step": 3854
+    },
+    {
+      "epoch": 0.7416313966910351,
+      "grad_norm": 2.454988718032837,
+      "learning_rate": 3.301597012307034e-05,
+      "loss": 0.8538,
+      "step": 3855
+    },
+    {
+      "epoch": 0.7418237783762985,
+      "grad_norm": 3.0438475608825684,
+      "learning_rate": 3.296971877169605e-05,
+      "loss": 0.9755,
+      "step": 3856
+    },
+    {
+      "epoch": 0.7420161600615621,
+      "grad_norm": 2.860647439956665,
+      "learning_rate": 3.292349344378486e-05,
+      "loss": 0.8977,
+      "step": 3857
+    },
+    {
+      "epoch": 0.7422085417468257,
+      "grad_norm": 2.5906081199645996,
+      "learning_rate": 3.287729415728298e-05,
+      "loss": 0.933,
+      "step": 3858
+    },
+    {
+      "epoch": 0.7424009234320893,
+      "grad_norm": 2.1425957679748535,
+      "learning_rate": 3.283112093012669e-05,
+      "loss": 0.9481,
+      "step": 3859
+    },
+    {
+      "epoch": 0.7425933051173528,
+      "grad_norm": 1.1661136150360107,
+      "learning_rate": 3.278497378024187e-05,
+      "loss": 0.8974,
+      "step": 3860
+    },
+    {
+      "epoch": 0.7427856868026164,
+      "grad_norm": 2.3713860511779785,
+      "learning_rate": 3.2738852725544544e-05,
+      "loss": 1.0292,
+      "step": 3861
+    },
+    {
+      "epoch": 0.74297806848788,
+      "grad_norm": 2.53062105178833,
+      "learning_rate": 3.269275778394047e-05,
+      "loss": 0.9124,
+      "step": 3862
+    },
+    {
+      "epoch": 0.7431704501731435,
+      "grad_norm": 2.1706185340881348,
+      "learning_rate": 3.264668897332527e-05,
+      "loss": 0.9318,
+      "step": 3863
+    },
+    {
+      "epoch": 0.7433628318584071,
+      "grad_norm": 2.5651159286499023,
+      "learning_rate": 3.260064631158449e-05,
+      "loss": 0.921,
+      "step": 3864
+    },
+    {
+      "epoch": 0.7435552135436706,
+      "grad_norm": 2.280496120452881,
+      "learning_rate": 3.2554629816593376e-05,
+      "loss": 1.02,
+      "step": 3865
+    },
+    {
+      "epoch": 0.7437475952289342,
+      "grad_norm": 2.323133945465088,
+      "learning_rate": 3.250863950621721e-05,
+      "loss": 0.8604,
+      "step": 3866
+    },
+    {
+      "epoch": 0.7439399769141978,
+      "grad_norm": 2.34867525100708,
+      "learning_rate": 3.2462675398310984e-05,
+      "loss": 0.8532,
+      "step": 3867
+    },
+    {
+      "epoch": 0.7441323585994614,
+      "grad_norm": 2.3248634338378906,
+      "learning_rate": 3.2416737510719544e-05,
+      "loss": 0.8973,
+      "step": 3868
+    },
+    {
+      "epoch": 0.7443247402847248,
+      "grad_norm": 2.282707691192627,
+      "learning_rate": 3.237082586127757e-05,
+      "loss": 0.9405,
+      "step": 3869
+    },
+    {
+      "epoch": 0.7445171219699884,
+      "grad_norm": 2.539942502975464,
+      "learning_rate": 3.232494046780953e-05,
+      "loss": 0.949,
+      "step": 3870
+    },
+    {
+      "epoch": 0.744709503655252,
+      "grad_norm": 2.481483221054077,
+      "learning_rate": 3.227908134812972e-05,
+      "loss": 0.9553,
+      "step": 3871
+    },
+    {
+      "epoch": 0.7449018853405156,
+      "grad_norm": 2.455275058746338,
+      "learning_rate": 3.223324852004219e-05,
+      "loss": 0.882,
+      "step": 3872
+    },
+    {
+      "epoch": 0.7450942670257792,
+      "grad_norm": 1.1716479063034058,
+      "learning_rate": 3.218744200134094e-05,
+      "loss": 0.9763,
+      "step": 3873
+    },
+    {
+      "epoch": 0.7452866487110427,
+      "grad_norm": 2.6570403575897217,
+      "learning_rate": 3.21416618098095e-05,
+      "loss": 0.977,
+      "step": 3874
+    },
+    {
+      "epoch": 0.7454790303963063,
+      "grad_norm": 2.6615257263183594,
+      "learning_rate": 3.2095907963221396e-05,
+      "loss": 0.8963,
+      "step": 3875
+    },
+    {
+      "epoch": 0.7456714120815698,
+      "grad_norm": 1.5858724117279053,
+      "learning_rate": 3.205018047933987e-05,
+      "loss": 0.7867,
+      "step": 3876
+    },
+    {
+      "epoch": 0.7458637937668334,
+      "grad_norm": 2.2716503143310547,
+      "learning_rate": 3.200447937591779e-05,
+      "loss": 0.9339,
+      "step": 3877
+    },
+    {
+      "epoch": 0.746056175452097,
+      "grad_norm": 1.446292519569397,
+      "learning_rate": 3.195880467069801e-05,
+      "loss": 0.8075,
+      "step": 3878
+    },
+    {
+      "epoch": 0.7462485571373605,
+      "grad_norm": 2.8510217666625977,
+      "learning_rate": 3.191315638141297e-05,
+      "loss": 0.9402,
+      "step": 3879
+    },
+    {
+      "epoch": 0.7464409388226241,
+      "grad_norm": 1.8792798519134521,
+      "learning_rate": 3.1867534525784934e-05,
+      "loss": 0.9047,
+      "step": 3880
+    },
+    {
+      "epoch": 0.7466333205078877,
+      "grad_norm": 1.349548578262329,
+      "learning_rate": 3.182193912152586e-05,
+      "loss": 0.9536,
+      "step": 3881
+    },
+    {
+      "epoch": 0.7468257021931513,
+      "grad_norm": 1.7199147939682007,
+      "learning_rate": 3.177637018633746e-05,
+      "loss": 0.973,
+      "step": 3882
+    },
+    {
+      "epoch": 0.7470180838784147,
+      "grad_norm": 1.6600416898727417,
+      "learning_rate": 3.173082773791116e-05,
+      "loss": 0.935,
+      "step": 3883
+    },
+    {
+      "epoch": 0.7472104655636783,
+      "grad_norm": 2.348742961883545,
+      "learning_rate": 3.168531179392808e-05,
+      "loss": 0.9703,
+      "step": 3884
+    },
+    {
+      "epoch": 0.7474028472489419,
+      "grad_norm": 2.4814839363098145,
+      "learning_rate": 3.163982237205917e-05,
+      "loss": 0.8572,
+      "step": 3885
+    },
+    {
+      "epoch": 0.7475952289342055,
+      "grad_norm": 2.324592113494873,
+      "learning_rate": 3.159435948996485e-05,
+      "loss": 0.9986,
+      "step": 3886
+    },
+    {
+      "epoch": 0.7477876106194691,
+      "grad_norm": 0.8735331296920776,
+      "learning_rate": 3.154892316529549e-05,
+      "loss": 0.9328,
+      "step": 3887
+    },
+    {
+      "epoch": 0.7479799923047326,
+      "grad_norm": 2.517611026763916,
+      "learning_rate": 3.150351341569101e-05,
+      "loss": 0.9151,
+      "step": 3888
+    },
+    {
+      "epoch": 0.7481723739899961,
+      "grad_norm": 2.131922721862793,
+      "learning_rate": 3.1458130258781005e-05,
+      "loss": 1.0244,
+      "step": 3889
+    },
+    {
+      "epoch": 0.7483647556752597,
+      "grad_norm": 2.6205451488494873,
+      "learning_rate": 3.141277371218484e-05,
+      "loss": 0.9457,
+      "step": 3890
+    },
+    {
+      "epoch": 0.7485571373605233,
+      "grad_norm": 1.7203587293624878,
+      "learning_rate": 3.1367443793511386e-05,
+      "loss": 0.9322,
+      "step": 3891
+    },
+    {
+      "epoch": 0.7487495190457868,
+      "grad_norm": 1.2446550130844116,
+      "learning_rate": 3.132214052035937e-05,
+      "loss": 0.8171,
+      "step": 3892
+    },
+    {
+      "epoch": 0.7489419007310504,
+      "grad_norm": 2.903545379638672,
+      "learning_rate": 3.127686391031705e-05,
+      "loss": 0.9627,
+      "step": 3893
+    },
+    {
+      "epoch": 0.749134282416314,
+      "grad_norm": 2.31449556350708,
+      "learning_rate": 3.123161398096237e-05,
+      "loss": 0.9374,
+      "step": 3894
+    },
+    {
+      "epoch": 0.7493266641015776,
+      "grad_norm": 2.2818844318389893,
+      "learning_rate": 3.11863907498629e-05,
+      "loss": 0.9909,
+      "step": 3895
+    },
+    {
+      "epoch": 0.7495190457868411,
+      "grad_norm": 2.518904447555542,
+      "learning_rate": 3.114119423457588e-05,
+      "loss": 0.9166,
+      "step": 3896
+    },
+    {
+      "epoch": 0.7497114274721046,
+      "grad_norm": 1.7459636926651,
+      "learning_rate": 3.109602445264812e-05,
+      "loss": 0.8989,
+      "step": 3897
+    },
+    {
+      "epoch": 0.7499038091573682,
+      "grad_norm": 2.8489131927490234,
+      "learning_rate": 3.1050881421616076e-05,
+      "loss": 0.994,
+      "step": 3898
+    },
+    {
+      "epoch": 0.7500961908426318,
+      "grad_norm": 2.170553207397461,
+      "learning_rate": 3.1005765159005904e-05,
+      "loss": 0.8835,
+      "step": 3899
+    },
+    {
+      "epoch": 0.7502885725278954,
+      "grad_norm": 2.782179355621338,
+      "learning_rate": 3.0960675682333186e-05,
+      "loss": 0.9957,
+      "step": 3900
+    },
+    {
+      "epoch": 0.7504809542131589,
+      "grad_norm": 2.709005355834961,
+      "learning_rate": 3.0915613009103295e-05,
+      "loss": 0.9676,
+      "step": 3901
+    },
+    {
+      "epoch": 0.7506733358984224,
+      "grad_norm": 2.217729330062866,
+      "learning_rate": 3.087057715681107e-05,
+      "loss": 0.8881,
+      "step": 3902
+    },
+    {
+      "epoch": 0.750865717583686,
+      "grad_norm": 2.8491084575653076,
+      "learning_rate": 3.0825568142940995e-05,
+      "loss": 0.9383,
+      "step": 3903
+    },
+    {
+      "epoch": 0.7510580992689496,
+      "grad_norm": 2.451951265335083,
+      "learning_rate": 3.078058598496711e-05,
+      "loss": 0.9179,
+      "step": 3904
+    },
+    {
+      "epoch": 0.7512504809542132,
+      "grad_norm": 1.9296663999557495,
+      "learning_rate": 3.0735630700353044e-05,
+      "loss": 0.9769,
+      "step": 3905
+    },
+    {
+      "epoch": 0.7514428626394767,
+      "grad_norm": 2.705906867980957,
+      "learning_rate": 3.069070230655198e-05,
+      "loss": 0.9493,
+      "step": 3906
+    },
+    {
+      "epoch": 0.7516352443247403,
+      "grad_norm": 2.3528995513916016,
+      "learning_rate": 3.0645800821006665e-05,
+      "loss": 0.9404,
+      "step": 3907
+    },
+    {
+      "epoch": 0.7518276260100039,
+      "grad_norm": 2.380770444869995,
+      "learning_rate": 3.060092626114941e-05,
+      "loss": 0.9042,
+      "step": 3908
+    },
+    {
+      "epoch": 0.7520200076952674,
+      "grad_norm": 2.310396432876587,
+      "learning_rate": 3.055607864440206e-05,
+      "loss": 0.9643,
+      "step": 3909
+    },
+    {
+      "epoch": 0.7522123893805309,
+      "grad_norm": 1.9172768592834473,
+      "learning_rate": 3.0511257988175978e-05,
+      "loss": 0.9032,
+      "step": 3910
+    },
+    {
+      "epoch": 0.7524047710657945,
+      "grad_norm": 2.3006343841552734,
+      "learning_rate": 3.0466464309872166e-05,
+      "loss": 0.8933,
+      "step": 3911
+    },
+    {
+      "epoch": 0.7525971527510581,
+      "grad_norm": 2.4233264923095703,
+      "learning_rate": 3.0421697626880962e-05,
+      "loss": 0.8537,
+      "step": 3912
+    },
+    {
+      "epoch": 0.7527895344363217,
+      "grad_norm": 2.2756433486938477,
+      "learning_rate": 3.0376957956582452e-05,
+      "loss": 0.9822,
+      "step": 3913
+    },
+    {
+      "epoch": 0.7529819161215853,
+      "grad_norm": 2.630666494369507,
+      "learning_rate": 3.0332245316345996e-05,
+      "loss": 1.0183,
+      "step": 3914
+    },
+    {
+      "epoch": 0.7531742978068487,
+      "grad_norm": 2.5131397247314453,
+      "learning_rate": 3.0287559723530667e-05,
+      "loss": 1.0105,
+      "step": 3915
+    },
+    {
+      "epoch": 0.7533666794921123,
+      "grad_norm": 2.502271890640259,
+      "learning_rate": 3.0242901195484953e-05,
+      "loss": 0.9568,
+      "step": 3916
+    },
+    {
+      "epoch": 0.7535590611773759,
+      "grad_norm": 2.267944097518921,
+      "learning_rate": 3.0198269749546737e-05,
+      "loss": 0.862,
+      "step": 3917
+    },
+    {
+      "epoch": 0.7537514428626395,
+      "grad_norm": 1.603880524635315,
+      "learning_rate": 3.0153665403043586e-05,
+      "loss": 0.9488,
+      "step": 3918
+    },
+    {
+      "epoch": 0.753943824547903,
+      "grad_norm": 2.36745548248291,
+      "learning_rate": 3.01090881732924e-05,
+      "loss": 0.9327,
+      "step": 3919
+    },
+    {
+      "epoch": 0.7541362062331666,
+      "grad_norm": 2.7339625358581543,
+      "learning_rate": 3.0064538077599603e-05,
+      "loss": 0.8855,
+      "step": 3920
+    },
+    {
+      "epoch": 0.7543285879184302,
+      "grad_norm": 2.161695957183838,
+      "learning_rate": 3.002001513326107e-05,
+      "loss": 0.8698,
+      "step": 3921
+    },
+    {
+      "epoch": 0.7545209696036937,
+      "grad_norm": 1.9461236000061035,
+      "learning_rate": 2.9975519357562155e-05,
+      "loss": 0.9144,
+      "step": 3922
+    },
+    {
+      "epoch": 0.7547133512889573,
+      "grad_norm": 2.2538444995880127,
+      "learning_rate": 2.9931050767777625e-05,
+      "loss": 0.9342,
+      "step": 3923
+    },
+    {
+      "epoch": 0.7549057329742208,
+      "grad_norm": 2.127108573913574,
+      "learning_rate": 2.9886609381171705e-05,
+      "loss": 0.8145,
+      "step": 3924
+    },
+    {
+      "epoch": 0.7550981146594844,
+      "grad_norm": 1.2924288511276245,
+      "learning_rate": 2.9842195214998158e-05,
+      "loss": 0.8959,
+      "step": 3925
+    },
+    {
+      "epoch": 0.755290496344748,
+      "grad_norm": 1.6531801223754883,
+      "learning_rate": 2.9797808286499973e-05,
+      "loss": 0.9298,
+      "step": 3926
+    },
+    {
+      "epoch": 0.7554828780300116,
+      "grad_norm": 1.204675316810608,
+      "learning_rate": 2.9753448612909775e-05,
+      "loss": 0.734,
+      "step": 3927
+    },
+    {
+      "epoch": 0.7556752597152752,
+      "grad_norm": 1.131517767906189,
+      "learning_rate": 2.9709116211449483e-05,
+      "loss": 0.9418,
+      "step": 3928
+    },
+    {
+      "epoch": 0.7558676414005386,
+      "grad_norm": 2.166621446609497,
+      "learning_rate": 2.966481109933047e-05,
+      "loss": 0.9737,
+      "step": 3929
+    },
+    {
+      "epoch": 0.7560600230858022,
+      "grad_norm": 2.510349750518799,
+      "learning_rate": 2.9620533293753495e-05,
+      "loss": 0.8874,
+      "step": 3930
+    },
+    {
+      "epoch": 0.7562524047710658,
+      "grad_norm": 2.333470106124878,
+      "learning_rate": 2.9576282811908728e-05,
+      "loss": 0.8927,
+      "step": 3931
+    },
+    {
+      "epoch": 0.7564447864563294,
+      "grad_norm": 2.365734577178955,
+      "learning_rate": 2.9532059670975732e-05,
+      "loss": 0.9678,
+      "step": 3932
+    },
+    {
+      "epoch": 0.7566371681415929,
+      "grad_norm": 2.725245714187622,
+      "learning_rate": 2.9487863888123456e-05,
+      "loss": 0.9001,
+      "step": 3933
+    },
+    {
+      "epoch": 0.7568295498268565,
+      "grad_norm": 2.6744444370269775,
+      "learning_rate": 2.9443695480510224e-05,
+      "loss": 0.9265,
+      "step": 3934
+    },
+    {
+      "epoch": 0.75702193151212,
+      "grad_norm": 1.498961091041565,
+      "learning_rate": 2.939955446528374e-05,
+      "loss": 0.7098,
+      "step": 3935
+    },
+    {
+      "epoch": 0.7572143131973836,
+      "grad_norm": 2.8316962718963623,
+      "learning_rate": 2.9355440859581018e-05,
+      "loss": 0.8977,
+      "step": 3936
+    },
+    {
+      "epoch": 0.7574066948826472,
+      "grad_norm": 1.5112662315368652,
+      "learning_rate": 2.931135468052858e-05,
+      "loss": 0.9958,
+      "step": 3937
+    },
+    {
+      "epoch": 0.7575990765679107,
+      "grad_norm": 2.6001625061035156,
+      "learning_rate": 2.9267295945242067e-05,
+      "loss": 0.9566,
+      "step": 3938
+    },
+    {
+      "epoch": 0.7577914582531743,
+      "grad_norm": 1.9037305116653442,
+      "learning_rate": 2.9223264670826743e-05,
+      "loss": 0.9716,
+      "step": 3939
+    },
+    {
+      "epoch": 0.7579838399384379,
+      "grad_norm": 2.339718818664551,
+      "learning_rate": 2.9179260874376913e-05,
+      "loss": 0.9672,
+      "step": 3940
+    },
+    {
+      "epoch": 0.7581762216237015,
+      "grad_norm": 1.49025559425354,
+      "learning_rate": 2.9135284572976486e-05,
+      "loss": 0.9425,
+      "step": 3941
+    },
+    {
+      "epoch": 0.7583686033089649,
+      "grad_norm": 2.907559871673584,
+      "learning_rate": 2.9091335783698515e-05,
+      "loss": 0.9227,
+      "step": 3942
+    },
+    {
+      "epoch": 0.7585609849942285,
+      "grad_norm": 2.6355481147766113,
+      "learning_rate": 2.9047414523605465e-05,
+      "loss": 0.9617,
+      "step": 3943
+    },
+    {
+      "epoch": 0.7587533666794921,
+      "grad_norm": 2.86198091506958,
+      "learning_rate": 2.9003520809749053e-05,
+      "loss": 0.9629,
+      "step": 3944
+    },
+    {
+      "epoch": 0.7589457483647557,
+      "grad_norm": 2.245133399963379,
+      "learning_rate": 2.8959654659170354e-05,
+      "loss": 1.0002,
+      "step": 3945
+    },
+    {
+      "epoch": 0.7591381300500193,
+      "grad_norm": 2.220444440841675,
+      "learning_rate": 2.8915816088899695e-05,
+      "loss": 0.9737,
+      "step": 3946
+    },
+    {
+      "epoch": 0.7593305117352828,
+      "grad_norm": 2.073909282684326,
+      "learning_rate": 2.8872005115956745e-05,
+      "loss": 0.964,
+      "step": 3947
+    },
+    {
+      "epoch": 0.7595228934205464,
+      "grad_norm": 1.3193913698196411,
+      "learning_rate": 2.8828221757350404e-05,
+      "loss": 0.7704,
+      "step": 3948
+    },
+    {
+      "epoch": 0.7597152751058099,
+      "grad_norm": 2.743717908859253,
+      "learning_rate": 2.8784466030078905e-05,
+      "loss": 0.9345,
+      "step": 3949
+    },
+    {
+      "epoch": 0.7599076567910735,
+      "grad_norm": 2.343154191970825,
+      "learning_rate": 2.874073795112967e-05,
+      "loss": 0.8783,
+      "step": 3950
+    },
+    {
+      "epoch": 0.760100038476337,
+      "grad_norm": 2.8090946674346924,
+      "learning_rate": 2.8697037537479564e-05,
+      "loss": 0.9713,
+      "step": 3951
+    },
+    {
+      "epoch": 0.7602924201616006,
+      "grad_norm": 1.3954426050186157,
+      "learning_rate": 2.8653364806094453e-05,
+      "loss": 0.7722,
+      "step": 3952
+    },
+    {
+      "epoch": 0.7604848018468642,
+      "grad_norm": 1.9566173553466797,
+      "learning_rate": 2.86097197739297e-05,
+      "loss": 1.0151,
+      "step": 3953
+    },
+    {
+      "epoch": 0.7606771835321278,
+      "grad_norm": 1.9745495319366455,
+      "learning_rate": 2.8566102457929755e-05,
+      "loss": 0.9778,
+      "step": 3954
+    },
+    {
+      "epoch": 0.7608695652173914,
+      "grad_norm": 2.720402240753174,
+      "learning_rate": 2.8522512875028395e-05,
+      "loss": 0.9779,
+      "step": 3955
+    },
+    {
+      "epoch": 0.7610619469026548,
+      "grad_norm": 2.722998857498169,
+      "learning_rate": 2.847895104214856e-05,
+      "loss": 0.9373,
+      "step": 3956
+    },
+    {
+      "epoch": 0.7612543285879184,
+      "grad_norm": 2.0419063568115234,
+      "learning_rate": 2.843541697620249e-05,
+      "loss": 0.8784,
+      "step": 3957
+    },
+    {
+      "epoch": 0.761446710273182,
+      "grad_norm": 1.7099955081939697,
+      "learning_rate": 2.8391910694091583e-05,
+      "loss": 0.9075,
+      "step": 3958
+    },
+    {
+      "epoch": 0.7616390919584456,
+      "grad_norm": 1.9524343013763428,
+      "learning_rate": 2.8348432212706445e-05,
+      "loss": 0.9574,
+      "step": 3959
+    },
+    {
+      "epoch": 0.7618314736437091,
+      "grad_norm": 1.8425785303115845,
+      "learning_rate": 2.8304981548927022e-05,
+      "loss": 0.888,
+      "step": 3960
+    },
+    {
+      "epoch": 0.7620238553289727,
+      "grad_norm": 1.7421965599060059,
+      "learning_rate": 2.826155871962227e-05,
+      "loss": 0.9975,
+      "step": 3961
+    },
+    {
+      "epoch": 0.7622162370142362,
+      "grad_norm": 2.444035053253174,
+      "learning_rate": 2.8218163741650416e-05,
+      "loss": 1.003,
+      "step": 3962
+    },
+    {
+      "epoch": 0.7624086186994998,
+      "grad_norm": 2.3485662937164307,
+      "learning_rate": 2.8174796631858978e-05,
+      "loss": 0.9621,
+      "step": 3963
+    },
+    {
+      "epoch": 0.7626010003847634,
+      "grad_norm": 2.3852202892303467,
+      "learning_rate": 2.813145740708445e-05,
+      "loss": 0.8888,
+      "step": 3964
+    },
+    {
+      "epoch": 0.7627933820700269,
+      "grad_norm": 1.3336941003799438,
+      "learning_rate": 2.8088146084152712e-05,
+      "loss": 1.02,
+      "step": 3965
+    },
+    {
+      "epoch": 0.7629857637552905,
+      "grad_norm": 1.597904086112976,
+      "learning_rate": 2.8044862679878604e-05,
+      "loss": 1.0363,
+      "step": 3966
+    },
+    {
+      "epoch": 0.7631781454405541,
+      "grad_norm": 2.667262554168701,
+      "learning_rate": 2.800160721106633e-05,
+      "loss": 0.9033,
+      "step": 3967
+    },
+    {
+      "epoch": 0.7633705271258177,
+      "grad_norm": 3.3380446434020996,
+      "learning_rate": 2.7958379694509106e-05,
+      "loss": 0.9534,
+      "step": 3968
+    },
+    {
+      "epoch": 0.7635629088110811,
+      "grad_norm": 2.4452826976776123,
+      "learning_rate": 2.791518014698935e-05,
+      "loss": 0.9457,
+      "step": 3969
+    },
+    {
+      "epoch": 0.7637552904963447,
+      "grad_norm": 2.6828453540802,
+      "learning_rate": 2.787200858527862e-05,
+      "loss": 0.9167,
+      "step": 3970
+    },
+    {
+      "epoch": 0.7639476721816083,
+      "grad_norm": 1.4249169826507568,
+      "learning_rate": 2.7828865026137585e-05,
+      "loss": 0.7628,
+      "step": 3971
+    },
+    {
+      "epoch": 0.7641400538668719,
+      "grad_norm": 2.527337074279785,
+      "learning_rate": 2.778574948631608e-05,
+      "loss": 0.9423,
+      "step": 3972
+    },
+    {
+      "epoch": 0.7643324355521355,
+      "grad_norm": 1.9848651885986328,
+      "learning_rate": 2.774266198255303e-05,
+      "loss": 0.9253,
+      "step": 3973
+    },
+    {
+      "epoch": 0.764524817237399,
+      "grad_norm": 2.316781520843506,
+      "learning_rate": 2.7699602531576496e-05,
+      "loss": 0.905,
+      "step": 3974
+    },
+    {
+      "epoch": 0.7647171989226625,
+      "grad_norm": 1.6554776430130005,
+      "learning_rate": 2.765657115010364e-05,
+      "loss": 0.933,
+      "step": 3975
+    },
+    {
+      "epoch": 0.7649095806079261,
+      "grad_norm": 2.183151960372925,
+      "learning_rate": 2.7613567854840683e-05,
+      "loss": 0.9728,
+      "step": 3976
+    },
+    {
+      "epoch": 0.7651019622931897,
+      "grad_norm": 2.1691460609436035,
+      "learning_rate": 2.7570592662483087e-05,
+      "loss": 0.9791,
+      "step": 3977
+    },
+    {
+      "epoch": 0.7652943439784533,
+      "grad_norm": 1.9097628593444824,
+      "learning_rate": 2.7527645589715167e-05,
+      "loss": 0.9948,
+      "step": 3978
+    },
+    {
+      "epoch": 0.7654867256637168,
+      "grad_norm": 2.947443962097168,
+      "learning_rate": 2.7484726653210558e-05,
+      "loss": 1.0444,
+      "step": 3979
+    },
+    {
+      "epoch": 0.7656791073489804,
+      "grad_norm": 2.2775774002075195,
+      "learning_rate": 2.744183586963185e-05,
+      "loss": 0.9175,
+      "step": 3980
+    },
+    {
+      "epoch": 0.765871489034244,
+      "grad_norm": 2.4279911518096924,
+      "learning_rate": 2.739897325563069e-05,
+      "loss": 0.9158,
+      "step": 3981
+    },
+    {
+      "epoch": 0.7660638707195075,
+      "grad_norm": 2.076107978820801,
+      "learning_rate": 2.7356138827847855e-05,
+      "loss": 0.9515,
+      "step": 3982
+    },
+    {
+      "epoch": 0.766256252404771,
+      "grad_norm": 2.146001100540161,
+      "learning_rate": 2.731333260291311e-05,
+      "loss": 0.9914,
+      "step": 3983
+    },
+    {
+      "epoch": 0.7664486340900346,
+      "grad_norm": 2.775479555130005,
+      "learning_rate": 2.7270554597445342e-05,
+      "loss": 0.9717,
+      "step": 3984
+    },
+    {
+      "epoch": 0.7666410157752982,
+      "grad_norm": 2.400527238845825,
+      "learning_rate": 2.7227804828052384e-05,
+      "loss": 0.8587,
+      "step": 3985
+    },
+    {
+      "epoch": 0.7668333974605618,
+      "grad_norm": 2.3386173248291016,
+      "learning_rate": 2.718508331133128e-05,
+      "loss": 0.9728,
+      "step": 3986
+    },
+    {
+      "epoch": 0.7670257791458254,
+      "grad_norm": 2.8929688930511475,
+      "learning_rate": 2.7142390063867896e-05,
+      "loss": 0.9907,
+      "step": 3987
+    },
+    {
+      "epoch": 0.7672181608310888,
+      "grad_norm": 1.4141018390655518,
+      "learning_rate": 2.709972510223725e-05,
+      "loss": 0.805,
+      "step": 3988
+    },
+    {
+      "epoch": 0.7674105425163524,
+      "grad_norm": 2.769606828689575,
+      "learning_rate": 2.7057088443003343e-05,
+      "loss": 0.9734,
+      "step": 3989
+    },
+    {
+      "epoch": 0.767602924201616,
+      "grad_norm": 1.9840471744537354,
+      "learning_rate": 2.7014480102719176e-05,
+      "loss": 0.9618,
+      "step": 3990
+    },
+    {
+      "epoch": 0.7677953058868796,
+      "grad_norm": 2.6989047527313232,
+      "learning_rate": 2.697190009792685e-05,
+      "loss": 0.9949,
+      "step": 3991
+    },
+    {
+      "epoch": 0.7679876875721431,
+      "grad_norm": 2.056713581085205,
+      "learning_rate": 2.692934844515729e-05,
+      "loss": 0.9496,
+      "step": 3992
+    },
+    {
+      "epoch": 0.7681800692574067,
+      "grad_norm": 0.8964336514472961,
+      "learning_rate": 2.6886825160930585e-05,
+      "loss": 0.9694,
+      "step": 3993
+    },
+    {
+      "epoch": 0.7683724509426703,
+      "grad_norm": 2.4035158157348633,
+      "learning_rate": 2.6844330261755714e-05,
+      "loss": 0.9781,
+      "step": 3994
+    },
+    {
+      "epoch": 0.7685648326279338,
+      "grad_norm": 2.715745210647583,
+      "learning_rate": 2.6801863764130652e-05,
+      "loss": 0.9243,
+      "step": 3995
+    },
+    {
+      "epoch": 0.7687572143131974,
+      "grad_norm": 2.227614402770996,
+      "learning_rate": 2.675942568454236e-05,
+      "loss": 0.9479,
+      "step": 3996
+    },
+    {
+      "epoch": 0.7689495959984609,
+      "grad_norm": 1.3948478698730469,
+      "learning_rate": 2.671701603946678e-05,
+      "loss": 0.747,
+      "step": 3997
+    },
+    {
+      "epoch": 0.7691419776837245,
+      "grad_norm": 2.5429768562316895,
+      "learning_rate": 2.667463484536876e-05,
+      "loss": 0.9268,
+      "step": 3998
+    },
+    {
+      "epoch": 0.7693343593689881,
+      "grad_norm": 2.2863423824310303,
+      "learning_rate": 2.6632282118702147e-05,
+      "loss": 0.9957,
+      "step": 3999
+    },
+    {
+      "epoch": 0.7695267410542517,
+      "grad_norm": 2.3798229694366455,
+      "learning_rate": 2.6589957875909798e-05,
+      "loss": 0.9702,
+      "step": 4000
+    },
+    {
+      "epoch": 0.7697191227395151,
+      "grad_norm": 2.1510541439056396,
+      "learning_rate": 2.6547662133423345e-05,
+      "loss": 0.9943,
+      "step": 4001
+    },
+    {
+      "epoch": 0.7699115044247787,
+      "grad_norm": 1.5700985193252563,
+      "learning_rate": 2.650539490766346e-05,
+      "loss": 0.9351,
+      "step": 4002
+    },
+    {
+      "epoch": 0.7701038861100423,
+      "grad_norm": 2.5398435592651367,
+      "learning_rate": 2.646315621503983e-05,
+      "loss": 0.9539,
+      "step": 4003
+    },
+    {
+      "epoch": 0.7702962677953059,
+      "grad_norm": 2.65816593170166,
+      "learning_rate": 2.6420946071950847e-05,
+      "loss": 0.9662,
+      "step": 4004
+    },
+    {
+      "epoch": 0.7704886494805695,
+      "grad_norm": 2.455376148223877,
+      "learning_rate": 2.6378764494784024e-05,
+      "loss": 0.966,
+      "step": 4005
+    },
+    {
+      "epoch": 0.770681031165833,
+      "grad_norm": 2.653364419937134,
+      "learning_rate": 2.6336611499915686e-05,
+      "loss": 0.9656,
+      "step": 4006
+    },
+    {
+      "epoch": 0.7708734128510966,
+      "grad_norm": 1.5277166366577148,
+      "learning_rate": 2.6294487103711065e-05,
+      "loss": 0.9709,
+      "step": 4007
+    },
+    {
+      "epoch": 0.7710657945363601,
+      "grad_norm": 2.6942484378814697,
+      "learning_rate": 2.6252391322524293e-05,
+      "loss": 0.9588,
+      "step": 4008
+    },
+    {
+      "epoch": 0.7712581762216237,
+      "grad_norm": 2.374521255493164,
+      "learning_rate": 2.621032417269843e-05,
+      "loss": 1.0117,
+      "step": 4009
+    },
+    {
+      "epoch": 0.7714505579068872,
+      "grad_norm": 2.5562245845794678,
+      "learning_rate": 2.6168285670565373e-05,
+      "loss": 1.0063,
+      "step": 4010
+    },
+    {
+      "epoch": 0.7716429395921508,
+      "grad_norm": 1.4743361473083496,
+      "learning_rate": 2.612627583244589e-05,
+      "loss": 0.8195,
+      "step": 4011
+    },
+    {
+      "epoch": 0.7718353212774144,
+      "grad_norm": 2.4765584468841553,
+      "learning_rate": 2.6084294674649735e-05,
+      "loss": 1.0268,
+      "step": 4012
+    },
+    {
+      "epoch": 0.772027702962678,
+      "grad_norm": 3.172699213027954,
+      "learning_rate": 2.6042342213475347e-05,
+      "loss": 0.9447,
+      "step": 4013
+    },
+    {
+      "epoch": 0.7722200846479416,
+      "grad_norm": 1.8838920593261719,
+      "learning_rate": 2.600041846521014e-05,
+      "loss": 0.9383,
+      "step": 4014
+    },
+    {
+      "epoch": 0.772412466333205,
+      "grad_norm": 1.338615894317627,
+      "learning_rate": 2.5958523446130378e-05,
+      "loss": 0.9258,
+      "step": 4015
+    },
+    {
+      "epoch": 0.7726048480184686,
+      "grad_norm": 2.0330257415771484,
+      "learning_rate": 2.5916657172501103e-05,
+      "loss": 0.901,
+      "step": 4016
+    },
+    {
+      "epoch": 0.7727972297037322,
+      "grad_norm": 2.285463571548462,
+      "learning_rate": 2.5874819660576334e-05,
+      "loss": 0.8217,
+      "step": 4017
+    },
+    {
+      "epoch": 0.7729896113889958,
+      "grad_norm": 3.088120937347412,
+      "learning_rate": 2.5833010926598722e-05,
+      "loss": 0.9759,
+      "step": 4018
+    },
+    {
+      "epoch": 0.7731819930742593,
+      "grad_norm": 2.8093323707580566,
+      "learning_rate": 2.579123098679994e-05,
+      "loss": 0.9032,
+      "step": 4019
+    },
+    {
+      "epoch": 0.7733743747595229,
+      "grad_norm": 2.1862008571624756,
+      "learning_rate": 2.5749479857400384e-05,
+      "loss": 0.9609,
+      "step": 4020
+    },
+    {
+      "epoch": 0.7735667564447865,
+      "grad_norm": 2.5371551513671875,
+      "learning_rate": 2.5707757554609246e-05,
+      "loss": 0.9081,
+      "step": 4021
+    },
+    {
+      "epoch": 0.77375913813005,
+      "grad_norm": 1.1032763719558716,
+      "learning_rate": 2.5666064094624598e-05,
+      "loss": 0.9871,
+      "step": 4022
+    },
+    {
+      "epoch": 0.7739515198153136,
+      "grad_norm": 1.2981103658676147,
+      "learning_rate": 2.5624399493633257e-05,
+      "loss": 0.7595,
+      "step": 4023
+    },
+    {
+      "epoch": 0.7741439015005771,
+      "grad_norm": 2.3944530487060547,
+      "learning_rate": 2.5582763767810857e-05,
+      "loss": 0.9252,
+      "step": 4024
+    },
+    {
+      "epoch": 0.7743362831858407,
+      "grad_norm": 2.2111759185791016,
+      "learning_rate": 2.5541156933321798e-05,
+      "loss": 0.9874,
+      "step": 4025
+    },
+    {
+      "epoch": 0.7745286648711043,
+      "grad_norm": 2.217884063720703,
+      "learning_rate": 2.5499579006319362e-05,
+      "loss": 0.935,
+      "step": 4026
+    },
+    {
+      "epoch": 0.7747210465563679,
+      "grad_norm": 2.882704019546509,
+      "learning_rate": 2.5458030002945453e-05,
+      "loss": 0.9294,
+      "step": 4027
+    },
+    {
+      "epoch": 0.7749134282416313,
+      "grad_norm": 2.4253594875335693,
+      "learning_rate": 2.5416509939330835e-05,
+      "loss": 0.9467,
+      "step": 4028
+    },
+    {
+      "epoch": 0.7751058099268949,
+      "grad_norm": 3.028942584991455,
+      "learning_rate": 2.5375018831595087e-05,
+      "loss": 0.9719,
+      "step": 4029
+    },
+    {
+      "epoch": 0.7752981916121585,
+      "grad_norm": 2.688896894454956,
+      "learning_rate": 2.5333556695846383e-05,
+      "loss": 0.9489,
+      "step": 4030
+    },
+    {
+      "epoch": 0.7754905732974221,
+      "grad_norm": 1.0037918090820312,
+      "learning_rate": 2.5292123548181845e-05,
+      "loss": 0.9697,
+      "step": 4031
+    },
+    {
+      "epoch": 0.7756829549826857,
+      "grad_norm": 3.074557065963745,
+      "learning_rate": 2.5250719404687216e-05,
+      "loss": 0.9874,
+      "step": 4032
+    },
+    {
+      "epoch": 0.7758753366679492,
+      "grad_norm": 2.1378819942474365,
+      "learning_rate": 2.520934428143701e-05,
+      "loss": 0.9746,
+      "step": 4033
+    },
+    {
+      "epoch": 0.7760677183532128,
+      "grad_norm": 2.4830586910247803,
+      "learning_rate": 2.516799819449447e-05,
+      "loss": 0.8952,
+      "step": 4034
+    },
+    {
+      "epoch": 0.7762601000384763,
+      "grad_norm": 2.335955858230591,
+      "learning_rate": 2.512668115991156e-05,
+      "loss": 0.9119,
+      "step": 4035
+    },
+    {
+      "epoch": 0.7764524817237399,
+      "grad_norm": 2.558234453201294,
+      "learning_rate": 2.5085393193729e-05,
+      "loss": 0.9796,
+      "step": 4036
+    },
+    {
+      "epoch": 0.7766448634090035,
+      "grad_norm": 2.1571204662323,
+      "learning_rate": 2.5044134311976152e-05,
+      "loss": 0.931,
+      "step": 4037
+    },
+    {
+      "epoch": 0.776837245094267,
+      "grad_norm": 2.5708346366882324,
+      "learning_rate": 2.5002904530671236e-05,
+      "loss": 0.8615,
+      "step": 4038
+    },
+    {
+      "epoch": 0.7770296267795306,
+      "grad_norm": 2.325291156768799,
+      "learning_rate": 2.4961703865820975e-05,
+      "loss": 0.9757,
+      "step": 4039
+    },
+    {
+      "epoch": 0.7772220084647942,
+      "grad_norm": 2.135955333709717,
+      "learning_rate": 2.492053233342091e-05,
+      "loss": 0.9522,
+      "step": 4040
+    },
+    {
+      "epoch": 0.7774143901500578,
+      "grad_norm": 1.9000520706176758,
+      "learning_rate": 2.4879389949455268e-05,
+      "loss": 0.8599,
+      "step": 4041
+    },
+    {
+      "epoch": 0.7776067718353212,
+      "grad_norm": 2.8894762992858887,
+      "learning_rate": 2.483827672989688e-05,
+      "loss": 0.8964,
+      "step": 4042
+    },
+    {
+      "epoch": 0.7777991535205848,
+      "grad_norm": 2.448740243911743,
+      "learning_rate": 2.4797192690707428e-05,
+      "loss": 0.9368,
+      "step": 4043
+    },
+    {
+      "epoch": 0.7779915352058484,
+      "grad_norm": 2.5536506175994873,
+      "learning_rate": 2.4756137847837025e-05,
+      "loss": 0.9472,
+      "step": 4044
+    },
+    {
+      "epoch": 0.778183916891112,
+      "grad_norm": 2.2995455265045166,
+      "learning_rate": 2.4715112217224655e-05,
+      "loss": 0.9358,
+      "step": 4045
+    },
+    {
+      "epoch": 0.7783762985763756,
+      "grad_norm": 2.2877469062805176,
+      "learning_rate": 2.467411581479786e-05,
+      "loss": 0.9426,
+      "step": 4046
+    },
+    {
+      "epoch": 0.7785686802616391,
+      "grad_norm": 2.0260703563690186,
+      "learning_rate": 2.463314865647286e-05,
+      "loss": 0.8386,
+      "step": 4047
+    },
+    {
+      "epoch": 0.7787610619469026,
+      "grad_norm": 2.713252305984497,
+      "learning_rate": 2.45922107581545e-05,
+      "loss": 0.8939,
+      "step": 4048
+    },
+    {
+      "epoch": 0.7789534436321662,
+      "grad_norm": 2.1695399284362793,
+      "learning_rate": 2.4551302135736286e-05,
+      "loss": 1.029,
+      "step": 4049
+    },
+    {
+      "epoch": 0.7791458253174298,
+      "grad_norm": 1.9467989206314087,
+      "learning_rate": 2.4510422805100363e-05,
+      "loss": 0.9362,
+      "step": 4050
+    },
+    {
+      "epoch": 0.7793382070026933,
+      "grad_norm": 2.2051024436950684,
+      "learning_rate": 2.4469572782117457e-05,
+      "loss": 0.8549,
+      "step": 4051
+    },
+    {
+      "epoch": 0.7795305886879569,
+      "grad_norm": 3.0180978775024414,
+      "learning_rate": 2.4428752082647045e-05,
+      "loss": 0.9465,
+      "step": 4052
+    },
+    {
+      "epoch": 0.7797229703732205,
+      "grad_norm": 2.229774236679077,
+      "learning_rate": 2.438796072253704e-05,
+      "loss": 0.9366,
+      "step": 4053
+    },
+    {
+      "epoch": 0.7799153520584841,
+      "grad_norm": 2.1465611457824707,
+      "learning_rate": 2.4347198717624054e-05,
+      "loss": 0.8726,
+      "step": 4054
+    },
+    {
+      "epoch": 0.7801077337437476,
+      "grad_norm": 2.771143913269043,
+      "learning_rate": 2.4306466083733393e-05,
+      "loss": 0.9705,
+      "step": 4055
+    },
+    {
+      "epoch": 0.7803001154290111,
+      "grad_norm": 2.804051637649536,
+      "learning_rate": 2.426576283667873e-05,
+      "loss": 1.0168,
+      "step": 4056
+    },
+    {
+      "epoch": 0.7804924971142747,
+      "grad_norm": 1.319234848022461,
+      "learning_rate": 2.422508899226258e-05,
+      "loss": 0.7832,
+      "step": 4057
+    },
+    {
+      "epoch": 0.7806848787995383,
+      "grad_norm": 2.418410301208496,
+      "learning_rate": 2.4184444566275886e-05,
+      "loss": 0.8958,
+      "step": 4058
+    },
+    {
+      "epoch": 0.7808772604848019,
+      "grad_norm": 1.529638409614563,
+      "learning_rate": 2.4143829574498223e-05,
+      "loss": 0.7766,
+      "step": 4059
+    },
+    {
+      "epoch": 0.7810696421700654,
+      "grad_norm": 2.3470025062561035,
+      "learning_rate": 2.4103244032697715e-05,
+      "loss": 0.9732,
+      "step": 4060
+    },
+    {
+      "epoch": 0.781262023855329,
+      "grad_norm": 1.4041332006454468,
+      "learning_rate": 2.406268795663108e-05,
+      "loss": 0.9608,
+      "step": 4061
+    },
+    {
+      "epoch": 0.7814544055405925,
+      "grad_norm": 2.164180040359497,
+      "learning_rate": 2.402216136204357e-05,
+      "loss": 0.9544,
+      "step": 4062
+    },
+    {
+      "epoch": 0.7816467872258561,
+      "grad_norm": 1.1260701417922974,
+      "learning_rate": 2.3981664264669023e-05,
+      "loss": 0.983,
+      "step": 4063
+    },
+    {
+      "epoch": 0.7818391689111197,
+      "grad_norm": 2.4947891235351562,
+      "learning_rate": 2.3941196680229794e-05,
+      "loss": 0.9732,
+      "step": 4064
+    },
+    {
+      "epoch": 0.7820315505963832,
+      "grad_norm": 2.030629873275757,
+      "learning_rate": 2.390075862443677e-05,
+      "loss": 0.9458,
+      "step": 4065
+    },
+    {
+      "epoch": 0.7822239322816468,
+      "grad_norm": 2.569977045059204,
+      "learning_rate": 2.386035011298947e-05,
+      "loss": 1.0019,
+      "step": 4066
+    },
+    {
+      "epoch": 0.7824163139669104,
+      "grad_norm": 1.2547601461410522,
+      "learning_rate": 2.3819971161575803e-05,
+      "loss": 0.8816,
+      "step": 4067
+    },
+    {
+      "epoch": 0.782608695652174,
+      "grad_norm": 1.2456977367401123,
+      "learning_rate": 2.377962178587225e-05,
+      "loss": 0.9527,
+      "step": 4068
+    },
+    {
+      "epoch": 0.7828010773374374,
+      "grad_norm": 2.291666269302368,
+      "learning_rate": 2.3739302001543918e-05,
+      "loss": 0.8813,
+      "step": 4069
+    },
+    {
+      "epoch": 0.782993459022701,
+      "grad_norm": 2.0913002490997314,
+      "learning_rate": 2.3699011824244232e-05,
+      "loss": 0.9544,
+      "step": 4070
+    },
+    {
+      "epoch": 0.7831858407079646,
+      "grad_norm": 2.4771316051483154,
+      "learning_rate": 2.365875126961531e-05,
+      "loss": 0.9037,
+      "step": 4071
+    },
+    {
+      "epoch": 0.7833782223932282,
+      "grad_norm": 1.7776479721069336,
+      "learning_rate": 2.3618520353287643e-05,
+      "loss": 0.9496,
+      "step": 4072
+    },
+    {
+      "epoch": 0.7835706040784918,
+      "grad_norm": 2.5017120838165283,
+      "learning_rate": 2.357831909088026e-05,
+      "loss": 0.9603,
+      "step": 4073
+    },
+    {
+      "epoch": 0.7837629857637552,
+      "grad_norm": 2.449716091156006,
+      "learning_rate": 2.3538147498000695e-05,
+      "loss": 0.9608,
+      "step": 4074
+    },
+    {
+      "epoch": 0.7839553674490188,
+      "grad_norm": 2.52034330368042,
+      "learning_rate": 2.349800559024492e-05,
+      "loss": 0.876,
+      "step": 4075
+    },
+    {
+      "epoch": 0.7841477491342824,
+      "grad_norm": 1.2002882957458496,
+      "learning_rate": 2.3457893383197416e-05,
+      "loss": 0.9463,
+      "step": 4076
+    },
+    {
+      "epoch": 0.784340130819546,
+      "grad_norm": 2.5053467750549316,
+      "learning_rate": 2.3417810892431103e-05,
+      "loss": 0.8402,
+      "step": 4077
+    },
+    {
+      "epoch": 0.7845325125048095,
+      "grad_norm": 1.7265993356704712,
+      "learning_rate": 2.3377758133507456e-05,
+      "loss": 0.9139,
+      "step": 4078
+    },
+    {
+      "epoch": 0.7847248941900731,
+      "grad_norm": 2.238858461380005,
+      "learning_rate": 2.3337735121976245e-05,
+      "loss": 1.0005,
+      "step": 4079
+    },
+    {
+      "epoch": 0.7849172758753367,
+      "grad_norm": 2.6887314319610596,
+      "learning_rate": 2.3297741873375802e-05,
+      "loss": 0.9969,
+      "step": 4080
+    },
+    {
+      "epoch": 0.7851096575606002,
+      "grad_norm": 2.6177725791931152,
+      "learning_rate": 2.3257778403232954e-05,
+      "loss": 0.895,
+      "step": 4081
+    },
+    {
+      "epoch": 0.7853020392458638,
+      "grad_norm": 2.499579668045044,
+      "learning_rate": 2.321784472706279e-05,
+      "loss": 0.9439,
+      "step": 4082
+    },
+    {
+      "epoch": 0.7854944209311273,
+      "grad_norm": 2.1406164169311523,
+      "learning_rate": 2.3177940860369007e-05,
+      "loss": 0.9335,
+      "step": 4083
+    },
+    {
+      "epoch": 0.7856868026163909,
+      "grad_norm": 2.2176897525787354,
+      "learning_rate": 2.3138066818643644e-05,
+      "loss": 0.9595,
+      "step": 4084
+    },
+    {
+      "epoch": 0.7858791843016545,
+      "grad_norm": 2.484381914138794,
+      "learning_rate": 2.3098222617367182e-05,
+      "loss": 0.8518,
+      "step": 4085
+    },
+    {
+      "epoch": 0.7860715659869181,
+      "grad_norm": 1.7014670372009277,
+      "learning_rate": 2.30584082720085e-05,
+      "loss": 0.9292,
+      "step": 4086
+    },
+    {
+      "epoch": 0.7862639476721817,
+      "grad_norm": 2.7603659629821777,
+      "learning_rate": 2.3018623798024918e-05,
+      "loss": 0.9866,
+      "step": 4087
+    },
+    {
+      "epoch": 0.7864563293574451,
+      "grad_norm": 2.6706831455230713,
+      "learning_rate": 2.297886921086211e-05,
+      "loss": 0.9153,
+      "step": 4088
+    },
+    {
+      "epoch": 0.7866487110427087,
+      "grad_norm": 2.3881795406341553,
+      "learning_rate": 2.293914452595419e-05,
+      "loss": 0.9339,
+      "step": 4089
+    },
+    {
+      "epoch": 0.7868410927279723,
+      "grad_norm": 2.4757652282714844,
+      "learning_rate": 2.2899449758723657e-05,
+      "loss": 0.9334,
+      "step": 4090
+    },
+    {
+      "epoch": 0.7870334744132359,
+      "grad_norm": 2.706226348876953,
+      "learning_rate": 2.285978492458134e-05,
+      "loss": 0.8475,
+      "step": 4091
+    },
+    {
+      "epoch": 0.7872258560984994,
+      "grad_norm": 2.406001329421997,
+      "learning_rate": 2.282015003892659e-05,
+      "loss": 0.9511,
+      "step": 4092
+    },
+    {
+      "epoch": 0.787418237783763,
+      "grad_norm": 2.1562609672546387,
+      "learning_rate": 2.2780545117146947e-05,
+      "loss": 0.9164,
+      "step": 4093
+    },
+    {
+      "epoch": 0.7876106194690266,
+      "grad_norm": 2.0879106521606445,
+      "learning_rate": 2.2740970174618403e-05,
+      "loss": 0.9232,
+      "step": 4094
+    },
+    {
+      "epoch": 0.7878030011542901,
+      "grad_norm": 2.920936346054077,
+      "learning_rate": 2.2701425226705408e-05,
+      "loss": 0.9062,
+      "step": 4095
+    },
+    {
+      "epoch": 0.7879953828395537,
+      "grad_norm": 3.0133068561553955,
+      "learning_rate": 2.2661910288760547e-05,
+      "loss": 0.9154,
+      "step": 4096
+    },
+    {
+      "epoch": 0.7881877645248172,
+      "grad_norm": 2.549840211868286,
+      "learning_rate": 2.2622425376124967e-05,
+      "loss": 0.9757,
+      "step": 4097
+    },
+    {
+      "epoch": 0.7883801462100808,
+      "grad_norm": 2.100926399230957,
+      "learning_rate": 2.2582970504128042e-05,
+      "loss": 0.8567,
+      "step": 4098
+    },
+    {
+      "epoch": 0.7885725278953444,
+      "grad_norm": 2.4787261486053467,
+      "learning_rate": 2.2543545688087518e-05,
+      "loss": 0.9612,
+      "step": 4099
+    },
+    {
+      "epoch": 0.788764909580608,
+      "grad_norm": 2.0038506984710693,
+      "learning_rate": 2.2504150943309453e-05,
+      "loss": 0.8618,
+      "step": 4100
+    },
+    {
+      "epoch": 0.7889572912658714,
+      "grad_norm": 2.2204761505126953,
+      "learning_rate": 2.2464786285088268e-05,
+      "loss": 0.8958,
+      "step": 4101
+    },
+    {
+      "epoch": 0.789149672951135,
+      "grad_norm": 2.473811626434326,
+      "learning_rate": 2.242545172870665e-05,
+      "loss": 0.9734,
+      "step": 4102
+    },
+    {
+      "epoch": 0.7893420546363986,
+      "grad_norm": 1.3858364820480347,
+      "learning_rate": 2.238614728943561e-05,
+      "loss": 0.7769,
+      "step": 4103
+    },
+    {
+      "epoch": 0.7895344363216622,
+      "grad_norm": 3.3545031547546387,
+      "learning_rate": 2.2346872982534584e-05,
+      "loss": 0.894,
+      "step": 4104
+    },
+    {
+      "epoch": 0.7897268180069258,
+      "grad_norm": 2.38153338432312,
+      "learning_rate": 2.230762882325108e-05,
+      "loss": 0.9436,
+      "step": 4105
+    },
+    {
+      "epoch": 0.7899191996921893,
+      "grad_norm": 1.4914954900741577,
+      "learning_rate": 2.2268414826821115e-05,
+      "loss": 0.7536,
+      "step": 4106
+    },
+    {
+      "epoch": 0.7901115813774529,
+      "grad_norm": 2.6259353160858154,
+      "learning_rate": 2.2229231008468932e-05,
+      "loss": 0.9805,
+      "step": 4107
+    },
+    {
+      "epoch": 0.7903039630627164,
+      "grad_norm": 2.397392988204956,
+      "learning_rate": 2.2190077383406936e-05,
+      "loss": 0.9356,
+      "step": 4108
+    },
+    {
+      "epoch": 0.79049634474798,
+      "grad_norm": 2.6146912574768066,
+      "learning_rate": 2.2150953966835996e-05,
+      "loss": 0.9266,
+      "step": 4109
+    },
+    {
+      "epoch": 0.7906887264332435,
+      "grad_norm": 2.4009480476379395,
+      "learning_rate": 2.211186077394516e-05,
+      "loss": 0.9057,
+      "step": 4110
+    },
+    {
+      "epoch": 0.7908811081185071,
+      "grad_norm": 2.21252179145813,
+      "learning_rate": 2.2072797819911726e-05,
+      "loss": 0.8811,
+      "step": 4111
+    },
+    {
+      "epoch": 0.7910734898037707,
+      "grad_norm": 2.410244941711426,
+      "learning_rate": 2.2033765119901295e-05,
+      "loss": 0.9499,
+      "step": 4112
+    },
+    {
+      "epoch": 0.7912658714890343,
+      "grad_norm": 2.647566080093384,
+      "learning_rate": 2.1994762689067704e-05,
+      "loss": 0.95,
+      "step": 4113
+    },
+    {
+      "epoch": 0.7914582531742979,
+      "grad_norm": 2.070547580718994,
+      "learning_rate": 2.1955790542553033e-05,
+      "loss": 0.9236,
+      "step": 4114
+    },
+    {
+      "epoch": 0.7916506348595613,
+      "grad_norm": 2.7970974445343018,
+      "learning_rate": 2.1916848695487614e-05,
+      "loss": 0.9532,
+      "step": 4115
+    },
+    {
+      "epoch": 0.7918430165448249,
+      "grad_norm": 2.382617950439453,
+      "learning_rate": 2.1877937162990014e-05,
+      "loss": 0.9037,
+      "step": 4116
+    },
+    {
+      "epoch": 0.7920353982300885,
+      "grad_norm": 2.9630191326141357,
+      "learning_rate": 2.1839055960166997e-05,
+      "loss": 0.8971,
+      "step": 4117
+    },
+    {
+      "epoch": 0.7922277799153521,
+      "grad_norm": 2.1682932376861572,
+      "learning_rate": 2.180020510211367e-05,
+      "loss": 0.8649,
+      "step": 4118
+    },
+    {
+      "epoch": 0.7924201616006156,
+      "grad_norm": 2.3968353271484375,
+      "learning_rate": 2.1761384603913205e-05,
+      "loss": 0.9939,
+      "step": 4119
+    },
+    {
+      "epoch": 0.7926125432858792,
+      "grad_norm": 2.631134271621704,
+      "learning_rate": 2.1722594480637036e-05,
+      "loss": 1.0175,
+      "step": 4120
+    },
+    {
+      "epoch": 0.7928049249711427,
+      "grad_norm": 1.9364196062088013,
+      "learning_rate": 2.168383474734491e-05,
+      "loss": 1.0213,
+      "step": 4121
+    },
+    {
+      "epoch": 0.7929973066564063,
+      "grad_norm": 2.5944271087646484,
+      "learning_rate": 2.1645105419084588e-05,
+      "loss": 0.8849,
+      "step": 4122
+    },
+    {
+      "epoch": 0.7931896883416699,
+      "grad_norm": 1.0300955772399902,
+      "learning_rate": 2.1606406510892208e-05,
+      "loss": 1.0106,
+      "step": 4123
+    },
+    {
+      "epoch": 0.7933820700269334,
+      "grad_norm": 2.154510974884033,
+      "learning_rate": 2.1567738037791995e-05,
+      "loss": 0.9615,
+      "step": 4124
+    },
+    {
+      "epoch": 0.793574451712197,
+      "grad_norm": 1.8587439060211182,
+      "learning_rate": 2.152910001479638e-05,
+      "loss": 0.9697,
+      "step": 4125
+    },
+    {
+      "epoch": 0.7937668333974606,
+      "grad_norm": 1.1648999452590942,
+      "learning_rate": 2.1490492456905965e-05,
+      "loss": 0.9306,
+      "step": 4126
+    },
+    {
+      "epoch": 0.7939592150827242,
+      "grad_norm": 1.3787832260131836,
+      "learning_rate": 2.1451915379109543e-05,
+      "loss": 0.9758,
+      "step": 4127
+    },
+    {
+      "epoch": 0.7941515967679876,
+      "grad_norm": 1.8640903234481812,
+      "learning_rate": 2.141336879638406e-05,
+      "loss": 0.9881,
+      "step": 4128
+    },
+    {
+      "epoch": 0.7943439784532512,
+      "grad_norm": 2.48803448677063,
+      "learning_rate": 2.1374852723694592e-05,
+      "loss": 0.9233,
+      "step": 4129
+    },
+    {
+      "epoch": 0.7945363601385148,
+      "grad_norm": 1.3624851703643799,
+      "learning_rate": 2.1336367175994508e-05,
+      "loss": 0.9908,
+      "step": 4130
+    },
+    {
+      "epoch": 0.7947287418237784,
+      "grad_norm": 2.3645989894866943,
+      "learning_rate": 2.1297912168225088e-05,
+      "loss": 0.8985,
+      "step": 4131
+    },
+    {
+      "epoch": 0.794921123509042,
+      "grad_norm": 2.040205478668213,
+      "learning_rate": 2.1259487715315997e-05,
+      "loss": 0.9258,
+      "step": 4132
+    },
+    {
+      "epoch": 0.7951135051943055,
+      "grad_norm": 1.304015874862671,
+      "learning_rate": 2.12210938321849e-05,
+      "loss": 0.6812,
+      "step": 4133
+    },
+    {
+      "epoch": 0.795305886879569,
+      "grad_norm": 2.156813621520996,
+      "learning_rate": 2.1182730533737573e-05,
+      "loss": 0.9304,
+      "step": 4134
+    },
+    {
+      "epoch": 0.7954982685648326,
+      "grad_norm": 1.2749477624893188,
+      "learning_rate": 2.1144397834868035e-05,
+      "loss": 0.7732,
+      "step": 4135
+    },
+    {
+      "epoch": 0.7956906502500962,
+      "grad_norm": 2.547126054763794,
+      "learning_rate": 2.110609575045833e-05,
+      "loss": 0.9094,
+      "step": 4136
+    },
+    {
+      "epoch": 0.7958830319353598,
+      "grad_norm": 1.7835642099380493,
+      "learning_rate": 2.1067824295378658e-05,
+      "loss": 0.9641,
+      "step": 4137
+    },
+    {
+      "epoch": 0.7960754136206233,
+      "grad_norm": 1.0006458759307861,
+      "learning_rate": 2.1029583484487314e-05,
+      "loss": 0.9553,
+      "step": 4138
+    },
+    {
+      "epoch": 0.7962677953058869,
+      "grad_norm": 2.6105000972747803,
+      "learning_rate": 2.0991373332630682e-05,
+      "loss": 0.9588,
+      "step": 4139
+    },
+    {
+      "epoch": 0.7964601769911505,
+      "grad_norm": 2.5964558124542236,
+      "learning_rate": 2.0953193854643273e-05,
+      "loss": 0.9286,
+      "step": 4140
+    },
+    {
+      "epoch": 0.796652558676414,
+      "grad_norm": 2.1522305011749268,
+      "learning_rate": 2.0915045065347672e-05,
+      "loss": 0.9144,
+      "step": 4141
+    },
+    {
+      "epoch": 0.7968449403616775,
+      "grad_norm": 2.2384660243988037,
+      "learning_rate": 2.0876926979554546e-05,
+      "loss": 0.9437,
+      "step": 4142
+    },
+    {
+      "epoch": 0.7970373220469411,
+      "grad_norm": 2.68320631980896,
+      "learning_rate": 2.0838839612062634e-05,
+      "loss": 0.9257,
+      "step": 4143
+    },
+    {
+      "epoch": 0.7972297037322047,
+      "grad_norm": 2.100607395172119,
+      "learning_rate": 2.0800782977658838e-05,
+      "loss": 0.9762,
+      "step": 4144
+    },
+    {
+      "epoch": 0.7974220854174683,
+      "grad_norm": 2.5114986896514893,
+      "learning_rate": 2.0762757091117936e-05,
+      "loss": 1.0009,
+      "step": 4145
+    },
+    {
+      "epoch": 0.7976144671027319,
+      "grad_norm": 0.9870612621307373,
+      "learning_rate": 2.0724761967202987e-05,
+      "loss": 0.9772,
+      "step": 4146
+    },
+    {
+      "epoch": 0.7978068487879953,
+      "grad_norm": 1.5597513914108276,
+      "learning_rate": 2.068679762066499e-05,
+      "loss": 0.9409,
+      "step": 4147
+    },
+    {
+      "epoch": 0.7979992304732589,
+      "grad_norm": 2.4630954265594482,
+      "learning_rate": 2.0648864066242933e-05,
+      "loss": 0.9563,
+      "step": 4148
+    },
+    {
+      "epoch": 0.7981916121585225,
+      "grad_norm": 1.1928784847259521,
+      "learning_rate": 2.0610961318664013e-05,
+      "loss": 0.9712,
+      "step": 4149
+    },
+    {
+      "epoch": 0.7983839938437861,
+      "grad_norm": 2.009357213973999,
+      "learning_rate": 2.0573089392643364e-05,
+      "loss": 0.9468,
+      "step": 4150
+    },
+    {
+      "epoch": 0.7985763755290496,
+      "grad_norm": 2.678090810775757,
+      "learning_rate": 2.0535248302884146e-05,
+      "loss": 0.9253,
+      "step": 4151
+    },
+    {
+      "epoch": 0.7987687572143132,
+      "grad_norm": 1.9405308961868286,
+      "learning_rate": 2.04974380640776e-05,
+      "loss": 0.9152,
+      "step": 4152
+    },
+    {
+      "epoch": 0.7989611388995768,
+      "grad_norm": 2.093592882156372,
+      "learning_rate": 2.045965869090295e-05,
+      "loss": 0.931,
+      "step": 4153
+    },
+    {
+      "epoch": 0.7991535205848403,
+      "grad_norm": 1.8132206201553345,
+      "learning_rate": 2.042191019802745e-05,
+      "loss": 0.9004,
+      "step": 4154
+    },
+    {
+      "epoch": 0.7993459022701039,
+      "grad_norm": 2.5165622234344482,
+      "learning_rate": 2.0384192600106335e-05,
+      "loss": 0.8656,
+      "step": 4155
+    },
+    {
+      "epoch": 0.7995382839553674,
+      "grad_norm": 1.0048422813415527,
+      "learning_rate": 2.0346505911782954e-05,
+      "loss": 0.9905,
+      "step": 4156
+    },
+    {
+      "epoch": 0.799730665640631,
+      "grad_norm": 2.5559239387512207,
+      "learning_rate": 2.0308850147688486e-05,
+      "loss": 0.8404,
+      "step": 4157
+    },
+    {
+      "epoch": 0.7999230473258946,
+      "grad_norm": 1.4039019346237183,
+      "learning_rate": 2.0271225322442256e-05,
+      "loss": 0.9634,
+      "step": 4158
+    },
+    {
+      "epoch": 0.8001154290111582,
+      "grad_norm": 1.8913276195526123,
+      "learning_rate": 2.0233631450651523e-05,
+      "loss": 0.9119,
+      "step": 4159
+    },
+    {
+      "epoch": 0.8003078106964217,
+      "grad_norm": 2.2880797386169434,
+      "learning_rate": 2.019606854691145e-05,
+      "loss": 0.9262,
+      "step": 4160
+    },
+    {
+      "epoch": 0.8005001923816852,
+      "grad_norm": 2.776689052581787,
+      "learning_rate": 2.0158536625805325e-05,
+      "loss": 0.9344,
+      "step": 4161
+    },
+    {
+      "epoch": 0.8006925740669488,
+      "grad_norm": 1.609510064125061,
+      "learning_rate": 2.01210357019043e-05,
+      "loss": 0.9403,
+      "step": 4162
+    },
+    {
+      "epoch": 0.8008849557522124,
+      "grad_norm": 2.4704506397247314,
+      "learning_rate": 2.0083565789767523e-05,
+      "loss": 0.9016,
+      "step": 4163
+    },
+    {
+      "epoch": 0.801077337437476,
+      "grad_norm": 1.8703029155731201,
+      "learning_rate": 2.004612690394212e-05,
+      "loss": 0.962,
+      "step": 4164
+    },
+    {
+      "epoch": 0.8012697191227395,
+      "grad_norm": 2.6986777782440186,
+      "learning_rate": 2.0008719058963145e-05,
+      "loss": 0.9343,
+      "step": 4165
+    },
+    {
+      "epoch": 0.8014621008080031,
+      "grad_norm": 1.9690980911254883,
+      "learning_rate": 1.997134226935361e-05,
+      "loss": 0.9442,
+      "step": 4166
+    },
+    {
+      "epoch": 0.8016544824932667,
+      "grad_norm": 1.7580958604812622,
+      "learning_rate": 1.993399654962447e-05,
+      "loss": 0.9496,
+      "step": 4167
+    },
+    {
+      "epoch": 0.8018468641785302,
+      "grad_norm": 2.999990224838257,
+      "learning_rate": 1.9896681914274617e-05,
+      "loss": 0.9202,
+      "step": 4168
+    },
+    {
+      "epoch": 0.8020392458637937,
+      "grad_norm": 2.0478904247283936,
+      "learning_rate": 1.9859398377790873e-05,
+      "loss": 0.9027,
+      "step": 4169
+    },
+    {
+      "epoch": 0.8022316275490573,
+      "grad_norm": 1.075636625289917,
+      "learning_rate": 1.982214595464804e-05,
+      "loss": 0.7889,
+      "step": 4170
+    },
+    {
+      "epoch": 0.8024240092343209,
+      "grad_norm": 0.700276792049408,
+      "learning_rate": 1.97849246593087e-05,
+      "loss": 0.9718,
+      "step": 4171
+    },
+    {
+      "epoch": 0.8026163909195845,
+      "grad_norm": 2.190044641494751,
+      "learning_rate": 1.9747734506223525e-05,
+      "loss": 0.9029,
+      "step": 4172
+    },
+    {
+      "epoch": 0.8028087726048481,
+      "grad_norm": 1.2782174348831177,
+      "learning_rate": 1.9710575509831007e-05,
+      "loss": 0.8232,
+      "step": 4173
+    },
+    {
+      "epoch": 0.8030011542901115,
+      "grad_norm": 2.8348186016082764,
+      "learning_rate": 1.9673447684557468e-05,
+      "loss": 1.0407,
+      "step": 4174
+    },
+    {
+      "epoch": 0.8031935359753751,
+      "grad_norm": 1.318998098373413,
+      "learning_rate": 1.9636351044817292e-05,
+      "loss": 0.888,
+      "step": 4175
+    },
+    {
+      "epoch": 0.8033859176606387,
+      "grad_norm": 2.5481300354003906,
+      "learning_rate": 1.9599285605012642e-05,
+      "loss": 0.9939,
+      "step": 4176
+    },
+    {
+      "epoch": 0.8035782993459023,
+      "grad_norm": 1.7297855615615845,
+      "learning_rate": 1.956225137953359e-05,
+      "loss": 0.9386,
+      "step": 4177
+    },
+    {
+      "epoch": 0.8037706810311658,
+      "grad_norm": 1.0881316661834717,
+      "learning_rate": 1.952524838275811e-05,
+      "loss": 0.9858,
+      "step": 4178
+    },
+    {
+      "epoch": 0.8039630627164294,
+      "grad_norm": 2.7757019996643066,
+      "learning_rate": 1.9488276629052026e-05,
+      "loss": 0.904,
+      "step": 4179
+    },
+    {
+      "epoch": 0.804155444401693,
+      "grad_norm": 1.4193774461746216,
+      "learning_rate": 1.945133613276907e-05,
+      "loss": 0.7449,
+      "step": 4180
+    },
+    {
+      "epoch": 0.8043478260869565,
+      "grad_norm": 1.6563955545425415,
+      "learning_rate": 1.941442690825076e-05,
+      "loss": 0.909,
+      "step": 4181
+    },
+    {
+      "epoch": 0.8045402077722201,
+      "grad_norm": 2.4005062580108643,
+      "learning_rate": 1.9377548969826633e-05,
+      "loss": 0.9181,
+      "step": 4182
+    },
+    {
+      "epoch": 0.8047325894574836,
+      "grad_norm": 2.108905076980591,
+      "learning_rate": 1.9340702331813842e-05,
+      "loss": 0.8605,
+      "step": 4183
+    },
+    {
+      "epoch": 0.8049249711427472,
+      "grad_norm": 2.5710134506225586,
+      "learning_rate": 1.930388700851762e-05,
+      "loss": 0.9115,
+      "step": 4184
+    },
+    {
+      "epoch": 0.8051173528280108,
+      "grad_norm": 2.240041732788086,
+      "learning_rate": 1.9267103014230935e-05,
+      "loss": 0.9164,
+      "step": 4185
+    },
+    {
+      "epoch": 0.8053097345132744,
+      "grad_norm": 2.262782573699951,
+      "learning_rate": 1.9230350363234516e-05,
+      "loss": 0.9016,
+      "step": 4186
+    },
+    {
+      "epoch": 0.805502116198538,
+      "grad_norm": 2.4640893936157227,
+      "learning_rate": 1.9193629069797102e-05,
+      "loss": 0.9497,
+      "step": 4187
+    },
+    {
+      "epoch": 0.8056944978838014,
+      "grad_norm": 2.203021287918091,
+      "learning_rate": 1.9156939148175125e-05,
+      "loss": 0.9959,
+      "step": 4188
+    },
+    {
+      "epoch": 0.805886879569065,
+      "grad_norm": 2.323892593383789,
+      "learning_rate": 1.912028061261287e-05,
+      "loss": 0.8904,
+      "step": 4189
+    },
+    {
+      "epoch": 0.8060792612543286,
+      "grad_norm": 2.252453327178955,
+      "learning_rate": 1.9083653477342466e-05,
+      "loss": 0.9527,
+      "step": 4190
+    },
+    {
+      "epoch": 0.8062716429395922,
+      "grad_norm": 2.61650013923645,
+      "learning_rate": 1.904705775658381e-05,
+      "loss": 0.9425,
+      "step": 4191
+    },
+    {
+      "epoch": 0.8064640246248557,
+      "grad_norm": 3.06697416305542,
+      "learning_rate": 1.901049346454462e-05,
+      "loss": 0.9466,
+      "step": 4192
+    },
+    {
+      "epoch": 0.8066564063101193,
+      "grad_norm": 2.068510055541992,
+      "learning_rate": 1.8973960615420416e-05,
+      "loss": 0.9532,
+      "step": 4193
+    },
+    {
+      "epoch": 0.8068487879953828,
+      "grad_norm": 1.6090667247772217,
+      "learning_rate": 1.8937459223394515e-05,
+      "loss": 0.957,
+      "step": 4194
+    },
+    {
+      "epoch": 0.8070411696806464,
+      "grad_norm": 1.87867271900177,
+      "learning_rate": 1.8900989302637985e-05,
+      "loss": 0.9538,
+      "step": 4195
+    },
+    {
+      "epoch": 0.80723355136591,
+      "grad_norm": 1.751249074935913,
+      "learning_rate": 1.8864550867309773e-05,
+      "loss": 0.8957,
+      "step": 4196
+    },
+    {
+      "epoch": 0.8074259330511735,
+      "grad_norm": 2.439894437789917,
+      "learning_rate": 1.8828143931556442e-05,
+      "loss": 0.959,
+      "step": 4197
+    },
+    {
+      "epoch": 0.8076183147364371,
+      "grad_norm": 2.2593541145324707,
+      "learning_rate": 1.8791768509512487e-05,
+      "loss": 0.9182,
+      "step": 4198
+    },
+    {
+      "epoch": 0.8078106964217007,
+      "grad_norm": 1.978951096534729,
+      "learning_rate": 1.875542461530011e-05,
+      "loss": 0.9279,
+      "step": 4199
+    },
+    {
+      "epoch": 0.8080030781069643,
+      "grad_norm": 2.0878772735595703,
+      "learning_rate": 1.871911226302917e-05,
+      "loss": 0.9497,
+      "step": 4200
+    },
+    {
+      "epoch": 0.8081954597922277,
+      "grad_norm": 2.766744375228882,
+      "learning_rate": 1.8682831466797467e-05,
+      "loss": 0.9192,
+      "step": 4201
+    },
+    {
+      "epoch": 0.8083878414774913,
+      "grad_norm": 2.559074640274048,
+      "learning_rate": 1.8646582240690412e-05,
+      "loss": 0.8901,
+      "step": 4202
+    },
+    {
+      "epoch": 0.8085802231627549,
+      "grad_norm": 2.359983444213867,
+      "learning_rate": 1.8610364598781226e-05,
+      "loss": 0.8394,
+      "step": 4203
+    },
+    {
+      "epoch": 0.8087726048480185,
+      "grad_norm": 2.0965492725372314,
+      "learning_rate": 1.8574178555130817e-05,
+      "loss": 1.0074,
+      "step": 4204
+    },
+    {
+      "epoch": 0.8089649865332821,
+      "grad_norm": 2.105022430419922,
+      "learning_rate": 1.8538024123787868e-05,
+      "loss": 1.0163,
+      "step": 4205
+    },
+    {
+      "epoch": 0.8091573682185456,
+      "grad_norm": 1.121229648590088,
+      "learning_rate": 1.8501901318788774e-05,
+      "loss": 1.0081,
+      "step": 4206
+    },
+    {
+      "epoch": 0.8093497499038091,
+      "grad_norm": 2.742797613143921,
+      "learning_rate": 1.8465810154157624e-05,
+      "loss": 0.9028,
+      "step": 4207
+    },
+    {
+      "epoch": 0.8095421315890727,
+      "grad_norm": 2.530022621154785,
+      "learning_rate": 1.842975064390633e-05,
+      "loss": 0.883,
+      "step": 4208
+    },
+    {
+      "epoch": 0.8097345132743363,
+      "grad_norm": 2.1644723415374756,
+      "learning_rate": 1.839372280203433e-05,
+      "loss": 0.9232,
+      "step": 4209
+    },
+    {
+      "epoch": 0.8099268949595998,
+      "grad_norm": 2.60211181640625,
+      "learning_rate": 1.835772664252895e-05,
+      "loss": 0.9089,
+      "step": 4210
+    },
+    {
+      "epoch": 0.8101192766448634,
+      "grad_norm": 1.2634164094924927,
+      "learning_rate": 1.832176217936511e-05,
+      "loss": 0.9102,
+      "step": 4211
+    },
+    {
+      "epoch": 0.810311658330127,
+      "grad_norm": 2.5788841247558594,
+      "learning_rate": 1.8285829426505453e-05,
+      "loss": 0.962,
+      "step": 4212
+    },
+    {
+      "epoch": 0.8105040400153906,
+      "grad_norm": 2.624084234237671,
+      "learning_rate": 1.824992839790035e-05,
+      "loss": 0.9856,
+      "step": 4213
+    },
+    {
+      "epoch": 0.8106964217006541,
+      "grad_norm": 2.4779651165008545,
+      "learning_rate": 1.8214059107487726e-05,
+      "loss": 0.9897,
+      "step": 4214
+    },
+    {
+      "epoch": 0.8108888033859176,
+      "grad_norm": 2.5709478855133057,
+      "learning_rate": 1.8178221569193343e-05,
+      "loss": 0.9367,
+      "step": 4215
+    },
+    {
+      "epoch": 0.8110811850711812,
+      "grad_norm": 2.6285085678100586,
+      "learning_rate": 1.8142415796930568e-05,
+      "loss": 0.9763,
+      "step": 4216
+    },
+    {
+      "epoch": 0.8112735667564448,
+      "grad_norm": 1.6224567890167236,
+      "learning_rate": 1.810664180460041e-05,
+      "loss": 0.8003,
+      "step": 4217
+    },
+    {
+      "epoch": 0.8114659484417084,
+      "grad_norm": 2.374595880508423,
+      "learning_rate": 1.8070899606091583e-05,
+      "loss": 0.9508,
+      "step": 4218
+    },
+    {
+      "epoch": 0.8116583301269719,
+      "grad_norm": 2.9575963020324707,
+      "learning_rate": 1.8035189215280423e-05,
+      "loss": 0.9243,
+      "step": 4219
+    },
+    {
+      "epoch": 0.8118507118122354,
+      "grad_norm": 2.6404449939727783,
+      "learning_rate": 1.7999510646030948e-05,
+      "loss": 0.909,
+      "step": 4220
+    },
+    {
+      "epoch": 0.812043093497499,
+      "grad_norm": 1.577078104019165,
+      "learning_rate": 1.7963863912194766e-05,
+      "loss": 1.0086,
+      "step": 4221
+    },
+    {
+      "epoch": 0.8122354751827626,
+      "grad_norm": 2.379591703414917,
+      "learning_rate": 1.7928249027611254e-05,
+      "loss": 0.9397,
+      "step": 4222
+    },
+    {
+      "epoch": 0.8124278568680262,
+      "grad_norm": 1.1735278367996216,
+      "learning_rate": 1.789266600610724e-05,
+      "loss": 0.8031,
+      "step": 4223
+    },
+    {
+      "epoch": 0.8126202385532897,
+      "grad_norm": 2.7916769981384277,
+      "learning_rate": 1.7857114861497336e-05,
+      "loss": 0.9084,
+      "step": 4224
+    },
+    {
+      "epoch": 0.8128126202385533,
+      "grad_norm": 1.3423105478286743,
+      "learning_rate": 1.782159560758373e-05,
+      "loss": 0.7782,
+      "step": 4225
+    },
+    {
+      "epoch": 0.8130050019238169,
+      "grad_norm": 3.1374948024749756,
+      "learning_rate": 1.7786108258156155e-05,
+      "loss": 1.0264,
+      "step": 4226
+    },
+    {
+      "epoch": 0.8131973836090804,
+      "grad_norm": 2.6141481399536133,
+      "learning_rate": 1.7750652826992075e-05,
+      "loss": 0.91,
+      "step": 4227
+    },
+    {
+      "epoch": 0.8133897652943439,
+      "grad_norm": 2.505141019821167,
+      "learning_rate": 1.7715229327856498e-05,
+      "loss": 0.952,
+      "step": 4228
+    },
+    {
+      "epoch": 0.8135821469796075,
+      "grad_norm": 2.1579246520996094,
+      "learning_rate": 1.767983777450205e-05,
+      "loss": 0.9057,
+      "step": 4229
+    },
+    {
+      "epoch": 0.8137745286648711,
+      "grad_norm": 2.44008207321167,
+      "learning_rate": 1.7644478180668943e-05,
+      "loss": 0.9795,
+      "step": 4230
+    },
+    {
+      "epoch": 0.8139669103501347,
+      "grad_norm": 2.1410136222839355,
+      "learning_rate": 1.7609150560084985e-05,
+      "loss": 0.844,
+      "step": 4231
+    },
+    {
+      "epoch": 0.8141592920353983,
+      "grad_norm": 2.304213285446167,
+      "learning_rate": 1.757385492646558e-05,
+      "loss": 0.9558,
+      "step": 4232
+    },
+    {
+      "epoch": 0.8143516737206618,
+      "grad_norm": 2.1675398349761963,
+      "learning_rate": 1.7538591293513685e-05,
+      "loss": 0.9472,
+      "step": 4233
+    },
+    {
+      "epoch": 0.8145440554059253,
+      "grad_norm": 1.3325634002685547,
+      "learning_rate": 1.750335967491993e-05,
+      "loss": 0.9626,
+      "step": 4234
+    },
+    {
+      "epoch": 0.8147364370911889,
+      "grad_norm": 2.784323215484619,
+      "learning_rate": 1.746816008436234e-05,
+      "loss": 0.9892,
+      "step": 4235
+    },
+    {
+      "epoch": 0.8149288187764525,
+      "grad_norm": 2.6772968769073486,
+      "learning_rate": 1.7432992535506686e-05,
+      "loss": 0.9246,
+      "step": 4236
+    },
+    {
+      "epoch": 0.8151212004617161,
+      "grad_norm": 2.2723872661590576,
+      "learning_rate": 1.7397857042006194e-05,
+      "loss": 0.9253,
+      "step": 4237
+    },
+    {
+      "epoch": 0.8153135821469796,
+      "grad_norm": 1.3481782674789429,
+      "learning_rate": 1.736275361750167e-05,
+      "loss": 0.7712,
+      "step": 4238
+    },
+    {
+      "epoch": 0.8155059638322432,
+      "grad_norm": 3.060006856918335,
+      "learning_rate": 1.7327682275621503e-05,
+      "loss": 0.9546,
+      "step": 4239
+    },
+    {
+      "epoch": 0.8156983455175068,
+      "grad_norm": 2.4290425777435303,
+      "learning_rate": 1.7292643029981526e-05,
+      "loss": 0.998,
+      "step": 4240
+    },
+    {
+      "epoch": 0.8158907272027703,
+      "grad_norm": 2.4697844982147217,
+      "learning_rate": 1.725763589418523e-05,
+      "loss": 0.9335,
+      "step": 4241
+    },
+    {
+      "epoch": 0.8160831088880338,
+      "grad_norm": 1.612105369567871,
+      "learning_rate": 1.7222660881823593e-05,
+      "loss": 0.967,
+      "step": 4242
+    },
+    {
+      "epoch": 0.8162754905732974,
+      "grad_norm": 2.7191519737243652,
+      "learning_rate": 1.7187718006475117e-05,
+      "loss": 0.919,
+      "step": 4243
+    },
+    {
+      "epoch": 0.816467872258561,
+      "grad_norm": 1.2318203449249268,
+      "learning_rate": 1.7152807281705806e-05,
+      "loss": 0.7169,
+      "step": 4244
+    },
+    {
+      "epoch": 0.8166602539438246,
+      "grad_norm": 2.5941884517669678,
+      "learning_rate": 1.7117928721069232e-05,
+      "loss": 1.006,
+      "step": 4245
+    },
+    {
+      "epoch": 0.8168526356290882,
+      "grad_norm": 1.3960570096969604,
+      "learning_rate": 1.7083082338106438e-05,
+      "loss": 0.7343,
+      "step": 4246
+    },
+    {
+      "epoch": 0.8170450173143516,
+      "grad_norm": 2.5316834449768066,
+      "learning_rate": 1.7048268146345968e-05,
+      "loss": 0.9803,
+      "step": 4247
+    },
+    {
+      "epoch": 0.8172373989996152,
+      "grad_norm": 1.8889247179031372,
+      "learning_rate": 1.701348615930397e-05,
+      "loss": 0.9516,
+      "step": 4248
+    },
+    {
+      "epoch": 0.8174297806848788,
+      "grad_norm": 2.244135618209839,
+      "learning_rate": 1.6978736390483895e-05,
+      "loss": 0.907,
+      "step": 4249
+    },
+    {
+      "epoch": 0.8176221623701424,
+      "grad_norm": 2.0699093341827393,
+      "learning_rate": 1.6944018853376897e-05,
+      "loss": 0.9086,
+      "step": 4250
+    },
+    {
+      "epoch": 0.8178145440554059,
+      "grad_norm": 2.4169623851776123,
+      "learning_rate": 1.6909333561461472e-05,
+      "loss": 0.9281,
+      "step": 4251
+    },
+    {
+      "epoch": 0.8180069257406695,
+      "grad_norm": 2.809767961502075,
+      "learning_rate": 1.6874680528203656e-05,
+      "loss": 0.9896,
+      "step": 4252
+    },
+    {
+      "epoch": 0.818199307425933,
+      "grad_norm": 2.5491902828216553,
+      "learning_rate": 1.684005976705695e-05,
+      "loss": 0.9631,
+      "step": 4253
+    },
+    {
+      "epoch": 0.8183916891111966,
+      "grad_norm": 2.227562189102173,
+      "learning_rate": 1.6805471291462317e-05,
+      "loss": 0.9012,
+      "step": 4254
+    },
+    {
+      "epoch": 0.8185840707964602,
+      "grad_norm": 2.2272160053253174,
+      "learning_rate": 1.6770915114848194e-05,
+      "loss": 0.906,
+      "step": 4255
+    },
+    {
+      "epoch": 0.8187764524817237,
+      "grad_norm": 2.151442050933838,
+      "learning_rate": 1.67363912506305e-05,
+      "loss": 0.9191,
+      "step": 4256
+    },
+    {
+      "epoch": 0.8189688341669873,
+      "grad_norm": 2.574570655822754,
+      "learning_rate": 1.6701899712212564e-05,
+      "loss": 0.9236,
+      "step": 4257
+    },
+    {
+      "epoch": 0.8191612158522509,
+      "grad_norm": 1.2457756996154785,
+      "learning_rate": 1.66674405129852e-05,
+      "loss": 0.7019,
+      "step": 4258
+    },
+    {
+      "epoch": 0.8193535975375145,
+      "grad_norm": 1.7321438789367676,
+      "learning_rate": 1.6633013666326637e-05,
+      "loss": 0.9879,
+      "step": 4259
+    },
+    {
+      "epoch": 0.8195459792227779,
+      "grad_norm": 1.9647603034973145,
+      "learning_rate": 1.6598619185602614e-05,
+      "loss": 0.9821,
+      "step": 4260
+    },
+    {
+      "epoch": 0.8197383609080415,
+      "grad_norm": 2.3184876441955566,
+      "learning_rate": 1.656425708416617e-05,
+      "loss": 0.9548,
+      "step": 4261
+    },
+    {
+      "epoch": 0.8199307425933051,
+      "grad_norm": 2.5084686279296875,
+      "learning_rate": 1.6529927375357958e-05,
+      "loss": 0.9633,
+      "step": 4262
+    },
+    {
+      "epoch": 0.8201231242785687,
+      "grad_norm": 1.9338574409484863,
+      "learning_rate": 1.649563007250584e-05,
+      "loss": 0.9788,
+      "step": 4263
+    },
+    {
+      "epoch": 0.8203155059638323,
+      "grad_norm": 1.7269905805587769,
+      "learning_rate": 1.6461365188925304e-05,
+      "loss": 0.9263,
+      "step": 4264
+    },
+    {
+      "epoch": 0.8205078876490958,
+      "grad_norm": 2.1949126720428467,
+      "learning_rate": 1.6427132737919137e-05,
+      "loss": 0.9097,
+      "step": 4265
+    },
+    {
+      "epoch": 0.8207002693343594,
+      "grad_norm": 2.0645716190338135,
+      "learning_rate": 1.6392932732777487e-05,
+      "loss": 0.8379,
+      "step": 4266
+    },
+    {
+      "epoch": 0.8208926510196229,
+      "grad_norm": 2.3977224826812744,
+      "learning_rate": 1.6358765186778057e-05,
+      "loss": 0.9257,
+      "step": 4267
+    },
+    {
+      "epoch": 0.8210850327048865,
+      "grad_norm": 2.3104326725006104,
+      "learning_rate": 1.6324630113185836e-05,
+      "loss": 0.9694,
+      "step": 4268
+    },
+    {
+      "epoch": 0.82127741439015,
+      "grad_norm": 2.022033929824829,
+      "learning_rate": 1.6290527525253228e-05,
+      "loss": 0.9283,
+      "step": 4269
+    },
+    {
+      "epoch": 0.8214697960754136,
+      "grad_norm": 2.5913190841674805,
+      "learning_rate": 1.6256457436220028e-05,
+      "loss": 0.9778,
+      "step": 4270
+    },
+    {
+      "epoch": 0.8216621777606772,
+      "grad_norm": 2.3777847290039062,
+      "learning_rate": 1.6222419859313442e-05,
+      "loss": 1.0186,
+      "step": 4271
+    },
+    {
+      "epoch": 0.8218545594459408,
+      "grad_norm": 1.2237138748168945,
+      "learning_rate": 1.6188414807747997e-05,
+      "loss": 0.9387,
+      "step": 4272
+    },
+    {
+      "epoch": 0.8220469411312044,
+      "grad_norm": 1.1584200859069824,
+      "learning_rate": 1.6154442294725636e-05,
+      "loss": 1.0038,
+      "step": 4273
+    },
+    {
+      "epoch": 0.8222393228164678,
+      "grad_norm": 1.3378217220306396,
+      "learning_rate": 1.6120502333435693e-05,
+      "loss": 1.0344,
+      "step": 4274
+    },
+    {
+      "epoch": 0.8224317045017314,
+      "grad_norm": 2.5985190868377686,
+      "learning_rate": 1.6086594937054765e-05,
+      "loss": 1.0157,
+      "step": 4275
+    },
+    {
+      "epoch": 0.822624086186995,
+      "grad_norm": 2.7500429153442383,
+      "learning_rate": 1.6052720118746922e-05,
+      "loss": 0.9331,
+      "step": 4276
+    },
+    {
+      "epoch": 0.8228164678722586,
+      "grad_norm": 1.1627689599990845,
+      "learning_rate": 1.601887789166352e-05,
+      "loss": 0.9203,
+      "step": 4277
+    },
+    {
+      "epoch": 0.8230088495575221,
+      "grad_norm": 2.839909315109253,
+      "learning_rate": 1.598506826894328e-05,
+      "loss": 0.8917,
+      "step": 4278
+    },
+    {
+      "epoch": 0.8232012312427857,
+      "grad_norm": 2.4829134941101074,
+      "learning_rate": 1.5951291263712252e-05,
+      "loss": 0.9541,
+      "step": 4279
+    },
+    {
+      "epoch": 0.8233936129280492,
+      "grad_norm": 2.039595603942871,
+      "learning_rate": 1.5917546889083834e-05,
+      "loss": 0.8275,
+      "step": 4280
+    },
+    {
+      "epoch": 0.8235859946133128,
+      "grad_norm": 2.2778995037078857,
+      "learning_rate": 1.5883835158158765e-05,
+      "loss": 0.9266,
+      "step": 4281
+    },
+    {
+      "epoch": 0.8237783762985764,
+      "grad_norm": 1.5303864479064941,
+      "learning_rate": 1.585015608402509e-05,
+      "loss": 0.9304,
+      "step": 4282
+    },
+    {
+      "epoch": 0.8239707579838399,
+      "grad_norm": 1.1347392797470093,
+      "learning_rate": 1.5816509679758185e-05,
+      "loss": 0.7514,
+      "step": 4283
+    },
+    {
+      "epoch": 0.8241631396691035,
+      "grad_norm": 2.687572479248047,
+      "learning_rate": 1.578289595842074e-05,
+      "loss": 0.9014,
+      "step": 4284
+    },
+    {
+      "epoch": 0.8243555213543671,
+      "grad_norm": 1.5734219551086426,
+      "learning_rate": 1.5749314933062753e-05,
+      "loss": 0.8052,
+      "step": 4285
+    },
+    {
+      "epoch": 0.8245479030396307,
+      "grad_norm": 2.1503922939300537,
+      "learning_rate": 1.5715766616721582e-05,
+      "loss": 0.9564,
+      "step": 4286
+    },
+    {
+      "epoch": 0.8247402847248941,
+      "grad_norm": 2.342648506164551,
+      "learning_rate": 1.5682251022421758e-05,
+      "loss": 0.8893,
+      "step": 4287
+    },
+    {
+      "epoch": 0.8249326664101577,
+      "grad_norm": 2.3593180179595947,
+      "learning_rate": 1.5648768163175277e-05,
+      "loss": 0.927,
+      "step": 4288
+    },
+    {
+      "epoch": 0.8251250480954213,
+      "grad_norm": 2.4213733673095703,
+      "learning_rate": 1.561531805198124e-05,
+      "loss": 0.9783,
+      "step": 4289
+    },
+    {
+      "epoch": 0.8253174297806849,
+      "grad_norm": 2.5601162910461426,
+      "learning_rate": 1.5581900701826225e-05,
+      "loss": 0.8905,
+      "step": 4290
+    },
+    {
+      "epoch": 0.8255098114659485,
+      "grad_norm": 2.0548651218414307,
+      "learning_rate": 1.5548516125683978e-05,
+      "loss": 0.9923,
+      "step": 4291
+    },
+    {
+      "epoch": 0.825702193151212,
+      "grad_norm": 2.2252628803253174,
+      "learning_rate": 1.5515164336515465e-05,
+      "loss": 0.8971,
+      "step": 4292
+    },
+    {
+      "epoch": 0.8258945748364755,
+      "grad_norm": 2.246380567550659,
+      "learning_rate": 1.5481845347269076e-05,
+      "loss": 0.9483,
+      "step": 4293
+    },
+    {
+      "epoch": 0.8260869565217391,
+      "grad_norm": 2.576582908630371,
+      "learning_rate": 1.5448559170880374e-05,
+      "loss": 0.9076,
+      "step": 4294
+    },
+    {
+      "epoch": 0.8262793382070027,
+      "grad_norm": 2.332869052886963,
+      "learning_rate": 1.5415305820272197e-05,
+      "loss": 0.9617,
+      "step": 4295
+    },
+    {
+      "epoch": 0.8264717198922663,
+      "grad_norm": 2.06768798828125,
+      "learning_rate": 1.538208530835463e-05,
+      "loss": 0.9605,
+      "step": 4296
+    },
+    {
+      "epoch": 0.8266641015775298,
+      "grad_norm": 2.2764174938201904,
+      "learning_rate": 1.534889764802503e-05,
+      "loss": 0.8606,
+      "step": 4297
+    },
+    {
+      "epoch": 0.8268564832627934,
+      "grad_norm": 2.216350793838501,
+      "learning_rate": 1.531574285216799e-05,
+      "loss": 0.9894,
+      "step": 4298
+    },
+    {
+      "epoch": 0.827048864948057,
+      "grad_norm": 1.3343639373779297,
+      "learning_rate": 1.528262093365531e-05,
+      "loss": 0.9614,
+      "step": 4299
+    },
+    {
+      "epoch": 0.8272412466333205,
+      "grad_norm": 1.1215732097625732,
+      "learning_rate": 1.5249531905346138e-05,
+      "loss": 0.9881,
+      "step": 4300
+    },
+    {
+      "epoch": 0.827433628318584,
+      "grad_norm": 1.6420294046401978,
+      "learning_rate": 1.5216475780086669e-05,
+      "loss": 0.9664,
+      "step": 4301
+    },
+    {
+      "epoch": 0.8276260100038476,
+      "grad_norm": 2.4027271270751953,
+      "learning_rate": 1.5183452570710522e-05,
+      "loss": 0.9117,
+      "step": 4302
+    },
+    {
+      "epoch": 0.8278183916891112,
+      "grad_norm": 1.9258999824523926,
+      "learning_rate": 1.5150462290038392e-05,
+      "loss": 0.9525,
+      "step": 4303
+    },
+    {
+      "epoch": 0.8280107733743748,
+      "grad_norm": 2.534813642501831,
+      "learning_rate": 1.511750495087827e-05,
+      "loss": 0.8927,
+      "step": 4304
+    },
+    {
+      "epoch": 0.8282031550596384,
+      "grad_norm": 2.3109819889068604,
+      "learning_rate": 1.5084580566025308e-05,
+      "loss": 0.9711,
+      "step": 4305
+    },
+    {
+      "epoch": 0.8283955367449019,
+      "grad_norm": 2.3718042373657227,
+      "learning_rate": 1.5051689148261894e-05,
+      "loss": 0.8973,
+      "step": 4306
+    },
+    {
+      "epoch": 0.8285879184301654,
+      "grad_norm": 2.200052261352539,
+      "learning_rate": 1.5018830710357611e-05,
+      "loss": 0.9455,
+      "step": 4307
+    },
+    {
+      "epoch": 0.828780300115429,
+      "grad_norm": 1.5570647716522217,
+      "learning_rate": 1.4986005265069204e-05,
+      "loss": 0.9007,
+      "step": 4308
+    },
+    {
+      "epoch": 0.8289726818006926,
+      "grad_norm": 2.2958667278289795,
+      "learning_rate": 1.4953212825140727e-05,
+      "loss": 0.9385,
+      "step": 4309
+    },
+    {
+      "epoch": 0.8291650634859561,
+      "grad_norm": 2.412257671356201,
+      "learning_rate": 1.4920453403303247e-05,
+      "loss": 0.935,
+      "step": 4310
+    },
+    {
+      "epoch": 0.8293574451712197,
+      "grad_norm": 2.202838659286499,
+      "learning_rate": 1.488772701227511e-05,
+      "loss": 0.9399,
+      "step": 4311
+    },
+    {
+      "epoch": 0.8295498268564833,
+      "grad_norm": 1.0792138576507568,
+      "learning_rate": 1.4855033664761897e-05,
+      "loss": 0.944,
+      "step": 4312
+    },
+    {
+      "epoch": 0.8297422085417469,
+      "grad_norm": 2.8197145462036133,
+      "learning_rate": 1.48223733734562e-05,
+      "loss": 0.9032,
+      "step": 4313
+    },
+    {
+      "epoch": 0.8299345902270104,
+      "grad_norm": 1.6927437782287598,
+      "learning_rate": 1.4789746151037942e-05,
+      "loss": 0.8742,
+      "step": 4314
+    },
+    {
+      "epoch": 0.8301269719122739,
+      "grad_norm": 2.4867351055145264,
+      "learning_rate": 1.4757152010174069e-05,
+      "loss": 0.9739,
+      "step": 4315
+    },
+    {
+      "epoch": 0.8303193535975375,
+      "grad_norm": 2.443977117538452,
+      "learning_rate": 1.4724590963518803e-05,
+      "loss": 1.0197,
+      "step": 4316
+    },
+    {
+      "epoch": 0.8305117352828011,
+      "grad_norm": 1.1042571067810059,
+      "learning_rate": 1.4692063023713443e-05,
+      "loss": 0.9774,
+      "step": 4317
+    },
+    {
+      "epoch": 0.8307041169680647,
+      "grad_norm": 2.6829397678375244,
+      "learning_rate": 1.4659568203386465e-05,
+      "loss": 0.8818,
+      "step": 4318
+    },
+    {
+      "epoch": 0.8308964986533282,
+      "grad_norm": 2.577235698699951,
+      "learning_rate": 1.4627106515153455e-05,
+      "loss": 0.9397,
+      "step": 4319
+    },
+    {
+      "epoch": 0.8310888803385917,
+      "grad_norm": 2.0848371982574463,
+      "learning_rate": 1.4594677971617177e-05,
+      "loss": 0.9219,
+      "step": 4320
+    },
+    {
+      "epoch": 0.8312812620238553,
+      "grad_norm": 1.657609224319458,
+      "learning_rate": 1.4562282585367492e-05,
+      "loss": 0.9767,
+      "step": 4321
+    },
+    {
+      "epoch": 0.8314736437091189,
+      "grad_norm": 1.7219592332839966,
+      "learning_rate": 1.452992036898142e-05,
+      "loss": 0.8558,
+      "step": 4322
+    },
+    {
+      "epoch": 0.8316660253943825,
+      "grad_norm": 2.3813529014587402,
+      "learning_rate": 1.4497591335023086e-05,
+      "loss": 0.9497,
+      "step": 4323
+    },
+    {
+      "epoch": 0.831858407079646,
+      "grad_norm": 2.024906873703003,
+      "learning_rate": 1.4465295496043729e-05,
+      "loss": 0.8935,
+      "step": 4324
+    },
+    {
+      "epoch": 0.8320507887649096,
+      "grad_norm": 2.5948026180267334,
+      "learning_rate": 1.4433032864581686e-05,
+      "loss": 0.9659,
+      "step": 4325
+    },
+    {
+      "epoch": 0.8322431704501732,
+      "grad_norm": 1.3252772092819214,
+      "learning_rate": 1.4400803453162482e-05,
+      "loss": 0.9329,
+      "step": 4326
+    },
+    {
+      "epoch": 0.8324355521354367,
+      "grad_norm": 1.3588662147521973,
+      "learning_rate": 1.4368607274298595e-05,
+      "loss": 0.9991,
+      "step": 4327
+    },
+    {
+      "epoch": 0.8326279338207002,
+      "grad_norm": 1.0293331146240234,
+      "learning_rate": 1.4336444340489775e-05,
+      "loss": 0.9194,
+      "step": 4328
+    },
+    {
+      "epoch": 0.8328203155059638,
+      "grad_norm": 2.511300802230835,
+      "learning_rate": 1.4304314664222728e-05,
+      "loss": 0.9517,
+      "step": 4329
+    },
+    {
+      "epoch": 0.8330126971912274,
+      "grad_norm": 2.2753684520721436,
+      "learning_rate": 1.4272218257971326e-05,
+      "loss": 0.9764,
+      "step": 4330
+    },
+    {
+      "epoch": 0.833205078876491,
+      "grad_norm": 2.594447135925293,
+      "learning_rate": 1.4240155134196499e-05,
+      "loss": 1.0027,
+      "step": 4331
+    },
+    {
+      "epoch": 0.8333974605617546,
+      "grad_norm": 1.9887388944625854,
+      "learning_rate": 1.4208125305346232e-05,
+      "loss": 0.9511,
+      "step": 4332
+    },
+    {
+      "epoch": 0.833589842247018,
+      "grad_norm": 2.5889508724212646,
+      "learning_rate": 1.4176128783855636e-05,
+      "loss": 0.936,
+      "step": 4333
+    },
+    {
+      "epoch": 0.8337822239322816,
+      "grad_norm": 1.8996127843856812,
+      "learning_rate": 1.4144165582146817e-05,
+      "loss": 0.9556,
+      "step": 4334
+    },
+    {
+      "epoch": 0.8339746056175452,
+      "grad_norm": 2.404356002807617,
+      "learning_rate": 1.4112235712629064e-05,
+      "loss": 0.9497,
+      "step": 4335
+    },
+    {
+      "epoch": 0.8341669873028088,
+      "grad_norm": 1.5808215141296387,
+      "learning_rate": 1.40803391876986e-05,
+      "loss": 0.9718,
+      "step": 4336
+    },
+    {
+      "epoch": 0.8343593689880723,
+      "grad_norm": 2.748476028442383,
+      "learning_rate": 1.4048476019738755e-05,
+      "loss": 0.9619,
+      "step": 4337
+    },
+    {
+      "epoch": 0.8345517506733359,
+      "grad_norm": 2.4150967597961426,
+      "learning_rate": 1.401664622111991e-05,
+      "loss": 0.9075,
+      "step": 4338
+    },
+    {
+      "epoch": 0.8347441323585995,
+      "grad_norm": 1.6867778301239014,
+      "learning_rate": 1.3984849804199485e-05,
+      "loss": 0.8889,
+      "step": 4339
+    },
+    {
+      "epoch": 0.834936514043863,
+      "grad_norm": 2.390503406524658,
+      "learning_rate": 1.3953086781321988e-05,
+      "loss": 0.9645,
+      "step": 4340
+    },
+    {
+      "epoch": 0.8351288957291266,
+      "grad_norm": 2.35526704788208,
+      "learning_rate": 1.3921357164818848e-05,
+      "loss": 0.9125,
+      "step": 4341
+    },
+    {
+      "epoch": 0.8353212774143901,
+      "grad_norm": 1.088305115699768,
+      "learning_rate": 1.3889660967008655e-05,
+      "loss": 0.8961,
+      "step": 4342
+    },
+    {
+      "epoch": 0.8355136590996537,
+      "grad_norm": 2.3180525302886963,
+      "learning_rate": 1.3857998200196941e-05,
+      "loss": 0.9193,
+      "step": 4343
+    },
+    {
+      "epoch": 0.8357060407849173,
+      "grad_norm": 1.787286400794983,
+      "learning_rate": 1.3826368876676278e-05,
+      "loss": 0.9281,
+      "step": 4344
+    },
+    {
+      "epoch": 0.8358984224701809,
+      "grad_norm": 1.9564844369888306,
+      "learning_rate": 1.3794773008726259e-05,
+      "loss": 0.9662,
+      "step": 4345
+    },
+    {
+      "epoch": 0.8360908041554445,
+      "grad_norm": 2.2608401775360107,
+      "learning_rate": 1.3763210608613498e-05,
+      "loss": 0.9428,
+      "step": 4346
+    },
+    {
+      "epoch": 0.8362831858407079,
+      "grad_norm": 2.680612802505493,
+      "learning_rate": 1.3731681688591591e-05,
+      "loss": 0.9627,
+      "step": 4347
+    },
+    {
+      "epoch": 0.8364755675259715,
+      "grad_norm": 2.189525604248047,
+      "learning_rate": 1.370018626090116e-05,
+      "loss": 0.9128,
+      "step": 4348
+    },
+    {
+      "epoch": 0.8366679492112351,
+      "grad_norm": 2.4549551010131836,
+      "learning_rate": 1.3668724337769823e-05,
+      "loss": 1.0046,
+      "step": 4349
+    },
+    {
+      "epoch": 0.8368603308964987,
+      "grad_norm": 2.5571107864379883,
+      "learning_rate": 1.3637295931412152e-05,
+      "loss": 0.9788,
+      "step": 4350
+    },
+    {
+      "epoch": 0.8370527125817622,
+      "grad_norm": 2.3024494647979736,
+      "learning_rate": 1.3605901054029746e-05,
+      "loss": 0.9753,
+      "step": 4351
+    },
+    {
+      "epoch": 0.8372450942670258,
+      "grad_norm": 2.0618364810943604,
+      "learning_rate": 1.357453971781123e-05,
+      "loss": 0.9384,
+      "step": 4352
+    },
+    {
+      "epoch": 0.8374374759522893,
+      "grad_norm": 2.163374423980713,
+      "learning_rate": 1.3543211934932065e-05,
+      "loss": 0.9513,
+      "step": 4353
+    },
+    {
+      "epoch": 0.8376298576375529,
+      "grad_norm": 2.2278599739074707,
+      "learning_rate": 1.3511917717554846e-05,
+      "loss": 0.932,
+      "step": 4354
+    },
+    {
+      "epoch": 0.8378222393228165,
+      "grad_norm": 2.1292471885681152,
+      "learning_rate": 1.348065707782904e-05,
+      "loss": 0.9171,
+      "step": 4355
+    },
+    {
+      "epoch": 0.83801462100808,
+      "grad_norm": 2.364267587661743,
+      "learning_rate": 1.3449430027891097e-05,
+      "loss": 0.9503,
+      "step": 4356
+    },
+    {
+      "epoch": 0.8382070026933436,
+      "grad_norm": 2.976205587387085,
+      "learning_rate": 1.341823657986445e-05,
+      "loss": 1.0243,
+      "step": 4357
+    },
+    {
+      "epoch": 0.8383993843786072,
+      "grad_norm": 2.4288723468780518,
+      "learning_rate": 1.338707674585945e-05,
+      "loss": 0.9528,
+      "step": 4358
+    },
+    {
+      "epoch": 0.8385917660638708,
+      "grad_norm": 2.553147554397583,
+      "learning_rate": 1.335595053797344e-05,
+      "loss": 0.9796,
+      "step": 4359
+    },
+    {
+      "epoch": 0.8387841477491342,
+      "grad_norm": 2.679713010787964,
+      "learning_rate": 1.332485796829065e-05,
+      "loss": 1.0127,
+      "step": 4360
+    },
+    {
+      "epoch": 0.8389765294343978,
+      "grad_norm": 1.4238145351409912,
+      "learning_rate": 1.329379904888235e-05,
+      "loss": 0.7397,
+      "step": 4361
+    },
+    {
+      "epoch": 0.8391689111196614,
+      "grad_norm": 2.0439443588256836,
+      "learning_rate": 1.3262773791806615e-05,
+      "loss": 0.9638,
+      "step": 4362
+    },
+    {
+      "epoch": 0.839361292804925,
+      "grad_norm": 1.191663384437561,
+      "learning_rate": 1.3231782209108545e-05,
+      "loss": 0.908,
+      "step": 4363
+    },
+    {
+      "epoch": 0.8395536744901886,
+      "grad_norm": 2.3901965618133545,
+      "learning_rate": 1.3200824312820136e-05,
+      "loss": 0.9603,
+      "step": 4364
+    },
+    {
+      "epoch": 0.8397460561754521,
+      "grad_norm": 2.1114790439605713,
+      "learning_rate": 1.3169900114960298e-05,
+      "loss": 0.9629,
+      "step": 4365
+    },
+    {
+      "epoch": 0.8399384378607156,
+      "grad_norm": 1.483425498008728,
+      "learning_rate": 1.3139009627534927e-05,
+      "loss": 1.0347,
+      "step": 4366
+    },
+    {
+      "epoch": 0.8401308195459792,
+      "grad_norm": 0.9786659479141235,
+      "learning_rate": 1.3108152862536682e-05,
+      "loss": 0.9548,
+      "step": 4367
+    },
+    {
+      "epoch": 0.8403232012312428,
+      "grad_norm": 2.5085232257843018,
+      "learning_rate": 1.3077329831945295e-05,
+      "loss": 0.9284,
+      "step": 4368
+    },
+    {
+      "epoch": 0.8405155829165063,
+      "grad_norm": 2.4877822399139404,
+      "learning_rate": 1.3046540547727303e-05,
+      "loss": 0.9157,
+      "step": 4369
+    },
+    {
+      "epoch": 0.8407079646017699,
+      "grad_norm": 1.7775264978408813,
+      "learning_rate": 1.3015785021836158e-05,
+      "loss": 1.0222,
+      "step": 4370
+    },
+    {
+      "epoch": 0.8409003462870335,
+      "grad_norm": 2.4562439918518066,
+      "learning_rate": 1.2985063266212227e-05,
+      "loss": 0.9201,
+      "step": 4371
+    },
+    {
+      "epoch": 0.8410927279722971,
+      "grad_norm": 2.239586353302002,
+      "learning_rate": 1.295437529278275e-05,
+      "loss": 0.8924,
+      "step": 4372
+    },
+    {
+      "epoch": 0.8412851096575606,
+      "grad_norm": 1.3646639585494995,
+      "learning_rate": 1.2923721113461851e-05,
+      "loss": 0.7645,
+      "step": 4373
+    },
+    {
+      "epoch": 0.8414774913428241,
+      "grad_norm": 2.1150832176208496,
+      "learning_rate": 1.289310074015052e-05,
+      "loss": 0.9005,
+      "step": 4374
+    },
+    {
+      "epoch": 0.8416698730280877,
+      "grad_norm": 2.169037342071533,
+      "learning_rate": 1.2862514184736696e-05,
+      "loss": 0.9004,
+      "step": 4375
+    },
+    {
+      "epoch": 0.8418622547133513,
+      "grad_norm": 2.292607069015503,
+      "learning_rate": 1.2831961459095088e-05,
+      "loss": 0.9149,
+      "step": 4376
+    },
+    {
+      "epoch": 0.8420546363986149,
+      "grad_norm": 2.052917957305908,
+      "learning_rate": 1.2801442575087298e-05,
+      "loss": 0.8102,
+      "step": 4377
+    },
+    {
+      "epoch": 0.8422470180838784,
+      "grad_norm": 2.8715569972991943,
+      "learning_rate": 1.2770957544561869e-05,
+      "loss": 0.9721,
+      "step": 4378
+    },
+    {
+      "epoch": 0.842439399769142,
+      "grad_norm": 2.424457550048828,
+      "learning_rate": 1.2740506379354077e-05,
+      "loss": 0.9445,
+      "step": 4379
+    },
+    {
+      "epoch": 0.8426317814544055,
+      "grad_norm": 2.566615104675293,
+      "learning_rate": 1.2710089091286149e-05,
+      "loss": 1.0207,
+      "step": 4380
+    },
+    {
+      "epoch": 0.8428241631396691,
+      "grad_norm": 2.450044631958008,
+      "learning_rate": 1.267970569216712e-05,
+      "loss": 0.9195,
+      "step": 4381
+    },
+    {
+      "epoch": 0.8430165448249327,
+      "grad_norm": 2.221393346786499,
+      "learning_rate": 1.2649356193792872e-05,
+      "loss": 0.9453,
+      "step": 4382
+    },
+    {
+      "epoch": 0.8432089265101962,
+      "grad_norm": 2.325629711151123,
+      "learning_rate": 1.2619040607946119e-05,
+      "loss": 0.9106,
+      "step": 4383
+    },
+    {
+      "epoch": 0.8434013081954598,
+      "grad_norm": 2.2289535999298096,
+      "learning_rate": 1.2588758946396418e-05,
+      "loss": 0.9866,
+      "step": 4384
+    },
+    {
+      "epoch": 0.8435936898807234,
+      "grad_norm": 2.299389123916626,
+      "learning_rate": 1.2558511220900138e-05,
+      "loss": 0.9813,
+      "step": 4385
+    },
+    {
+      "epoch": 0.843786071565987,
+      "grad_norm": 2.171234607696533,
+      "learning_rate": 1.2528297443200488e-05,
+      "loss": 0.9103,
+      "step": 4386
+    },
+    {
+      "epoch": 0.8439784532512504,
+      "grad_norm": 1.3782275915145874,
+      "learning_rate": 1.2498117625027561e-05,
+      "loss": 0.7598,
+      "step": 4387
+    },
+    {
+      "epoch": 0.844170834936514,
+      "grad_norm": 2.3873608112335205,
+      "learning_rate": 1.246797177809812e-05,
+      "loss": 0.9181,
+      "step": 4388
+    },
+    {
+      "epoch": 0.8443632166217776,
+      "grad_norm": 2.6363329887390137,
+      "learning_rate": 1.2437859914115845e-05,
+      "loss": 0.923,
+      "step": 4389
+    },
+    {
+      "epoch": 0.8445555983070412,
+      "grad_norm": 2.3539412021636963,
+      "learning_rate": 1.2407782044771221e-05,
+      "loss": 0.9337,
+      "step": 4390
+    },
+    {
+      "epoch": 0.8447479799923048,
+      "grad_norm": 2.196791410446167,
+      "learning_rate": 1.237773818174146e-05,
+      "loss": 0.8564,
+      "step": 4391
+    },
+    {
+      "epoch": 0.8449403616775683,
+      "grad_norm": 1.8900405168533325,
+      "learning_rate": 1.2347728336690701e-05,
+      "loss": 0.9201,
+      "step": 4392
+    },
+    {
+      "epoch": 0.8451327433628318,
+      "grad_norm": 2.3956637382507324,
+      "learning_rate": 1.2317752521269722e-05,
+      "loss": 0.8257,
+      "step": 4393
+    },
+    {
+      "epoch": 0.8453251250480954,
+      "grad_norm": 1.5139740705490112,
+      "learning_rate": 1.2287810747116224e-05,
+      "loss": 0.9556,
+      "step": 4394
+    },
+    {
+      "epoch": 0.845517506733359,
+      "grad_norm": 2.5123817920684814,
+      "learning_rate": 1.2257903025854612e-05,
+      "loss": 0.8132,
+      "step": 4395
+    },
+    {
+      "epoch": 0.8457098884186226,
+      "grad_norm": 1.0498409271240234,
+      "learning_rate": 1.2228029369096094e-05,
+      "loss": 0.9805,
+      "step": 4396
+    },
+    {
+      "epoch": 0.8459022701038861,
+      "grad_norm": 2.490847110748291,
+      "learning_rate": 1.2198189788438651e-05,
+      "loss": 0.9312,
+      "step": 4397
+    },
+    {
+      "epoch": 0.8460946517891497,
+      "grad_norm": 2.1021411418914795,
+      "learning_rate": 1.2168384295467038e-05,
+      "loss": 0.843,
+      "step": 4398
+    },
+    {
+      "epoch": 0.8462870334744133,
+      "grad_norm": 2.276226043701172,
+      "learning_rate": 1.2138612901752777e-05,
+      "loss": 0.9409,
+      "step": 4399
+    },
+    {
+      "epoch": 0.8464794151596768,
+      "grad_norm": 1.3779385089874268,
+      "learning_rate": 1.2108875618854121e-05,
+      "loss": 0.8144,
+      "step": 4400
+    },
+    {
+      "epoch": 0.8466717968449403,
+      "grad_norm": 2.3484747409820557,
+      "learning_rate": 1.2079172458316168e-05,
+      "loss": 0.8927,
+      "step": 4401
+    },
+    {
+      "epoch": 0.8468641785302039,
+      "grad_norm": 2.7813339233398438,
+      "learning_rate": 1.204950343167065e-05,
+      "loss": 0.9732,
+      "step": 4402
+    },
+    {
+      "epoch": 0.8470565602154675,
+      "grad_norm": 1.3916268348693848,
+      "learning_rate": 1.2019868550436098e-05,
+      "loss": 1.0145,
+      "step": 4403
+    },
+    {
+      "epoch": 0.8472489419007311,
+      "grad_norm": 2.6252758502960205,
+      "learning_rate": 1.1990267826117873e-05,
+      "loss": 0.9726,
+      "step": 4404
+    },
+    {
+      "epoch": 0.8474413235859947,
+      "grad_norm": 1.9145727157592773,
+      "learning_rate": 1.1960701270207886e-05,
+      "loss": 0.9033,
+      "step": 4405
+    },
+    {
+      "epoch": 0.8476337052712581,
+      "grad_norm": 2.14038348197937,
+      "learning_rate": 1.1931168894184974e-05,
+      "loss": 0.9164,
+      "step": 4406
+    },
+    {
+      "epoch": 0.8478260869565217,
+      "grad_norm": 2.046553373336792,
+      "learning_rate": 1.19016707095146e-05,
+      "loss": 0.972,
+      "step": 4407
+    },
+    {
+      "epoch": 0.8480184686417853,
+      "grad_norm": 2.4932920932769775,
+      "learning_rate": 1.1872206727648972e-05,
+      "loss": 0.9216,
+      "step": 4408
+    },
+    {
+      "epoch": 0.8482108503270489,
+      "grad_norm": 2.8300633430480957,
+      "learning_rate": 1.1842776960027013e-05,
+      "loss": 0.9363,
+      "step": 4409
+    },
+    {
+      "epoch": 0.8484032320123124,
+      "grad_norm": 2.1694018840789795,
+      "learning_rate": 1.1813381418074388e-05,
+      "loss": 0.947,
+      "step": 4410
+    },
+    {
+      "epoch": 0.848595613697576,
+      "grad_norm": 1.3088219165802002,
+      "learning_rate": 1.1784020113203454e-05,
+      "loss": 0.784,
+      "step": 4411
+    },
+    {
+      "epoch": 0.8487879953828396,
+      "grad_norm": 2.4190971851348877,
+      "learning_rate": 1.175469305681327e-05,
+      "loss": 0.8579,
+      "step": 4412
+    },
+    {
+      "epoch": 0.8489803770681031,
+      "grad_norm": 2.8524909019470215,
+      "learning_rate": 1.1725400260289621e-05,
+      "loss": 0.9138,
+      "step": 4413
+    },
+    {
+      "epoch": 0.8491727587533667,
+      "grad_norm": 2.153306245803833,
+      "learning_rate": 1.1696141735004939e-05,
+      "loss": 0.8669,
+      "step": 4414
+    },
+    {
+      "epoch": 0.8493651404386302,
+      "grad_norm": 2.8521556854248047,
+      "learning_rate": 1.1666917492318485e-05,
+      "loss": 0.9016,
+      "step": 4415
+    },
+    {
+      "epoch": 0.8495575221238938,
+      "grad_norm": 2.3301868438720703,
+      "learning_rate": 1.1637727543576027e-05,
+      "loss": 0.9397,
+      "step": 4416
+    },
+    {
+      "epoch": 0.8497499038091574,
+      "grad_norm": 2.2714102268218994,
+      "learning_rate": 1.160857190011012e-05,
+      "loss": 0.9372,
+      "step": 4417
+    },
+    {
+      "epoch": 0.849942285494421,
+      "grad_norm": 2.2788472175598145,
+      "learning_rate": 1.1579450573240058e-05,
+      "loss": 0.8766,
+      "step": 4418
+    },
+    {
+      "epoch": 0.8501346671796844,
+      "grad_norm": 2.2984507083892822,
+      "learning_rate": 1.1550363574271639e-05,
+      "loss": 0.9883,
+      "step": 4419
+    },
+    {
+      "epoch": 0.850327048864948,
+      "grad_norm": 2.4327518939971924,
+      "learning_rate": 1.1521310914497519e-05,
+      "loss": 0.9771,
+      "step": 4420
+    },
+    {
+      "epoch": 0.8505194305502116,
+      "grad_norm": 2.1820316314697266,
+      "learning_rate": 1.1492292605196907e-05,
+      "loss": 0.9056,
+      "step": 4421
+    },
+    {
+      "epoch": 0.8507118122354752,
+      "grad_norm": 2.1739048957824707,
+      "learning_rate": 1.1463308657635719e-05,
+      "loss": 0.9994,
+      "step": 4422
+    },
+    {
+      "epoch": 0.8509041939207388,
+      "grad_norm": 1.8901264667510986,
+      "learning_rate": 1.1434359083066515e-05,
+      "loss": 0.9722,
+      "step": 4423
+    },
+    {
+      "epoch": 0.8510965756060023,
+      "grad_norm": 2.5158982276916504,
+      "learning_rate": 1.1405443892728528e-05,
+      "loss": 0.9691,
+      "step": 4424
+    },
+    {
+      "epoch": 0.8512889572912659,
+      "grad_norm": 2.4533090591430664,
+      "learning_rate": 1.1376563097847614e-05,
+      "loss": 0.8847,
+      "step": 4425
+    },
+    {
+      "epoch": 0.8514813389765294,
+      "grad_norm": 2.22841477394104,
+      "learning_rate": 1.134771670963628e-05,
+      "loss": 0.895,
+      "step": 4426
+    },
+    {
+      "epoch": 0.851673720661793,
+      "grad_norm": 2.507352590560913,
+      "learning_rate": 1.1318904739293745e-05,
+      "loss": 0.8811,
+      "step": 4427
+    },
+    {
+      "epoch": 0.8518661023470565,
+      "grad_norm": 2.3455753326416016,
+      "learning_rate": 1.1290127198005752e-05,
+      "loss": 0.9392,
+      "step": 4428
+    },
+    {
+      "epoch": 0.8520584840323201,
+      "grad_norm": 2.1985623836517334,
+      "learning_rate": 1.1261384096944727e-05,
+      "loss": 0.891,
+      "step": 4429
+    },
+    {
+      "epoch": 0.8522508657175837,
+      "grad_norm": 2.8043370246887207,
+      "learning_rate": 1.1232675447269803e-05,
+      "loss": 0.9097,
+      "step": 4430
+    },
+    {
+      "epoch": 0.8524432474028473,
+      "grad_norm": 0.9900025129318237,
+      "learning_rate": 1.1204001260126574e-05,
+      "loss": 0.9138,
+      "step": 4431
+    },
+    {
+      "epoch": 0.8526356290881109,
+      "grad_norm": 3.063666343688965,
+      "learning_rate": 1.1175361546647412e-05,
+      "loss": 0.9126,
+      "step": 4432
+    },
+    {
+      "epoch": 0.8528280107733743,
+      "grad_norm": 1.4358506202697754,
+      "learning_rate": 1.1146756317951224e-05,
+      "loss": 0.9331,
+      "step": 4433
+    },
+    {
+      "epoch": 0.8530203924586379,
+      "grad_norm": 1.521584391593933,
+      "learning_rate": 1.1118185585143537e-05,
+      "loss": 0.9567,
+      "step": 4434
+    },
+    {
+      "epoch": 0.8532127741439015,
+      "grad_norm": 2.4567880630493164,
+      "learning_rate": 1.1089649359316501e-05,
+      "loss": 0.9706,
+      "step": 4435
+    },
+    {
+      "epoch": 0.8534051558291651,
+      "grad_norm": 2.4966611862182617,
+      "learning_rate": 1.1061147651548853e-05,
+      "loss": 0.9589,
+      "step": 4436
+    },
+    {
+      "epoch": 0.8535975375144286,
+      "grad_norm": 2.852184772491455,
+      "learning_rate": 1.1032680472905932e-05,
+      "loss": 0.8944,
+      "step": 4437
+    },
+    {
+      "epoch": 0.8537899191996922,
+      "grad_norm": 2.510829448699951,
+      "learning_rate": 1.1004247834439696e-05,
+      "loss": 0.9547,
+      "step": 4438
+    },
+    {
+      "epoch": 0.8539823008849557,
+      "grad_norm": 1.3945900201797485,
+      "learning_rate": 1.097584974718866e-05,
+      "loss": 0.9611,
+      "step": 4439
+    },
+    {
+      "epoch": 0.8541746825702193,
+      "grad_norm": 1.640638828277588,
+      "learning_rate": 1.0947486222177928e-05,
+      "loss": 0.9339,
+      "step": 4440
+    },
+    {
+      "epoch": 0.8543670642554829,
+      "grad_norm": 2.621061325073242,
+      "learning_rate": 1.0919157270419256e-05,
+      "loss": 0.8844,
+      "step": 4441
+    },
+    {
+      "epoch": 0.8545594459407464,
+      "grad_norm": 2.3708505630493164,
+      "learning_rate": 1.0890862902910847e-05,
+      "loss": 0.9452,
+      "step": 4442
+    },
+    {
+      "epoch": 0.85475182762601,
+      "grad_norm": 2.138733148574829,
+      "learning_rate": 1.0862603130637561e-05,
+      "loss": 0.9208,
+      "step": 4443
+    },
+    {
+      "epoch": 0.8549442093112736,
+      "grad_norm": 2.381007194519043,
+      "learning_rate": 1.0834377964570864e-05,
+      "loss": 0.8859,
+      "step": 4444
+    },
+    {
+      "epoch": 0.8551365909965372,
+      "grad_norm": 2.316962957382202,
+      "learning_rate": 1.0806187415668667e-05,
+      "loss": 0.9067,
+      "step": 4445
+    },
+    {
+      "epoch": 0.8553289726818007,
+      "grad_norm": 1.4686713218688965,
+      "learning_rate": 1.0778031494875573e-05,
+      "loss": 0.9022,
+      "step": 4446
+    },
+    {
+      "epoch": 0.8555213543670642,
+      "grad_norm": 2.431527853012085,
+      "learning_rate": 1.0749910213122649e-05,
+      "loss": 0.8804,
+      "step": 4447
+    },
+    {
+      "epoch": 0.8557137360523278,
+      "grad_norm": 2.3304505348205566,
+      "learning_rate": 1.072182358132755e-05,
+      "loss": 0.8723,
+      "step": 4448
+    },
+    {
+      "epoch": 0.8559061177375914,
+      "grad_norm": 2.4242939949035645,
+      "learning_rate": 1.0693771610394476e-05,
+      "loss": 0.9201,
+      "step": 4449
+    },
+    {
+      "epoch": 0.856098499422855,
+      "grad_norm": 2.9617056846618652,
+      "learning_rate": 1.0665754311214172e-05,
+      "loss": 0.9663,
+      "step": 4450
+    },
+    {
+      "epoch": 0.8562908811081185,
+      "grad_norm": 2.5705480575561523,
+      "learning_rate": 1.0637771694663901e-05,
+      "loss": 0.9759,
+      "step": 4451
+    },
+    {
+      "epoch": 0.856483262793382,
+      "grad_norm": 2.011035680770874,
+      "learning_rate": 1.0609823771607485e-05,
+      "loss": 0.9055,
+      "step": 4452
+    },
+    {
+      "epoch": 0.8566756444786456,
+      "grad_norm": 0.8969613313674927,
+      "learning_rate": 1.0581910552895302e-05,
+      "loss": 0.9672,
+      "step": 4453
+    },
+    {
+      "epoch": 0.8568680261639092,
+      "grad_norm": 1.853776216506958,
+      "learning_rate": 1.055403204936416e-05,
+      "loss": 0.9723,
+      "step": 4454
+    },
+    {
+      "epoch": 0.8570604078491728,
+      "grad_norm": 2.794917345046997,
+      "learning_rate": 1.0526188271837511e-05,
+      "loss": 0.9447,
+      "step": 4455
+    },
+    {
+      "epoch": 0.8572527895344363,
+      "grad_norm": 1.539707064628601,
+      "learning_rate": 1.0498379231125278e-05,
+      "loss": 0.9904,
+      "step": 4456
+    },
+    {
+      "epoch": 0.8574451712196999,
+      "grad_norm": 2.083130121231079,
+      "learning_rate": 1.0470604938023808e-05,
+      "loss": 0.9606,
+      "step": 4457
+    },
+    {
+      "epoch": 0.8576375529049635,
+      "grad_norm": 2.548099994659424,
+      "learning_rate": 1.0442865403316116e-05,
+      "loss": 0.8777,
+      "step": 4458
+    },
+    {
+      "epoch": 0.857829934590227,
+      "grad_norm": 1.3576269149780273,
+      "learning_rate": 1.0415160637771603e-05,
+      "loss": 0.8368,
+      "step": 4459
+    },
+    {
+      "epoch": 0.8580223162754905,
+      "grad_norm": 2.4227514266967773,
+      "learning_rate": 1.0387490652146237e-05,
+      "loss": 1.0295,
+      "step": 4460
+    },
+    {
+      "epoch": 0.8582146979607541,
+      "grad_norm": 2.358242988586426,
+      "learning_rate": 1.0359855457182454e-05,
+      "loss": 0.8927,
+      "step": 4461
+    },
+    {
+      "epoch": 0.8584070796460177,
+      "grad_norm": 2.098923444747925,
+      "learning_rate": 1.0332255063609175e-05,
+      "loss": 0.9456,
+      "step": 4462
+    },
+    {
+      "epoch": 0.8585994613312813,
+      "grad_norm": 2.7473461627960205,
+      "learning_rate": 1.0304689482141838e-05,
+      "loss": 0.9817,
+      "step": 4463
+    },
+    {
+      "epoch": 0.8587918430165449,
+      "grad_norm": 2.412815570831299,
+      "learning_rate": 1.0277158723482339e-05,
+      "loss": 0.9249,
+      "step": 4464
+    },
+    {
+      "epoch": 0.8589842247018084,
+      "grad_norm": 2.234729528427124,
+      "learning_rate": 1.024966279831907e-05,
+      "loss": 0.9171,
+      "step": 4465
+    },
+    {
+      "epoch": 0.8591766063870719,
+      "grad_norm": 2.4091410636901855,
+      "learning_rate": 1.0222201717326885e-05,
+      "loss": 0.8992,
+      "step": 4466
+    },
+    {
+      "epoch": 0.8593689880723355,
+      "grad_norm": 2.3546090126037598,
+      "learning_rate": 1.0194775491167163e-05,
+      "loss": 0.9312,
+      "step": 4467
+    },
+    {
+      "epoch": 0.8595613697575991,
+      "grad_norm": 1.4361937046051025,
+      "learning_rate": 1.0167384130487667e-05,
+      "loss": 0.9558,
+      "step": 4468
+    },
+    {
+      "epoch": 0.8597537514428626,
+      "grad_norm": 2.4838643074035645,
+      "learning_rate": 1.0140027645922657e-05,
+      "loss": 0.885,
+      "step": 4469
+    },
+    {
+      "epoch": 0.8599461331281262,
+      "grad_norm": 2.828441619873047,
+      "learning_rate": 1.0112706048092923e-05,
+      "loss": 0.9382,
+      "step": 4470
+    },
+    {
+      "epoch": 0.8601385148133898,
+      "grad_norm": 2.320748805999756,
+      "learning_rate": 1.0085419347605574e-05,
+      "loss": 0.946,
+      "step": 4471
+    },
+    {
+      "epoch": 0.8603308964986534,
+      "grad_norm": 2.543241262435913,
+      "learning_rate": 1.0058167555054298e-05,
+      "loss": 0.9593,
+      "step": 4472
+    },
+    {
+      "epoch": 0.8605232781839169,
+      "grad_norm": 2.465984344482422,
+      "learning_rate": 1.003095068101917e-05,
+      "loss": 0.8652,
+      "step": 4473
+    },
+    {
+      "epoch": 0.8607156598691804,
+      "grad_norm": 2.8797361850738525,
+      "learning_rate": 1.0003768736066722e-05,
+      "loss": 0.879,
+      "step": 4474
+    },
+    {
+      "epoch": 0.860908041554444,
+      "grad_norm": 2.80220627784729,
+      "learning_rate": 9.97662173074989e-06,
+      "loss": 0.9652,
+      "step": 4475
+    },
+    {
+      "epoch": 0.8611004232397076,
+      "grad_norm": 2.627016544342041,
+      "learning_rate": 9.949509675608115e-06,
+      "loss": 0.9064,
+      "step": 4476
+    },
+    {
+      "epoch": 0.8612928049249712,
+      "grad_norm": 2.3447325229644775,
+      "learning_rate": 9.922432581167207e-06,
+      "loss": 0.9933,
+      "step": 4477
+    },
+    {
+      "epoch": 0.8614851866102347,
+      "grad_norm": 3.059931516647339,
+      "learning_rate": 9.895390457939413e-06,
+      "loss": 0.9627,
+      "step": 4478
+    },
+    {
+      "epoch": 0.8616775682954982,
+      "grad_norm": 1.4842373132705688,
+      "learning_rate": 9.86838331642348e-06,
+      "loss": 0.9275,
+      "step": 4479
+    },
+    {
+      "epoch": 0.8618699499807618,
+      "grad_norm": 2.7123260498046875,
+      "learning_rate": 9.84141116710442e-06,
+      "loss": 0.9309,
+      "step": 4480
+    },
+    {
+      "epoch": 0.8620623316660254,
+      "grad_norm": 2.355614185333252,
+      "learning_rate": 9.814474020453822e-06,
+      "loss": 0.7986,
+      "step": 4481
+    },
+    {
+      "epoch": 0.862254713351289,
+      "grad_norm": 2.489675283432007,
+      "learning_rate": 9.787571886929603e-06,
+      "loss": 0.9686,
+      "step": 4482
+    },
+    {
+      "epoch": 0.8624470950365525,
+      "grad_norm": 2.136120557785034,
+      "learning_rate": 9.760704776976049e-06,
+      "loss": 0.8912,
+      "step": 4483
+    },
+    {
+      "epoch": 0.8626394767218161,
+      "grad_norm": 2.2709498405456543,
+      "learning_rate": 9.733872701023939e-06,
+      "loss": 0.9556,
+      "step": 4484
+    },
+    {
+      "epoch": 0.8628318584070797,
+      "grad_norm": 2.3334782123565674,
+      "learning_rate": 9.707075669490407e-06,
+      "loss": 0.9539,
+      "step": 4485
+    },
+    {
+      "epoch": 0.8630242400923432,
+      "grad_norm": 0.7851255536079407,
+      "learning_rate": 9.680313692778974e-06,
+      "loss": 0.8978,
+      "step": 4486
+    },
+    {
+      "epoch": 0.8632166217776067,
+      "grad_norm": 2.2792305946350098,
+      "learning_rate": 9.653586781279567e-06,
+      "loss": 0.8573,
+      "step": 4487
+    },
+    {
+      "epoch": 0.8634090034628703,
+      "grad_norm": 2.1684741973876953,
+      "learning_rate": 9.626894945368491e-06,
+      "loss": 0.8956,
+      "step": 4488
+    },
+    {
+      "epoch": 0.8636013851481339,
+      "grad_norm": 2.6171746253967285,
+      "learning_rate": 9.600238195408428e-06,
+      "loss": 0.9519,
+      "step": 4489
+    },
+    {
+      "epoch": 0.8637937668333975,
+      "grad_norm": 2.311100721359253,
+      "learning_rate": 9.573616541748464e-06,
+      "loss": 0.9133,
+      "step": 4490
+    },
+    {
+      "epoch": 0.8639861485186611,
+      "grad_norm": 1.819903016090393,
+      "learning_rate": 9.547029994724022e-06,
+      "loss": 0.8752,
+      "step": 4491
+    },
+    {
+      "epoch": 0.8641785302039245,
+      "grad_norm": 2.2713980674743652,
+      "learning_rate": 9.520478564656899e-06,
+      "loss": 0.8909,
+      "step": 4492
+    },
+    {
+      "epoch": 0.8643709118891881,
+      "grad_norm": 1.864542841911316,
+      "learning_rate": 9.493962261855349e-06,
+      "loss": 0.9256,
+      "step": 4493
+    },
+    {
+      "epoch": 0.8645632935744517,
+      "grad_norm": 2.828012704849243,
+      "learning_rate": 9.467481096613828e-06,
+      "loss": 0.9438,
+      "step": 4494
+    },
+    {
+      "epoch": 0.8647556752597153,
+      "grad_norm": 2.5122358798980713,
+      "learning_rate": 9.441035079213267e-06,
+      "loss": 1.0013,
+      "step": 4495
+    },
+    {
+      "epoch": 0.8649480569449788,
+      "grad_norm": 2.2682783603668213,
+      "learning_rate": 9.414624219920953e-06,
+      "loss": 0.918,
+      "step": 4496
+    },
+    {
+      "epoch": 0.8651404386302424,
+      "grad_norm": 2.3026223182678223,
+      "learning_rate": 9.388248528990428e-06,
+      "loss": 0.8826,
+      "step": 4497
+    },
+    {
+      "epoch": 0.865332820315506,
+      "grad_norm": 2.5408215522766113,
+      "learning_rate": 9.361908016661702e-06,
+      "loss": 0.9418,
+      "step": 4498
+    },
+    {
+      "epoch": 0.8655252020007695,
+      "grad_norm": 1.5720374584197998,
+      "learning_rate": 9.335602693161038e-06,
+      "loss": 0.7664,
+      "step": 4499
+    },
+    {
+      "epoch": 0.8657175836860331,
+      "grad_norm": 2.451186180114746,
+      "learning_rate": 9.309332568701079e-06,
+      "loss": 0.9614,
+      "step": 4500
+    },
+    {
+      "epoch": 0.8659099653712966,
+      "grad_norm": 2.0709896087646484,
+      "learning_rate": 9.283097653480789e-06,
+      "loss": 0.9228,
+      "step": 4501
+    },
+    {
+      "epoch": 0.8661023470565602,
+      "grad_norm": 1.9987996816635132,
+      "learning_rate": 9.256897957685462e-06,
+      "loss": 1.0061,
+      "step": 4502
+    },
+    {
+      "epoch": 0.8662947287418238,
+      "grad_norm": 2.4063539505004883,
+      "learning_rate": 9.23073349148672e-06,
+      "loss": 0.9233,
+      "step": 4503
+    },
+    {
+      "epoch": 0.8664871104270874,
+      "grad_norm": 2.667308807373047,
+      "learning_rate": 9.204604265042505e-06,
+      "loss": 0.9463,
+      "step": 4504
+    },
+    {
+      "epoch": 0.866679492112351,
+      "grad_norm": 1.4577813148498535,
+      "learning_rate": 9.178510288497122e-06,
+      "loss": 0.8949,
+      "step": 4505
+    },
+    {
+      "epoch": 0.8668718737976144,
+      "grad_norm": 2.263188362121582,
+      "learning_rate": 9.15245157198108e-06,
+      "loss": 0.9207,
+      "step": 4506
+    },
+    {
+      "epoch": 0.867064255482878,
+      "grad_norm": 2.2464070320129395,
+      "learning_rate": 9.126428125611342e-06,
+      "loss": 0.9563,
+      "step": 4507
+    },
+    {
+      "epoch": 0.8672566371681416,
+      "grad_norm": 2.231600284576416,
+      "learning_rate": 9.10043995949108e-06,
+      "loss": 0.8751,
+      "step": 4508
+    },
+    {
+      "epoch": 0.8674490188534052,
+      "grad_norm": 2.5917584896087646,
+      "learning_rate": 9.07448708370976e-06,
+      "loss": 0.9713,
+      "step": 4509
+    },
+    {
+      "epoch": 0.8676414005386687,
+      "grad_norm": 2.82987380027771,
+      "learning_rate": 9.04856950834323e-06,
+      "loss": 0.9898,
+      "step": 4510
+    },
+    {
+      "epoch": 0.8678337822239323,
+      "grad_norm": 2.870833396911621,
+      "learning_rate": 9.022687243453554e-06,
+      "loss": 0.9534,
+      "step": 4511
+    },
+    {
+      "epoch": 0.8680261639091958,
+      "grad_norm": 1.1718508005142212,
+      "learning_rate": 8.996840299089149e-06,
+      "loss": 0.9051,
+      "step": 4512
+    },
+    {
+      "epoch": 0.8682185455944594,
+      "grad_norm": 2.3164238929748535,
+      "learning_rate": 8.971028685284655e-06,
+      "loss": 0.9445,
+      "step": 4513
+    },
+    {
+      "epoch": 0.868410927279723,
+      "grad_norm": 2.283665657043457,
+      "learning_rate": 8.945252412061055e-06,
+      "loss": 0.9232,
+      "step": 4514
+    },
+    {
+      "epoch": 0.8686033089649865,
+      "grad_norm": 2.128082513809204,
+      "learning_rate": 8.91951148942557e-06,
+      "loss": 0.936,
+      "step": 4515
+    },
+    {
+      "epoch": 0.8687956906502501,
+      "grad_norm": 3.131608009338379,
+      "learning_rate": 8.893805927371723e-06,
+      "loss": 0.9574,
+      "step": 4516
+    },
+    {
+      "epoch": 0.8689880723355137,
+      "grad_norm": 1.840084195137024,
+      "learning_rate": 8.86813573587929e-06,
+      "loss": 0.9893,
+      "step": 4517
+    },
+    {
+      "epoch": 0.8691804540207773,
+      "grad_norm": 2.4804816246032715,
+      "learning_rate": 8.842500924914298e-06,
+      "loss": 1.0052,
+      "step": 4518
+    },
+    {
+      "epoch": 0.8693728357060407,
+      "grad_norm": 2.1668593883514404,
+      "learning_rate": 8.816901504429143e-06,
+      "loss": 0.9203,
+      "step": 4519
+    },
+    {
+      "epoch": 0.8695652173913043,
+      "grad_norm": 2.242377758026123,
+      "learning_rate": 8.791337484362306e-06,
+      "loss": 0.9171,
+      "step": 4520
+    },
+    {
+      "epoch": 0.8697575990765679,
+      "grad_norm": 1.2146891355514526,
+      "learning_rate": 8.765808874638681e-06,
+      "loss": 0.8928,
+      "step": 4521
+    },
+    {
+      "epoch": 0.8699499807618315,
+      "grad_norm": 0.9996099472045898,
+      "learning_rate": 8.740315685169365e-06,
+      "loss": 0.9469,
+      "step": 4522
+    },
+    {
+      "epoch": 0.8701423624470951,
+      "grad_norm": 2.0653700828552246,
+      "learning_rate": 8.714857925851615e-06,
+      "loss": 0.8769,
+      "step": 4523
+    },
+    {
+      "epoch": 0.8703347441323586,
+      "grad_norm": 2.2442550659179688,
+      "learning_rate": 8.689435606569085e-06,
+      "loss": 1.0045,
+      "step": 4524
+    },
+    {
+      "epoch": 0.8705271258176221,
+      "grad_norm": 2.3329036235809326,
+      "learning_rate": 8.664048737191566e-06,
+      "loss": 0.9677,
+      "step": 4525
+    },
+    {
+      "epoch": 0.8707195075028857,
+      "grad_norm": 2.4511120319366455,
+      "learning_rate": 8.638697327575108e-06,
+      "loss": 0.9783,
+      "step": 4526
+    },
+    {
+      "epoch": 0.8709118891881493,
+      "grad_norm": 2.302006483078003,
+      "learning_rate": 8.613381387562015e-06,
+      "loss": 0.969,
+      "step": 4527
+    },
+    {
+      "epoch": 0.8711042708734128,
+      "grad_norm": 2.4764974117279053,
+      "learning_rate": 8.588100926980802e-06,
+      "loss": 0.9249,
+      "step": 4528
+    },
+    {
+      "epoch": 0.8712966525586764,
+      "grad_norm": 2.919574737548828,
+      "learning_rate": 8.56285595564621e-06,
+      "loss": 0.9679,
+      "step": 4529
+    },
+    {
+      "epoch": 0.87148903424394,
+      "grad_norm": 1.987478256225586,
+      "learning_rate": 8.537646483359186e-06,
+      "loss": 0.83,
+      "step": 4530
+    },
+    {
+      "epoch": 0.8716814159292036,
+      "grad_norm": 2.148763656616211,
+      "learning_rate": 8.512472519906978e-06,
+      "loss": 0.9858,
+      "step": 4531
+    },
+    {
+      "epoch": 0.8718737976144671,
+      "grad_norm": 3.176190137863159,
+      "learning_rate": 8.487334075062913e-06,
+      "loss": 0.9882,
+      "step": 4532
+    },
+    {
+      "epoch": 0.8720661792997306,
+      "grad_norm": 1.351284146308899,
+      "learning_rate": 8.462231158586653e-06,
+      "loss": 0.9151,
+      "step": 4533
+    },
+    {
+      "epoch": 0.8722585609849942,
+      "grad_norm": 1.3206980228424072,
+      "learning_rate": 8.437163780224011e-06,
+      "loss": 0.7718,
+      "step": 4534
+    },
+    {
+      "epoch": 0.8724509426702578,
+      "grad_norm": 2.0250988006591797,
+      "learning_rate": 8.41213194970696e-06,
+      "loss": 1.0056,
+      "step": 4535
+    },
+    {
+      "epoch": 0.8726433243555214,
+      "grad_norm": 1.8703514337539673,
+      "learning_rate": 8.387135676753755e-06,
+      "loss": 0.9159,
+      "step": 4536
+    },
+    {
+      "epoch": 0.8728357060407849,
+      "grad_norm": 2.7341437339782715,
+      "learning_rate": 8.362174971068803e-06,
+      "loss": 0.9277,
+      "step": 4537
+    },
+    {
+      "epoch": 0.8730280877260485,
+      "grad_norm": 2.442427158355713,
+      "learning_rate": 8.33724984234272e-06,
+      "loss": 0.9121,
+      "step": 4538
+    },
+    {
+      "epoch": 0.873220469411312,
+      "grad_norm": 2.355283260345459,
+      "learning_rate": 8.312360300252286e-06,
+      "loss": 0.9293,
+      "step": 4539
+    },
+    {
+      "epoch": 0.8734128510965756,
+      "grad_norm": 1.492513656616211,
+      "learning_rate": 8.287506354460484e-06,
+      "loss": 0.9731,
+      "step": 4540
+    },
+    {
+      "epoch": 0.8736052327818392,
+      "grad_norm": 2.3618388175964355,
+      "learning_rate": 8.26268801461646e-06,
+      "loss": 0.9365,
+      "step": 4541
+    },
+    {
+      "epoch": 0.8737976144671027,
+      "grad_norm": 2.504058837890625,
+      "learning_rate": 8.237905290355564e-06,
+      "loss": 0.921,
+      "step": 4542
+    },
+    {
+      "epoch": 0.8739899961523663,
+      "grad_norm": 2.481947183609009,
+      "learning_rate": 8.213158191299297e-06,
+      "loss": 0.9342,
+      "step": 4543
+    },
+    {
+      "epoch": 0.8741823778376299,
+      "grad_norm": 2.4897425174713135,
+      "learning_rate": 8.18844672705531e-06,
+      "loss": 0.8847,
+      "step": 4544
+    },
+    {
+      "epoch": 0.8743747595228935,
+      "grad_norm": 2.3275575637817383,
+      "learning_rate": 8.163770907217505e-06,
+      "loss": 0.9592,
+      "step": 4545
+    },
+    {
+      "epoch": 0.8745671412081569,
+      "grad_norm": 2.56964373588562,
+      "learning_rate": 8.13913074136582e-06,
+      "loss": 0.9615,
+      "step": 4546
+    },
+    {
+      "epoch": 0.8747595228934205,
+      "grad_norm": 1.7673465013504028,
+      "learning_rate": 8.114526239066455e-06,
+      "loss": 0.9599,
+      "step": 4547
+    },
+    {
+      "epoch": 0.8749519045786841,
+      "grad_norm": 2.277331829071045,
+      "learning_rate": 8.08995740987173e-06,
+      "loss": 0.9277,
+      "step": 4548
+    },
+    {
+      "epoch": 0.8751442862639477,
+      "grad_norm": 2.333125352859497,
+      "learning_rate": 8.065424263320054e-06,
+      "loss": 0.9145,
+      "step": 4549
+    },
+    {
+      "epoch": 0.8753366679492113,
+      "grad_norm": 2.806680202484131,
+      "learning_rate": 8.040926808936111e-06,
+      "loss": 0.9157,
+      "step": 4550
+    },
+    {
+      "epoch": 0.8755290496344748,
+      "grad_norm": 2.167412757873535,
+      "learning_rate": 8.016465056230615e-06,
+      "loss": 0.9439,
+      "step": 4551
+    },
+    {
+      "epoch": 0.8757214313197383,
+      "grad_norm": 2.225499153137207,
+      "learning_rate": 7.99203901470047e-06,
+      "loss": 0.9335,
+      "step": 4552
+    },
+    {
+      "epoch": 0.8759138130050019,
+      "grad_norm": 1.9210277795791626,
+      "learning_rate": 7.967648693828712e-06,
+      "loss": 0.8657,
+      "step": 4553
+    },
+    {
+      "epoch": 0.8761061946902655,
+      "grad_norm": 2.025977611541748,
+      "learning_rate": 7.943294103084487e-06,
+      "loss": 0.8974,
+      "step": 4554
+    },
+    {
+      "epoch": 0.8762985763755291,
+      "grad_norm": 2.2583673000335693,
+      "learning_rate": 7.918975251923099e-06,
+      "loss": 0.8594,
+      "step": 4555
+    },
+    {
+      "epoch": 0.8764909580607926,
+      "grad_norm": 2.3103601932525635,
+      "learning_rate": 7.894692149785954e-06,
+      "loss": 0.8989,
+      "step": 4556
+    },
+    {
+      "epoch": 0.8766833397460562,
+      "grad_norm": 2.3751931190490723,
+      "learning_rate": 7.87044480610062e-06,
+      "loss": 0.9253,
+      "step": 4557
+    },
+    {
+      "epoch": 0.8768757214313198,
+      "grad_norm": 1.0219385623931885,
+      "learning_rate": 7.846233230280697e-06,
+      "loss": 0.9008,
+      "step": 4558
+    },
+    {
+      "epoch": 0.8770681031165833,
+      "grad_norm": 2.1455883979797363,
+      "learning_rate": 7.822057431725993e-06,
+      "loss": 0.9698,
+      "step": 4559
+    },
+    {
+      "epoch": 0.8772604848018468,
+      "grad_norm": 2.842714309692383,
+      "learning_rate": 7.797917419822375e-06,
+      "loss": 0.9333,
+      "step": 4560
+    },
+    {
+      "epoch": 0.8774528664871104,
+      "grad_norm": 1.5275874137878418,
+      "learning_rate": 7.773813203941826e-06,
+      "loss": 0.9549,
+      "step": 4561
+    },
+    {
+      "epoch": 0.877645248172374,
+      "grad_norm": 2.3856003284454346,
+      "learning_rate": 7.749744793442448e-06,
+      "loss": 0.9403,
+      "step": 4562
+    },
+    {
+      "epoch": 0.8778376298576376,
+      "grad_norm": 2.4556643962860107,
+      "learning_rate": 7.725712197668378e-06,
+      "loss": 0.8706,
+      "step": 4563
+    },
+    {
+      "epoch": 0.8780300115429012,
+      "grad_norm": 2.2390754222869873,
+      "learning_rate": 7.70171542594995e-06,
+      "loss": 0.9103,
+      "step": 4564
+    },
+    {
+      "epoch": 0.8782223932281646,
+      "grad_norm": 1.142562747001648,
+      "learning_rate": 7.677754487603517e-06,
+      "loss": 0.884,
+      "step": 4565
+    },
+    {
+      "epoch": 0.8784147749134282,
+      "grad_norm": 2.3427906036376953,
+      "learning_rate": 7.653829391931534e-06,
+      "loss": 0.7802,
+      "step": 4566
+    },
+    {
+      "epoch": 0.8786071565986918,
+      "grad_norm": 2.531890630722046,
+      "learning_rate": 7.629940148222559e-06,
+      "loss": 0.8455,
+      "step": 4567
+    },
+    {
+      "epoch": 0.8787995382839554,
+      "grad_norm": 1.2361706495285034,
+      "learning_rate": 7.606086765751208e-06,
+      "loss": 0.9094,
+      "step": 4568
+    },
+    {
+      "epoch": 0.8789919199692189,
+      "grad_norm": 2.4541423320770264,
+      "learning_rate": 7.582269253778185e-06,
+      "loss": 0.9845,
+      "step": 4569
+    },
+    {
+      "epoch": 0.8791843016544825,
+      "grad_norm": 2.7851109504699707,
+      "learning_rate": 7.5584876215502695e-06,
+      "loss": 0.9352,
+      "step": 4570
+    },
+    {
+      "epoch": 0.8793766833397461,
+      "grad_norm": 1.698155164718628,
+      "learning_rate": 7.534741878300333e-06,
+      "loss": 0.885,
+      "step": 4571
+    },
+    {
+      "epoch": 0.8795690650250096,
+      "grad_norm": 2.303974151611328,
+      "learning_rate": 7.511032033247256e-06,
+      "loss": 0.9312,
+      "step": 4572
+    },
+    {
+      "epoch": 0.8797614467102732,
+      "grad_norm": 1.8082711696624756,
+      "learning_rate": 7.487358095596031e-06,
+      "loss": 1.0267,
+      "step": 4573
+    },
+    {
+      "epoch": 0.8799538283955367,
+      "grad_norm": 2.6159114837646484,
+      "learning_rate": 7.463720074537728e-06,
+      "loss": 0.9821,
+      "step": 4574
+    },
+    {
+      "epoch": 0.8801462100808003,
+      "grad_norm": 2.0157527923583984,
+      "learning_rate": 7.440117979249362e-06,
+      "loss": 0.9531,
+      "step": 4575
+    },
+    {
+      "epoch": 0.8803385917660639,
+      "grad_norm": 2.350834846496582,
+      "learning_rate": 7.416551818894157e-06,
+      "loss": 0.9765,
+      "step": 4576
+    },
+    {
+      "epoch": 0.8805309734513275,
+      "grad_norm": 2.2355549335479736,
+      "learning_rate": 7.393021602621264e-06,
+      "loss": 0.9355,
+      "step": 4577
+    },
+    {
+      "epoch": 0.880723355136591,
+      "grad_norm": 2.600553512573242,
+      "learning_rate": 7.3695273395659515e-06,
+      "loss": 0.9385,
+      "step": 4578
+    },
+    {
+      "epoch": 0.8809157368218545,
+      "grad_norm": 2.6584839820861816,
+      "learning_rate": 7.346069038849468e-06,
+      "loss": 0.9054,
+      "step": 4579
+    },
+    {
+      "epoch": 0.8811081185071181,
+      "grad_norm": 2.306091547012329,
+      "learning_rate": 7.322646709579173e-06,
+      "loss": 0.8807,
+      "step": 4580
+    },
+    {
+      "epoch": 0.8813005001923817,
+      "grad_norm": 2.133676052093506,
+      "learning_rate": 7.299260360848381e-06,
+      "loss": 0.9483,
+      "step": 4581
+    },
+    {
+      "epoch": 0.8814928818776453,
+      "grad_norm": 2.6830270290374756,
+      "learning_rate": 7.275910001736497e-06,
+      "loss": 0.9391,
+      "step": 4582
+    },
+    {
+      "epoch": 0.8816852635629088,
+      "grad_norm": 2.873530149459839,
+      "learning_rate": 7.2525956413089565e-06,
+      "loss": 1.0035,
+      "step": 4583
+    },
+    {
+      "epoch": 0.8818776452481724,
+      "grad_norm": 2.2993717193603516,
+      "learning_rate": 7.2293172886171435e-06,
+      "loss": 0.9105,
+      "step": 4584
+    },
+    {
+      "epoch": 0.882070026933436,
+      "grad_norm": 1.8310467004776,
+      "learning_rate": 7.206074952698561e-06,
+      "loss": 0.8455,
+      "step": 4585
+    },
+    {
+      "epoch": 0.8822624086186995,
+      "grad_norm": 1.1084327697753906,
+      "learning_rate": 7.1828686425766785e-06,
+      "loss": 0.8976,
+      "step": 4586
+    },
+    {
+      "epoch": 0.882454790303963,
+      "grad_norm": 2.5977439880371094,
+      "learning_rate": 7.15969836726097e-06,
+      "loss": 0.9192,
+      "step": 4587
+    },
+    {
+      "epoch": 0.8826471719892266,
+      "grad_norm": 1.1536214351654053,
+      "learning_rate": 7.136564135746959e-06,
+      "loss": 0.8916,
+      "step": 4588
+    },
+    {
+      "epoch": 0.8828395536744902,
+      "grad_norm": 2.268746852874756,
+      "learning_rate": 7.113465957016097e-06,
+      "loss": 0.9626,
+      "step": 4589
+    },
+    {
+      "epoch": 0.8830319353597538,
+      "grad_norm": 2.230236291885376,
+      "learning_rate": 7.090403840035942e-06,
+      "loss": 0.9254,
+      "step": 4590
+    },
+    {
+      "epoch": 0.8832243170450174,
+      "grad_norm": 2.9058539867401123,
+      "learning_rate": 7.067377793759999e-06,
+      "loss": 0.9237,
+      "step": 4591
+    },
+    {
+      "epoch": 0.8834166987302808,
+      "grad_norm": 2.3888657093048096,
+      "learning_rate": 7.044387827127752e-06,
+      "loss": 0.9372,
+      "step": 4592
+    },
+    {
+      "epoch": 0.8836090804155444,
+      "grad_norm": 2.5279276371002197,
+      "learning_rate": 7.0214339490647045e-06,
+      "loss": 0.9052,
+      "step": 4593
+    },
+    {
+      "epoch": 0.883801462100808,
+      "grad_norm": 2.393483877182007,
+      "learning_rate": 6.99851616848235e-06,
+      "loss": 0.9141,
+      "step": 4594
+    },
+    {
+      "epoch": 0.8839938437860716,
+      "grad_norm": 1.6774965524673462,
+      "learning_rate": 6.975634494278149e-06,
+      "loss": 0.8848,
+      "step": 4595
+    },
+    {
+      "epoch": 0.8841862254713351,
+      "grad_norm": 2.4659311771392822,
+      "learning_rate": 6.952788935335541e-06,
+      "loss": 0.8989,
+      "step": 4596
+    },
+    {
+      "epoch": 0.8843786071565987,
+      "grad_norm": 1.898004412651062,
+      "learning_rate": 6.92997950052402e-06,
+      "loss": 0.9456,
+      "step": 4597
+    },
+    {
+      "epoch": 0.8845709888418622,
+      "grad_norm": 2.3238508701324463,
+      "learning_rate": 6.907206198698913e-06,
+      "loss": 0.939,
+      "step": 4598
+    },
+    {
+      "epoch": 0.8847633705271258,
+      "grad_norm": 2.5695009231567383,
+      "learning_rate": 6.884469038701646e-06,
+      "loss": 0.9269,
+      "step": 4599
+    },
+    {
+      "epoch": 0.8849557522123894,
+      "grad_norm": 2.516818046569824,
+      "learning_rate": 6.861768029359594e-06,
+      "loss": 0.9237,
+      "step": 4600
+    },
+    {
+      "epoch": 0.8851481338976529,
+      "grad_norm": 1.753456473350525,
+      "learning_rate": 6.839103179485995e-06,
+      "loss": 0.9733,
+      "step": 4601
+    },
+    {
+      "epoch": 0.8853405155829165,
+      "grad_norm": 2.4091827869415283,
+      "learning_rate": 6.8164744978801765e-06,
+      "loss": 0.9364,
+      "step": 4602
+    },
+    {
+      "epoch": 0.8855328972681801,
+      "grad_norm": 2.3648312091827393,
+      "learning_rate": 6.793881993327367e-06,
+      "loss": 0.9282,
+      "step": 4603
+    },
+    {
+      "epoch": 0.8857252789534437,
+      "grad_norm": 2.2656002044677734,
+      "learning_rate": 6.771325674598749e-06,
+      "loss": 0.8956,
+      "step": 4604
+    },
+    {
+      "epoch": 0.8859176606387072,
+      "grad_norm": 2.4713385105133057,
+      "learning_rate": 6.748805550451454e-06,
+      "loss": 0.9445,
+      "step": 4605
+    },
+    {
+      "epoch": 0.8861100423239707,
+      "grad_norm": 1.9033927917480469,
+      "learning_rate": 6.726321629628585e-06,
+      "loss": 0.9859,
+      "step": 4606
+    },
+    {
+      "epoch": 0.8863024240092343,
+      "grad_norm": 2.286325693130493,
+      "learning_rate": 6.7038739208591605e-06,
+      "loss": 0.9008,
+      "step": 4607
+    },
+    {
+      "epoch": 0.8864948056944979,
+      "grad_norm": 2.030144691467285,
+      "learning_rate": 6.681462432858154e-06,
+      "loss": 0.9573,
+      "step": 4608
+    },
+    {
+      "epoch": 0.8866871873797615,
+      "grad_norm": 1.30076265335083,
+      "learning_rate": 6.659087174326506e-06,
+      "loss": 0.7715,
+      "step": 4609
+    },
+    {
+      "epoch": 0.886879569065025,
+      "grad_norm": 2.3072502613067627,
+      "learning_rate": 6.636748153951e-06,
+      "loss": 0.9278,
+      "step": 4610
+    },
+    {
+      "epoch": 0.8870719507502886,
+      "grad_norm": 1.2268966436386108,
+      "learning_rate": 6.614445380404477e-06,
+      "loss": 0.7717,
+      "step": 4611
+    },
+    {
+      "epoch": 0.8872643324355521,
+      "grad_norm": 2.3069519996643066,
+      "learning_rate": 6.5921788623456216e-06,
+      "loss": 0.8568,
+      "step": 4612
+    },
+    {
+      "epoch": 0.8874567141208157,
+      "grad_norm": 1.828891396522522,
+      "learning_rate": 6.569948608419041e-06,
+      "loss": 0.9429,
+      "step": 4613
+    },
+    {
+      "epoch": 0.8876490958060793,
+      "grad_norm": 0.8480768203735352,
+      "learning_rate": 6.547754627255331e-06,
+      "loss": 0.917,
+      "step": 4614
+    },
+    {
+      "epoch": 0.8878414774913428,
+      "grad_norm": 2.4693665504455566,
+      "learning_rate": 6.5255969274708896e-06,
+      "loss": 0.9453,
+      "step": 4615
+    },
+    {
+      "epoch": 0.8880338591766064,
+      "grad_norm": 2.8525123596191406,
+      "learning_rate": 6.503475517668167e-06,
+      "loss": 0.9404,
+      "step": 4616
+    },
+    {
+      "epoch": 0.88822624086187,
+      "grad_norm": 2.1710045337677,
+      "learning_rate": 6.481390406435417e-06,
+      "loss": 0.9195,
+      "step": 4617
+    },
+    {
+      "epoch": 0.8884186225471336,
+      "grad_norm": 2.6109776496887207,
+      "learning_rate": 6.4593416023468575e-06,
+      "loss": 0.9452,
+      "step": 4618
+    },
+    {
+      "epoch": 0.888611004232397,
+      "grad_norm": 0.9743319749832153,
+      "learning_rate": 6.437329113962576e-06,
+      "loss": 0.7619,
+      "step": 4619
+    },
+    {
+      "epoch": 0.8888033859176606,
+      "grad_norm": 1.9476319551467896,
+      "learning_rate": 6.4153529498286014e-06,
+      "loss": 0.9426,
+      "step": 4620
+    },
+    {
+      "epoch": 0.8889957676029242,
+      "grad_norm": 2.002272367477417,
+      "learning_rate": 6.393413118476821e-06,
+      "loss": 0.9417,
+      "step": 4621
+    },
+    {
+      "epoch": 0.8891881492881878,
+      "grad_norm": 2.228262186050415,
+      "learning_rate": 6.3715096284250205e-06,
+      "loss": 0.9411,
+      "step": 4622
+    },
+    {
+      "epoch": 0.8893805309734514,
+      "grad_norm": 2.584094285964966,
+      "learning_rate": 6.349642488176944e-06,
+      "loss": 0.9445,
+      "step": 4623
+    },
+    {
+      "epoch": 0.8895729126587149,
+      "grad_norm": 2.425111770629883,
+      "learning_rate": 6.327811706222098e-06,
+      "loss": 0.9589,
+      "step": 4624
+    },
+    {
+      "epoch": 0.8897652943439784,
+      "grad_norm": 2.4013679027557373,
+      "learning_rate": 6.3060172910359815e-06,
+      "loss": 0.9479,
+      "step": 4625
+    },
+    {
+      "epoch": 0.889957676029242,
+      "grad_norm": 2.3803131580352783,
+      "learning_rate": 6.284259251079938e-06,
+      "loss": 0.8589,
+      "step": 4626
+    },
+    {
+      "epoch": 0.8901500577145056,
+      "grad_norm": 1.7019749879837036,
+      "learning_rate": 6.262537594801177e-06,
+      "loss": 0.9225,
+      "step": 4627
+    },
+    {
+      "epoch": 0.8903424393997691,
+      "grad_norm": 2.442537307739258,
+      "learning_rate": 6.240852330632796e-06,
+      "loss": 0.9009,
+      "step": 4628
+    },
+    {
+      "epoch": 0.8905348210850327,
+      "grad_norm": 2.3495864868164062,
+      "learning_rate": 6.219203466993761e-06,
+      "loss": 0.9363,
+      "step": 4629
+    },
+    {
+      "epoch": 0.8907272027702963,
+      "grad_norm": 2.2748258113861084,
+      "learning_rate": 6.197591012288917e-06,
+      "loss": 0.9168,
+      "step": 4630
+    },
+    {
+      "epoch": 0.8909195844555599,
+      "grad_norm": 2.033209800720215,
+      "learning_rate": 6.17601497490895e-06,
+      "loss": 0.8901,
+      "step": 4631
+    },
+    {
+      "epoch": 0.8911119661408234,
+      "grad_norm": 2.4090476036071777,
+      "learning_rate": 6.1544753632304165e-06,
+      "loss": 0.9832,
+      "step": 4632
+    },
+    {
+      "epoch": 0.8913043478260869,
+      "grad_norm": 2.566701889038086,
+      "learning_rate": 6.132972185615749e-06,
+      "loss": 0.9474,
+      "step": 4633
+    },
+    {
+      "epoch": 0.8914967295113505,
+      "grad_norm": 1.867811679840088,
+      "learning_rate": 6.111505450413202e-06,
+      "loss": 0.9923,
+      "step": 4634
+    },
+    {
+      "epoch": 0.8916891111966141,
+      "grad_norm": 0.9578442573547363,
+      "learning_rate": 6.090075165956943e-06,
+      "loss": 0.9295,
+      "step": 4635
+    },
+    {
+      "epoch": 0.8918814928818777,
+      "grad_norm": 2.2916524410247803,
+      "learning_rate": 6.068681340566895e-06,
+      "loss": 0.8625,
+      "step": 4636
+    },
+    {
+      "epoch": 0.8920738745671412,
+      "grad_norm": 2.6651768684387207,
+      "learning_rate": 6.0473239825489244e-06,
+      "loss": 0.9488,
+      "step": 4637
+    },
+    {
+      "epoch": 0.8922662562524047,
+      "grad_norm": 2.368232011795044,
+      "learning_rate": 6.026003100194632e-06,
+      "loss": 0.9784,
+      "step": 4638
+    },
+    {
+      "epoch": 0.8924586379376683,
+      "grad_norm": 2.7270100116729736,
+      "learning_rate": 6.004718701781575e-06,
+      "loss": 0.9202,
+      "step": 4639
+    },
+    {
+      "epoch": 0.8926510196229319,
+      "grad_norm": 2.3074958324432373,
+      "learning_rate": 5.983470795573088e-06,
+      "loss": 0.9445,
+      "step": 4640
+    },
+    {
+      "epoch": 0.8928434013081955,
+      "grad_norm": 2.596534013748169,
+      "learning_rate": 5.962259389818292e-06,
+      "loss": 0.914,
+      "step": 4641
+    },
+    {
+      "epoch": 0.893035782993459,
+      "grad_norm": 1.4798157215118408,
+      "learning_rate": 5.941084492752235e-06,
+      "loss": 0.7956,
+      "step": 4642
+    },
+    {
+      "epoch": 0.8932281646787226,
+      "grad_norm": 2.2729697227478027,
+      "learning_rate": 5.91994611259572e-06,
+      "loss": 0.9236,
+      "step": 4643
+    },
+    {
+      "epoch": 0.8934205463639862,
+      "grad_norm": 2.3929994106292725,
+      "learning_rate": 5.898844257555391e-06,
+      "loss": 0.9726,
+      "step": 4644
+    },
+    {
+      "epoch": 0.8936129280492497,
+      "grad_norm": 2.884836435317993,
+      "learning_rate": 5.877778935823719e-06,
+      "loss": 0.8934,
+      "step": 4645
+    },
+    {
+      "epoch": 0.8938053097345132,
+      "grad_norm": 1.9811251163482666,
+      "learning_rate": 5.856750155578983e-06,
+      "loss": 0.9418,
+      "step": 4646
+    },
+    {
+      "epoch": 0.8939976914197768,
+      "grad_norm": 2.265118360519409,
+      "learning_rate": 5.835757924985286e-06,
+      "loss": 0.9654,
+      "step": 4647
+    },
+    {
+      "epoch": 0.8941900731050404,
+      "grad_norm": 2.5621795654296875,
+      "learning_rate": 5.8148022521925195e-06,
+      "loss": 0.9237,
+      "step": 4648
+    },
+    {
+      "epoch": 0.894382454790304,
+      "grad_norm": 2.034226655960083,
+      "learning_rate": 5.793883145336443e-06,
+      "loss": 0.9083,
+      "step": 4649
+    },
+    {
+      "epoch": 0.8945748364755676,
+      "grad_norm": 2.649937629699707,
+      "learning_rate": 5.773000612538504e-06,
+      "loss": 0.9896,
+      "step": 4650
+    },
+    {
+      "epoch": 0.894767218160831,
+      "grad_norm": 2.7265844345092773,
+      "learning_rate": 5.752154661906084e-06,
+      "loss": 0.9523,
+      "step": 4651
+    },
+    {
+      "epoch": 0.8949595998460946,
+      "grad_norm": 2.3456053733825684,
+      "learning_rate": 5.731345301532265e-06,
+      "loss": 0.9337,
+      "step": 4652
+    },
+    {
+      "epoch": 0.8951519815313582,
+      "grad_norm": 2.092815637588501,
+      "learning_rate": 5.710572539495962e-06,
+      "loss": 0.8792,
+      "step": 4653
+    },
+    {
+      "epoch": 0.8953443632166218,
+      "grad_norm": 2.142465114593506,
+      "learning_rate": 5.689836383861879e-06,
+      "loss": 0.9757,
+      "step": 4654
+    },
+    {
+      "epoch": 0.8955367449018854,
+      "grad_norm": 2.181342363357544,
+      "learning_rate": 5.6691368426805114e-06,
+      "loss": 0.9615,
+      "step": 4655
+    },
+    {
+      "epoch": 0.8957291265871489,
+      "grad_norm": 2.2682511806488037,
+      "learning_rate": 5.648473923988129e-06,
+      "loss": 0.8782,
+      "step": 4656
+    },
+    {
+      "epoch": 0.8959215082724125,
+      "grad_norm": 1.087551236152649,
+      "learning_rate": 5.627847635806771e-06,
+      "loss": 0.881,
+      "step": 4657
+    },
+    {
+      "epoch": 0.896113889957676,
+      "grad_norm": 2.3606317043304443,
+      "learning_rate": 5.607257986144321e-06,
+      "loss": 0.8972,
+      "step": 4658
+    },
+    {
+      "epoch": 0.8963062716429396,
+      "grad_norm": 1.9887170791625977,
+      "learning_rate": 5.5867049829943395e-06,
+      "loss": 0.9531,
+      "step": 4659
+    },
+    {
+      "epoch": 0.8964986533282031,
+      "grad_norm": 2.4711573123931885,
+      "learning_rate": 5.566188634336211e-06,
+      "loss": 0.9193,
+      "step": 4660
+    },
+    {
+      "epoch": 0.8966910350134667,
+      "grad_norm": 2.7099266052246094,
+      "learning_rate": 5.545708948135142e-06,
+      "loss": 0.9423,
+      "step": 4661
+    },
+    {
+      "epoch": 0.8968834166987303,
+      "grad_norm": 2.3589389324188232,
+      "learning_rate": 5.525265932341983e-06,
+      "loss": 0.9395,
+      "step": 4662
+    },
+    {
+      "epoch": 0.8970757983839939,
+      "grad_norm": 2.214674472808838,
+      "learning_rate": 5.504859594893474e-06,
+      "loss": 0.9744,
+      "step": 4663
+    },
+    {
+      "epoch": 0.8972681800692575,
+      "grad_norm": 2.393026351928711,
+      "learning_rate": 5.484489943712012e-06,
+      "loss": 0.8477,
+      "step": 4664
+    },
+    {
+      "epoch": 0.8974605617545209,
+      "grad_norm": 2.322995662689209,
+      "learning_rate": 5.464156986705826e-06,
+      "loss": 0.9573,
+      "step": 4665
+    },
+    {
+      "epoch": 0.8976529434397845,
+      "grad_norm": 1.1991783380508423,
+      "learning_rate": 5.443860731768868e-06,
+      "loss": 0.9808,
+      "step": 4666
+    },
+    {
+      "epoch": 0.8978453251250481,
+      "grad_norm": 2.122305393218994,
+      "learning_rate": 5.4236011867808355e-06,
+      "loss": 0.9512,
+      "step": 4667
+    },
+    {
+      "epoch": 0.8980377068103117,
+      "grad_norm": 2.2125654220581055,
+      "learning_rate": 5.403378359607181e-06,
+      "loss": 0.948,
+      "step": 4668
+    },
+    {
+      "epoch": 0.8982300884955752,
+      "grad_norm": 2.127154588699341,
+      "learning_rate": 5.383192258099112e-06,
+      "loss": 0.9523,
+      "step": 4669
+    },
+    {
+      "epoch": 0.8984224701808388,
+      "grad_norm": 2.2564857006073,
+      "learning_rate": 5.36304289009355e-06,
+      "loss": 0.8814,
+      "step": 4670
+    },
+    {
+      "epoch": 0.8986148518661023,
+      "grad_norm": 2.522014617919922,
+      "learning_rate": 5.342930263413193e-06,
+      "loss": 0.8757,
+      "step": 4671
+    },
+    {
+      "epoch": 0.8988072335513659,
+      "grad_norm": 1.2579835653305054,
+      "learning_rate": 5.322854385866438e-06,
+      "loss": 0.8797,
+      "step": 4672
+    },
+    {
+      "epoch": 0.8989996152366295,
+      "grad_norm": 1.2660350799560547,
+      "learning_rate": 5.3028152652474514e-06,
+      "loss": 0.9846,
+      "step": 4673
+    },
+    {
+      "epoch": 0.899191996921893,
+      "grad_norm": 2.148149013519287,
+      "learning_rate": 5.282812909336077e-06,
+      "loss": 0.8597,
+      "step": 4674
+    },
+    {
+      "epoch": 0.8993843786071566,
+      "grad_norm": 0.8170333504676819,
+      "learning_rate": 5.262847325897968e-06,
+      "loss": 0.8996,
+      "step": 4675
+    },
+    {
+      "epoch": 0.8995767602924202,
+      "grad_norm": 2.5525619983673096,
+      "learning_rate": 5.242918522684393e-06,
+      "loss": 0.9442,
+      "step": 4676
+    },
+    {
+      "epoch": 0.8997691419776838,
+      "grad_norm": 1.9026066064834595,
+      "learning_rate": 5.223026507432449e-06,
+      "loss": 0.9647,
+      "step": 4677
+    },
+    {
+      "epoch": 0.8999615236629472,
+      "grad_norm": 2.2992000579833984,
+      "learning_rate": 5.203171287864872e-06,
+      "loss": 0.9293,
+      "step": 4678
+    },
+    {
+      "epoch": 0.9001539053482108,
+      "grad_norm": 1.4033225774765015,
+      "learning_rate": 5.183352871690162e-06,
+      "loss": 0.8517,
+      "step": 4679
+    },
+    {
+      "epoch": 0.9003462870334744,
+      "grad_norm": 2.5606417655944824,
+      "learning_rate": 5.163571266602485e-06,
+      "loss": 0.9219,
+      "step": 4680
+    },
+    {
+      "epoch": 0.900538668718738,
+      "grad_norm": 2.326183557510376,
+      "learning_rate": 5.143826480281777e-06,
+      "loss": 0.8902,
+      "step": 4681
+    },
+    {
+      "epoch": 0.9007310504040016,
+      "grad_norm": 2.1615984439849854,
+      "learning_rate": 5.124118520393606e-06,
+      "loss": 0.9704,
+      "step": 4682
+    },
+    {
+      "epoch": 0.9009234320892651,
+      "grad_norm": 2.3686275482177734,
+      "learning_rate": 5.104447394589296e-06,
+      "loss": 0.8913,
+      "step": 4683
+    },
+    {
+      "epoch": 0.9011158137745287,
+      "grad_norm": 1.6360557079315186,
+      "learning_rate": 5.0848131105058704e-06,
+      "loss": 0.9695,
+      "step": 4684
+    },
+    {
+      "epoch": 0.9013081954597922,
+      "grad_norm": 2.2877213954925537,
+      "learning_rate": 5.0652156757660215e-06,
+      "loss": 0.9439,
+      "step": 4685
+    },
+    {
+      "epoch": 0.9015005771450558,
+      "grad_norm": 2.7830121517181396,
+      "learning_rate": 5.045655097978131e-06,
+      "loss": 0.9318,
+      "step": 4686
+    },
+    {
+      "epoch": 0.9016929588303193,
+      "grad_norm": 3.013922691345215,
+      "learning_rate": 5.026131384736321e-06,
+      "loss": 0.9543,
+      "step": 4687
+    },
+    {
+      "epoch": 0.9018853405155829,
+      "grad_norm": 2.195812225341797,
+      "learning_rate": 5.006644543620342e-06,
+      "loss": 0.9944,
+      "step": 4688
+    },
+    {
+      "epoch": 0.9020777222008465,
+      "grad_norm": 2.3121249675750732,
+      "learning_rate": 4.987194582195687e-06,
+      "loss": 0.9143,
+      "step": 4689
+    },
+    {
+      "epoch": 0.9022701038861101,
+      "grad_norm": 2.634197473526001,
+      "learning_rate": 4.967781508013458e-06,
+      "loss": 0.9283,
+      "step": 4690
+    },
+    {
+      "epoch": 0.9024624855713737,
+      "grad_norm": 1.7333359718322754,
+      "learning_rate": 4.9484053286105055e-06,
+      "loss": 0.9795,
+      "step": 4691
+    },
+    {
+      "epoch": 0.9026548672566371,
+      "grad_norm": 2.2729575634002686,
+      "learning_rate": 4.929066051509346e-06,
+      "loss": 1.0049,
+      "step": 4692
+    },
+    {
+      "epoch": 0.9028472489419007,
+      "grad_norm": 2.4983315467834473,
+      "learning_rate": 4.909763684218116e-06,
+      "loss": 0.8573,
+      "step": 4693
+    },
+    {
+      "epoch": 0.9030396306271643,
+      "grad_norm": 1.751395583152771,
+      "learning_rate": 4.89049823423069e-06,
+      "loss": 0.9363,
+      "step": 4694
+    },
+    {
+      "epoch": 0.9032320123124279,
+      "grad_norm": 2.171529769897461,
+      "learning_rate": 4.87126970902656e-06,
+      "loss": 0.9425,
+      "step": 4695
+    },
+    {
+      "epoch": 0.9034243939976914,
+      "grad_norm": 2.6281232833862305,
+      "learning_rate": 4.852078116070902e-06,
+      "loss": 0.8444,
+      "step": 4696
+    },
+    {
+      "epoch": 0.903616775682955,
+      "grad_norm": 1.7363502979278564,
+      "learning_rate": 4.832923462814565e-06,
+      "loss": 0.941,
+      "step": 4697
+    },
+    {
+      "epoch": 0.9038091573682185,
+      "grad_norm": 2.5878067016601562,
+      "learning_rate": 4.813805756694034e-06,
+      "loss": 0.9201,
+      "step": 4698
+    },
+    {
+      "epoch": 0.9040015390534821,
+      "grad_norm": 2.907254934310913,
+      "learning_rate": 4.794725005131462e-06,
+      "loss": 0.9245,
+      "step": 4699
+    },
+    {
+      "epoch": 0.9041939207387457,
+      "grad_norm": 2.689049005508423,
+      "learning_rate": 4.775681215534656e-06,
+      "loss": 0.9167,
+      "step": 4700
+    },
+    {
+      "epoch": 0.9043863024240092,
+      "grad_norm": 2.9908101558685303,
+      "learning_rate": 4.756674395297089e-06,
+      "loss": 0.9104,
+      "step": 4701
+    },
+    {
+      "epoch": 0.9045786841092728,
+      "grad_norm": 1.3902076482772827,
+      "learning_rate": 4.737704551797817e-06,
+      "loss": 0.726,
+      "step": 4702
+    },
+    {
+      "epoch": 0.9047710657945364,
+      "grad_norm": 2.441714286804199,
+      "learning_rate": 4.718771692401636e-06,
+      "loss": 0.9571,
+      "step": 4703
+    },
+    {
+      "epoch": 0.9049634474798,
+      "grad_norm": 2.527742624282837,
+      "learning_rate": 4.699875824458899e-06,
+      "loss": 0.8995,
+      "step": 4704
+    },
+    {
+      "epoch": 0.9051558291650635,
+      "grad_norm": 2.4551117420196533,
+      "learning_rate": 4.6810169553056615e-06,
+      "loss": 0.8309,
+      "step": 4705
+    },
+    {
+      "epoch": 0.905348210850327,
+      "grad_norm": 1.5882192850112915,
+      "learning_rate": 4.662195092263566e-06,
+      "loss": 0.899,
+      "step": 4706
+    },
+    {
+      "epoch": 0.9055405925355906,
+      "grad_norm": 2.5693013668060303,
+      "learning_rate": 4.643410242639911e-06,
+      "loss": 0.9198,
+      "step": 4707
+    },
+    {
+      "epoch": 0.9057329742208542,
+      "grad_norm": 1.6890672445297241,
+      "learning_rate": 4.624662413727621e-06,
+      "loss": 0.9839,
+      "step": 4708
+    },
+    {
+      "epoch": 0.9059253559061178,
+      "grad_norm": 2.106248617172241,
+      "learning_rate": 4.605951612805237e-06,
+      "loss": 0.9306,
+      "step": 4709
+    },
+    {
+      "epoch": 0.9061177375913813,
+      "grad_norm": 2.3968894481658936,
+      "learning_rate": 4.587277847136984e-06,
+      "loss": 0.8996,
+      "step": 4710
+    },
+    {
+      "epoch": 0.9063101192766448,
+      "grad_norm": 1.8357685804367065,
+      "learning_rate": 4.5686411239726055e-06,
+      "loss": 0.9547,
+      "step": 4711
+    },
+    {
+      "epoch": 0.9065025009619084,
+      "grad_norm": 2.6064836978912354,
+      "learning_rate": 4.550041450547548e-06,
+      "loss": 0.9199,
+      "step": 4712
+    },
+    {
+      "epoch": 0.906694882647172,
+      "grad_norm": 2.763138771057129,
+      "learning_rate": 4.531478834082836e-06,
+      "loss": 0.9326,
+      "step": 4713
+    },
+    {
+      "epoch": 0.9068872643324356,
+      "grad_norm": 2.321904420852661,
+      "learning_rate": 4.512953281785104e-06,
+      "loss": 0.9455,
+      "step": 4714
+    },
+    {
+      "epoch": 0.9070796460176991,
+      "grad_norm": 2.1747419834136963,
+      "learning_rate": 4.494464800846654e-06,
+      "loss": 0.9129,
+      "step": 4715
+    },
+    {
+      "epoch": 0.9072720277029627,
+      "grad_norm": 1.4032576084136963,
+      "learning_rate": 4.476013398445289e-06,
+      "loss": 0.8491,
+      "step": 4716
+    },
+    {
+      "epoch": 0.9074644093882263,
+      "grad_norm": 1.8318302631378174,
+      "learning_rate": 4.457599081744523e-06,
+      "loss": 1.0241,
+      "step": 4717
+    },
+    {
+      "epoch": 0.9076567910734898,
+      "grad_norm": 2.2286744117736816,
+      "learning_rate": 4.439221857893416e-06,
+      "loss": 0.963,
+      "step": 4718
+    },
+    {
+      "epoch": 0.9078491727587533,
+      "grad_norm": 2.3406221866607666,
+      "learning_rate": 4.420881734026639e-06,
+      "loss": 0.9565,
+      "step": 4719
+    },
+    {
+      "epoch": 0.9080415544440169,
+      "grad_norm": 2.5359816551208496,
+      "learning_rate": 4.402578717264449e-06,
+      "loss": 0.9881,
+      "step": 4720
+    },
+    {
+      "epoch": 0.9082339361292805,
+      "grad_norm": 2.99497652053833,
+      "learning_rate": 4.384312814712721e-06,
+      "loss": 0.8965,
+      "step": 4721
+    },
+    {
+      "epoch": 0.9084263178145441,
+      "grad_norm": 1.2814637422561646,
+      "learning_rate": 4.366084033462914e-06,
+      "loss": 0.9575,
+      "step": 4722
+    },
+    {
+      "epoch": 0.9086186994998077,
+      "grad_norm": 2.498748540878296,
+      "learning_rate": 4.347892380592034e-06,
+      "loss": 0.9348,
+      "step": 4723
+    },
+    {
+      "epoch": 0.9088110811850711,
+      "grad_norm": 1.5467443466186523,
+      "learning_rate": 4.329737863162752e-06,
+      "loss": 0.9073,
+      "step": 4724
+    },
+    {
+      "epoch": 0.9090034628703347,
+      "grad_norm": 1.4779081344604492,
+      "learning_rate": 4.311620488223256e-06,
+      "loss": 0.8763,
+      "step": 4725
+    },
+    {
+      "epoch": 0.9091958445555983,
+      "grad_norm": 1.5459269285202026,
+      "learning_rate": 4.293540262807317e-06,
+      "loss": 0.9648,
+      "step": 4726
+    },
+    {
+      "epoch": 0.9093882262408619,
+      "grad_norm": 2.6886167526245117,
+      "learning_rate": 4.27549719393433e-06,
+      "loss": 0.9352,
+      "step": 4727
+    },
+    {
+      "epoch": 0.9095806079261254,
+      "grad_norm": 2.898974657058716,
+      "learning_rate": 4.257491288609217e-06,
+      "loss": 0.9192,
+      "step": 4728
+    },
+    {
+      "epoch": 0.909772989611389,
+      "grad_norm": 2.9750025272369385,
+      "learning_rate": 4.239522553822495e-06,
+      "loss": 0.9644,
+      "step": 4729
+    },
+    {
+      "epoch": 0.9099653712966526,
+      "grad_norm": 2.2272541522979736,
+      "learning_rate": 4.221590996550251e-06,
+      "loss": 0.8943,
+      "step": 4730
+    },
+    {
+      "epoch": 0.9101577529819161,
+      "grad_norm": 1.728391170501709,
+      "learning_rate": 4.203696623754138e-06,
+      "loss": 0.9017,
+      "step": 4731
+    },
+    {
+      "epoch": 0.9103501346671797,
+      "grad_norm": 1.6193269491195679,
+      "learning_rate": 4.185839442381357e-06,
+      "loss": 0.9345,
+      "step": 4732
+    },
+    {
+      "epoch": 0.9105425163524432,
+      "grad_norm": 2.237632989883423,
+      "learning_rate": 4.168019459364669e-06,
+      "loss": 0.8981,
+      "step": 4733
+    },
+    {
+      "epoch": 0.9107348980377068,
+      "grad_norm": 2.6894500255584717,
+      "learning_rate": 4.150236681622433e-06,
+      "loss": 0.966,
+      "step": 4734
+    },
+    {
+      "epoch": 0.9109272797229704,
+      "grad_norm": 2.655308485031128,
+      "learning_rate": 4.132491116058501e-06,
+      "loss": 0.9719,
+      "step": 4735
+    },
+    {
+      "epoch": 0.911119661408234,
+      "grad_norm": 1.6360094547271729,
+      "learning_rate": 4.114782769562364e-06,
+      "loss": 0.9456,
+      "step": 4736
+    },
+    {
+      "epoch": 0.9113120430934974,
+      "grad_norm": 2.724212169647217,
+      "learning_rate": 4.097111649008967e-06,
+      "loss": 0.9536,
+      "step": 4737
+    },
+    {
+      "epoch": 0.911504424778761,
+      "grad_norm": 2.7976667881011963,
+      "learning_rate": 4.079477761258854e-06,
+      "loss": 0.9149,
+      "step": 4738
+    },
+    {
+      "epoch": 0.9116968064640246,
+      "grad_norm": 2.5596976280212402,
+      "learning_rate": 4.061881113158117e-06,
+      "loss": 0.9636,
+      "step": 4739
+    },
+    {
+      "epoch": 0.9118891881492882,
+      "grad_norm": 2.226968288421631,
+      "learning_rate": 4.044321711538368e-06,
+      "loss": 0.9619,
+      "step": 4740
+    },
+    {
+      "epoch": 0.9120815698345518,
+      "grad_norm": 2.5160202980041504,
+      "learning_rate": 4.0267995632168094e-06,
+      "loss": 0.9054,
+      "step": 4741
+    },
+    {
+      "epoch": 0.9122739515198153,
+      "grad_norm": 2.372066020965576,
+      "learning_rate": 4.00931467499609e-06,
+      "loss": 0.9576,
+      "step": 4742
+    },
+    {
+      "epoch": 0.9124663332050789,
+      "grad_norm": 3.037501573562622,
+      "learning_rate": 3.9918670536644775e-06,
+      "loss": 0.9741,
+      "step": 4743
+    },
+    {
+      "epoch": 0.9126587148903424,
+      "grad_norm": 1.6752413511276245,
+      "learning_rate": 3.974456705995733e-06,
+      "loss": 0.9353,
+      "step": 4744
+    },
+    {
+      "epoch": 0.912851096575606,
+      "grad_norm": 1.9947565793991089,
+      "learning_rate": 3.957083638749148e-06,
+      "loss": 0.9065,
+      "step": 4745
+    },
+    {
+      "epoch": 0.9130434782608695,
+      "grad_norm": 1.8987032175064087,
+      "learning_rate": 3.939747858669551e-06,
+      "loss": 0.9384,
+      "step": 4746
+    },
+    {
+      "epoch": 0.9132358599461331,
+      "grad_norm": 2.5511317253112793,
+      "learning_rate": 3.922449372487291e-06,
+      "loss": 0.9565,
+      "step": 4747
+    },
+    {
+      "epoch": 0.9134282416313967,
+      "grad_norm": 2.585596799850464,
+      "learning_rate": 3.905188186918229e-06,
+      "loss": 0.9044,
+      "step": 4748
+    },
+    {
+      "epoch": 0.9136206233166603,
+      "grad_norm": 2.5836355686187744,
+      "learning_rate": 3.887964308663739e-06,
+      "loss": 0.9576,
+      "step": 4749
+    },
+    {
+      "epoch": 0.9138130050019239,
+      "grad_norm": 2.6169803142547607,
+      "learning_rate": 3.8707777444107695e-06,
+      "loss": 0.9409,
+      "step": 4750
+    },
+    {
+      "epoch": 0.9140053866871873,
+      "grad_norm": 2.6214287281036377,
+      "learning_rate": 3.853628500831685e-06,
+      "loss": 1.0473,
+      "step": 4751
+    },
+    {
+      "epoch": 0.9141977683724509,
+      "grad_norm": 1.939430594444275,
+      "learning_rate": 3.836516584584426e-06,
+      "loss": 0.9728,
+      "step": 4752
+    },
+    {
+      "epoch": 0.9143901500577145,
+      "grad_norm": 2.6966447830200195,
+      "learning_rate": 3.819442002312457e-06,
+      "loss": 0.8833,
+      "step": 4753
+    },
+    {
+      "epoch": 0.9145825317429781,
+      "grad_norm": 1.702908992767334,
+      "learning_rate": 3.8024047606446735e-06,
+      "loss": 0.9322,
+      "step": 4754
+    },
+    {
+      "epoch": 0.9147749134282416,
+      "grad_norm": 2.487579584121704,
+      "learning_rate": 3.7854048661955522e-06,
+      "loss": 0.9019,
+      "step": 4755
+    },
+    {
+      "epoch": 0.9149672951135052,
+      "grad_norm": 2.475738286972046,
+      "learning_rate": 3.7684423255650357e-06,
+      "loss": 0.975,
+      "step": 4756
+    },
+    {
+      "epoch": 0.9151596767987688,
+      "grad_norm": 2.0815649032592773,
+      "learning_rate": 3.7515171453385458e-06,
+      "loss": 0.9439,
+      "step": 4757
+    },
+    {
+      "epoch": 0.9153520584840323,
+      "grad_norm": 2.5085997581481934,
+      "learning_rate": 3.7346293320870363e-06,
+      "loss": 0.9563,
+      "step": 4758
+    },
+    {
+      "epoch": 0.9155444401692959,
+      "grad_norm": 2.702038049697876,
+      "learning_rate": 3.717778892366941e-06,
+      "loss": 0.9516,
+      "step": 4759
+    },
+    {
+      "epoch": 0.9157368218545594,
+      "grad_norm": 1.3037793636322021,
+      "learning_rate": 3.700965832720171e-06,
+      "loss": 0.7493,
+      "step": 4760
+    },
+    {
+      "epoch": 0.915929203539823,
+      "grad_norm": 1.3766849040985107,
+      "learning_rate": 3.684190159674117e-06,
+      "loss": 0.7782,
+      "step": 4761
+    },
+    {
+      "epoch": 0.9161215852250866,
+      "grad_norm": 1.4517571926116943,
+      "learning_rate": 3.6674518797417235e-06,
+      "loss": 0.9259,
+      "step": 4762
+    },
+    {
+      "epoch": 0.9163139669103502,
+      "grad_norm": 2.4123575687408447,
+      "learning_rate": 3.650750999421315e-06,
+      "loss": 0.9324,
+      "step": 4763
+    },
+    {
+      "epoch": 0.9165063485956138,
+      "grad_norm": 2.1797683238983154,
+      "learning_rate": 3.634087525196794e-06,
+      "loss": 0.9955,
+      "step": 4764
+    },
+    {
+      "epoch": 0.9166987302808772,
+      "grad_norm": 2.298092842102051,
+      "learning_rate": 3.6174614635374637e-06,
+      "loss": 0.9388,
+      "step": 4765
+    },
+    {
+      "epoch": 0.9168911119661408,
+      "grad_norm": 2.3173415660858154,
+      "learning_rate": 3.6008728208981156e-06,
+      "loss": 0.9715,
+      "step": 4766
+    },
+    {
+      "epoch": 0.9170834936514044,
+      "grad_norm": 1.4017475843429565,
+      "learning_rate": 3.5843216037190876e-06,
+      "loss": 0.9709,
+      "step": 4767
+    },
+    {
+      "epoch": 0.917275875336668,
+      "grad_norm": 2.9613797664642334,
+      "learning_rate": 3.567807818426083e-06,
+      "loss": 0.912,
+      "step": 4768
+    },
+    {
+      "epoch": 0.9174682570219315,
+      "grad_norm": 1.2716662883758545,
+      "learning_rate": 3.5513314714303526e-06,
+      "loss": 0.8,
+      "step": 4769
+    },
+    {
+      "epoch": 0.917660638707195,
+      "grad_norm": 1.95364248752594,
+      "learning_rate": 3.534892569128567e-06,
+      "loss": 0.9574,
+      "step": 4770
+    },
+    {
+      "epoch": 0.9178530203924586,
+      "grad_norm": 1.0363378524780273,
+      "learning_rate": 3.518491117902878e-06,
+      "loss": 0.9653,
+      "step": 4771
+    },
+    {
+      "epoch": 0.9180454020777222,
+      "grad_norm": 0.9874885082244873,
+      "learning_rate": 3.5021271241208907e-06,
+      "loss": 0.9461,
+      "step": 4772
+    },
+    {
+      "epoch": 0.9182377837629858,
+      "grad_norm": 2.213238000869751,
+      "learning_rate": 3.48580059413568e-06,
+      "loss": 0.9992,
+      "step": 4773
+    },
+    {
+      "epoch": 0.9184301654482493,
+      "grad_norm": 1.9561426639556885,
+      "learning_rate": 3.469511534285752e-06,
+      "loss": 0.9263,
+      "step": 4774
+    },
+    {
+      "epoch": 0.9186225471335129,
+      "grad_norm": 1.2688592672348022,
+      "learning_rate": 3.4532599508950824e-06,
+      "loss": 0.9828,
+      "step": 4775
+    },
+    {
+      "epoch": 0.9188149288187765,
+      "grad_norm": 1.8123509883880615,
+      "learning_rate": 3.437045850273113e-06,
+      "loss": 0.8625,
+      "step": 4776
+    },
+    {
+      "epoch": 0.91900731050404,
+      "grad_norm": 2.4712536334991455,
+      "learning_rate": 3.420869238714708e-06,
+      "loss": 0.9617,
+      "step": 4777
+    },
+    {
+      "epoch": 0.9191996921893035,
+      "grad_norm": 2.404334783554077,
+      "learning_rate": 3.4047301225001547e-06,
+      "loss": 0.908,
+      "step": 4778
+    },
+    {
+      "epoch": 0.9193920738745671,
+      "grad_norm": 1.8304945230484009,
+      "learning_rate": 3.388628507895275e-06,
+      "loss": 0.8812,
+      "step": 4779
+    },
+    {
+      "epoch": 0.9195844555598307,
+      "grad_norm": 1.62456214427948,
+      "learning_rate": 3.3725644011512124e-06,
+      "loss": 0.7723,
+      "step": 4780
+    },
+    {
+      "epoch": 0.9197768372450943,
+      "grad_norm": 1.8838753700256348,
+      "learning_rate": 3.3565378085046337e-06,
+      "loss": 0.8889,
+      "step": 4781
+    },
+    {
+      "epoch": 0.9199692189303579,
+      "grad_norm": 2.0724902153015137,
+      "learning_rate": 3.340548736177618e-06,
+      "loss": 0.9365,
+      "step": 4782
+    },
+    {
+      "epoch": 0.9201616006156214,
+      "grad_norm": 2.8274621963500977,
+      "learning_rate": 3.324597190377665e-06,
+      "loss": 0.8825,
+      "step": 4783
+    },
+    {
+      "epoch": 0.9203539823008849,
+      "grad_norm": 2.327754497528076,
+      "learning_rate": 3.308683177297711e-06,
+      "loss": 0.9717,
+      "step": 4784
+    },
+    {
+      "epoch": 0.9205463639861485,
+      "grad_norm": 1.5262638330459595,
+      "learning_rate": 3.2928067031161247e-06,
+      "loss": 0.952,
+      "step": 4785
+    },
+    {
+      "epoch": 0.9207387456714121,
+      "grad_norm": 2.2462332248687744,
+      "learning_rate": 3.2769677739966973e-06,
+      "loss": 0.8775,
+      "step": 4786
+    },
+    {
+      "epoch": 0.9209311273566756,
+      "grad_norm": 2.5601673126220703,
+      "learning_rate": 3.2611663960886664e-06,
+      "loss": 0.9097,
+      "step": 4787
+    },
+    {
+      "epoch": 0.9211235090419392,
+      "grad_norm": 2.7896554470062256,
+      "learning_rate": 3.245402575526646e-06,
+      "loss": 0.9443,
+      "step": 4788
+    },
+    {
+      "epoch": 0.9213158907272028,
+      "grad_norm": 2.08781099319458,
+      "learning_rate": 3.229676318430697e-06,
+      "loss": 0.9393,
+      "step": 4789
+    },
+    {
+      "epoch": 0.9215082724124664,
+      "grad_norm": 2.268389940261841,
+      "learning_rate": 3.2139876309063234e-06,
+      "loss": 0.8729,
+      "step": 4790
+    },
+    {
+      "epoch": 0.9217006540977299,
+      "grad_norm": 1.2559328079223633,
+      "learning_rate": 3.198336519044376e-06,
+      "loss": 0.7726,
+      "step": 4791
+    },
+    {
+      "epoch": 0.9218930357829934,
+      "grad_norm": 2.397711753845215,
+      "learning_rate": 3.1827229889211606e-06,
+      "loss": 0.867,
+      "step": 4792
+    },
+    {
+      "epoch": 0.922085417468257,
+      "grad_norm": 2.5989599227905273,
+      "learning_rate": 3.1671470465984177e-06,
+      "loss": 0.898,
+      "step": 4793
+    },
+    {
+      "epoch": 0.9222777991535206,
+      "grad_norm": 2.4238219261169434,
+      "learning_rate": 3.151608698123232e-06,
+      "loss": 0.9523,
+      "step": 4794
+    },
+    {
+      "epoch": 0.9224701808387842,
+      "grad_norm": 1.8159488439559937,
+      "learning_rate": 3.1361079495281444e-06,
+      "loss": 0.927,
+      "step": 4795
+    },
+    {
+      "epoch": 0.9226625625240477,
+      "grad_norm": 1.7993288040161133,
+      "learning_rate": 3.1206448068310633e-06,
+      "loss": 0.8748,
+      "step": 4796
+    },
+    {
+      "epoch": 0.9228549442093112,
+      "grad_norm": 1.7850666046142578,
+      "learning_rate": 3.1052192760353314e-06,
+      "loss": 0.987,
+      "step": 4797
+    },
+    {
+      "epoch": 0.9230473258945748,
+      "grad_norm": 2.37784743309021,
+      "learning_rate": 3.0898313631296583e-06,
+      "loss": 0.9844,
+      "step": 4798
+    },
+    {
+      "epoch": 0.9232397075798384,
+      "grad_norm": 1.9331392049789429,
+      "learning_rate": 3.074481074088165e-06,
+      "loss": 0.8869,
+      "step": 4799
+    },
+    {
+      "epoch": 0.923432089265102,
+      "grad_norm": 2.3100647926330566,
+      "learning_rate": 3.0591684148703614e-06,
+      "loss": 0.9625,
+      "step": 4800
+    },
+    {
+      "epoch": 0.9236244709503655,
+      "grad_norm": 2.4676907062530518,
+      "learning_rate": 3.0438933914211486e-06,
+      "loss": 0.8446,
+      "step": 4801
+    },
+    {
+      "epoch": 0.9238168526356291,
+      "grad_norm": 2.0618162155151367,
+      "learning_rate": 3.028656009670827e-06,
+      "loss": 0.9697,
+      "step": 4802
+    },
+    {
+      "epoch": 0.9240092343208927,
+      "grad_norm": 1.6505862474441528,
+      "learning_rate": 3.0134562755350537e-06,
+      "loss": 0.926,
+      "step": 4803
+    },
+    {
+      "epoch": 0.9242016160061562,
+      "grad_norm": 2.38851261138916,
+      "learning_rate": 2.998294194914897e-06,
+      "loss": 0.8032,
+      "step": 4804
+    },
+    {
+      "epoch": 0.9243939976914197,
+      "grad_norm": 2.8340373039245605,
+      "learning_rate": 2.9831697736968146e-06,
+      "loss": 0.9066,
+      "step": 4805
+    },
+    {
+      "epoch": 0.9245863793766833,
+      "grad_norm": 2.3643784523010254,
+      "learning_rate": 2.9680830177525986e-06,
+      "loss": 1.0492,
+      "step": 4806
+    },
+    {
+      "epoch": 0.9247787610619469,
+      "grad_norm": 2.1756980419158936,
+      "learning_rate": 2.953033932939464e-06,
+      "loss": 0.8795,
+      "step": 4807
+    },
+    {
+      "epoch": 0.9249711427472105,
+      "grad_norm": 2.505967378616333,
+      "learning_rate": 2.9380225250999816e-06,
+      "loss": 0.9743,
+      "step": 4808
+    },
+    {
+      "epoch": 0.9251635244324741,
+      "grad_norm": 1.7989166975021362,
+      "learning_rate": 2.9230488000621004e-06,
+      "loss": 0.9654,
+      "step": 4809
+    },
+    {
+      "epoch": 0.9253559061177375,
+      "grad_norm": 2.4374067783355713,
+      "learning_rate": 2.908112763639137e-06,
+      "loss": 0.9052,
+      "step": 4810
+    },
+    {
+      "epoch": 0.9255482878030011,
+      "grad_norm": 1.44603431224823,
+      "learning_rate": 2.893214421629764e-06,
+      "loss": 0.7778,
+      "step": 4811
+    },
+    {
+      "epoch": 0.9257406694882647,
+      "grad_norm": 2.3537375926971436,
+      "learning_rate": 2.8783537798180437e-06,
+      "loss": 0.9439,
+      "step": 4812
+    },
+    {
+      "epoch": 0.9259330511735283,
+      "grad_norm": 2.1617796421051025,
+      "learning_rate": 2.863530843973372e-06,
+      "loss": 0.958,
+      "step": 4813
+    },
+    {
+      "epoch": 0.9261254328587919,
+      "grad_norm": 2.291513442993164,
+      "learning_rate": 2.848745619850546e-06,
+      "loss": 0.8897,
+      "step": 4814
+    },
+    {
+      "epoch": 0.9263178145440554,
+      "grad_norm": 2.1925830841064453,
+      "learning_rate": 2.833998113189662e-06,
+      "loss": 0.9663,
+      "step": 4815
+    },
+    {
+      "epoch": 0.926510196229319,
+      "grad_norm": 1.8293066024780273,
+      "learning_rate": 2.819288329716263e-06,
+      "loss": 0.9528,
+      "step": 4816
+    },
+    {
+      "epoch": 0.9267025779145825,
+      "grad_norm": 2.2707526683807373,
+      "learning_rate": 2.8046162751411475e-06,
+      "loss": 0.9655,
+      "step": 4817
+    },
+    {
+      "epoch": 0.9268949595998461,
+      "grad_norm": 2.3133177757263184,
+      "learning_rate": 2.7899819551605257e-06,
+      "loss": 0.9407,
+      "step": 4818
+    },
+    {
+      "epoch": 0.9270873412851096,
+      "grad_norm": 2.2705743312835693,
+      "learning_rate": 2.7753853754559635e-06,
+      "loss": 0.9582,
+      "step": 4819
+    },
+    {
+      "epoch": 0.9272797229703732,
+      "grad_norm": 2.3612172603607178,
+      "learning_rate": 2.760826541694328e-06,
+      "loss": 0.9759,
+      "step": 4820
+    },
+    {
+      "epoch": 0.9274721046556368,
+      "grad_norm": 2.5905814170837402,
+      "learning_rate": 2.746305459527876e-06,
+      "loss": 0.8824,
+      "step": 4821
+    },
+    {
+      "epoch": 0.9276644863409004,
+      "grad_norm": 1.252577304840088,
+      "learning_rate": 2.7318221345941863e-06,
+      "loss": 0.8274,
+      "step": 4822
+    },
+    {
+      "epoch": 0.927856868026164,
+      "grad_norm": 2.2394373416900635,
+      "learning_rate": 2.7173765725161837e-06,
+      "loss": 0.9356,
+      "step": 4823
+    },
+    {
+      "epoch": 0.9280492497114274,
+      "grad_norm": 1.473252534866333,
+      "learning_rate": 2.7029687789021373e-06,
+      "loss": 0.8498,
+      "step": 4824
+    },
+    {
+      "epoch": 0.928241631396691,
+      "grad_norm": 2.4784035682678223,
+      "learning_rate": 2.6885987593456507e-06,
+      "loss": 0.9516,
+      "step": 4825
+    },
+    {
+      "epoch": 0.9284340130819546,
+      "grad_norm": 2.353128433227539,
+      "learning_rate": 2.67426651942565e-06,
+      "loss": 0.9434,
+      "step": 4826
+    },
+    {
+      "epoch": 0.9286263947672182,
+      "grad_norm": 2.288614273071289,
+      "learning_rate": 2.6599720647064062e-06,
+      "loss": 0.9709,
+      "step": 4827
+    },
+    {
+      "epoch": 0.9288187764524817,
+      "grad_norm": 2.2655999660491943,
+      "learning_rate": 2.645715400737536e-06,
+      "loss": 0.913,
+      "step": 4828
+    },
+    {
+      "epoch": 0.9290111581377453,
+      "grad_norm": 1.1774840354919434,
+      "learning_rate": 2.631496533053934e-06,
+      "loss": 0.9802,
+      "step": 4829
+    },
+    {
+      "epoch": 0.9292035398230089,
+      "grad_norm": 1.374403476715088,
+      "learning_rate": 2.617315467175885e-06,
+      "loss": 0.7305,
+      "step": 4830
+    },
+    {
+      "epoch": 0.9293959215082724,
+      "grad_norm": 1.2269548177719116,
+      "learning_rate": 2.6031722086089615e-06,
+      "loss": 0.7555,
+      "step": 4831
+    },
+    {
+      "epoch": 0.929588303193536,
+      "grad_norm": 2.4357473850250244,
+      "learning_rate": 2.589066762844039e-06,
+      "loss": 0.9848,
+      "step": 4832
+    },
+    {
+      "epoch": 0.9297806848787995,
+      "grad_norm": 1.8485374450683594,
+      "learning_rate": 2.57499913535737e-06,
+      "loss": 0.9248,
+      "step": 4833
+    },
+    {
+      "epoch": 0.9299730665640631,
+      "grad_norm": 1.7403756380081177,
+      "learning_rate": 2.5609693316104744e-06,
+      "loss": 0.9111,
+      "step": 4834
+    },
+    {
+      "epoch": 0.9301654482493267,
+      "grad_norm": 2.3715896606445312,
+      "learning_rate": 2.5469773570502065e-06,
+      "loss": 0.9508,
+      "step": 4835
+    },
+    {
+      "epoch": 0.9303578299345903,
+      "grad_norm": 1.1186500787734985,
+      "learning_rate": 2.5330232171087433e-06,
+      "loss": 0.7716,
+      "step": 4836
+    },
+    {
+      "epoch": 0.9305502116198537,
+      "grad_norm": 1.6027231216430664,
+      "learning_rate": 2.5191069172035396e-06,
+      "loss": 0.9389,
+      "step": 4837
+    },
+    {
+      "epoch": 0.9307425933051173,
+      "grad_norm": 1.6103405952453613,
+      "learning_rate": 2.5052284627374077e-06,
+      "loss": 0.928,
+      "step": 4838
+    },
+    {
+      "epoch": 0.9309349749903809,
+      "grad_norm": 1.7877336740493774,
+      "learning_rate": 2.491387859098426e-06,
+      "loss": 0.9827,
+      "step": 4839
+    },
+    {
+      "epoch": 0.9311273566756445,
+      "grad_norm": 1.899251103401184,
+      "learning_rate": 2.4775851116599967e-06,
+      "loss": 0.8746,
+      "step": 4840
+    },
+    {
+      "epoch": 0.9313197383609081,
+      "grad_norm": 2.410978078842163,
+      "learning_rate": 2.4638202257808107e-06,
+      "loss": 0.9365,
+      "step": 4841
+    },
+    {
+      "epoch": 0.9315121200461716,
+      "grad_norm": 1.6276390552520752,
+      "learning_rate": 2.4500932068049043e-06,
+      "loss": 0.9209,
+      "step": 4842
+    },
+    {
+      "epoch": 0.9317045017314352,
+      "grad_norm": 3.1989290714263916,
+      "learning_rate": 2.436404060061548e-06,
+      "loss": 0.8991,
+      "step": 4843
+    },
+    {
+      "epoch": 0.9318968834166987,
+      "grad_norm": 2.544271945953369,
+      "learning_rate": 2.4227527908653458e-06,
+      "loss": 0.8847,
+      "step": 4844
+    },
+    {
+      "epoch": 0.9320892651019623,
+      "grad_norm": 2.5869076251983643,
+      "learning_rate": 2.409139404516203e-06,
+      "loss": 0.9333,
+      "step": 4845
+    },
+    {
+      "epoch": 0.9322816467872258,
+      "grad_norm": 2.2039899826049805,
+      "learning_rate": 2.3955639062992696e-06,
+      "loss": 0.9404,
+      "step": 4846
+    },
+    {
+      "epoch": 0.9324740284724894,
+      "grad_norm": 2.9237308502197266,
+      "learning_rate": 2.382026301485074e-06,
+      "loss": 0.9509,
+      "step": 4847
+    },
+    {
+      "epoch": 0.932666410157753,
+      "grad_norm": 2.341932773590088,
+      "learning_rate": 2.3685265953293343e-06,
+      "loss": 0.9509,
+      "step": 4848
+    },
+    {
+      "epoch": 0.9328587918430166,
+      "grad_norm": 2.009359121322632,
+      "learning_rate": 2.3550647930731364e-06,
+      "loss": 0.9073,
+      "step": 4849
+    },
+    {
+      "epoch": 0.9330511735282802,
+      "grad_norm": 1.6933027505874634,
+      "learning_rate": 2.3416408999427875e-06,
+      "loss": 0.9517,
+      "step": 4850
+    },
+    {
+      "epoch": 0.9332435552135436,
+      "grad_norm": 1.5422406196594238,
+      "learning_rate": 2.3282549211499304e-06,
+      "loss": 0.9836,
+      "step": 4851
+    },
+    {
+      "epoch": 0.9334359368988072,
+      "grad_norm": 2.669349193572998,
+      "learning_rate": 2.3149068618914415e-06,
+      "loss": 0.9235,
+      "step": 4852
+    },
+    {
+      "epoch": 0.9336283185840708,
+      "grad_norm": 2.339726448059082,
+      "learning_rate": 2.3015967273494864e-06,
+      "loss": 0.945,
+      "step": 4853
+    },
+    {
+      "epoch": 0.9338207002693344,
+      "grad_norm": 1.478136420249939,
+      "learning_rate": 2.288324522691565e-06,
+      "loss": 0.7716,
+      "step": 4854
+    },
+    {
+      "epoch": 0.9340130819545979,
+      "grad_norm": 2.0843100547790527,
+      "learning_rate": 2.2750902530703666e-06,
+      "loss": 0.936,
+      "step": 4855
+    },
+    {
+      "epoch": 0.9342054636398615,
+      "grad_norm": 2.498220205307007,
+      "learning_rate": 2.2618939236238923e-06,
+      "loss": 0.8723,
+      "step": 4856
+    },
+    {
+      "epoch": 0.934397845325125,
+      "grad_norm": 2.692112922668457,
+      "learning_rate": 2.248735539475433e-06,
+      "loss": 0.9775,
+      "step": 4857
+    },
+    {
+      "epoch": 0.9345902270103886,
+      "grad_norm": 1.7786952257156372,
+      "learning_rate": 2.2356151057334906e-06,
+      "loss": 0.928,
+      "step": 4858
+    },
+    {
+      "epoch": 0.9347826086956522,
+      "grad_norm": 1.3894731998443604,
+      "learning_rate": 2.2225326274919133e-06,
+      "loss": 0.7779,
+      "step": 4859
+    },
+    {
+      "epoch": 0.9349749903809157,
+      "grad_norm": 2.182222604751587,
+      "learning_rate": 2.209488109829727e-06,
+      "loss": 0.928,
+      "step": 4860
+    },
+    {
+      "epoch": 0.9351673720661793,
+      "grad_norm": 2.625750780105591,
+      "learning_rate": 2.196481557811303e-06,
+      "loss": 0.8794,
+      "step": 4861
+    },
+    {
+      "epoch": 0.9353597537514429,
+      "grad_norm": 2.423037528991699,
+      "learning_rate": 2.1835129764861906e-06,
+      "loss": 0.9924,
+      "step": 4862
+    },
+    {
+      "epoch": 0.9355521354367065,
+      "grad_norm": 1.9406262636184692,
+      "learning_rate": 2.1705823708892734e-06,
+      "loss": 0.9469,
+      "step": 4863
+    },
+    {
+      "epoch": 0.93574451712197,
+      "grad_norm": 1.307350516319275,
+      "learning_rate": 2.1576897460406474e-06,
+      "loss": 0.8902,
+      "step": 4864
+    },
+    {
+      "epoch": 0.9359368988072335,
+      "grad_norm": 2.5325324535369873,
+      "learning_rate": 2.144835106945664e-06,
+      "loss": 0.9011,
+      "step": 4865
+    },
+    {
+      "epoch": 0.9361292804924971,
+      "grad_norm": 2.8517704010009766,
+      "learning_rate": 2.1320184585949533e-06,
+      "loss": 0.9841,
+      "step": 4866
+    },
+    {
+      "epoch": 0.9363216621777607,
+      "grad_norm": 2.5311458110809326,
+      "learning_rate": 2.119239805964357e-06,
+      "loss": 0.9457,
+      "step": 4867
+    },
+    {
+      "epoch": 0.9365140438630243,
+      "grad_norm": 1.5483862161636353,
+      "learning_rate": 2.106499154015018e-06,
+      "loss": 0.7969,
+      "step": 4868
+    },
+    {
+      "epoch": 0.9367064255482878,
+      "grad_norm": 2.7507429122924805,
+      "learning_rate": 2.0937965076932574e-06,
+      "loss": 0.9051,
+      "step": 4869
+    },
+    {
+      "epoch": 0.9368988072335513,
+      "grad_norm": 2.185966730117798,
+      "learning_rate": 2.0811318719307193e-06,
+      "loss": 0.939,
+      "step": 4870
+    },
+    {
+      "epoch": 0.9370911889188149,
+      "grad_norm": 2.385812997817993,
+      "learning_rate": 2.068505251644237e-06,
+      "loss": 0.8937,
+      "step": 4871
+    },
+    {
+      "epoch": 0.9372835706040785,
+      "grad_norm": 1.545632004737854,
+      "learning_rate": 2.0559166517358785e-06,
+      "loss": 0.897,
+      "step": 4872
+    },
+    {
+      "epoch": 0.9374759522893421,
+      "grad_norm": 2.4496986865997314,
+      "learning_rate": 2.0433660770930007e-06,
+      "loss": 0.9467,
+      "step": 4873
+    },
+    {
+      "epoch": 0.9376683339746056,
+      "grad_norm": 2.3655261993408203,
+      "learning_rate": 2.030853532588162e-06,
+      "loss": 0.9329,
+      "step": 4874
+    },
+    {
+      "epoch": 0.9378607156598692,
+      "grad_norm": 3.021181106567383,
+      "learning_rate": 2.0183790230791534e-06,
+      "loss": 0.9442,
+      "step": 4875
+    },
+    {
+      "epoch": 0.9380530973451328,
+      "grad_norm": 1.2944889068603516,
+      "learning_rate": 2.0059425534090125e-06,
+      "loss": 0.9107,
+      "step": 4876
+    },
+    {
+      "epoch": 0.9382454790303963,
+      "grad_norm": 2.4987781047821045,
+      "learning_rate": 1.993544128406e-06,
+      "loss": 0.9913,
+      "step": 4877
+    },
+    {
+      "epoch": 0.9384378607156598,
+      "grad_norm": 2.482832908630371,
+      "learning_rate": 1.981183752883631e-06,
+      "loss": 0.91,
+      "step": 4878
+    },
+    {
+      "epoch": 0.9386302424009234,
+      "grad_norm": 2.8146917819976807,
+      "learning_rate": 1.9688614316406007e-06,
+      "loss": 0.9148,
+      "step": 4879
+    },
+    {
+      "epoch": 0.938822624086187,
+      "grad_norm": 2.680894613265991,
+      "learning_rate": 1.9565771694608936e-06,
+      "loss": 0.9067,
+      "step": 4880
+    },
+    {
+      "epoch": 0.9390150057714506,
+      "grad_norm": 0.8912118077278137,
+      "learning_rate": 1.944330971113639e-06,
+      "loss": 0.9345,
+      "step": 4881
+    },
+    {
+      "epoch": 0.9392073874567142,
+      "grad_norm": 0.9303460717201233,
+      "learning_rate": 1.9321228413532787e-06,
+      "loss": 0.8909,
+      "step": 4882
+    },
+    {
+      "epoch": 0.9393997691419776,
+      "grad_norm": 2.4433329105377197,
+      "learning_rate": 1.91995278491941e-06,
+      "loss": 0.9105,
+      "step": 4883
+    },
+    {
+      "epoch": 0.9395921508272412,
+      "grad_norm": 1.5665040016174316,
+      "learning_rate": 1.907820806536842e-06,
+      "loss": 0.7964,
+      "step": 4884
+    },
+    {
+      "epoch": 0.9397845325125048,
+      "grad_norm": 2.393232822418213,
+      "learning_rate": 1.8957269109156628e-06,
+      "loss": 0.8662,
+      "step": 4885
+    },
+    {
+      "epoch": 0.9399769141977684,
+      "grad_norm": 1.9949501752853394,
+      "learning_rate": 1.8836711027511277e-06,
+      "loss": 0.9655,
+      "step": 4886
+    },
+    {
+      "epoch": 0.9401692958830319,
+      "grad_norm": 1.5361274480819702,
+      "learning_rate": 1.8716533867237152e-06,
+      "loss": 0.7921,
+      "step": 4887
+    },
+    {
+      "epoch": 0.9403616775682955,
+      "grad_norm": 1.1726634502410889,
+      "learning_rate": 1.859673767499115e-06,
+      "loss": 1.0048,
+      "step": 4888
+    },
+    {
+      "epoch": 0.9405540592535591,
+      "grad_norm": 1.2900404930114746,
+      "learning_rate": 1.8477322497282178e-06,
+      "loss": 0.7383,
+      "step": 4889
+    },
+    {
+      "epoch": 0.9407464409388226,
+      "grad_norm": 2.349045991897583,
+      "learning_rate": 1.83582883804716e-06,
+      "loss": 0.9671,
+      "step": 4890
+    },
+    {
+      "epoch": 0.9409388226240862,
+      "grad_norm": 2.5451619625091553,
+      "learning_rate": 1.8239635370772223e-06,
+      "loss": 0.8754,
+      "step": 4891
+    },
+    {
+      "epoch": 0.9411312043093497,
+      "grad_norm": 1.6198445558547974,
+      "learning_rate": 1.8121363514249535e-06,
+      "loss": 0.7804,
+      "step": 4892
+    },
+    {
+      "epoch": 0.9413235859946133,
+      "grad_norm": 2.612311840057373,
+      "learning_rate": 1.8003472856820469e-06,
+      "loss": 0.89,
+      "step": 4893
+    },
+    {
+      "epoch": 0.9415159676798769,
+      "grad_norm": 2.359758138656616,
+      "learning_rate": 1.7885963444254528e-06,
+      "loss": 0.9278,
+      "step": 4894
+    },
+    {
+      "epoch": 0.9417083493651405,
+      "grad_norm": 2.0599098205566406,
+      "learning_rate": 1.776883532217255e-06,
+      "loss": 0.9094,
+      "step": 4895
+    },
+    {
+      "epoch": 0.941900731050404,
+      "grad_norm": 1.3053240776062012,
+      "learning_rate": 1.7652088536048051e-06,
+      "loss": 0.9921,
+      "step": 4896
+    },
+    {
+      "epoch": 0.9420931127356675,
+      "grad_norm": 2.798172950744629,
+      "learning_rate": 1.7535723131206106e-06,
+      "loss": 0.9178,
+      "step": 4897
+    },
+    {
+      "epoch": 0.9422854944209311,
+      "grad_norm": 2.7331323623657227,
+      "learning_rate": 1.7419739152823467e-06,
+      "loss": 0.9975,
+      "step": 4898
+    },
+    {
+      "epoch": 0.9424778761061947,
+      "grad_norm": 2.534885883331299,
+      "learning_rate": 1.7304136645929447e-06,
+      "loss": 0.955,
+      "step": 4899
+    },
+    {
+      "epoch": 0.9426702577914583,
+      "grad_norm": 2.3095791339874268,
+      "learning_rate": 1.7188915655404813e-06,
+      "loss": 0.9202,
+      "step": 4900
+    },
+    {
+      "epoch": 0.9428626394767218,
+      "grad_norm": 2.61446475982666,
+      "learning_rate": 1.7074076225982228e-06,
+      "loss": 0.8588,
+      "step": 4901
+    },
+    {
+      "epoch": 0.9430550211619854,
+      "grad_norm": 2.445775270462036,
+      "learning_rate": 1.6959618402246357e-06,
+      "loss": 0.9396,
+      "step": 4902
+    },
+    {
+      "epoch": 0.943247402847249,
+      "grad_norm": 2.4564831256866455,
+      "learning_rate": 1.684554222863377e-06,
+      "loss": 0.9057,
+      "step": 4903
+    },
+    {
+      "epoch": 0.9434397845325125,
+      "grad_norm": 2.0799195766448975,
+      "learning_rate": 1.6731847749432705e-06,
+      "loss": 0.9089,
+      "step": 4904
+    },
+    {
+      "epoch": 0.943632166217776,
+      "grad_norm": 1.9029091596603394,
+      "learning_rate": 1.6618535008783075e-06,
+      "loss": 0.9506,
+      "step": 4905
+    },
+    {
+      "epoch": 0.9438245479030396,
+      "grad_norm": 1.9384286403656006,
+      "learning_rate": 1.650560405067725e-06,
+      "loss": 0.9815,
+      "step": 4906
+    },
+    {
+      "epoch": 0.9440169295883032,
+      "grad_norm": 2.0288031101226807,
+      "learning_rate": 1.6393054918958373e-06,
+      "loss": 0.9296,
+      "step": 4907
+    },
+    {
+      "epoch": 0.9442093112735668,
+      "grad_norm": 2.080385446548462,
+      "learning_rate": 1.6280887657322275e-06,
+      "loss": 0.8808,
+      "step": 4908
+    },
+    {
+      "epoch": 0.9444016929588304,
+      "grad_norm": 2.2836031913757324,
+      "learning_rate": 1.616910230931612e-06,
+      "loss": 0.9352,
+      "step": 4909
+    },
+    {
+      "epoch": 0.9445940746440938,
+      "grad_norm": 2.020735025405884,
+      "learning_rate": 1.6057698918338526e-06,
+      "loss": 0.9593,
+      "step": 4910
+    },
+    {
+      "epoch": 0.9447864563293574,
+      "grad_norm": 2.260990858078003,
+      "learning_rate": 1.5946677527640564e-06,
+      "loss": 0.9363,
+      "step": 4911
+    },
+    {
+      "epoch": 0.944978838014621,
+      "grad_norm": 2.479508638381958,
+      "learning_rate": 1.5836038180324198e-06,
+      "loss": 0.9051,
+      "step": 4912
+    },
+    {
+      "epoch": 0.9451712196998846,
+      "grad_norm": 1.206002950668335,
+      "learning_rate": 1.5725780919343625e-06,
+      "loss": 0.7606,
+      "step": 4913
+    },
+    {
+      "epoch": 0.9453636013851482,
+      "grad_norm": 1.781956672668457,
+      "learning_rate": 1.561590578750438e-06,
+      "loss": 1.0372,
+      "step": 4914
+    },
+    {
+      "epoch": 0.9455559830704117,
+      "grad_norm": 1.2197638750076294,
+      "learning_rate": 1.55064128274639e-06,
+      "loss": 0.8091,
+      "step": 4915
+    },
+    {
+      "epoch": 0.9457483647556753,
+      "grad_norm": 2.5718655586242676,
+      "learning_rate": 1.5397302081731068e-06,
+      "loss": 0.9447,
+      "step": 4916
+    },
+    {
+      "epoch": 0.9459407464409388,
+      "grad_norm": 1.9351884126663208,
+      "learning_rate": 1.5288573592666444e-06,
+      "loss": 0.9478,
+      "step": 4917
+    },
+    {
+      "epoch": 0.9461331281262024,
+      "grad_norm": 2.313593864440918,
+      "learning_rate": 1.518022740248215e-06,
+      "loss": 0.9166,
+      "step": 4918
+    },
+    {
+      "epoch": 0.9463255098114659,
+      "grad_norm": 2.533297538757324,
+      "learning_rate": 1.5072263553241872e-06,
+      "loss": 0.887,
+      "step": 4919
+    },
+    {
+      "epoch": 0.9465178914967295,
+      "grad_norm": 2.0024261474609375,
+      "learning_rate": 1.4964682086861082e-06,
+      "loss": 0.9415,
+      "step": 4920
+    },
+    {
+      "epoch": 0.9467102731819931,
+      "grad_norm": 2.5764756202697754,
+      "learning_rate": 1.4857483045106257e-06,
+      "loss": 0.9962,
+      "step": 4921
+    },
+    {
+      "epoch": 0.9469026548672567,
+      "grad_norm": 2.339301347732544,
+      "learning_rate": 1.4750666469596108e-06,
+      "loss": 1.0223,
+      "step": 4922
+    },
+    {
+      "epoch": 0.9470950365525203,
+      "grad_norm": 2.4751837253570557,
+      "learning_rate": 1.464423240180035e-06,
+      "loss": 0.8928,
+      "step": 4923
+    },
+    {
+      "epoch": 0.9472874182377837,
+      "grad_norm": 2.318168878555298,
+      "learning_rate": 1.4538180883040265e-06,
+      "loss": 0.937,
+      "step": 4924
+    },
+    {
+      "epoch": 0.9474797999230473,
+      "grad_norm": 1.700100064277649,
+      "learning_rate": 1.4432511954488914e-06,
+      "loss": 0.9203,
+      "step": 4925
+    },
+    {
+      "epoch": 0.9476721816083109,
+      "grad_norm": 2.192725896835327,
+      "learning_rate": 1.4327225657170484e-06,
+      "loss": 0.9654,
+      "step": 4926
+    },
+    {
+      "epoch": 0.9478645632935745,
+      "grad_norm": 2.288020610809326,
+      "learning_rate": 1.4222322031960722e-06,
+      "loss": 0.9106,
+      "step": 4927
+    },
+    {
+      "epoch": 0.948056944978838,
+      "grad_norm": 2.4799911975860596,
+      "learning_rate": 1.4117801119586938e-06,
+      "loss": 0.8882,
+      "step": 4928
+    },
+    {
+      "epoch": 0.9482493266641016,
+      "grad_norm": 1.0839301347732544,
+      "learning_rate": 1.401366296062756e-06,
+      "loss": 0.9637,
+      "step": 4929
+    },
+    {
+      "epoch": 0.9484417083493651,
+      "grad_norm": 2.5828514099121094,
+      "learning_rate": 1.3909907595512805e-06,
+      "loss": 0.9152,
+      "step": 4930
+    },
+    {
+      "epoch": 0.9486340900346287,
+      "grad_norm": 2.5637991428375244,
+      "learning_rate": 1.3806535064524006e-06,
+      "loss": 0.9125,
+      "step": 4931
+    },
+    {
+      "epoch": 0.9488264717198923,
+      "grad_norm": 2.5850422382354736,
+      "learning_rate": 1.370354540779395e-06,
+      "loss": 0.9916,
+      "step": 4932
+    },
+    {
+      "epoch": 0.9490188534051558,
+      "grad_norm": 2.117042303085327,
+      "learning_rate": 1.360093866530665e-06,
+      "loss": 1.0099,
+      "step": 4933
+    },
+    {
+      "epoch": 0.9492112350904194,
+      "grad_norm": 2.755326747894287,
+      "learning_rate": 1.34987148768978e-06,
+      "loss": 0.9236,
+      "step": 4934
+    },
+    {
+      "epoch": 0.949403616775683,
+      "grad_norm": 1.6244795322418213,
+      "learning_rate": 1.3396874082253986e-06,
+      "loss": 0.9617,
+      "step": 4935
+    },
+    {
+      "epoch": 0.9495959984609466,
+      "grad_norm": 2.133923053741455,
+      "learning_rate": 1.3295416320913357e-06,
+      "loss": 0.953,
+      "step": 4936
+    },
+    {
+      "epoch": 0.94978838014621,
+      "grad_norm": 2.77656888961792,
+      "learning_rate": 1.3194341632265516e-06,
+      "loss": 0.9432,
+      "step": 4937
+    },
+    {
+      "epoch": 0.9499807618314736,
+      "grad_norm": 1.184454321861267,
+      "learning_rate": 1.3093650055550854e-06,
+      "loss": 0.9969,
+      "step": 4938
+    },
+    {
+      "epoch": 0.9501731435167372,
+      "grad_norm": 2.4864296913146973,
+      "learning_rate": 1.2993341629861433e-06,
+      "loss": 0.9469,
+      "step": 4939
+    },
+    {
+      "epoch": 0.9503655252020008,
+      "grad_norm": 2.219656229019165,
+      "learning_rate": 1.2893416394140323e-06,
+      "loss": 0.9154,
+      "step": 4940
+    },
+    {
+      "epoch": 0.9505579068872644,
+      "grad_norm": 1.8871357440948486,
+      "learning_rate": 1.279387438718216e-06,
+      "loss": 0.9554,
+      "step": 4941
+    },
+    {
+      "epoch": 0.9507502885725279,
+      "grad_norm": 2.2224366664886475,
+      "learning_rate": 1.269471564763247e-06,
+      "loss": 0.9258,
+      "step": 4942
+    },
+    {
+      "epoch": 0.9509426702577914,
+      "grad_norm": 2.273153066635132,
+      "learning_rate": 1.2595940213988022e-06,
+      "loss": 0.915,
+      "step": 4943
+    },
+    {
+      "epoch": 0.951135051943055,
+      "grad_norm": 1.9906115531921387,
+      "learning_rate": 1.2497548124597024e-06,
+      "loss": 0.9433,
+      "step": 4944
+    },
+    {
+      "epoch": 0.9513274336283186,
+      "grad_norm": 2.088672637939453,
+      "learning_rate": 1.2399539417658369e-06,
+      "loss": 0.9358,
+      "step": 4945
+    },
+    {
+      "epoch": 0.9515198153135821,
+      "grad_norm": 2.006319761276245,
+      "learning_rate": 1.2301914131222726e-06,
+      "loss": 0.926,
+      "step": 4946
+    },
+    {
+      "epoch": 0.9517121969988457,
+      "grad_norm": 2.4258556365966797,
+      "learning_rate": 1.2204672303191334e-06,
+      "loss": 0.8946,
+      "step": 4947
+    },
+    {
+      "epoch": 0.9519045786841093,
+      "grad_norm": 2.443539619445801,
+      "learning_rate": 1.2107813971317105e-06,
+      "loss": 0.8551,
+      "step": 4948
+    },
+    {
+      "epoch": 0.9520969603693729,
+      "grad_norm": 2.145639419555664,
+      "learning_rate": 1.2011339173203628e-06,
+      "loss": 0.9031,
+      "step": 4949
+    },
+    {
+      "epoch": 0.9522893420546364,
+      "grad_norm": 2.2897133827209473,
+      "learning_rate": 1.1915247946305496e-06,
+      "loss": 0.9873,
+      "step": 4950
+    },
+    {
+      "epoch": 0.9524817237398999,
+      "grad_norm": 2.2235584259033203,
+      "learning_rate": 1.181954032792909e-06,
+      "loss": 0.9177,
+      "step": 4951
+    },
+    {
+      "epoch": 0.9526741054251635,
+      "grad_norm": 2.8551974296569824,
+      "learning_rate": 1.1724216355231022e-06,
+      "loss": 0.9432,
+      "step": 4952
+    },
+    {
+      "epoch": 0.9528664871104271,
+      "grad_norm": 2.8836395740509033,
+      "learning_rate": 1.1629276065219575e-06,
+      "loss": 0.9231,
+      "step": 4953
+    },
+    {
+      "epoch": 0.9530588687956907,
+      "grad_norm": 2.5728259086608887,
+      "learning_rate": 1.153471949475382e-06,
+      "loss": 0.991,
+      "step": 4954
+    },
+    {
+      "epoch": 0.9532512504809542,
+      "grad_norm": 2.415388822555542,
+      "learning_rate": 1.1440546680543728e-06,
+      "loss": 0.9491,
+      "step": 4955
+    },
+    {
+      "epoch": 0.9534436321662177,
+      "grad_norm": 2.348055839538574,
+      "learning_rate": 1.1346757659150497e-06,
+      "loss": 0.9413,
+      "step": 4956
+    },
+    {
+      "epoch": 0.9536360138514813,
+      "grad_norm": 2.6180596351623535,
+      "learning_rate": 1.1253352466986334e-06,
+      "loss": 0.9374,
+      "step": 4957
+    },
+    {
+      "epoch": 0.9538283955367449,
+      "grad_norm": 2.095935344696045,
+      "learning_rate": 1.1160331140314339e-06,
+      "loss": 0.9461,
+      "step": 4958
+    },
+    {
+      "epoch": 0.9540207772220085,
+      "grad_norm": 1.177519679069519,
+      "learning_rate": 1.1067693715248406e-06,
+      "loss": 0.9093,
+      "step": 4959
+    },
+    {
+      "epoch": 0.954213158907272,
+      "grad_norm": 2.3667423725128174,
+      "learning_rate": 1.0975440227753764e-06,
+      "loss": 0.9501,
+      "step": 4960
+    },
+    {
+      "epoch": 0.9544055405925356,
+      "grad_norm": 2.156726837158203,
+      "learning_rate": 1.0883570713646319e-06,
+      "loss": 0.9211,
+      "step": 4961
+    },
+    {
+      "epoch": 0.9545979222777992,
+      "grad_norm": 2.122968912124634,
+      "learning_rate": 1.0792085208593094e-06,
+      "loss": 0.947,
+      "step": 4962
+    },
+    {
+      "epoch": 0.9547903039630627,
+      "grad_norm": 2.5612072944641113,
+      "learning_rate": 1.0700983748111793e-06,
+      "loss": 0.8838,
+      "step": 4963
+    },
+    {
+      "epoch": 0.9549826856483262,
+      "grad_norm": 2.9023077487945557,
+      "learning_rate": 1.0610266367571009e-06,
+      "loss": 0.9288,
+      "step": 4964
+    },
+    {
+      "epoch": 0.9551750673335898,
+      "grad_norm": 2.415797233581543,
+      "learning_rate": 1.0519933102190683e-06,
+      "loss": 0.9389,
+      "step": 4965
+    },
+    {
+      "epoch": 0.9553674490188534,
+      "grad_norm": 2.1494102478027344,
+      "learning_rate": 1.0429983987041092e-06,
+      "loss": 0.9478,
+      "step": 4966
+    },
+    {
+      "epoch": 0.955559830704117,
+      "grad_norm": 2.7154459953308105,
+      "learning_rate": 1.0340419057043526e-06,
+      "loss": 0.9416,
+      "step": 4967
+    },
+    {
+      "epoch": 0.9557522123893806,
+      "grad_norm": 1.974350929260254,
+      "learning_rate": 1.0251238346970393e-06,
+      "loss": 0.9512,
+      "step": 4968
+    },
+    {
+      "epoch": 0.955944594074644,
+      "grad_norm": 2.5912985801696777,
+      "learning_rate": 1.0162441891444441e-06,
+      "loss": 0.9524,
+      "step": 4969
+    },
+    {
+      "epoch": 0.9561369757599076,
+      "grad_norm": 2.5755326747894287,
+      "learning_rate": 1.007402972493976e-06,
+      "loss": 0.9146,
+      "step": 4970
+    },
+    {
+      "epoch": 0.9563293574451712,
+      "grad_norm": 2.501891613006592,
+      "learning_rate": 9.986001881780782e-07,
+      "loss": 0.8897,
+      "step": 4971
+    },
+    {
+      "epoch": 0.9565217391304348,
+      "grad_norm": 1.5602517127990723,
+      "learning_rate": 9.898358396143171e-07,
+      "loss": 0.7885,
+      "step": 4972
+    },
+    {
+      "epoch": 0.9567141208156984,
+      "grad_norm": 2.212331771850586,
+      "learning_rate": 9.811099302052929e-07,
+      "loss": 0.9002,
+      "step": 4973
+    },
+    {
+      "epoch": 0.9569065025009619,
+      "grad_norm": 2.2407288551330566,
+      "learning_rate": 9.72422463338718e-07,
+      "loss": 0.8987,
+      "step": 4974
+    },
+    {
+      "epoch": 0.9570988841862255,
+      "grad_norm": 2.250364303588867,
+      "learning_rate": 9.637734423873613e-07,
+      "loss": 0.8848,
+      "step": 4975
+    },
+    {
+      "epoch": 0.957291265871489,
+      "grad_norm": 2.405254602432251,
+      "learning_rate": 9.551628707090809e-07,
+      "loss": 0.8747,
+      "step": 4976
+    },
+    {
+      "epoch": 0.9574836475567526,
+      "grad_norm": 2.696540117263794,
+      "learning_rate": 9.465907516467698e-07,
+      "loss": 0.9548,
+      "step": 4977
+    },
+    {
+      "epoch": 0.9576760292420161,
+      "grad_norm": 1.5306111574172974,
+      "learning_rate": 9.380570885284545e-07,
+      "loss": 0.8097,
+      "step": 4978
+    },
+    {
+      "epoch": 0.9578684109272797,
+      "grad_norm": 1.1005927324295044,
+      "learning_rate": 9.295618846671738e-07,
+      "loss": 1.0048,
+      "step": 4979
+    },
+    {
+      "epoch": 0.9580607926125433,
+      "grad_norm": 1.2212982177734375,
+      "learning_rate": 9.211051433610674e-07,
+      "loss": 0.9602,
+      "step": 4980
+    },
+    {
+      "epoch": 0.9582531742978069,
+      "grad_norm": 1.7547104358673096,
+      "learning_rate": 9.126868678933198e-07,
+      "loss": 0.9242,
+      "step": 4981
+    },
+    {
+      "epoch": 0.9584455559830705,
+      "grad_norm": 1.925552248954773,
+      "learning_rate": 9.043070615322169e-07,
+      "loss": 0.9101,
+      "step": 4982
+    },
+    {
+      "epoch": 0.9586379376683339,
+      "grad_norm": 2.3609254360198975,
+      "learning_rate": 8.959657275310674e-07,
+      "loss": 0.9452,
+      "step": 4983
+    },
+    {
+      "epoch": 0.9588303193535975,
+      "grad_norm": 2.653265953063965,
+      "learning_rate": 8.876628691282918e-07,
+      "loss": 0.9663,
+      "step": 4984
+    },
+    {
+      "epoch": 0.9590227010388611,
+      "grad_norm": 0.8292410969734192,
+      "learning_rate": 8.793984895473117e-07,
+      "loss": 0.8929,
+      "step": 4985
+    },
+    {
+      "epoch": 0.9592150827241247,
+      "grad_norm": 2.686002254486084,
+      "learning_rate": 8.711725919966718e-07,
+      "loss": 0.928,
+      "step": 4986
+    },
+    {
+      "epoch": 0.9594074644093882,
+      "grad_norm": 2.3383848667144775,
+      "learning_rate": 8.629851796699284e-07,
+      "loss": 0.8423,
+      "step": 4987
+    },
+    {
+      "epoch": 0.9595998460946518,
+      "grad_norm": 2.385464906692505,
+      "learning_rate": 8.548362557457279e-07,
+      "loss": 0.8945,
+      "step": 4988
+    },
+    {
+      "epoch": 0.9597922277799154,
+      "grad_norm": 2.9829630851745605,
+      "learning_rate": 8.467258233877728e-07,
+      "loss": 0.8882,
+      "step": 4989
+    },
+    {
+      "epoch": 0.9599846094651789,
+      "grad_norm": 2.640317916870117,
+      "learning_rate": 8.386538857447779e-07,
+      "loss": 0.9964,
+      "step": 4990
+    },
+    {
+      "epoch": 0.9601769911504425,
+      "grad_norm": 2.3079147338867188,
+      "learning_rate": 8.306204459505806e-07,
+      "loss": 0.9099,
+      "step": 4991
+    },
+    {
+      "epoch": 0.960369372835706,
+      "grad_norm": 2.1737005710601807,
+      "learning_rate": 8.226255071240308e-07,
+      "loss": 0.9089,
+      "step": 4992
+    },
+    {
+      "epoch": 0.9605617545209696,
+      "grad_norm": 2.8816497325897217,
+      "learning_rate": 8.146690723690342e-07,
+      "loss": 0.9662,
+      "step": 4993
+    },
+    {
+      "epoch": 0.9607541362062332,
+      "grad_norm": 2.073428153991699,
+      "learning_rate": 8.067511447745535e-07,
+      "loss": 0.9812,
+      "step": 4994
+    },
+    {
+      "epoch": 0.9609465178914968,
+      "grad_norm": 2.4275424480438232,
+      "learning_rate": 7.988717274146074e-07,
+      "loss": 0.8859,
+      "step": 4995
+    },
+    {
+      "epoch": 0.9611388995767602,
+      "grad_norm": 1.1688021421432495,
+      "learning_rate": 7.910308233482488e-07,
+      "loss": 0.9627,
+      "step": 4996
+    },
+    {
+      "epoch": 0.9613312812620238,
+      "grad_norm": 2.465407133102417,
+      "learning_rate": 7.832284356195762e-07,
+      "loss": 0.9498,
+      "step": 4997
+    },
+    {
+      "epoch": 0.9615236629472874,
+      "grad_norm": 2.182365894317627,
+      "learning_rate": 7.754645672577776e-07,
+      "loss": 0.8852,
+      "step": 4998
+    },
+    {
+      "epoch": 0.961716044632551,
+      "grad_norm": 1.9663063287734985,
+      "learning_rate": 7.677392212770196e-07,
+      "loss": 0.842,
+      "step": 4999
+    },
+    {
+      "epoch": 0.9619084263178146,
+      "grad_norm": 1.7530165910720825,
+      "learning_rate": 7.600524006765808e-07,
+      "loss": 0.9089,
+      "step": 5000
+    },
+    {
+      "epoch": 0.9621008080030781,
+      "grad_norm": 2.201231002807617,
+      "learning_rate": 7.524041084407185e-07,
+      "loss": 0.8553,
+      "step": 5001
+    },
+    {
+      "epoch": 0.9622931896883417,
+      "grad_norm": 1.437628149986267,
+      "learning_rate": 7.447943475387797e-07,
+      "loss": 0.9551,
+      "step": 5002
+    },
+    {
+      "epoch": 0.9624855713736052,
+      "grad_norm": 2.2981760501861572,
+      "learning_rate": 7.372231209251346e-07,
+      "loss": 0.9158,
+      "step": 5003
+    },
+    {
+      "epoch": 0.9626779530588688,
+      "grad_norm": 2.220031261444092,
+      "learning_rate": 7.296904315391873e-07,
+      "loss": 0.959,
+      "step": 5004
+    },
+    {
+      "epoch": 0.9628703347441323,
+      "grad_norm": 2.844444513320923,
+      "learning_rate": 7.221962823053874e-07,
+      "loss": 0.9567,
+      "step": 5005
+    },
+    {
+      "epoch": 0.9630627164293959,
+      "grad_norm": 2.3197004795074463,
+      "learning_rate": 7.147406761332298e-07,
+      "loss": 0.874,
+      "step": 5006
+    },
+    {
+      "epoch": 0.9632550981146595,
+      "grad_norm": 2.263298749923706,
+      "learning_rate": 7.073236159172325e-07,
+      "loss": 0.9206,
+      "step": 5007
+    },
+    {
+      "epoch": 0.9634474797999231,
+      "grad_norm": 1.9995230436325073,
+      "learning_rate": 6.999451045369587e-07,
+      "loss": 0.9242,
+      "step": 5008
+    },
+    {
+      "epoch": 0.9636398614851867,
+      "grad_norm": 1.842055082321167,
+      "learning_rate": 6.926051448569948e-07,
+      "loss": 0.9927,
+      "step": 5009
+    },
+    {
+      "epoch": 0.9638322431704501,
+      "grad_norm": 2.469576597213745,
+      "learning_rate": 6.853037397269724e-07,
+      "loss": 0.8878,
+      "step": 5010
+    },
+    {
+      "epoch": 0.9640246248557137,
+      "grad_norm": 2.1380419731140137,
+      "learning_rate": 6.78040891981524e-07,
+      "loss": 0.958,
+      "step": 5011
+    },
+    {
+      "epoch": 0.9642170065409773,
+      "grad_norm": 1.3453749418258667,
+      "learning_rate": 6.70816604440383e-07,
+      "loss": 0.9945,
+      "step": 5012
+    },
+    {
+      "epoch": 0.9644093882262409,
+      "grad_norm": 2.9267704486846924,
+      "learning_rate": 6.63630879908217e-07,
+      "loss": 0.9741,
+      "step": 5013
+    },
+    {
+      "epoch": 0.9646017699115044,
+      "grad_norm": 2.2856533527374268,
+      "learning_rate": 6.564837211748054e-07,
+      "loss": 0.8716,
+      "step": 5014
+    },
+    {
+      "epoch": 0.964794151596768,
+      "grad_norm": 2.1723334789276123,
+      "learning_rate": 6.493751310149176e-07,
+      "loss": 0.9699,
+      "step": 5015
+    },
+    {
+      "epoch": 0.9649865332820315,
+      "grad_norm": 2.52671217918396,
+      "learning_rate": 6.42305112188335e-07,
+      "loss": 0.9826,
+      "step": 5016
+    },
+    {
+      "epoch": 0.9651789149672951,
+      "grad_norm": 2.6194190979003906,
+      "learning_rate": 6.352736674398951e-07,
+      "loss": 0.9326,
+      "step": 5017
+    },
+    {
+      "epoch": 0.9653712966525587,
+      "grad_norm": 1.4757956266403198,
+      "learning_rate": 6.282807994994478e-07,
+      "loss": 0.9029,
+      "step": 5018
+    },
+    {
+      "epoch": 0.9655636783378222,
+      "grad_norm": 2.274352550506592,
+      "learning_rate": 6.213265110818655e-07,
+      "loss": 0.9536,
+      "step": 5019
+    },
+    {
+      "epoch": 0.9657560600230858,
+      "grad_norm": 2.5611472129821777,
+      "learning_rate": 6.144108048870333e-07,
+      "loss": 0.9368,
+      "step": 5020
+    },
+    {
+      "epoch": 0.9659484417083494,
+      "grad_norm": 2.4604039192199707,
+      "learning_rate": 6.075336835998812e-07,
+      "loss": 0.9475,
+      "step": 5021
+    },
+    {
+      "epoch": 0.966140823393613,
+      "grad_norm": 1.4330271482467651,
+      "learning_rate": 6.00695149890329e-07,
+      "loss": 0.7842,
+      "step": 5022
+    },
+    {
+      "epoch": 0.9663332050788765,
+      "grad_norm": 2.692432403564453,
+      "learning_rate": 5.938952064133418e-07,
+      "loss": 0.8978,
+      "step": 5023
+    },
+    {
+      "epoch": 0.96652558676414,
+      "grad_norm": 2.278808832168579,
+      "learning_rate": 5.871338558088857e-07,
+      "loss": 0.9757,
+      "step": 5024
+    },
+    {
+      "epoch": 0.9667179684494036,
+      "grad_norm": 2.1004724502563477,
+      "learning_rate": 5.804111007019497e-07,
+      "loss": 0.9456,
+      "step": 5025
+    },
+    {
+      "epoch": 0.9669103501346672,
+      "grad_norm": 2.3045122623443604,
+      "learning_rate": 5.737269437025461e-07,
+      "loss": 0.9824,
+      "step": 5026
+    },
+    {
+      "epoch": 0.9671027318199308,
+      "grad_norm": 2.5523853302001953,
+      "learning_rate": 5.67081387405688e-07,
+      "loss": 0.8491,
+      "step": 5027
+    },
+    {
+      "epoch": 0.9672951135051943,
+      "grad_norm": 2.437584161758423,
+      "learning_rate": 5.604744343914115e-07,
+      "loss": 0.9335,
+      "step": 5028
+    },
+    {
+      "epoch": 0.9674874951904578,
+      "grad_norm": 2.2010138034820557,
+      "learning_rate": 5.539060872247537e-07,
+      "loss": 0.9405,
+      "step": 5029
+    },
+    {
+      "epoch": 0.9676798768757214,
+      "grad_norm": 2.280245065689087,
+      "learning_rate": 5.47376348455797e-07,
+      "loss": 0.9275,
+      "step": 5030
+    },
+    {
+      "epoch": 0.967872258560985,
+      "grad_norm": 1.3288160562515259,
+      "learning_rate": 5.408852206195913e-07,
+      "loss": 0.7517,
+      "step": 5031
+    },
+    {
+      "epoch": 0.9680646402462486,
+      "grad_norm": 2.4154670238494873,
+      "learning_rate": 5.344327062362098e-07,
+      "loss": 0.8593,
+      "step": 5032
+    },
+    {
+      "epoch": 0.9682570219315121,
+      "grad_norm": 1.511285662651062,
+      "learning_rate": 5.280188078107595e-07,
+      "loss": 0.7665,
+      "step": 5033
+    },
+    {
+      "epoch": 0.9684494036167757,
+      "grad_norm": 2.278862476348877,
+      "learning_rate": 5.216435278333376e-07,
+      "loss": 0.8955,
+      "step": 5034
+    },
+    {
+      "epoch": 0.9686417853020393,
+      "grad_norm": 1.1495243310928345,
+      "learning_rate": 5.153068687790197e-07,
+      "loss": 0.9731,
+      "step": 5035
+    },
+    {
+      "epoch": 0.9688341669873028,
+      "grad_norm": 2.6274821758270264,
+      "learning_rate": 5.09008833107949e-07,
+      "loss": 0.9626,
+      "step": 5036
+    },
+    {
+      "epoch": 0.9690265486725663,
+      "grad_norm": 1.9402812719345093,
+      "learning_rate": 5.027494232652142e-07,
+      "loss": 0.9159,
+      "step": 5037
+    },
+    {
+      "epoch": 0.9692189303578299,
+      "grad_norm": 2.4957430362701416,
+      "learning_rate": 4.9652864168096e-07,
+      "loss": 0.8784,
+      "step": 5038
+    },
+    {
+      "epoch": 0.9694113120430935,
+      "grad_norm": 2.588534355163574,
+      "learning_rate": 4.90346490770277e-07,
+      "loss": 0.8885,
+      "step": 5039
+    },
+    {
+      "epoch": 0.9696036937283571,
+      "grad_norm": 1.1987226009368896,
+      "learning_rate": 4.84202972933312e-07,
+      "loss": 0.9561,
+      "step": 5040
+    },
+    {
+      "epoch": 0.9697960754136207,
+      "grad_norm": 1.2796447277069092,
+      "learning_rate": 4.78098090555179e-07,
+      "loss": 0.9972,
+      "step": 5041
+    },
+    {
+      "epoch": 0.9699884570988841,
+      "grad_norm": 2.3980064392089844,
+      "learning_rate": 4.720318460060047e-07,
+      "loss": 0.9917,
+      "step": 5042
+    },
+    {
+      "epoch": 0.9701808387841477,
+      "grad_norm": 2.541720390319824,
+      "learning_rate": 4.6600424164091605e-07,
+      "loss": 0.9233,
+      "step": 5043
+    },
+    {
+      "epoch": 0.9703732204694113,
+      "grad_norm": 2.373028516769409,
+      "learning_rate": 4.6001527980004124e-07,
+      "loss": 0.8695,
+      "step": 5044
+    },
+    {
+      "epoch": 0.9705656021546749,
+      "grad_norm": 2.2226803302764893,
+      "learning_rate": 4.54064962808487e-07,
+      "loss": 0.9133,
+      "step": 5045
+    },
+    {
+      "epoch": 0.9707579838399384,
+      "grad_norm": 1.9714189767837524,
+      "learning_rate": 4.481532929763943e-07,
+      "loss": 0.9298,
+      "step": 5046
+    },
+    {
+      "epoch": 0.970950365525202,
+      "grad_norm": 2.95004940032959,
+      "learning_rate": 4.422802725988606e-07,
+      "loss": 0.9124,
+      "step": 5047
+    },
+    {
+      "epoch": 0.9711427472104656,
+      "grad_norm": 3.1210529804229736,
+      "learning_rate": 4.364459039559843e-07,
+      "loss": 0.9116,
+      "step": 5048
+    },
+    {
+      "epoch": 0.9713351288957291,
+      "grad_norm": 2.382300615310669,
+      "learning_rate": 4.306501893128978e-07,
+      "loss": 0.9726,
+      "step": 5049
+    },
+    {
+      "epoch": 0.9715275105809927,
+      "grad_norm": 2.32368803024292,
+      "learning_rate": 4.2489313091967906e-07,
+      "loss": 0.9367,
+      "step": 5050
+    },
+    {
+      "epoch": 0.9717198922662562,
+      "grad_norm": 1.8623225688934326,
+      "learning_rate": 4.191747310114069e-07,
+      "loss": 0.8959,
+      "step": 5051
+    },
+    {
+      "epoch": 0.9719122739515198,
+      "grad_norm": 2.280378818511963,
+      "learning_rate": 4.134949918081832e-07,
+      "loss": 0.8915,
+      "step": 5052
+    },
+    {
+      "epoch": 0.9721046556367834,
+      "grad_norm": 1.8166823387145996,
+      "learning_rate": 4.0785391551506623e-07,
+      "loss": 0.9075,
+      "step": 5053
+    },
+    {
+      "epoch": 0.972297037322047,
+      "grad_norm": 2.836430549621582,
+      "learning_rate": 4.0225150432211535e-07,
+      "loss": 0.9204,
+      "step": 5054
+    },
+    {
+      "epoch": 0.9724894190073105,
+      "grad_norm": 2.801870822906494,
+      "learning_rate": 3.9668776040437947e-07,
+      "loss": 0.8874,
+      "step": 5055
+    },
+    {
+      "epoch": 0.972681800692574,
+      "grad_norm": 2.686314582824707,
+      "learning_rate": 3.9116268592189755e-07,
+      "loss": 0.9433,
+      "step": 5056
+    },
+    {
+      "epoch": 0.9728741823778376,
+      "grad_norm": 1.9863234758377075,
+      "learning_rate": 3.8567628301969803e-07,
+      "loss": 0.9699,
+      "step": 5057
+    },
+    {
+      "epoch": 0.9730665640631012,
+      "grad_norm": 1.9622169733047485,
+      "learning_rate": 3.8022855382777724e-07,
+      "loss": 0.9559,
+      "step": 5058
+    },
+    {
+      "epoch": 0.9732589457483648,
+      "grad_norm": 1.1960387229919434,
+      "learning_rate": 3.7481950046115433e-07,
+      "loss": 0.8773,
+      "step": 5059
+    },
+    {
+      "epoch": 0.9734513274336283,
+      "grad_norm": 1.8682823181152344,
+      "learning_rate": 3.69449125019794e-07,
+      "loss": 0.8939,
+      "step": 5060
+    },
+    {
+      "epoch": 0.9736437091188919,
+      "grad_norm": 2.137162208557129,
+      "learning_rate": 3.641174295886618e-07,
+      "loss": 0.8997,
+      "step": 5061
+    },
+    {
+      "epoch": 0.9738360908041555,
+      "grad_norm": 2.608020305633545,
+      "learning_rate": 3.588244162377019e-07,
+      "loss": 0.9559,
+      "step": 5062
+    },
+    {
+      "epoch": 0.974028472489419,
+      "grad_norm": 1.4021375179290771,
+      "learning_rate": 3.5357008702185944e-07,
+      "loss": 0.7749,
+      "step": 5063
+    },
+    {
+      "epoch": 0.9742208541746825,
+      "grad_norm": 1.1271320581436157,
+      "learning_rate": 3.483544439810249e-07,
+      "loss": 0.8353,
+      "step": 5064
+    },
+    {
+      "epoch": 0.9744132358599461,
+      "grad_norm": 2.215437889099121,
+      "learning_rate": 3.431774891401118e-07,
+      "loss": 0.8815,
+      "step": 5065
+    },
+    {
+      "epoch": 0.9746056175452097,
+      "grad_norm": 1.8049120903015137,
+      "learning_rate": 3.380392245089792e-07,
+      "loss": 0.8818,
+      "step": 5066
+    },
+    {
+      "epoch": 0.9747979992304733,
+      "grad_norm": 2.391051769256592,
+      "learning_rate": 3.329396520824757e-07,
+      "loss": 0.941,
+      "step": 5067
+    },
+    {
+      "epoch": 0.9749903809157369,
+      "grad_norm": 2.55454421043396,
+      "learning_rate": 3.2787877384045096e-07,
+      "loss": 0.8701,
+      "step": 5068
+    },
+    {
+      "epoch": 0.9751827626010003,
+      "grad_norm": 2.324885845184326,
+      "learning_rate": 3.228565917476889e-07,
+      "loss": 0.952,
+      "step": 5069
+    },
+    {
+      "epoch": 0.9753751442862639,
+      "grad_norm": 2.6207950115203857,
+      "learning_rate": 3.178731077539743e-07,
+      "loss": 0.9621,
+      "step": 5070
+    },
+    {
+      "epoch": 0.9755675259715275,
+      "grad_norm": 2.277364492416382,
+      "learning_rate": 3.1292832379409274e-07,
+      "loss": 0.9682,
+      "step": 5071
+    },
+    {
+      "epoch": 0.9757599076567911,
+      "grad_norm": 2.52128529548645,
+      "learning_rate": 3.080222417877421e-07,
+      "loss": 0.9069,
+      "step": 5072
+    },
+    {
+      "epoch": 0.9759522893420547,
+      "grad_norm": 2.8291821479797363,
+      "learning_rate": 3.031548636396764e-07,
+      "loss": 0.9373,
+      "step": 5073
+    },
+    {
+      "epoch": 0.9761446710273182,
+      "grad_norm": 2.3888540267944336,
+      "learning_rate": 2.9832619123953965e-07,
+      "loss": 0.9559,
+      "step": 5074
+    },
+    {
+      "epoch": 0.9763370527125818,
+      "grad_norm": 3.0157101154327393,
+      "learning_rate": 2.9353622646199897e-07,
+      "loss": 0.9084,
+      "step": 5075
+    },
+    {
+      "epoch": 0.9765294343978453,
+      "grad_norm": 2.341093063354492,
+      "learning_rate": 2.8878497116671123e-07,
+      "loss": 0.9073,
+      "step": 5076
+    },
+    {
+      "epoch": 0.9767218160831089,
+      "grad_norm": 2.0678749084472656,
+      "learning_rate": 2.8407242719823424e-07,
+      "loss": 0.9955,
+      "step": 5077
+    },
+    {
+      "epoch": 0.9769141977683724,
+      "grad_norm": 2.7600972652435303,
+      "learning_rate": 2.793985963861712e-07,
+      "loss": 0.9611,
+      "step": 5078
+    },
+    {
+      "epoch": 0.977106579453636,
+      "grad_norm": 2.7944655418395996,
+      "learning_rate": 2.747634805450483e-07,
+      "loss": 0.918,
+      "step": 5079
+    },
+    {
+      "epoch": 0.9772989611388996,
+      "grad_norm": 1.290582537651062,
+      "learning_rate": 2.7016708147439286e-07,
+      "loss": 0.931,
+      "step": 5080
+    },
+    {
+      "epoch": 0.9774913428241632,
+      "grad_norm": 2.391841173171997,
+      "learning_rate": 2.656094009586663e-07,
+      "loss": 0.9787,
+      "step": 5081
+    },
+    {
+      "epoch": 0.9776837245094268,
+      "grad_norm": 1.416449785232544,
+      "learning_rate": 2.6109044076733093e-07,
+      "loss": 0.7832,
+      "step": 5082
+    },
+    {
+      "epoch": 0.9778761061946902,
+      "grad_norm": 1.9877935647964478,
+      "learning_rate": 2.5661020265479454e-07,
+      "loss": 0.9537,
+      "step": 5083
+    },
+    {
+      "epoch": 0.9780684878799538,
+      "grad_norm": 1.8608965873718262,
+      "learning_rate": 2.521686883604324e-07,
+      "loss": 0.928,
+      "step": 5084
+    },
+    {
+      "epoch": 0.9782608695652174,
+      "grad_norm": 2.526538848876953,
+      "learning_rate": 2.4776589960862075e-07,
+      "loss": 0.9576,
+      "step": 5085
+    },
+    {
+      "epoch": 0.978453251250481,
+      "grad_norm": 2.496432065963745,
+      "learning_rate": 2.434018381086589e-07,
+      "loss": 0.9343,
+      "step": 5086
+    },
+    {
+      "epoch": 0.9786456329357445,
+      "grad_norm": 1.0915898084640503,
+      "learning_rate": 2.3907650555481385e-07,
+      "loss": 0.8994,
+      "step": 5087
+    },
+    {
+      "epoch": 0.9788380146210081,
+      "grad_norm": 1.9284701347351074,
+      "learning_rate": 2.3478990362634234e-07,
+      "loss": 0.9094,
+      "step": 5088
+    },
+    {
+      "epoch": 0.9790303963062716,
+      "grad_norm": 1.1647907495498657,
+      "learning_rate": 2.3054203398743534e-07,
+      "loss": 0.9548,
+      "step": 5089
+    },
+    {
+      "epoch": 0.9792227779915352,
+      "grad_norm": 1.7298787832260132,
+      "learning_rate": 2.2633289828729586e-07,
+      "loss": 0.9843,
+      "step": 5090
+    },
+    {
+      "epoch": 0.9794151596767988,
+      "grad_norm": 2.6084158420562744,
+      "learning_rate": 2.221624981600168e-07,
+      "loss": 0.9695,
+      "step": 5091
+    },
+    {
+      "epoch": 0.9796075413620623,
+      "grad_norm": 0.9629896283149719,
+      "learning_rate": 2.1803083522471402e-07,
+      "loss": 0.9752,
+      "step": 5092
+    },
+    {
+      "epoch": 0.9797999230473259,
+      "grad_norm": 2.3044626712799072,
+      "learning_rate": 2.139379110854267e-07,
+      "loss": 0.9712,
+      "step": 5093
+    },
+    {
+      "epoch": 0.9799923047325895,
+      "grad_norm": 2.4599239826202393,
+      "learning_rate": 2.0988372733118379e-07,
+      "loss": 0.9504,
+      "step": 5094
+    },
+    {
+      "epoch": 0.9801846864178531,
+      "grad_norm": 2.458465576171875,
+      "learning_rate": 2.058682855359595e-07,
+      "loss": 0.9742,
+      "step": 5095
+    },
+    {
+      "epoch": 0.9803770681031165,
+      "grad_norm": 1.2329429388046265,
+      "learning_rate": 2.0189158725867352e-07,
+      "loss": 0.9185,
+      "step": 5096
+    },
+    {
+      "epoch": 0.9805694497883801,
+      "grad_norm": 2.201972723007202,
+      "learning_rate": 1.979536340432131e-07,
+      "loss": 0.9105,
+      "step": 5097
+    },
+    {
+      "epoch": 0.9807618314736437,
+      "grad_norm": 2.3730764389038086,
+      "learning_rate": 1.9405442741844416e-07,
+      "loss": 0.8162,
+      "step": 5098
+    },
+    {
+      "epoch": 0.9809542131589073,
+      "grad_norm": 3.067450761795044,
+      "learning_rate": 1.9019396889816688e-07,
+      "loss": 0.8951,
+      "step": 5099
+    },
+    {
+      "epoch": 0.9811465948441709,
+      "grad_norm": 2.8725860118865967,
+      "learning_rate": 1.8637225998114906e-07,
+      "loss": 0.9174,
+      "step": 5100
+    },
+    {
+      "epoch": 0.9813389765294344,
+      "grad_norm": 2.6475207805633545,
+      "learning_rate": 1.825893021510927e-07,
+      "loss": 0.9146,
+      "step": 5101
+    },
+    {
+      "epoch": 0.981531358214698,
+      "grad_norm": 2.7061760425567627,
+      "learning_rate": 1.788450968766897e-07,
+      "loss": 0.9648,
+      "step": 5102
+    },
+    {
+      "epoch": 0.9817237398999615,
+      "grad_norm": 2.3510098457336426,
+      "learning_rate": 1.7513964561156615e-07,
+      "loss": 0.9369,
+      "step": 5103
+    },
+    {
+      "epoch": 0.9819161215852251,
+      "grad_norm": 2.377934694290161,
+      "learning_rate": 1.7147294979429352e-07,
+      "loss": 0.9334,
+      "step": 5104
+    },
+    {
+      "epoch": 0.9821085032704886,
+      "grad_norm": 2.4377057552337646,
+      "learning_rate": 1.6784501084843308e-07,
+      "loss": 0.9823,
+      "step": 5105
+    },
+    {
+      "epoch": 0.9823008849557522,
+      "grad_norm": 2.4161105155944824,
+      "learning_rate": 1.6425583018244706e-07,
+      "loss": 0.9236,
+      "step": 5106
+    },
+    {
+      "epoch": 0.9824932666410158,
+      "grad_norm": 2.9955811500549316,
+      "learning_rate": 1.607054091897986e-07,
+      "loss": 0.8495,
+      "step": 5107
+    },
+    {
+      "epoch": 0.9826856483262794,
+      "grad_norm": 1.1155765056610107,
+      "learning_rate": 1.57193749248874e-07,
+      "loss": 0.9443,
+      "step": 5108
+    },
+    {
+      "epoch": 0.982878030011543,
+      "grad_norm": 2.3350820541381836,
+      "learning_rate": 1.5372085172302708e-07,
+      "loss": 0.9831,
+      "step": 5109
+    },
+    {
+      "epoch": 0.9830704116968064,
+      "grad_norm": 2.4474868774414062,
+      "learning_rate": 1.5028671796055715e-07,
+      "loss": 0.9492,
+      "step": 5110
+    },
+    {
+      "epoch": 0.98326279338207,
+      "grad_norm": 2.2202701568603516,
+      "learning_rate": 1.4689134929470882e-07,
+      "loss": 0.961,
+      "step": 5111
+    },
+    {
+      "epoch": 0.9834551750673336,
+      "grad_norm": 2.6017067432403564,
+      "learning_rate": 1.4353474704368319e-07,
+      "loss": 0.9269,
+      "step": 5112
+    },
+    {
+      "epoch": 0.9836475567525972,
+      "grad_norm": 1.6204805374145508,
+      "learning_rate": 1.4021691251062673e-07,
+      "loss": 0.898,
+      "step": 5113
+    },
+    {
+      "epoch": 0.9838399384378607,
+      "grad_norm": 2.5977931022644043,
+      "learning_rate": 1.3693784698363133e-07,
+      "loss": 0.848,
+      "step": 5114
+    },
+    {
+      "epoch": 0.9840323201231242,
+      "grad_norm": 2.0220143795013428,
+      "learning_rate": 1.3369755173575637e-07,
+      "loss": 0.9064,
+      "step": 5115
+    },
+    {
+      "epoch": 0.9842247018083878,
+      "grad_norm": 1.9875364303588867,
+      "learning_rate": 1.304960280249845e-07,
+      "loss": 0.9906,
+      "step": 5116
+    },
+    {
+      "epoch": 0.9844170834936514,
+      "grad_norm": 1.422049880027771,
+      "learning_rate": 1.273332770942659e-07,
+      "loss": 0.957,
+      "step": 5117
+    },
+    {
+      "epoch": 0.984609465178915,
+      "grad_norm": 1.2446074485778809,
+      "learning_rate": 1.2420930017148502e-07,
+      "loss": 0.7705,
+      "step": 5118
+    },
+    {
+      "epoch": 0.9848018468641785,
+      "grad_norm": 1.8918547630310059,
+      "learning_rate": 1.211240984694717e-07,
+      "loss": 0.9478,
+      "step": 5119
+    },
+    {
+      "epoch": 0.9849942285494421,
+      "grad_norm": 1.4438732862472534,
+      "learning_rate": 1.1807767318602336e-07,
+      "loss": 0.9414,
+      "step": 5120
+    },
+    {
+      "epoch": 0.9851866102347057,
+      "grad_norm": 1.3502094745635986,
+      "learning_rate": 1.1507002550386059e-07,
+      "loss": 0.9148,
+      "step": 5121
+    },
+    {
+      "epoch": 0.9853789919199692,
+      "grad_norm": 1.8079582452774048,
+      "learning_rate": 1.1210115659063825e-07,
+      "loss": 0.8751,
+      "step": 5122
+    },
+    {
+      "epoch": 0.9855713736052328,
+      "grad_norm": 2.0031282901763916,
+      "learning_rate": 1.0917106759900097e-07,
+      "loss": 0.8508,
+      "step": 5123
+    },
+    {
+      "epoch": 0.9857637552904963,
+      "grad_norm": 3.066633939743042,
+      "learning_rate": 1.062797596664944e-07,
+      "loss": 0.9443,
+      "step": 5124
+    },
+    {
+      "epoch": 0.9859561369757599,
+      "grad_norm": 2.234318971633911,
+      "learning_rate": 1.034272339156206e-07,
+      "loss": 0.951,
+      "step": 5125
+    },
+    {
+      "epoch": 0.9861485186610235,
+      "grad_norm": 1.1173598766326904,
+      "learning_rate": 1.0061349145383813e-07,
+      "loss": 0.9158,
+      "step": 5126
+    },
+    {
+      "epoch": 0.9863409003462871,
+      "grad_norm": 2.578056812286377,
+      "learning_rate": 9.783853337353988e-08,
+      "loss": 0.9713,
+      "step": 5127
+    },
+    {
+      "epoch": 0.9865332820315506,
+      "grad_norm": 1.0793486833572388,
+      "learning_rate": 9.510236075205292e-08,
+      "loss": 0.772,
+      "step": 5128
+    },
+    {
+      "epoch": 0.9867256637168141,
+      "grad_norm": 1.8480361700057983,
+      "learning_rate": 9.240497465164976e-08,
+      "loss": 0.8963,
+      "step": 5129
+    },
+    {
+      "epoch": 0.9869180454020777,
+      "grad_norm": 1.993288516998291,
+      "learning_rate": 8.974637611955938e-08,
+      "loss": 0.9807,
+      "step": 5130
+    },
+    {
+      "epoch": 0.9871104270873413,
+      "grad_norm": 2.120173215866089,
+      "learning_rate": 8.712656618793391e-08,
+      "loss": 0.937,
+      "step": 5131
+    },
+    {
+      "epoch": 0.9873028087726049,
+      "grad_norm": 1.9954184293746948,
+      "learning_rate": 8.454554587388197e-08,
+      "loss": 0.8982,
+      "step": 5132
+    },
+    {
+      "epoch": 0.9874951904578684,
+      "grad_norm": 1.3559672832489014,
+      "learning_rate": 8.200331617943536e-08,
+      "loss": 0.986,
+      "step": 5133
+    },
+    {
+      "epoch": 0.987687572143132,
+      "grad_norm": 2.4161288738250732,
+      "learning_rate": 7.949987809158233e-08,
+      "loss": 0.9383,
+      "step": 5134
+    },
+    {
+      "epoch": 0.9878799538283956,
+      "grad_norm": 1.3261014223098755,
+      "learning_rate": 7.703523258223432e-08,
+      "loss": 0.781,
+      "step": 5135
+    },
+    {
+      "epoch": 0.9880723355136591,
+      "grad_norm": 2.279492139816284,
+      "learning_rate": 7.460938060825929e-08,
+      "loss": 0.9496,
+      "step": 5136
+    },
+    {
+      "epoch": 0.9882647171989226,
+      "grad_norm": 1.18631112575531,
+      "learning_rate": 7.222232311145937e-08,
+      "loss": 0.9751,
+      "step": 5137
+    },
+    {
+      "epoch": 0.9884570988841862,
+      "grad_norm": 2.295994281768799,
+      "learning_rate": 6.987406101855998e-08,
+      "loss": 0.9367,
+      "step": 5138
+    },
+    {
+      "epoch": 0.9886494805694498,
+      "grad_norm": 1.8331881761550903,
+      "learning_rate": 6.756459524125402e-08,
+      "loss": 0.9875,
+      "step": 5139
+    },
+    {
+      "epoch": 0.9888418622547134,
+      "grad_norm": 2.581237316131592,
+      "learning_rate": 6.529392667613543e-08,
+      "loss": 1.0343,
+      "step": 5140
+    },
+    {
+      "epoch": 0.989034243939977,
+      "grad_norm": 1.9359954595565796,
+      "learning_rate": 6.306205620477678e-08,
+      "loss": 0.9129,
+      "step": 5141
+    },
+    {
+      "epoch": 0.9892266256252404,
+      "grad_norm": 2.5234875679016113,
+      "learning_rate": 6.086898469365165e-08,
+      "loss": 0.9372,
+      "step": 5142
+    },
+    {
+      "epoch": 0.989419007310504,
+      "grad_norm": 2.139219284057617,
+      "learning_rate": 5.8714712994190067e-08,
+      "loss": 0.9664,
+      "step": 5143
+    },
+    {
+      "epoch": 0.9896113889957676,
+      "grad_norm": 2.5497467517852783,
+      "learning_rate": 5.659924194276745e-08,
+      "loss": 0.9457,
+      "step": 5144
+    },
+    {
+      "epoch": 0.9898037706810312,
+      "grad_norm": 2.2948012351989746,
+      "learning_rate": 5.4522572360660164e-08,
+      "loss": 0.9318,
+      "step": 5145
+    },
+    {
+      "epoch": 0.9899961523662947,
+      "grad_norm": 1.3281975984573364,
+      "learning_rate": 5.2484705054123286e-08,
+      "loss": 0.7834,
+      "step": 5146
+    },
+    {
+      "epoch": 0.9901885340515583,
+      "grad_norm": 2.4619827270507812,
+      "learning_rate": 5.048564081431284e-08,
+      "loss": 0.997,
+      "step": 5147
+    },
+    {
+      "epoch": 0.9903809157368219,
+      "grad_norm": 2.2308316230773926,
+      "learning_rate": 4.8525380417330236e-08,
+      "loss": 0.9402,
+      "step": 5148
+    },
+    {
+      "epoch": 0.9905732974220854,
+      "grad_norm": 1.8684043884277344,
+      "learning_rate": 4.6603924624244456e-08,
+      "loss": 0.9177,
+      "step": 5149
+    },
+    {
+      "epoch": 0.990765679107349,
+      "grad_norm": 2.598182201385498,
+      "learning_rate": 4.472127418099214e-08,
+      "loss": 0.9272,
+      "step": 5150
+    },
+    {
+      "epoch": 0.9909580607926125,
+      "grad_norm": 2.5570342540740967,
+      "learning_rate": 4.287742981851084e-08,
+      "loss": 0.9314,
+      "step": 5151
+    },
+    {
+      "epoch": 0.9911504424778761,
+      "grad_norm": 1.9099490642547607,
+      "learning_rate": 4.1072392252639036e-08,
+      "loss": 0.9069,
+      "step": 5152
+    },
+    {
+      "epoch": 0.9913428241631397,
+      "grad_norm": 1.185688853263855,
+      "learning_rate": 3.930616218414951e-08,
+      "loss": 0.9401,
+      "step": 5153
+    },
+    {
+      "epoch": 0.9915352058484033,
+      "grad_norm": 1.105296015739441,
+      "learning_rate": 3.7578740298749306e-08,
+      "loss": 0.9932,
+      "step": 5154
+    },
+    {
+      "epoch": 0.9917275875336667,
+      "grad_norm": 2.89141845703125,
+      "learning_rate": 3.5890127267090846e-08,
+      "loss": 0.9575,
+      "step": 5155
+    },
+    {
+      "epoch": 0.9919199692189303,
+      "grad_norm": 2.7442612648010254,
+      "learning_rate": 3.424032374476083e-08,
+      "loss": 0.8948,
+      "step": 5156
+    },
+    {
+      "epoch": 0.9921123509041939,
+      "grad_norm": 1.5300356149673462,
+      "learning_rate": 3.262933037224691e-08,
+      "loss": 0.941,
+      "step": 5157
+    },
+    {
+      "epoch": 0.9923047325894575,
+      "grad_norm": 2.3971714973449707,
+      "learning_rate": 3.105714777501545e-08,
+      "loss": 0.8829,
+      "step": 5158
+    },
+    {
+      "epoch": 0.9924971142747211,
+      "grad_norm": 2.2095775604248047,
+      "learning_rate": 2.9523776563422644e-08,
+      "loss": 0.9199,
+      "step": 5159
+    },
+    {
+      "epoch": 0.9926894959599846,
+      "grad_norm": 2.2314536571502686,
+      "learning_rate": 2.8029217332781188e-08,
+      "loss": 0.9381,
+      "step": 5160
+    },
+    {
+      "epoch": 0.9928818776452482,
+      "grad_norm": 2.41172456741333,
+      "learning_rate": 2.657347066333804e-08,
+      "loss": 0.9327,
+      "step": 5161
+    },
+    {
+      "epoch": 0.9930742593305117,
+      "grad_norm": 2.577918529510498,
+      "learning_rate": 2.5156537120263335e-08,
+      "loss": 0.9646,
+      "step": 5162
+    },
+    {
+      "epoch": 0.9932666410157753,
+      "grad_norm": 2.2619481086730957,
+      "learning_rate": 2.3778417253650376e-08,
+      "loss": 1.0094,
+      "step": 5163
+    },
+    {
+      "epoch": 0.9934590227010388,
+      "grad_norm": 1.2540929317474365,
+      "learning_rate": 2.243911159853784e-08,
+      "loss": 0.9668,
+      "step": 5164
+    },
+    {
+      "epoch": 0.9936514043863024,
+      "grad_norm": 1.9408282041549683,
+      "learning_rate": 2.113862067488759e-08,
+      "loss": 1.0108,
+      "step": 5165
+    },
+    {
+      "epoch": 0.993843786071566,
+      "grad_norm": 2.4251434803009033,
+      "learning_rate": 1.9876944987606838e-08,
+      "loss": 0.9225,
+      "step": 5166
+    },
+    {
+      "epoch": 0.9940361677568296,
+      "grad_norm": 2.609710216522217,
+      "learning_rate": 1.8654085026503787e-08,
+      "loss": 0.9115,
+      "step": 5167
+    },
+    {
+      "epoch": 0.9942285494420932,
+      "grad_norm": 2.7624571323394775,
+      "learning_rate": 1.747004126635421e-08,
+      "loss": 0.9077,
+      "step": 5168
+    },
+    {
+      "epoch": 0.9944209311273566,
+      "grad_norm": 2.2174770832061768,
+      "learning_rate": 1.6324814166823744e-08,
+      "loss": 0.9343,
+      "step": 5169
+    },
+    {
+      "epoch": 0.9946133128126202,
+      "grad_norm": 2.20049786567688,
+      "learning_rate": 1.5218404172545608e-08,
+      "loss": 0.9341,
+      "step": 5170
+    },
+    {
+      "epoch": 0.9948056944978838,
+      "grad_norm": 2.698138952255249,
+      "learning_rate": 1.415081171305399e-08,
+      "loss": 0.9352,
+      "step": 5171
+    },
+    {
+      "epoch": 0.9949980761831474,
+      "grad_norm": 2.7882730960845947,
+      "learning_rate": 1.3122037202828452e-08,
+      "loss": 0.9798,
+      "step": 5172
+    },
+    {
+      "epoch": 0.995190457868411,
+      "grad_norm": 1.43471360206604,
+      "learning_rate": 1.2132081041282827e-08,
+      "loss": 0.9821,
+      "step": 5173
+    },
+    {
+      "epoch": 0.9953828395536745,
+      "grad_norm": 2.6286542415618896,
+      "learning_rate": 1.1180943612754124e-08,
+      "loss": 0.9393,
+      "step": 5174
+    },
+    {
+      "epoch": 0.995575221238938,
+      "grad_norm": 2.9027037620544434,
+      "learning_rate": 1.0268625286491418e-08,
+      "loss": 0.9578,
+      "step": 5175
+    },
+    {
+      "epoch": 0.9957676029242016,
+      "grad_norm": 2.4967503547668457,
+      "learning_rate": 9.39512641668916e-09,
+      "loss": 0.8892,
+      "step": 5176
+    },
+    {
+      "epoch": 0.9959599846094652,
+      "grad_norm": 1.1843159198760986,
+      "learning_rate": 8.560447342487176e-09,
+      "loss": 0.9091,
+      "step": 5177
+    },
+    {
+      "epoch": 0.9961523662947287,
+      "grad_norm": 2.213186740875244,
+      "learning_rate": 7.76458838791516e-09,
+      "loss": 0.9211,
+      "step": 5178
+    },
+    {
+      "epoch": 0.9963447479799923,
+      "grad_norm": 2.051454544067383,
+      "learning_rate": 7.0075498619703865e-09,
+      "loss": 0.8462,
+      "step": 5179
+    },
+    {
+      "epoch": 0.9965371296652559,
+      "grad_norm": 2.7767257690429688,
+      "learning_rate": 6.28933205855109e-09,
+      "loss": 0.9999,
+      "step": 5180
+    },
+    {
+      "epoch": 0.9967295113505195,
+      "grad_norm": 2.277855634689331,
+      "learning_rate": 5.609935256500886e-09,
+      "loss": 0.9425,
+      "step": 5181
+    },
+    {
+      "epoch": 0.996921893035783,
+      "grad_norm": 2.571744918823242,
+      "learning_rate": 4.9693597195865634e-09,
+      "loss": 0.8881,
+      "step": 5182
+    },
+    {
+      "epoch": 0.9971142747210465,
+      "grad_norm": 2.3167688846588135,
+      "learning_rate": 4.367605696486976e-09,
+      "loss": 0.9155,
+      "step": 5183
+    },
+    {
+      "epoch": 0.9973066564063101,
+      "grad_norm": 1.8949366807937622,
+      "learning_rate": 3.8046734208374565e-09,
+      "loss": 0.9577,
+      "step": 5184
+    },
+    {
+      "epoch": 0.9974990380915737,
+      "grad_norm": 2.2429091930389404,
+      "learning_rate": 3.2805631111743062e-09,
+      "loss": 0.9322,
+      "step": 5185
+    },
+    {
+      "epoch": 0.9976914197768373,
+      "grad_norm": 2.2102115154266357,
+      "learning_rate": 2.7952749710014047e-09,
+      "loss": 0.8772,
+      "step": 5186
+    },
+    {
+      "epoch": 0.9978838014621008,
+      "grad_norm": 2.844273805618286,
+      "learning_rate": 2.348809188690293e-09,
+      "loss": 0.9381,
+      "step": 5187
+    },
+    {
+      "epoch": 0.9980761831473643,
+      "grad_norm": 2.582653284072876,
+      "learning_rate": 1.941165937602296e-09,
+      "loss": 0.9737,
+      "step": 5188
+    },
+    {
+      "epoch": 0.9982685648326279,
+      "grad_norm": 1.0740894079208374,
+      "learning_rate": 1.5723453759886041e-09,
+      "loss": 1.0279,
+      "step": 5189
+    },
+    {
+      "epoch": 0.9984609465178915,
+      "grad_norm": 2.4375388622283936,
+      "learning_rate": 1.2423476470346806e-09,
+      "loss": 0.8745,
+      "step": 5190
+    },
+    {
+      "epoch": 0.9986533282031551,
+      "grad_norm": 3.017707586288452,
+      "learning_rate": 9.511728788602625e-10,
+      "loss": 0.9596,
+      "step": 5191
+    },
+    {
+      "epoch": 0.9988457098884186,
+      "grad_norm": 2.202202320098877,
+      "learning_rate": 6.988211845082582e-10,
+      "loss": 0.9924,
+      "step": 5192
+    },
+    {
+      "epoch": 0.9990380915736822,
+      "grad_norm": 1.76643967628479,
+      "learning_rate": 4.852926619447473e-10,
+      "loss": 1.0092,
+      "step": 5193
+    },
+    {
+      "epoch": 0.9992304732589458,
+      "grad_norm": 2.7879955768585205,
+      "learning_rate": 3.1058739408118544e-10,
+      "loss": 0.9827,
+      "step": 5194
+    },
+    {
+      "epoch": 0.9994228549442093,
+      "grad_norm": 2.811441659927368,
+      "learning_rate": 1.7470544874109706e-10,
+      "loss": 0.8846,
+      "step": 5195
+    },
+    {
+      "epoch": 0.9996152366294728,
+      "grad_norm": 1.1568018198013306,
+      "learning_rate": 7.764687866007592e-11,
+      "loss": 0.9188,
+      "step": 5196
+    },
+    {
+      "epoch": 0.9998076183147364,
+      "grad_norm": 2.7783005237579346,
+      "learning_rate": 1.941172155239812e-11,
+      "loss": 0.7327,
+      "step": 5197
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2672364711761475,
+      "learning_rate": 0.0,
+      "loss": 0.8682,
+      "step": 5198
+    },
+    {
+      "epoch": 1.0,
+      "step": 5198,
+      "total_flos": 1.989730174198874e+19,
+      "train_loss": 0.981559110196933,
+      "train_runtime": 19799.2832,
+      "train_samples_per_second": 33.602,
+      "train_steps_per_second": 0.263
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 5198,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 50000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": false,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.989730174198874e+19,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}