diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,32872 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 5.0,
+  "eval_steps": 500,
+  "global_step": 4690,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0010660980810234541,
+      "grad_norm": 5.4528142214861735,
+      "learning_rate": 1.7057569296375268e-07,
+      "loss": 0.837,
+      "step": 1
+    },
+    {
+      "epoch": 0.0021321961620469083,
+      "grad_norm": 5.511675498609771,
+      "learning_rate": 3.4115138592750537e-07,
+      "loss": 0.8508,
+      "step": 2
+    },
+    {
+      "epoch": 0.0031982942430703624,
+      "grad_norm": 5.464511849599797,
+      "learning_rate": 5.11727078891258e-07,
+      "loss": 0.8447,
+      "step": 3
+    },
+    {
+      "epoch": 0.0042643923240938165,
+      "grad_norm": 5.447268946229452,
+      "learning_rate": 6.823027718550107e-07,
+      "loss": 0.836,
+      "step": 4
+    },
+    {
+      "epoch": 0.005330490405117271,
+      "grad_norm": 5.381379257626523,
+      "learning_rate": 8.528784648187634e-07,
+      "loss": 0.8463,
+      "step": 5
+    },
+    {
+      "epoch": 0.006396588486140725,
+      "grad_norm": 5.4441126548183725,
+      "learning_rate": 1.023454157782516e-06,
+      "loss": 0.8514,
+      "step": 6
+    },
+    {
+      "epoch": 0.007462686567164179,
+      "grad_norm": 5.008014360316408,
+      "learning_rate": 1.1940298507462686e-06,
+      "loss": 0.8215,
+      "step": 7
+    },
+    {
+      "epoch": 0.008528784648187633,
+      "grad_norm": 5.000283670760778,
+      "learning_rate": 1.3646055437100215e-06,
+      "loss": 0.8361,
+      "step": 8
+    },
+    {
+      "epoch": 0.009594882729211088,
+      "grad_norm": 3.964243964413472,
+      "learning_rate": 1.5351812366737743e-06,
+      "loss": 0.7919,
+      "step": 9
+    },
+    {
+      "epoch": 0.010660980810234541,
+      "grad_norm": 3.7864879773612823,
+      "learning_rate": 1.7057569296375267e-06,
+      "loss": 0.7889,
+      "step": 10
+    },
+    {
+      "epoch": 0.011727078891257996,
+      "grad_norm": 2.512068820205037,
+      "learning_rate": 1.8763326226012796e-06,
+      "loss": 0.7586,
+      "step": 11
+    },
+    {
+      "epoch": 0.01279317697228145,
+      "grad_norm": 2.259289045541819,
+      "learning_rate": 2.046908315565032e-06,
+      "loss": 0.7527,
+      "step": 12
+    },
+    {
+      "epoch": 0.013859275053304905,
+      "grad_norm": 2.138291149917742,
+      "learning_rate": 2.217484008528785e-06,
+      "loss": 0.7504,
+      "step": 13
+    },
+    {
+      "epoch": 0.014925373134328358,
+      "grad_norm": 1.9179684528050671,
+      "learning_rate": 2.3880597014925373e-06,
+      "loss": 0.7449,
+      "step": 14
+    },
+    {
+      "epoch": 0.015991471215351813,
+      "grad_norm": 2.0680127043179235,
+      "learning_rate": 2.55863539445629e-06,
+      "loss": 0.7272,
+      "step": 15
+    },
+    {
+      "epoch": 0.017057569296375266,
+      "grad_norm": 3.1604717469534447,
+      "learning_rate": 2.729211087420043e-06,
+      "loss": 0.7213,
+      "step": 16
+    },
+    {
+      "epoch": 0.01812366737739872,
+      "grad_norm": 3.278566086414656,
+      "learning_rate": 2.8997867803837954e-06,
+      "loss": 0.7202,
+      "step": 17
+    },
+    {
+      "epoch": 0.019189765458422176,
+      "grad_norm": 3.1700600356728033,
+      "learning_rate": 3.0703624733475486e-06,
+      "loss": 0.7189,
+      "step": 18
+    },
+    {
+      "epoch": 0.02025586353944563,
+      "grad_norm": 3.0282075577668857,
+      "learning_rate": 3.240938166311301e-06,
+      "loss": 0.7239,
+      "step": 19
+    },
+    {
+      "epoch": 0.021321961620469083,
+      "grad_norm": 2.652594115580126,
+      "learning_rate": 3.4115138592750535e-06,
+      "loss": 0.701,
+      "step": 20
+    },
+    {
+      "epoch": 0.022388059701492536,
+      "grad_norm": 1.9034175357194445,
+      "learning_rate": 3.582089552238806e-06,
+      "loss": 0.6933,
+      "step": 21
+    },
+    {
+      "epoch": 0.023454157782515993,
+      "grad_norm": 1.5538529247302069,
+      "learning_rate": 3.752665245202559e-06,
+      "loss": 0.6571,
+      "step": 22
+    },
+    {
+      "epoch": 0.024520255863539446,
+      "grad_norm": 1.2849096876637858,
+      "learning_rate": 3.9232409381663116e-06,
+      "loss": 0.6563,
+      "step": 23
+    },
+    {
+      "epoch": 0.0255863539445629,
+      "grad_norm": 1.0948864208805895,
+      "learning_rate": 4.093816631130064e-06,
+      "loss": 0.6466,
+      "step": 24
+    },
+    {
+      "epoch": 0.026652452025586353,
+      "grad_norm": 1.0624300872894943,
+      "learning_rate": 4.264392324093816e-06,
+      "loss": 0.6424,
+      "step": 25
+    },
+    {
+      "epoch": 0.02771855010660981,
+      "grad_norm": 1.131312650641839,
+      "learning_rate": 4.43496801705757e-06,
+      "loss": 0.6369,
+      "step": 26
+    },
+    {
+      "epoch": 0.028784648187633263,
+      "grad_norm": 1.0990409054108197,
+      "learning_rate": 4.605543710021322e-06,
+      "loss": 0.624,
+      "step": 27
+    },
+    {
+      "epoch": 0.029850746268656716,
+      "grad_norm": 0.9649930973741134,
+      "learning_rate": 4.7761194029850745e-06,
+      "loss": 0.6149,
+      "step": 28
+    },
+    {
+      "epoch": 0.03091684434968017,
+      "grad_norm": 0.7025445114737157,
+      "learning_rate": 4.946695095948828e-06,
+      "loss": 0.6143,
+      "step": 29
+    },
+    {
+      "epoch": 0.031982942430703626,
+      "grad_norm": 0.7153092735896356,
+      "learning_rate": 5.11727078891258e-06,
+      "loss": 0.6087,
+      "step": 30
+    },
+    {
+      "epoch": 0.03304904051172708,
+      "grad_norm": 0.8673451563666814,
+      "learning_rate": 5.2878464818763335e-06,
+      "loss": 0.6062,
+      "step": 31
+    },
+    {
+      "epoch": 0.03411513859275053,
+      "grad_norm": 0.8051803897020771,
+      "learning_rate": 5.458422174840086e-06,
+      "loss": 0.5961,
+      "step": 32
+    },
+    {
+      "epoch": 0.035181236673773986,
+      "grad_norm": 0.6282947287043709,
+      "learning_rate": 5.628997867803838e-06,
+      "loss": 0.5945,
+      "step": 33
+    },
+    {
+      "epoch": 0.03624733475479744,
+      "grad_norm": 0.5284036506183948,
+      "learning_rate": 5.799573560767591e-06,
+      "loss": 0.5873,
+      "step": 34
+    },
+    {
+      "epoch": 0.03731343283582089,
+      "grad_norm": 0.6560915453306406,
+      "learning_rate": 5.970149253731343e-06,
+      "loss": 0.5926,
+      "step": 35
+    },
+    {
+      "epoch": 0.03837953091684435,
+      "grad_norm": 0.6254389118129263,
+      "learning_rate": 6.140724946695097e-06,
+      "loss": 0.5925,
+      "step": 36
+    },
+    {
+      "epoch": 0.039445628997867806,
+      "grad_norm": 0.4747128371854551,
+      "learning_rate": 6.31130063965885e-06,
+      "loss": 0.5789,
+      "step": 37
+    },
+    {
+      "epoch": 0.04051172707889126,
+      "grad_norm": 0.43328964711656565,
+      "learning_rate": 6.481876332622602e-06,
+      "loss": 0.5777,
+      "step": 38
+    },
+    {
+      "epoch": 0.04157782515991471,
+      "grad_norm": 0.5041187156835155,
+      "learning_rate": 6.6524520255863545e-06,
+      "loss": 0.5731,
+      "step": 39
+    },
+    {
+      "epoch": 0.042643923240938165,
+      "grad_norm": 0.4996573987096877,
+      "learning_rate": 6.823027718550107e-06,
+      "loss": 0.5704,
+      "step": 40
+    },
+    {
+      "epoch": 0.04371002132196162,
+      "grad_norm": 0.40178653209472504,
+      "learning_rate": 6.993603411513859e-06,
+      "loss": 0.5687,
+      "step": 41
+    },
+    {
+      "epoch": 0.04477611940298507,
+      "grad_norm": 0.39678565068179716,
+      "learning_rate": 7.164179104477612e-06,
+      "loss": 0.5606,
+      "step": 42
+    },
+    {
+      "epoch": 0.04584221748400853,
+      "grad_norm": 0.3838477339215644,
+      "learning_rate": 7.334754797441366e-06,
+      "loss": 0.5551,
+      "step": 43
+    },
+    {
+      "epoch": 0.046908315565031986,
+      "grad_norm": 0.39376377958875114,
+      "learning_rate": 7.505330490405118e-06,
+      "loss": 0.5586,
+      "step": 44
+    },
+    {
+      "epoch": 0.04797441364605544,
+      "grad_norm": 0.4089383682673747,
+      "learning_rate": 7.67590618336887e-06,
+      "loss": 0.544,
+      "step": 45
+    },
+    {
+      "epoch": 0.04904051172707889,
+      "grad_norm": 0.3674905464499231,
+      "learning_rate": 7.846481876332623e-06,
+      "loss": 0.5562,
+      "step": 46
+    },
+    {
+      "epoch": 0.050106609808102345,
+      "grad_norm": 0.3391811295227044,
+      "learning_rate": 8.017057569296376e-06,
+      "loss": 0.5489,
+      "step": 47
+    },
+    {
+      "epoch": 0.0511727078891258,
+      "grad_norm": 0.3771487479148055,
+      "learning_rate": 8.187633262260128e-06,
+      "loss": 0.5553,
+      "step": 48
+    },
+    {
+      "epoch": 0.05223880597014925,
+      "grad_norm": 0.3264015224929875,
+      "learning_rate": 8.35820895522388e-06,
+      "loss": 0.5505,
+      "step": 49
+    },
+    {
+      "epoch": 0.053304904051172705,
+      "grad_norm": 0.3331723647408018,
+      "learning_rate": 8.528784648187633e-06,
+      "loss": 0.5357,
+      "step": 50
+    },
+    {
+      "epoch": 0.054371002132196165,
+      "grad_norm": 0.37485261764922745,
+      "learning_rate": 8.699360341151387e-06,
+      "loss": 0.5441,
+      "step": 51
+    },
+    {
+      "epoch": 0.05543710021321962,
+      "grad_norm": 0.9120774222301191,
+      "learning_rate": 8.86993603411514e-06,
+      "loss": 0.5324,
+      "step": 52
+    },
+    {
+      "epoch": 0.05650319829424307,
+      "grad_norm": 0.3263354859744661,
+      "learning_rate": 9.040511727078892e-06,
+      "loss": 0.5352,
+      "step": 53
+    },
+    {
+      "epoch": 0.057569296375266525,
+      "grad_norm": 0.35885753984901503,
+      "learning_rate": 9.211087420042644e-06,
+      "loss": 0.5405,
+      "step": 54
+    },
+    {
+      "epoch": 0.05863539445628998,
+      "grad_norm": 0.29649061716771913,
+      "learning_rate": 9.381663113006397e-06,
+      "loss": 0.5361,
+      "step": 55
+    },
+    {
+      "epoch": 0.05970149253731343,
+      "grad_norm": 0.24951946598497796,
+      "learning_rate": 9.552238805970149e-06,
+      "loss": 0.5301,
+      "step": 56
+    },
+    {
+      "epoch": 0.060767590618336885,
+      "grad_norm": 0.3772047271710351,
+      "learning_rate": 9.722814498933903e-06,
+      "loss": 0.5448,
+      "step": 57
+    },
+    {
+      "epoch": 0.06183368869936034,
+      "grad_norm": 0.2493874964259867,
+      "learning_rate": 9.893390191897656e-06,
+      "loss": 0.522,
+      "step": 58
+    },
+    {
+      "epoch": 0.0628997867803838,
+      "grad_norm": 0.24173737961013422,
+      "learning_rate": 1.0063965884861408e-05,
+      "loss": 0.5295,
+      "step": 59
+    },
+    {
+      "epoch": 0.06396588486140725,
+      "grad_norm": 0.3214864858366852,
+      "learning_rate": 1.023454157782516e-05,
+      "loss": 0.534,
+      "step": 60
+    },
+    {
+      "epoch": 0.0650319829424307,
+      "grad_norm": 0.22660214109074844,
+      "learning_rate": 1.0405117270788913e-05,
+      "loss": 0.5207,
+      "step": 61
+    },
+    {
+      "epoch": 0.06609808102345416,
+      "grad_norm": 0.22692046089904563,
+      "learning_rate": 1.0575692963752667e-05,
+      "loss": 0.524,
+      "step": 62
+    },
+    {
+      "epoch": 0.06716417910447761,
+      "grad_norm": 0.267684560903461,
+      "learning_rate": 1.074626865671642e-05,
+      "loss": 0.5241,
+      "step": 63
+    },
+    {
+      "epoch": 0.06823027718550106,
+      "grad_norm": 0.2145100827755043,
+      "learning_rate": 1.0916844349680172e-05,
+      "loss": 0.5267,
+      "step": 64
+    },
+    {
+      "epoch": 0.06929637526652452,
+      "grad_norm": 0.2581454974388823,
+      "learning_rate": 1.1087420042643924e-05,
+      "loss": 0.5263,
+      "step": 65
+    },
+    {
+      "epoch": 0.07036247334754797,
+      "grad_norm": 0.23991967383468807,
+      "learning_rate": 1.1257995735607677e-05,
+      "loss": 0.5165,
+      "step": 66
+    },
+    {
+      "epoch": 0.07142857142857142,
+      "grad_norm": 0.19942140384569496,
+      "learning_rate": 1.1428571428571429e-05,
+      "loss": 0.5168,
+      "step": 67
+    },
+    {
+      "epoch": 0.07249466950959488,
+      "grad_norm": 0.2412614596538889,
+      "learning_rate": 1.1599147121535181e-05,
+      "loss": 0.5137,
+      "step": 68
+    },
+    {
+      "epoch": 0.07356076759061833,
+      "grad_norm": 0.213075163701682,
+      "learning_rate": 1.1769722814498934e-05,
+      "loss": 0.5177,
+      "step": 69
+    },
+    {
+      "epoch": 0.07462686567164178,
+      "grad_norm": 0.2078353315515298,
+      "learning_rate": 1.1940298507462686e-05,
+      "loss": 0.5203,
+      "step": 70
+    },
+    {
+      "epoch": 0.07569296375266525,
+      "grad_norm": 0.24803061207482333,
+      "learning_rate": 1.2110874200426442e-05,
+      "loss": 0.5127,
+      "step": 71
+    },
+    {
+      "epoch": 0.0767590618336887,
+      "grad_norm": 0.200388439144567,
+      "learning_rate": 1.2281449893390195e-05,
+      "loss": 0.524,
+      "step": 72
+    },
+    {
+      "epoch": 0.07782515991471216,
+      "grad_norm": 0.26609196507744726,
+      "learning_rate": 1.2452025586353947e-05,
+      "loss": 0.5166,
+      "step": 73
+    },
+    {
+      "epoch": 0.07889125799573561,
+      "grad_norm": 0.2746212953481589,
+      "learning_rate": 1.26226012793177e-05,
+      "loss": 0.5097,
+      "step": 74
+    },
+    {
+      "epoch": 0.07995735607675906,
+      "grad_norm": 0.25408109337829365,
+      "learning_rate": 1.2793176972281452e-05,
+      "loss": 0.5124,
+      "step": 75
+    },
+    {
+      "epoch": 0.08102345415778252,
+      "grad_norm": 0.23041327392909428,
+      "learning_rate": 1.2963752665245204e-05,
+      "loss": 0.5114,
+      "step": 76
+    },
+    {
+      "epoch": 0.08208955223880597,
+      "grad_norm": 0.25359601440739266,
+      "learning_rate": 1.3134328358208957e-05,
+      "loss": 0.5126,
+      "step": 77
+    },
+    {
+      "epoch": 0.08315565031982942,
+      "grad_norm": 0.26693374625690547,
+      "learning_rate": 1.3304904051172709e-05,
+      "loss": 0.5026,
+      "step": 78
+    },
+    {
+      "epoch": 0.08422174840085288,
+      "grad_norm": 0.32689327987808636,
+      "learning_rate": 1.3475479744136461e-05,
+      "loss": 0.51,
+      "step": 79
+    },
+    {
+      "epoch": 0.08528784648187633,
+      "grad_norm": 0.3372086808231497,
+      "learning_rate": 1.3646055437100214e-05,
+      "loss": 0.5026,
+      "step": 80
+    },
+    {
+      "epoch": 0.08635394456289978,
+      "grad_norm": 0.3813830415514882,
+      "learning_rate": 1.3816631130063966e-05,
+      "loss": 0.5066,
+      "step": 81
+    },
+    {
+      "epoch": 0.08742004264392324,
+      "grad_norm": 0.5438093416524072,
+      "learning_rate": 1.3987206823027719e-05,
+      "loss": 0.5164,
+      "step": 82
+    },
+    {
+      "epoch": 0.08848614072494669,
+      "grad_norm": 0.6495392831189992,
+      "learning_rate": 1.4157782515991471e-05,
+      "loss": 0.5043,
+      "step": 83
+    },
+    {
+      "epoch": 0.08955223880597014,
+      "grad_norm": 0.6977147730974188,
+      "learning_rate": 1.4328358208955224e-05,
+      "loss": 0.5036,
+      "step": 84
+    },
+    {
+      "epoch": 0.0906183368869936,
+      "grad_norm": 0.6222679447369283,
+      "learning_rate": 1.4498933901918976e-05,
+      "loss": 0.5098,
+      "step": 85
+    },
+    {
+      "epoch": 0.09168443496801706,
+      "grad_norm": 0.47152661245187916,
+      "learning_rate": 1.4669509594882732e-05,
+      "loss": 0.5055,
+      "step": 86
+    },
+    {
+      "epoch": 0.09275053304904052,
+      "grad_norm": 0.3720647971705006,
+      "learning_rate": 1.4840085287846484e-05,
+      "loss": 0.5093,
+      "step": 87
+    },
+    {
+      "epoch": 0.09381663113006397,
+      "grad_norm": 0.429318588389719,
+      "learning_rate": 1.5010660980810237e-05,
+      "loss": 0.5012,
+      "step": 88
+    },
+    {
+      "epoch": 0.09488272921108742,
+      "grad_norm": 0.5114262103736821,
+      "learning_rate": 1.5181236673773989e-05,
+      "loss": 0.5057,
+      "step": 89
+    },
+    {
+      "epoch": 0.09594882729211088,
+      "grad_norm": 0.4452652595986177,
+      "learning_rate": 1.535181236673774e-05,
+      "loss": 0.5044,
+      "step": 90
+    },
+    {
+      "epoch": 0.09701492537313433,
+      "grad_norm": 0.30399112681775725,
+      "learning_rate": 1.5522388059701494e-05,
+      "loss": 0.5039,
+      "step": 91
+    },
+    {
+      "epoch": 0.09808102345415778,
+      "grad_norm": 0.44638035180999586,
+      "learning_rate": 1.5692963752665246e-05,
+      "loss": 0.5088,
+      "step": 92
+    },
+    {
+      "epoch": 0.09914712153518124,
+      "grad_norm": 0.5355105792075635,
+      "learning_rate": 1.5863539445629e-05,
+      "loss": 0.5068,
+      "step": 93
+    },
+    {
+      "epoch": 0.10021321961620469,
+      "grad_norm": 1.6435340842916062,
+      "learning_rate": 1.603411513859275e-05,
+      "loss": 0.5044,
+      "step": 94
+    },
+    {
+      "epoch": 0.10127931769722814,
+      "grad_norm": 0.41709741899745906,
+      "learning_rate": 1.6204690831556504e-05,
+      "loss": 0.4993,
+      "step": 95
+    },
+    {
+      "epoch": 0.1023454157782516,
+      "grad_norm": 0.8489238666616422,
+      "learning_rate": 1.6375266524520256e-05,
+      "loss": 0.4982,
+      "step": 96
+    },
+    {
+      "epoch": 0.10341151385927505,
+      "grad_norm": 0.7950895739539379,
+      "learning_rate": 1.654584221748401e-05,
+      "loss": 0.4955,
+      "step": 97
+    },
+    {
+      "epoch": 0.1044776119402985,
+      "grad_norm": 0.3524964920301634,
+      "learning_rate": 1.671641791044776e-05,
+      "loss": 0.4896,
+      "step": 98
+    },
+    {
+      "epoch": 0.10554371002132196,
+      "grad_norm": 0.48403087660030325,
+      "learning_rate": 1.6886993603411513e-05,
+      "loss": 0.4938,
+      "step": 99
+    },
+    {
+      "epoch": 0.10660980810234541,
+      "grad_norm": 0.8862953843822765,
+      "learning_rate": 1.7057569296375266e-05,
+      "loss": 0.5003,
+      "step": 100
+    },
+    {
+      "epoch": 0.10767590618336886,
+      "grad_norm": 0.84687190937384,
+      "learning_rate": 1.7228144989339018e-05,
+      "loss": 0.498,
+      "step": 101
+    },
+    {
+      "epoch": 0.10874200426439233,
+      "grad_norm": 0.7104553983279852,
+      "learning_rate": 1.7398720682302774e-05,
+      "loss": 0.5002,
+      "step": 102
+    },
+    {
+      "epoch": 0.10980810234541578,
+      "grad_norm": 1.1434719823658013,
+      "learning_rate": 1.7569296375266526e-05,
+      "loss": 0.4992,
+      "step": 103
+    },
+    {
+      "epoch": 0.11087420042643924,
+      "grad_norm": 0.8704506872447706,
+      "learning_rate": 1.773987206823028e-05,
+      "loss": 0.4951,
+      "step": 104
+    },
+    {
+      "epoch": 0.11194029850746269,
+      "grad_norm": 0.897490279226691,
+      "learning_rate": 1.791044776119403e-05,
+      "loss": 0.4916,
+      "step": 105
+    },
+    {
+      "epoch": 0.11300639658848614,
+      "grad_norm": 1.484307049969091,
+      "learning_rate": 1.8081023454157784e-05,
+      "loss": 0.4943,
+      "step": 106
+    },
+    {
+      "epoch": 0.1140724946695096,
+      "grad_norm": 0.6421017701917927,
+      "learning_rate": 1.8251599147121536e-05,
+      "loss": 0.4906,
+      "step": 107
+    },
+    {
+      "epoch": 0.11513859275053305,
+      "grad_norm": 1.6387082014606487,
+      "learning_rate": 1.842217484008529e-05,
+      "loss": 0.4976,
+      "step": 108
+    },
+    {
+      "epoch": 0.1162046908315565,
+      "grad_norm": 0.7508434661676452,
+      "learning_rate": 1.859275053304904e-05,
+      "loss": 0.4893,
+      "step": 109
+    },
+    {
+      "epoch": 0.11727078891257996,
+      "grad_norm": 1.5878659815817608,
+      "learning_rate": 1.8763326226012793e-05,
+      "loss": 0.4936,
+      "step": 110
+    },
+    {
+      "epoch": 0.11833688699360341,
+      "grad_norm": 1.2700012445274647,
+      "learning_rate": 1.8933901918976546e-05,
+      "loss": 0.4976,
+      "step": 111
+    },
+    {
+      "epoch": 0.11940298507462686,
+      "grad_norm": 1.4468232793600246,
+      "learning_rate": 1.9104477611940298e-05,
+      "loss": 0.4916,
+      "step": 112
+    },
+    {
+      "epoch": 0.12046908315565032,
+      "grad_norm": 1.2073925718708678,
+      "learning_rate": 1.9275053304904054e-05,
+      "loss": 0.4924,
+      "step": 113
+    },
+    {
+      "epoch": 0.12153518123667377,
+      "grad_norm": 1.382482149756203,
+      "learning_rate": 1.9445628997867806e-05,
+      "loss": 0.495,
+      "step": 114
+    },
+    {
+      "epoch": 0.12260127931769722,
+      "grad_norm": 1.1681250872190037,
+      "learning_rate": 1.961620469083156e-05,
+      "loss": 0.4929,
+      "step": 115
+    },
+    {
+      "epoch": 0.12366737739872068,
+      "grad_norm": 1.1644844991590728,
+      "learning_rate": 1.978678038379531e-05,
+      "loss": 0.487,
+      "step": 116
+    },
+    {
+      "epoch": 0.12473347547974413,
+      "grad_norm": 1.1590131735118943,
+      "learning_rate": 1.9957356076759064e-05,
+      "loss": 0.491,
+      "step": 117
+    },
+    {
+      "epoch": 0.1257995735607676,
+      "grad_norm": 0.8845517269114772,
+      "learning_rate": 2.0127931769722816e-05,
+      "loss": 0.4899,
+      "step": 118
+    },
+    {
+      "epoch": 0.12686567164179105,
+      "grad_norm": 0.8720137285450001,
+      "learning_rate": 2.029850746268657e-05,
+      "loss": 0.4948,
+      "step": 119
+    },
+    {
+      "epoch": 0.1279317697228145,
+      "grad_norm": 0.8270825293355472,
+      "learning_rate": 2.046908315565032e-05,
+      "loss": 0.4873,
+      "step": 120
+    },
+    {
+      "epoch": 0.12899786780383796,
+      "grad_norm": 0.7112733086734145,
+      "learning_rate": 2.0639658848614073e-05,
+      "loss": 0.4852,
+      "step": 121
+    },
+    {
+      "epoch": 0.1300639658848614,
+      "grad_norm": 0.63087466396226,
+      "learning_rate": 2.0810234541577826e-05,
+      "loss": 0.4842,
+      "step": 122
+    },
+    {
+      "epoch": 0.13113006396588486,
+      "grad_norm": 0.6768967900200205,
+      "learning_rate": 2.098081023454158e-05,
+      "loss": 0.4892,
+      "step": 123
+    },
+    {
+      "epoch": 0.13219616204690832,
+      "grad_norm": 0.4441105086375026,
+      "learning_rate": 2.1151385927505334e-05,
+      "loss": 0.49,
+      "step": 124
+    },
+    {
+      "epoch": 0.13326226012793177,
+      "grad_norm": 0.7871223305057534,
+      "learning_rate": 2.1321961620469086e-05,
+      "loss": 0.4861,
+      "step": 125
+    },
+    {
+      "epoch": 0.13432835820895522,
+      "grad_norm": 0.7473471277752516,
+      "learning_rate": 2.149253731343284e-05,
+      "loss": 0.4859,
+      "step": 126
+    },
+    {
+      "epoch": 0.13539445628997868,
+      "grad_norm": 0.6860378579705178,
+      "learning_rate": 2.166311300639659e-05,
+      "loss": 0.483,
+      "step": 127
+    },
+    {
+      "epoch": 0.13646055437100213,
+      "grad_norm": 0.6877050783906112,
+      "learning_rate": 2.1833688699360344e-05,
+      "loss": 0.4778,
+      "step": 128
+    },
+    {
+      "epoch": 0.13752665245202558,
+      "grad_norm": 0.7802953077555831,
+      "learning_rate": 2.2004264392324096e-05,
+      "loss": 0.4841,
+      "step": 129
+    },
+    {
+      "epoch": 0.13859275053304904,
+      "grad_norm": 0.6105421403859265,
+      "learning_rate": 2.217484008528785e-05,
+      "loss": 0.4888,
+      "step": 130
+    },
+    {
+      "epoch": 0.1396588486140725,
+      "grad_norm": 0.8085739454604434,
+      "learning_rate": 2.23454157782516e-05,
+      "loss": 0.4844,
+      "step": 131
+    },
+    {
+      "epoch": 0.14072494669509594,
+      "grad_norm": 0.599568911512429,
+      "learning_rate": 2.2515991471215353e-05,
+      "loss": 0.4859,
+      "step": 132
+    },
+    {
+      "epoch": 0.1417910447761194,
+      "grad_norm": 0.7298411511687625,
+      "learning_rate": 2.2686567164179106e-05,
+      "loss": 0.4839,
+      "step": 133
+    },
+    {
+      "epoch": 0.14285714285714285,
+      "grad_norm": 0.5022968348693084,
+      "learning_rate": 2.2857142857142858e-05,
+      "loss": 0.4799,
+      "step": 134
+    },
+    {
+      "epoch": 0.1439232409381663,
+      "grad_norm": 0.7782509559217928,
+      "learning_rate": 2.302771855010661e-05,
+      "loss": 0.4811,
+      "step": 135
+    },
+    {
+      "epoch": 0.14498933901918976,
+      "grad_norm": 0.8576912283821915,
+      "learning_rate": 2.3198294243070363e-05,
+      "loss": 0.4855,
+      "step": 136
+    },
+    {
+      "epoch": 0.1460554371002132,
+      "grad_norm": 0.75095345536952,
+      "learning_rate": 2.3368869936034115e-05,
+      "loss": 0.4831,
+      "step": 137
+    },
+    {
+      "epoch": 0.14712153518123666,
+      "grad_norm": 1.1021094209824167,
+      "learning_rate": 2.3539445628997868e-05,
+      "loss": 0.485,
+      "step": 138
+    },
+    {
+      "epoch": 0.14818763326226012,
+      "grad_norm": 1.3850416902760037,
+      "learning_rate": 2.371002132196162e-05,
+      "loss": 0.478,
+      "step": 139
+    },
+    {
+      "epoch": 0.14925373134328357,
+      "grad_norm": 0.5373300308852162,
+      "learning_rate": 2.3880597014925373e-05,
+      "loss": 0.4743,
+      "step": 140
+    },
+    {
+      "epoch": 0.15031982942430705,
+      "grad_norm": 1.2965983051499783,
+      "learning_rate": 2.405117270788913e-05,
+      "loss": 0.4877,
+      "step": 141
+    },
+    {
+      "epoch": 0.1513859275053305,
+      "grad_norm": 0.9983577194279637,
+      "learning_rate": 2.4221748400852884e-05,
+      "loss": 0.4814,
+      "step": 142
+    },
+    {
+      "epoch": 0.15245202558635396,
+      "grad_norm": 0.720371040224616,
+      "learning_rate": 2.4392324093816637e-05,
+      "loss": 0.4809,
+      "step": 143
+    },
+    {
+      "epoch": 0.1535181236673774,
+      "grad_norm": 0.8215017978389719,
+      "learning_rate": 2.456289978678039e-05,
+      "loss": 0.4751,
+      "step": 144
+    },
+    {
+      "epoch": 0.15458422174840086,
+      "grad_norm": 0.8563504324634484,
+      "learning_rate": 2.473347547974414e-05,
+      "loss": 0.4788,
+      "step": 145
+    },
+    {
+      "epoch": 0.15565031982942432,
+      "grad_norm": 1.0397801044745016,
+      "learning_rate": 2.4904051172707894e-05,
+      "loss": 0.4828,
+      "step": 146
+    },
+    {
+      "epoch": 0.15671641791044777,
+      "grad_norm": 0.8499495288588741,
+      "learning_rate": 2.5074626865671646e-05,
+      "loss": 0.4823,
+      "step": 147
+    },
+    {
+      "epoch": 0.15778251599147122,
+      "grad_norm": 0.7506494495443314,
+      "learning_rate": 2.52452025586354e-05,
+      "loss": 0.4719,
+      "step": 148
+    },
+    {
+      "epoch": 0.15884861407249468,
+      "grad_norm": 0.8572573387032592,
+      "learning_rate": 2.541577825159915e-05,
+      "loss": 0.4779,
+      "step": 149
+    },
+    {
+      "epoch": 0.15991471215351813,
+      "grad_norm": 0.8980484862912236,
+      "learning_rate": 2.5586353944562904e-05,
+      "loss": 0.4669,
+      "step": 150
+    },
+    {
+      "epoch": 0.16098081023454158,
+      "grad_norm": 1.3008931725410442,
+      "learning_rate": 2.5756929637526656e-05,
+      "loss": 0.4778,
+      "step": 151
+    },
+    {
+      "epoch": 0.16204690831556504,
+      "grad_norm": 0.5404519887326864,
+      "learning_rate": 2.592750533049041e-05,
+      "loss": 0.4731,
+      "step": 152
+    },
+    {
+      "epoch": 0.1631130063965885,
+      "grad_norm": 1.0716812716539528,
+      "learning_rate": 2.609808102345416e-05,
+      "loss": 0.4798,
+      "step": 153
+    },
+    {
+      "epoch": 0.16417910447761194,
+      "grad_norm": 1.2709697777670215,
+      "learning_rate": 2.6268656716417913e-05,
+      "loss": 0.4871,
+      "step": 154
+    },
+    {
+      "epoch": 0.1652452025586354,
+      "grad_norm": 0.537320261984439,
+      "learning_rate": 2.6439232409381666e-05,
+      "loss": 0.4749,
+      "step": 155
+    },
+    {
+      "epoch": 0.16631130063965885,
+      "grad_norm": 1.1514684368533086,
+      "learning_rate": 2.6609808102345418e-05,
+      "loss": 0.4753,
+      "step": 156
+    },
+    {
+      "epoch": 0.1673773987206823,
+      "grad_norm": 0.9432669540338412,
+      "learning_rate": 2.678038379530917e-05,
+      "loss": 0.4807,
+      "step": 157
+    },
+    {
+      "epoch": 0.16844349680170576,
+      "grad_norm": 0.642515332753071,
+      "learning_rate": 2.6950959488272923e-05,
+      "loss": 0.477,
+      "step": 158
+    },
+    {
+      "epoch": 0.1695095948827292,
+      "grad_norm": 0.8211012768546435,
+      "learning_rate": 2.7121535181236675e-05,
+      "loss": 0.4766,
+      "step": 159
+    },
+    {
+      "epoch": 0.17057569296375266,
+      "grad_norm": 0.8776070226460816,
+      "learning_rate": 2.7292110874200428e-05,
+      "loss": 0.4717,
+      "step": 160
+    },
+    {
+      "epoch": 0.17164179104477612,
+      "grad_norm": 0.7718966074497755,
+      "learning_rate": 2.746268656716418e-05,
+      "loss": 0.4687,
+      "step": 161
+    },
+    {
+      "epoch": 0.17270788912579957,
+      "grad_norm": 0.7779381456798092,
+      "learning_rate": 2.7633262260127933e-05,
+      "loss": 0.4679,
+      "step": 162
+    },
+    {
+      "epoch": 0.17377398720682302,
+      "grad_norm": 0.5901156093872161,
+      "learning_rate": 2.7803837953091685e-05,
+      "loss": 0.4751,
+      "step": 163
+    },
+    {
+      "epoch": 0.17484008528784648,
+      "grad_norm": 0.7154596410625235,
+      "learning_rate": 2.7974413646055437e-05,
+      "loss": 0.4728,
+      "step": 164
+    },
+    {
+      "epoch": 0.17590618336886993,
+      "grad_norm": 0.7000348159687018,
+      "learning_rate": 2.814498933901919e-05,
+      "loss": 0.4725,
+      "step": 165
+    },
+    {
+      "epoch": 0.17697228144989338,
+      "grad_norm": 0.7657735939109871,
+      "learning_rate": 2.8315565031982942e-05,
+      "loss": 0.4784,
+      "step": 166
+    },
+    {
+      "epoch": 0.17803837953091683,
+      "grad_norm": 0.8475989590871966,
+      "learning_rate": 2.8486140724946695e-05,
+      "loss": 0.4711,
+      "step": 167
+    },
+    {
+      "epoch": 0.1791044776119403,
+      "grad_norm": 1.0922356892583558,
+      "learning_rate": 2.8656716417910447e-05,
+      "loss": 0.478,
+      "step": 168
+    },
+    {
+      "epoch": 0.18017057569296374,
+      "grad_norm": 0.970068377938896,
+      "learning_rate": 2.88272921108742e-05,
+      "loss": 0.4653,
+      "step": 169
+    },
+    {
+      "epoch": 0.1812366737739872,
+      "grad_norm": 0.7059217860390857,
+      "learning_rate": 2.8997867803837952e-05,
+      "loss": 0.474,
+      "step": 170
+    },
+    {
+      "epoch": 0.18230277185501065,
+      "grad_norm": 1.0368427776415303,
+      "learning_rate": 2.9168443496801708e-05,
+      "loss": 0.4745,
+      "step": 171
+    },
+    {
+      "epoch": 0.18336886993603413,
+      "grad_norm": 1.1739249199645363,
+      "learning_rate": 2.9339019189765464e-05,
+      "loss": 0.4713,
+      "step": 172
+    },
+    {
+      "epoch": 0.18443496801705758,
+      "grad_norm": 0.772697311844866,
+      "learning_rate": 2.9509594882729216e-05,
+      "loss": 0.4718,
+      "step": 173
+    },
+    {
+      "epoch": 0.18550106609808104,
+      "grad_norm": 0.8330939252501977,
+      "learning_rate": 2.968017057569297e-05,
+      "loss": 0.4722,
+      "step": 174
+    },
+    {
+      "epoch": 0.1865671641791045,
+      "grad_norm": 0.831387960842226,
+      "learning_rate": 2.985074626865672e-05,
+      "loss": 0.472,
+      "step": 175
+    },
+    {
+      "epoch": 0.18763326226012794,
+      "grad_norm": 0.7976609685235961,
+      "learning_rate": 3.0021321961620473e-05,
+      "loss": 0.47,
+      "step": 176
+    },
+    {
+      "epoch": 0.1886993603411514,
+      "grad_norm": 0.8823919254760567,
+      "learning_rate": 3.0191897654584226e-05,
+      "loss": 0.4759,
+      "step": 177
+    },
+    {
+      "epoch": 0.18976545842217485,
+      "grad_norm": 0.8804820784317922,
+      "learning_rate": 3.0362473347547978e-05,
+      "loss": 0.4659,
+      "step": 178
+    },
+    {
+      "epoch": 0.1908315565031983,
+      "grad_norm": 0.9681666262840064,
+      "learning_rate": 3.053304904051173e-05,
+      "loss": 0.4729,
+      "step": 179
+    },
+    {
+      "epoch": 0.19189765458422176,
+      "grad_norm": 1.1048871958759772,
+      "learning_rate": 3.070362473347548e-05,
+      "loss": 0.4665,
+      "step": 180
+    },
+    {
+      "epoch": 0.1929637526652452,
+      "grad_norm": 0.8870408442356555,
+      "learning_rate": 3.0874200426439235e-05,
+      "loss": 0.4745,
+      "step": 181
+    },
+    {
+      "epoch": 0.19402985074626866,
+      "grad_norm": 0.9631165384838795,
+      "learning_rate": 3.104477611940299e-05,
+      "loss": 0.4698,
+      "step": 182
+    },
+    {
+      "epoch": 0.19509594882729211,
+      "grad_norm": 0.9385999582540799,
+      "learning_rate": 3.121535181236674e-05,
+      "loss": 0.4705,
+      "step": 183
+    },
+    {
+      "epoch": 0.19616204690831557,
+      "grad_norm": 0.7939219655274056,
+      "learning_rate": 3.138592750533049e-05,
+      "loss": 0.4732,
+      "step": 184
+    },
+    {
+      "epoch": 0.19722814498933902,
+      "grad_norm": 0.736357203214717,
+      "learning_rate": 3.1556503198294245e-05,
+      "loss": 0.4746,
+      "step": 185
+    },
+    {
+      "epoch": 0.19829424307036247,
+      "grad_norm": 0.8037466041799969,
+      "learning_rate": 3.1727078891258e-05,
+      "loss": 0.4636,
+      "step": 186
+    },
+    {
+      "epoch": 0.19936034115138593,
+      "grad_norm": 1.0193532786616932,
+      "learning_rate": 3.189765458422175e-05,
+      "loss": 0.4721,
+      "step": 187
+    },
+    {
+      "epoch": 0.20042643923240938,
+      "grad_norm": 1.4921054947249843,
+      "learning_rate": 3.20682302771855e-05,
+      "loss": 0.4718,
+      "step": 188
+    },
+    {
+      "epoch": 0.20149253731343283,
+      "grad_norm": 0.7621564304858446,
+      "learning_rate": 3.2238805970149255e-05,
+      "loss": 0.4653,
+      "step": 189
+    },
+    {
+      "epoch": 0.2025586353944563,
+      "grad_norm": 1.070754972082962,
+      "learning_rate": 3.240938166311301e-05,
+      "loss": 0.4654,
+      "step": 190
+    },
+    {
+      "epoch": 0.20362473347547974,
+      "grad_norm": 1.5652147505824308,
+      "learning_rate": 3.257995735607676e-05,
+      "loss": 0.472,
+      "step": 191
+    },
+    {
+      "epoch": 0.2046908315565032,
+      "grad_norm": 0.869980401463994,
+      "learning_rate": 3.275053304904051e-05,
+      "loss": 0.4646,
+      "step": 192
+    },
+    {
+      "epoch": 0.20575692963752665,
+      "grad_norm": 1.4005482830515732,
+      "learning_rate": 3.2921108742004264e-05,
+      "loss": 0.4697,
+      "step": 193
+    },
+    {
+      "epoch": 0.2068230277185501,
+      "grad_norm": 1.2810330542613626,
+      "learning_rate": 3.309168443496802e-05,
+      "loss": 0.4631,
+      "step": 194
+    },
+    {
+      "epoch": 0.20788912579957355,
+      "grad_norm": 1.036250299110791,
+      "learning_rate": 3.326226012793177e-05,
+      "loss": 0.4757,
+      "step": 195
+    },
+    {
+      "epoch": 0.208955223880597,
+      "grad_norm": 0.8072040874296945,
+      "learning_rate": 3.343283582089552e-05,
+      "loss": 0.4629,
+      "step": 196
+    },
+    {
+      "epoch": 0.21002132196162046,
+      "grad_norm": 1.4196573201874585,
+      "learning_rate": 3.3603411513859274e-05,
+      "loss": 0.4684,
+      "step": 197
+    },
+    {
+      "epoch": 0.21108742004264391,
+      "grad_norm": 0.6191497943271066,
+      "learning_rate": 3.3773987206823026e-05,
+      "loss": 0.4631,
+      "step": 198
+    },
+    {
+      "epoch": 0.21215351812366737,
+      "grad_norm": 1.2158443187059038,
+      "learning_rate": 3.394456289978678e-05,
+      "loss": 0.4671,
+      "step": 199
+    },
+    {
+      "epoch": 0.21321961620469082,
+      "grad_norm": 0.8129012874793476,
+      "learning_rate": 3.411513859275053e-05,
+      "loss": 0.4618,
+      "step": 200
+    },
+    {
+      "epoch": 0.21428571428571427,
+      "grad_norm": 1.0518786895781704,
+      "learning_rate": 3.4285714285714284e-05,
+      "loss": 0.4697,
+      "step": 201
+    },
+    {
+      "epoch": 0.21535181236673773,
+      "grad_norm": 0.9688679826055264,
+      "learning_rate": 3.4456289978678036e-05,
+      "loss": 0.4683,
+      "step": 202
+    },
+    {
+      "epoch": 0.21641791044776118,
+      "grad_norm": 1.3724882136741294,
+      "learning_rate": 3.462686567164179e-05,
+      "loss": 0.4718,
+      "step": 203
+    },
+    {
+      "epoch": 0.21748400852878466,
+      "grad_norm": 0.5641264907256553,
+      "learning_rate": 3.479744136460555e-05,
+      "loss": 0.4551,
+      "step": 204
+    },
+    {
+      "epoch": 0.21855010660980811,
+      "grad_norm": 1.0701546739161645,
+      "learning_rate": 3.49680170575693e-05,
+      "loss": 0.4612,
+      "step": 205
+    },
+    {
+      "epoch": 0.21961620469083157,
+      "grad_norm": 1.049942985757254,
+      "learning_rate": 3.513859275053305e-05,
+      "loss": 0.4696,
+      "step": 206
+    },
+    {
+      "epoch": 0.22068230277185502,
+      "grad_norm": 1.1283220333372375,
+      "learning_rate": 3.5309168443496805e-05,
+      "loss": 0.4632,
+      "step": 207
+    },
+    {
+      "epoch": 0.22174840085287847,
+      "grad_norm": 0.9308516582288439,
+      "learning_rate": 3.547974413646056e-05,
+      "loss": 0.4661,
+      "step": 208
+    },
+    {
+      "epoch": 0.22281449893390193,
+      "grad_norm": 1.1185548874871707,
+      "learning_rate": 3.565031982942431e-05,
+      "loss": 0.4641,
+      "step": 209
+    },
+    {
+      "epoch": 0.22388059701492538,
+      "grad_norm": 0.919634938330087,
+      "learning_rate": 3.582089552238806e-05,
+      "loss": 0.4707,
+      "step": 210
+    },
+    {
+      "epoch": 0.22494669509594883,
+      "grad_norm": 1.4776402736329493,
+      "learning_rate": 3.5991471215351815e-05,
+      "loss": 0.4652,
+      "step": 211
+    },
+    {
+      "epoch": 0.2260127931769723,
+      "grad_norm": 0.7167195593546435,
+      "learning_rate": 3.616204690831557e-05,
+      "loss": 0.4595,
+      "step": 212
+    },
+    {
+      "epoch": 0.22707889125799574,
+      "grad_norm": 1.2140834565384866,
+      "learning_rate": 3.633262260127932e-05,
+      "loss": 0.4711,
+      "step": 213
+    },
+    {
+      "epoch": 0.2281449893390192,
+      "grad_norm": 1.0029544450885874,
+      "learning_rate": 3.650319829424307e-05,
+      "loss": 0.4596,
+      "step": 214
+    },
+    {
+      "epoch": 0.22921108742004265,
+      "grad_norm": 1.5043815628272939,
+      "learning_rate": 3.6673773987206824e-05,
+      "loss": 0.4638,
+      "step": 215
+    },
+    {
+      "epoch": 0.2302771855010661,
+      "grad_norm": 0.8569767572345512,
+      "learning_rate": 3.684434968017058e-05,
+      "loss": 0.4598,
+      "step": 216
+    },
+    {
+      "epoch": 0.23134328358208955,
+      "grad_norm": 1.3925636332034894,
+      "learning_rate": 3.701492537313433e-05,
+      "loss": 0.4604,
+      "step": 217
+    },
+    {
+      "epoch": 0.232409381663113,
+      "grad_norm": 1.1816721080629815,
+      "learning_rate": 3.718550106609808e-05,
+      "loss": 0.4518,
+      "step": 218
+    },
+    {
+      "epoch": 0.23347547974413646,
+      "grad_norm": 0.8651413005301738,
+      "learning_rate": 3.7356076759061834e-05,
+      "loss": 0.465,
+      "step": 219
+    },
+    {
+      "epoch": 0.2345415778251599,
+      "grad_norm": 1.1820130745722497,
+      "learning_rate": 3.7526652452025586e-05,
+      "loss": 0.4637,
+      "step": 220
+    },
+    {
+      "epoch": 0.23560767590618337,
+      "grad_norm": 0.8427968342892499,
+      "learning_rate": 3.769722814498934e-05,
+      "loss": 0.4617,
+      "step": 221
+    },
+    {
+      "epoch": 0.23667377398720682,
+      "grad_norm": 1.11073913884723,
+      "learning_rate": 3.786780383795309e-05,
+      "loss": 0.4655,
+      "step": 222
+    },
+    {
+      "epoch": 0.23773987206823027,
+      "grad_norm": 0.9189934131123165,
+      "learning_rate": 3.8038379530916844e-05,
+      "loss": 0.4766,
+      "step": 223
+    },
+    {
+      "epoch": 0.23880597014925373,
+      "grad_norm": 1.1535170456681654,
+      "learning_rate": 3.8208955223880596e-05,
+      "loss": 0.4625,
+      "step": 224
+    },
+    {
+      "epoch": 0.23987206823027718,
+      "grad_norm": 0.9374181623692117,
+      "learning_rate": 3.8379530916844355e-05,
+      "loss": 0.4669,
+      "step": 225
+    },
+    {
+      "epoch": 0.24093816631130063,
+      "grad_norm": 1.0563319134423954,
+      "learning_rate": 3.855010660980811e-05,
+      "loss": 0.4611,
+      "step": 226
+    },
+    {
+      "epoch": 0.2420042643923241,
+      "grad_norm": 1.168785335326744,
+      "learning_rate": 3.872068230277186e-05,
+      "loss": 0.4634,
+      "step": 227
+    },
+    {
+      "epoch": 0.24307036247334754,
+      "grad_norm": 1.0248065818014773,
+      "learning_rate": 3.889125799573561e-05,
+      "loss": 0.4593,
+      "step": 228
+    },
+    {
+      "epoch": 0.244136460554371,
+      "grad_norm": 0.8637465871329612,
+      "learning_rate": 3.9061833688699365e-05,
+      "loss": 0.4632,
+      "step": 229
+    },
+    {
+      "epoch": 0.24520255863539445,
+      "grad_norm": 0.8460415005099113,
+      "learning_rate": 3.923240938166312e-05,
+      "loss": 0.4624,
+      "step": 230
+    },
+    {
+      "epoch": 0.2462686567164179,
+      "grad_norm": 0.8955587399970688,
+      "learning_rate": 3.940298507462687e-05,
+      "loss": 0.457,
+      "step": 231
+    },
+    {
+      "epoch": 0.24733475479744135,
+      "grad_norm": 1.1657262128526866,
+      "learning_rate": 3.957356076759062e-05,
+      "loss": 0.4645,
+      "step": 232
+    },
+    {
+      "epoch": 0.2484008528784648,
+      "grad_norm": 1.0166441196536635,
+      "learning_rate": 3.9744136460554375e-05,
+      "loss": 0.4634,
+      "step": 233
+    },
+    {
+      "epoch": 0.24946695095948826,
+      "grad_norm": 0.9826615671115314,
+      "learning_rate": 3.991471215351813e-05,
+      "loss": 0.4614,
+      "step": 234
+    },
+    {
+      "epoch": 0.2505330490405117,
+      "grad_norm": 1.0599689795674345,
+      "learning_rate": 4.008528784648188e-05,
+      "loss": 0.4731,
+      "step": 235
+    },
+    {
+      "epoch": 0.2515991471215352,
+      "grad_norm": 0.9093543668098233,
+      "learning_rate": 4.025586353944563e-05,
+      "loss": 0.4566,
+      "step": 236
+    },
+    {
+      "epoch": 0.2526652452025586,
+      "grad_norm": 1.013585300158083,
+      "learning_rate": 4.0426439232409384e-05,
+      "loss": 0.468,
+      "step": 237
+    },
+    {
+      "epoch": 0.2537313432835821,
+      "grad_norm": 1.036723508797436,
+      "learning_rate": 4.059701492537314e-05,
+      "loss": 0.4584,
+      "step": 238
+    },
+    {
+      "epoch": 0.2547974413646055,
+      "grad_norm": 1.316129392280748,
+      "learning_rate": 4.076759061833689e-05,
+      "loss": 0.4675,
+      "step": 239
+    },
+    {
+      "epoch": 0.255863539445629,
+      "grad_norm": 1.0896330595803854,
+      "learning_rate": 4.093816631130064e-05,
+      "loss": 0.4623,
+      "step": 240
+    },
+    {
+      "epoch": 0.25692963752665243,
+      "grad_norm": 1.1510385991266068,
+      "learning_rate": 4.1108742004264394e-05,
+      "loss": 0.4649,
+      "step": 241
+    },
+    {
+      "epoch": 0.2579957356076759,
+      "grad_norm": 0.8229957414542415,
+      "learning_rate": 4.1279317697228146e-05,
+      "loss": 0.4584,
+      "step": 242
+    },
+    {
+      "epoch": 0.25906183368869934,
+      "grad_norm": 0.9568171883082784,
+      "learning_rate": 4.14498933901919e-05,
+      "loss": 0.4628,
+      "step": 243
+    },
+    {
+      "epoch": 0.2601279317697228,
+      "grad_norm": 0.9570238151905792,
+      "learning_rate": 4.162046908315565e-05,
+      "loss": 0.4544,
+      "step": 244
+    },
+    {
+      "epoch": 0.26119402985074625,
+      "grad_norm": 1.1790086045153048,
+      "learning_rate": 4.1791044776119404e-05,
+      "loss": 0.4565,
+      "step": 245
+    },
+    {
+      "epoch": 0.2622601279317697,
+      "grad_norm": 0.772255483862789,
+      "learning_rate": 4.196162046908316e-05,
+      "loss": 0.4575,
+      "step": 246
+    },
+    {
+      "epoch": 0.26332622601279315,
+      "grad_norm": 1.1363053470021465,
+      "learning_rate": 4.213219616204691e-05,
+      "loss": 0.4612,
+      "step": 247
+    },
+    {
+      "epoch": 0.26439232409381663,
+      "grad_norm": 0.7565375099785713,
+      "learning_rate": 4.230277185501067e-05,
+      "loss": 0.456,
+      "step": 248
+    },
+    {
+      "epoch": 0.26545842217484006,
+      "grad_norm": 0.6892425990836365,
+      "learning_rate": 4.247334754797441e-05,
+      "loss": 0.4559,
+      "step": 249
+    },
+    {
+      "epoch": 0.26652452025586354,
+      "grad_norm": 0.7417047962613619,
+      "learning_rate": 4.264392324093817e-05,
+      "loss": 0.4508,
+      "step": 250
+    },
+    {
+      "epoch": 0.267590618336887,
+      "grad_norm": 0.7470072441966399,
+      "learning_rate": 4.2814498933901925e-05,
+      "loss": 0.461,
+      "step": 251
+    },
+    {
+      "epoch": 0.26865671641791045,
+      "grad_norm": 0.9617786110698198,
+      "learning_rate": 4.298507462686568e-05,
+      "loss": 0.4513,
+      "step": 252
+    },
+    {
+      "epoch": 0.2697228144989339,
+      "grad_norm": 1.327440065045842,
+      "learning_rate": 4.315565031982943e-05,
+      "loss": 0.4638,
+      "step": 253
+    },
+    {
+      "epoch": 0.27078891257995735,
+      "grad_norm": 0.7442017560708861,
+      "learning_rate": 4.332622601279318e-05,
+      "loss": 0.4557,
+      "step": 254
+    },
+    {
+      "epoch": 0.27185501066098083,
+      "grad_norm": 0.7202532764515195,
+      "learning_rate": 4.3496801705756935e-05,
+      "loss": 0.4543,
+      "step": 255
+    },
+    {
+      "epoch": 0.27292110874200426,
+      "grad_norm": 0.8784340624592276,
+      "learning_rate": 4.366737739872069e-05,
+      "loss": 0.4579,
+      "step": 256
+    },
+    {
+      "epoch": 0.27398720682302774,
+      "grad_norm": 0.8848099666534693,
+      "learning_rate": 4.383795309168444e-05,
+      "loss": 0.4527,
+      "step": 257
+    },
+    {
+      "epoch": 0.27505330490405117,
+      "grad_norm": 0.7690402407029407,
+      "learning_rate": 4.400852878464819e-05,
+      "loss": 0.4576,
+      "step": 258
+    },
+    {
+      "epoch": 0.27611940298507465,
+      "grad_norm": 0.828823436313651,
+      "learning_rate": 4.4179104477611944e-05,
+      "loss": 0.4601,
+      "step": 259
+    },
+    {
+      "epoch": 0.2771855010660981,
+      "grad_norm": 1.186866608247454,
+      "learning_rate": 4.43496801705757e-05,
+      "loss": 0.4606,
+      "step": 260
+    },
+    {
+      "epoch": 0.27825159914712155,
+      "grad_norm": 0.9191614369421419,
+      "learning_rate": 4.452025586353945e-05,
+      "loss": 0.457,
+      "step": 261
+    },
+    {
+      "epoch": 0.279317697228145,
+      "grad_norm": 0.9275759467424034,
+      "learning_rate": 4.46908315565032e-05,
+      "loss": 0.4531,
+      "step": 262
+    },
+    {
+      "epoch": 0.28038379530916846,
+      "grad_norm": 1.3018942692140634,
+      "learning_rate": 4.4861407249466954e-05,
+      "loss": 0.4519,
+      "step": 263
+    },
+    {
+      "epoch": 0.2814498933901919,
+      "grad_norm": 1.2828905543825992,
+      "learning_rate": 4.5031982942430706e-05,
+      "loss": 0.4517,
+      "step": 264
+    },
+    {
+      "epoch": 0.28251599147121537,
+      "grad_norm": 0.8444164365212923,
+      "learning_rate": 4.5202558635394466e-05,
+      "loss": 0.4579,
+      "step": 265
+    },
+    {
+      "epoch": 0.2835820895522388,
+      "grad_norm": 0.7390719742847955,
+      "learning_rate": 4.537313432835821e-05,
+      "loss": 0.4564,
+      "step": 266
+    },
+    {
+      "epoch": 0.2846481876332623,
+      "grad_norm": 0.804814881156404,
+      "learning_rate": 4.554371002132197e-05,
+      "loss": 0.4516,
+      "step": 267
+    },
+    {
+      "epoch": 0.2857142857142857,
+      "grad_norm": 1.232886488524498,
+      "learning_rate": 4.5714285714285716e-05,
+      "loss": 0.4552,
+      "step": 268
+    },
+    {
+      "epoch": 0.2867803837953092,
+      "grad_norm": 1.1058003798103109,
+      "learning_rate": 4.5884861407249475e-05,
+      "loss": 0.4531,
+      "step": 269
+    },
+    {
+      "epoch": 0.2878464818763326,
+      "grad_norm": 1.1891457338823357,
+      "learning_rate": 4.605543710021322e-05,
+      "loss": 0.4511,
+      "step": 270
+    },
+    {
+      "epoch": 0.2889125799573561,
+      "grad_norm": 1.0630830374386846,
+      "learning_rate": 4.622601279317698e-05,
+      "loss": 0.4506,
+      "step": 271
+    },
+    {
+      "epoch": 0.2899786780383795,
+      "grad_norm": 0.8558226373887493,
+      "learning_rate": 4.6396588486140726e-05,
+      "loss": 0.4533,
+      "step": 272
+    },
+    {
+      "epoch": 0.291044776119403,
+      "grad_norm": 0.8431544624031982,
+      "learning_rate": 4.6567164179104485e-05,
+      "loss": 0.4489,
+      "step": 273
+    },
+    {
+      "epoch": 0.2921108742004264,
+      "grad_norm": 1.0780225196774107,
+      "learning_rate": 4.673773987206823e-05,
+      "loss": 0.4535,
+      "step": 274
+    },
+    {
+      "epoch": 0.2931769722814499,
+      "grad_norm": 1.040484818452518,
+      "learning_rate": 4.690831556503199e-05,
+      "loss": 0.4526,
+      "step": 275
+    },
+    {
+      "epoch": 0.2942430703624733,
+      "grad_norm": 1.4317419430672074,
+      "learning_rate": 4.7078891257995735e-05,
+      "loss": 0.4491,
+      "step": 276
+    },
+    {
+      "epoch": 0.2953091684434968,
+      "grad_norm": 0.7350004141525754,
+      "learning_rate": 4.7249466950959495e-05,
+      "loss": 0.4508,
+      "step": 277
+    },
+    {
+      "epoch": 0.29637526652452023,
+      "grad_norm": 1.0597151973403933,
+      "learning_rate": 4.742004264392324e-05,
+      "loss": 0.4536,
+      "step": 278
+    },
+    {
+      "epoch": 0.2974413646055437,
+      "grad_norm": 1.7726271804234848,
+      "learning_rate": 4.7590618336887e-05,
+      "loss": 0.4533,
+      "step": 279
+    },
+    {
+      "epoch": 0.29850746268656714,
+      "grad_norm": 0.7058843328989329,
+      "learning_rate": 4.7761194029850745e-05,
+      "loss": 0.4546,
+      "step": 280
+    },
+    {
+      "epoch": 0.2995735607675906,
+      "grad_norm": 2.0525677674617597,
+      "learning_rate": 4.7931769722814504e-05,
+      "loss": 0.4591,
+      "step": 281
+    },
+    {
+      "epoch": 0.3006396588486141,
+      "grad_norm": 1.0197143782821263,
+      "learning_rate": 4.810234541577826e-05,
+      "loss": 0.4539,
+      "step": 282
+    },
+    {
+      "epoch": 0.3017057569296375,
+      "grad_norm": 2.682566079537771,
+      "learning_rate": 4.827292110874201e-05,
+      "loss": 0.4696,
+      "step": 283
+    },
+    {
+      "epoch": 0.302771855010661,
+      "grad_norm": 2.7347833899867395,
+      "learning_rate": 4.844349680170577e-05,
+      "loss": 0.4854,
+      "step": 284
+    },
+    {
+      "epoch": 0.30383795309168443,
+      "grad_norm": 1.4147621114370599,
+      "learning_rate": 4.8614072494669514e-05,
+      "loss": 0.4613,
+      "step": 285
+    },
+    {
+      "epoch": 0.3049040511727079,
+      "grad_norm": 1.4281612301500841,
+      "learning_rate": 4.878464818763327e-05,
+      "loss": 0.4609,
+      "step": 286
+    },
+    {
+      "epoch": 0.30597014925373134,
+      "grad_norm": 1.0989723638165922,
+      "learning_rate": 4.895522388059702e-05,
+      "loss": 0.4627,
+      "step": 287
+    },
+    {
+      "epoch": 0.3070362473347548,
+      "grad_norm": 1.025429892081604,
+      "learning_rate": 4.912579957356078e-05,
+      "loss": 0.4516,
+      "step": 288
+    },
+    {
+      "epoch": 0.30810234541577824,
+      "grad_norm": 1.1574096426201355,
+      "learning_rate": 4.9296375266524524e-05,
+      "loss": 0.463,
+      "step": 289
+    },
+    {
+      "epoch": 0.3091684434968017,
+      "grad_norm": 0.9660522145703263,
+      "learning_rate": 4.946695095948828e-05,
+      "loss": 0.4551,
+      "step": 290
+    },
+    {
+      "epoch": 0.31023454157782515,
+      "grad_norm": 0.7154708715646663,
+      "learning_rate": 4.963752665245203e-05,
+      "loss": 0.4592,
+      "step": 291
+    },
+    {
+      "epoch": 0.31130063965884863,
+      "grad_norm": 0.6930388295609867,
+      "learning_rate": 4.980810234541579e-05,
+      "loss": 0.4541,
+      "step": 292
+    },
+    {
+      "epoch": 0.31236673773987206,
+      "grad_norm": 0.9114648684646235,
+      "learning_rate": 4.997867803837953e-05,
+      "loss": 0.4574,
+      "step": 293
+    },
+    {
+      "epoch": 0.31343283582089554,
+      "grad_norm": 1.245129676815517,
+      "learning_rate": 5.014925373134329e-05,
+      "loss": 0.4542,
+      "step": 294
+    },
+    {
+      "epoch": 0.31449893390191896,
+      "grad_norm": 0.9624168019988285,
+      "learning_rate": 5.031982942430704e-05,
+      "loss": 0.4536,
+      "step": 295
+    },
+    {
+      "epoch": 0.31556503198294245,
+      "grad_norm": 0.8009138019519386,
+      "learning_rate": 5.04904051172708e-05,
+      "loss": 0.4512,
+      "step": 296
+    },
+    {
+      "epoch": 0.31663113006396587,
+      "grad_norm": 0.6547749384104303,
+      "learning_rate": 5.066098081023454e-05,
+      "loss": 0.4519,
+      "step": 297
+    },
+    {
+      "epoch": 0.31769722814498935,
+      "grad_norm": 0.83857949506653,
+      "learning_rate": 5.08315565031983e-05,
+      "loss": 0.4513,
+      "step": 298
+    },
+    {
+      "epoch": 0.3187633262260128,
+      "grad_norm": 0.934818630813169,
+      "learning_rate": 5.100213219616205e-05,
+      "loss": 0.4547,
+      "step": 299
+    },
+    {
+      "epoch": 0.31982942430703626,
+      "grad_norm": 1.1298040583791609,
+      "learning_rate": 5.117270788912581e-05,
+      "loss": 0.4573,
+      "step": 300
+    },
+    {
+      "epoch": 0.3208955223880597,
+      "grad_norm": 1.1242753661227198,
+      "learning_rate": 5.134328358208955e-05,
+      "loss": 0.4555,
+      "step": 301
+    },
+    {
+      "epoch": 0.32196162046908317,
+      "grad_norm": 0.8386711391896823,
+      "learning_rate": 5.151385927505331e-05,
+      "loss": 0.4454,
+      "step": 302
+    },
+    {
+      "epoch": 0.3230277185501066,
+      "grad_norm": 1.0644491026122258,
+      "learning_rate": 5.168443496801706e-05,
+      "loss": 0.4544,
+      "step": 303
+    },
+    {
+      "epoch": 0.32409381663113007,
+      "grad_norm": 1.0964593961058677,
+      "learning_rate": 5.185501066098082e-05,
+      "loss": 0.4556,
+      "step": 304
+    },
+    {
+      "epoch": 0.3251599147121535,
+      "grad_norm": 0.9831083339112521,
+      "learning_rate": 5.202558635394456e-05,
+      "loss": 0.4467,
+      "step": 305
+    },
+    {
+      "epoch": 0.326226012793177,
+      "grad_norm": 0.9900349752767852,
+      "learning_rate": 5.219616204690832e-05,
+      "loss": 0.4516,
+      "step": 306
+    },
+    {
+      "epoch": 0.3272921108742004,
+      "grad_norm": 1.0598608792836142,
+      "learning_rate": 5.236673773987207e-05,
+      "loss": 0.4498,
+      "step": 307
+    },
+    {
+      "epoch": 0.3283582089552239,
+      "grad_norm": 0.9889355725212928,
+      "learning_rate": 5.2537313432835826e-05,
+      "loss": 0.4514,
+      "step": 308
+    },
+    {
+      "epoch": 0.3294243070362473,
+      "grad_norm": 0.9865582250975005,
+      "learning_rate": 5.270788912579957e-05,
+      "loss": 0.4513,
+      "step": 309
+    },
+    {
+      "epoch": 0.3304904051172708,
+      "grad_norm": 1.252376286795114,
+      "learning_rate": 5.287846481876333e-05,
+      "loss": 0.4523,
+      "step": 310
+    },
+    {
+      "epoch": 0.3315565031982942,
+      "grad_norm": 0.6493116364566459,
+      "learning_rate": 5.304904051172708e-05,
+      "loss": 0.445,
+      "step": 311
+    },
+    {
+      "epoch": 0.3326226012793177,
+      "grad_norm": 1.0832555318476744,
+      "learning_rate": 5.3219616204690836e-05,
+      "loss": 0.4551,
+      "step": 312
+    },
+    {
+      "epoch": 0.3336886993603412,
+      "grad_norm": 1.0891590791019572,
+      "learning_rate": 5.3390191897654595e-05,
+      "loss": 0.4501,
+      "step": 313
+    },
+    {
+      "epoch": 0.3347547974413646,
+      "grad_norm": 0.5965442300858297,
+      "learning_rate": 5.356076759061834e-05,
+      "loss": 0.4433,
+      "step": 314
+    },
+    {
+      "epoch": 0.3358208955223881,
+      "grad_norm": 0.9796043883687924,
+      "learning_rate": 5.37313432835821e-05,
+      "loss": 0.4517,
+      "step": 315
+    },
+    {
+      "epoch": 0.3368869936034115,
+      "grad_norm": 0.993909895121157,
+      "learning_rate": 5.3901918976545846e-05,
+      "loss": 0.452,
+      "step": 316
+    },
+    {
+      "epoch": 0.337953091684435,
+      "grad_norm": 0.6471597791846867,
+      "learning_rate": 5.4072494669509605e-05,
+      "loss": 0.4489,
+      "step": 317
+    },
+    {
+      "epoch": 0.3390191897654584,
+      "grad_norm": 1.200049501099017,
+      "learning_rate": 5.424307036247335e-05,
+      "loss": 0.4519,
+      "step": 318
+    },
+    {
+      "epoch": 0.3400852878464819,
+      "grad_norm": 0.8547891891381943,
+      "learning_rate": 5.441364605543711e-05,
+      "loss": 0.4491,
+      "step": 319
+    },
+    {
+      "epoch": 0.3411513859275053,
+      "grad_norm": 0.8872211588826856,
+      "learning_rate": 5.4584221748400855e-05,
+      "loss": 0.4522,
+      "step": 320
+    },
+    {
+      "epoch": 0.3422174840085288,
+      "grad_norm": 1.114364988812215,
+      "learning_rate": 5.4754797441364615e-05,
+      "loss": 0.4503,
+      "step": 321
+    },
+    {
+      "epoch": 0.34328358208955223,
+      "grad_norm": 0.9431484820101103,
+      "learning_rate": 5.492537313432836e-05,
+      "loss": 0.4433,
+      "step": 322
+    },
+    {
+      "epoch": 0.3443496801705757,
+      "grad_norm": 1.0650041065819467,
+      "learning_rate": 5.509594882729212e-05,
+      "loss": 0.4484,
+      "step": 323
+    },
+    {
+      "epoch": 0.34541577825159914,
+      "grad_norm": 1.1014629307818748,
+      "learning_rate": 5.5266524520255865e-05,
+      "loss": 0.4439,
+      "step": 324
+    },
+    {
+      "epoch": 0.3464818763326226,
+      "grad_norm": 1.3284550365482601,
+      "learning_rate": 5.5437100213219624e-05,
+      "loss": 0.4489,
+      "step": 325
+    },
+    {
+      "epoch": 0.34754797441364604,
+      "grad_norm": 0.6740188587682626,
+      "learning_rate": 5.560767590618337e-05,
+      "loss": 0.4461,
+      "step": 326
+    },
+    {
+      "epoch": 0.3486140724946695,
+      "grad_norm": 0.8535552014500319,
+      "learning_rate": 5.577825159914713e-05,
+      "loss": 0.4469,
+      "step": 327
+    },
+    {
+      "epoch": 0.34968017057569295,
+      "grad_norm": 0.9344951335512115,
+      "learning_rate": 5.5948827292110875e-05,
+      "loss": 0.4403,
+      "step": 328
+    },
+    {
+      "epoch": 0.35074626865671643,
+      "grad_norm": 1.6493682506419003,
+      "learning_rate": 5.6119402985074634e-05,
+      "loss": 0.4471,
+      "step": 329
+    },
+    {
+      "epoch": 0.35181236673773986,
+      "grad_norm": 0.5980828549596681,
+      "learning_rate": 5.628997867803838e-05,
+      "loss": 0.4407,
+      "step": 330
+    },
+    {
+      "epoch": 0.35287846481876334,
+      "grad_norm": 1.3993066518146489,
+      "learning_rate": 5.646055437100214e-05,
+      "loss": 0.4463,
+      "step": 331
+    },
+    {
+      "epoch": 0.35394456289978676,
+      "grad_norm": 1.364713422133084,
+      "learning_rate": 5.6631130063965884e-05,
+      "loss": 0.4515,
+      "step": 332
+    },
+    {
+      "epoch": 0.35501066098081024,
+      "grad_norm": 0.6912706432665139,
+      "learning_rate": 5.6801705756929644e-05,
+      "loss": 0.4495,
+      "step": 333
+    },
+    {
+      "epoch": 0.35607675906183367,
+      "grad_norm": 1.0070213135643369,
+      "learning_rate": 5.697228144989339e-05,
+      "loss": 0.4444,
+      "step": 334
+    },
+    {
+      "epoch": 0.35714285714285715,
+      "grad_norm": 1.60072329010338,
+      "learning_rate": 5.714285714285715e-05,
+      "loss": 0.4491,
+      "step": 335
+    },
+    {
+      "epoch": 0.3582089552238806,
+      "grad_norm": 0.7418596944182502,
+      "learning_rate": 5.7313432835820894e-05,
+      "loss": 0.4426,
+      "step": 336
+    },
+    {
+      "epoch": 0.35927505330490406,
+      "grad_norm": 1.5682849618752128,
+      "learning_rate": 5.748400852878465e-05,
+      "loss": 0.4557,
+      "step": 337
+    },
+    {
+      "epoch": 0.3603411513859275,
+      "grad_norm": 0.9757128890719955,
+      "learning_rate": 5.76545842217484e-05,
+      "loss": 0.4434,
+      "step": 338
+    },
+    {
+      "epoch": 0.36140724946695096,
+      "grad_norm": 1.023456270027187,
+      "learning_rate": 5.782515991471216e-05,
+      "loss": 0.4493,
+      "step": 339
+    },
+    {
+      "epoch": 0.3624733475479744,
+      "grad_norm": 1.1069512289563967,
+      "learning_rate": 5.7995735607675904e-05,
+      "loss": 0.4487,
+      "step": 340
+    },
+    {
+      "epoch": 0.36353944562899787,
+      "grad_norm": 1.2762171846197148,
+      "learning_rate": 5.816631130063966e-05,
+      "loss": 0.4494,
+      "step": 341
+    },
+    {
+      "epoch": 0.3646055437100213,
+      "grad_norm": 0.9563297488524058,
+      "learning_rate": 5.8336886993603415e-05,
+      "loss": 0.4538,
+      "step": 342
+    },
+    {
+      "epoch": 0.3656716417910448,
+      "grad_norm": 0.8844708150282824,
+      "learning_rate": 5.850746268656717e-05,
+      "loss": 0.4599,
+      "step": 343
+    },
+    {
+      "epoch": 0.36673773987206826,
+      "grad_norm": 0.9248540817564378,
+      "learning_rate": 5.867803837953093e-05,
+      "loss": 0.4532,
+      "step": 344
+    },
+    {
+      "epoch": 0.3678038379530917,
+      "grad_norm": 0.897676935037186,
+      "learning_rate": 5.884861407249467e-05,
+      "loss": 0.4424,
+      "step": 345
+    },
+    {
+      "epoch": 0.36886993603411516,
+      "grad_norm": 1.3371410250762796,
+      "learning_rate": 5.901918976545843e-05,
+      "loss": 0.4528,
+      "step": 346
+    },
+    {
+      "epoch": 0.3699360341151386,
+      "grad_norm": 0.710536412391709,
+      "learning_rate": 5.918976545842218e-05,
+      "loss": 0.4496,
+      "step": 347
+    },
+    {
+      "epoch": 0.37100213219616207,
+      "grad_norm": 0.9252679189981835,
+      "learning_rate": 5.936034115138594e-05,
+      "loss": 0.4512,
+      "step": 348
+    },
+    {
+      "epoch": 0.3720682302771855,
+      "grad_norm": 1.2517055018731635,
+      "learning_rate": 5.953091684434968e-05,
+      "loss": 0.4473,
+      "step": 349
+    },
+    {
+      "epoch": 0.373134328358209,
+      "grad_norm": 0.9298497655115758,
+      "learning_rate": 5.970149253731344e-05,
+      "loss": 0.4411,
+      "step": 350
+    },
+    {
+      "epoch": 0.3742004264392324,
+      "grad_norm": 0.7999639695039473,
+      "learning_rate": 5.987206823027719e-05,
+      "loss": 0.4525,
+      "step": 351
+    },
+    {
+      "epoch": 0.3752665245202559,
+      "grad_norm": 0.9596155583258207,
+      "learning_rate": 6.0042643923240946e-05,
+      "loss": 0.4472,
+      "step": 352
+    },
+    {
+      "epoch": 0.3763326226012793,
+      "grad_norm": 1.4841986378845615,
+      "learning_rate": 6.021321961620469e-05,
+      "loss": 0.453,
+      "step": 353
+    },
+    {
+      "epoch": 0.3773987206823028,
+      "grad_norm": 0.704856727131918,
+      "learning_rate": 6.038379530916845e-05,
+      "loss": 0.4418,
+      "step": 354
+    },
+    {
+      "epoch": 0.3784648187633262,
+      "grad_norm": 1.1708028615880235,
+      "learning_rate": 6.05543710021322e-05,
+      "loss": 0.4483,
+      "step": 355
+    },
+    {
+      "epoch": 0.3795309168443497,
+      "grad_norm": 1.6336721778285583,
+      "learning_rate": 6.0724946695095956e-05,
+      "loss": 0.4523,
+      "step": 356
+    },
+    {
+      "epoch": 0.3805970149253731,
+      "grad_norm": 0.7634972844271057,
+      "learning_rate": 6.08955223880597e-05,
+      "loss": 0.438,
+      "step": 357
+    },
+    {
+      "epoch": 0.3816631130063966,
+      "grad_norm": 2.0551249748572276,
+      "learning_rate": 6.106609808102346e-05,
+      "loss": 0.4465,
+      "step": 358
+    },
+    {
+      "epoch": 0.38272921108742003,
+      "grad_norm": 1.1247068935515665,
+      "learning_rate": 6.123667377398721e-05,
+      "loss": 0.4528,
+      "step": 359
+    },
+    {
+      "epoch": 0.3837953091684435,
+      "grad_norm": 2.399781202126181,
+      "learning_rate": 6.140724946695097e-05,
+      "loss": 0.4638,
+      "step": 360
+    },
+    {
+      "epoch": 0.38486140724946694,
+      "grad_norm": 2.227691194378861,
+      "learning_rate": 6.157782515991472e-05,
+      "loss": 0.4605,
+      "step": 361
+    },
+    {
+      "epoch": 0.3859275053304904,
+      "grad_norm": 1.286105573757533,
+      "learning_rate": 6.174840085287847e-05,
+      "loss": 0.4549,
+      "step": 362
+    },
+    {
+      "epoch": 0.38699360341151384,
+      "grad_norm": 1.5661635396554183,
+      "learning_rate": 6.191897654584222e-05,
+      "loss": 0.4553,
+      "step": 363
+    },
+    {
+      "epoch": 0.3880597014925373,
+      "grad_norm": 1.1356847004536101,
+      "learning_rate": 6.208955223880598e-05,
+      "loss": 0.4551,
+      "step": 364
+    },
+    {
+      "epoch": 0.38912579957356075,
+      "grad_norm": 1.5226959517373695,
+      "learning_rate": 6.226012793176973e-05,
+      "loss": 0.4527,
+      "step": 365
+    },
+    {
+      "epoch": 0.39019189765458423,
+      "grad_norm": 1.2678150824491392,
+      "learning_rate": 6.243070362473348e-05,
+      "loss": 0.4535,
+      "step": 366
+    },
+    {
+      "epoch": 0.39125799573560766,
+      "grad_norm": 1.2767687308223976,
+      "learning_rate": 6.260127931769723e-05,
+      "loss": 0.4533,
+      "step": 367
+    },
+    {
+      "epoch": 0.39232409381663114,
+      "grad_norm": 1.2583112180411133,
+      "learning_rate": 6.277185501066099e-05,
+      "loss": 0.4479,
+      "step": 368
+    },
+    {
+      "epoch": 0.39339019189765456,
+      "grad_norm": 0.8712817882443821,
+      "learning_rate": 6.294243070362474e-05,
+      "loss": 0.4487,
+      "step": 369
+    },
+    {
+      "epoch": 0.39445628997867804,
+      "grad_norm": 1.367324065405987,
+      "learning_rate": 6.311300639658849e-05,
+      "loss": 0.4478,
+      "step": 370
+    },
+    {
+      "epoch": 0.39552238805970147,
+      "grad_norm": 0.8890072622408898,
+      "learning_rate": 6.328358208955224e-05,
+      "loss": 0.448,
+      "step": 371
+    },
+    {
+      "epoch": 0.39658848614072495,
+      "grad_norm": 1.3147387007679658,
+      "learning_rate": 6.3454157782516e-05,
+      "loss": 0.4462,
+      "step": 372
+    },
+    {
+      "epoch": 0.3976545842217484,
+      "grad_norm": 1.0068738388038798,
+      "learning_rate": 6.362473347547975e-05,
+      "loss": 0.4477,
+      "step": 373
+    },
+    {
+      "epoch": 0.39872068230277186,
+      "grad_norm": 0.8943784235638984,
+      "learning_rate": 6.37953091684435e-05,
+      "loss": 0.4386,
+      "step": 374
+    },
+    {
+      "epoch": 0.3997867803837953,
+      "grad_norm": 0.9000653302119703,
+      "learning_rate": 6.396588486140725e-05,
+      "loss": 0.4473,
+      "step": 375
+    },
+    {
+      "epoch": 0.40085287846481876,
+      "grad_norm": 0.9661659593224818,
+      "learning_rate": 6.4136460554371e-05,
+      "loss": 0.438,
+      "step": 376
+    },
+    {
+      "epoch": 0.40191897654584224,
+      "grad_norm": 1.2385940487697282,
+      "learning_rate": 6.430703624733477e-05,
+      "loss": 0.4466,
+      "step": 377
+    },
+    {
+      "epoch": 0.40298507462686567,
+      "grad_norm": 0.7438645095725858,
+      "learning_rate": 6.447761194029851e-05,
+      "loss": 0.4419,
+      "step": 378
+    },
+    {
+      "epoch": 0.40405117270788915,
+      "grad_norm": 0.7506871300171193,
+      "learning_rate": 6.464818763326228e-05,
+      "loss": 0.4448,
+      "step": 379
+    },
+    {
+      "epoch": 0.4051172707889126,
+      "grad_norm": 0.9063302132407942,
+      "learning_rate": 6.481876332622601e-05,
+      "loss": 0.443,
+      "step": 380
+    },
+    {
+      "epoch": 0.40618336886993606,
+      "grad_norm": 0.7598737346973297,
+      "learning_rate": 6.498933901918978e-05,
+      "loss": 0.4387,
+      "step": 381
+    },
+    {
+      "epoch": 0.4072494669509595,
+      "grad_norm": 0.7700124455332041,
+      "learning_rate": 6.515991471215352e-05,
+      "loss": 0.4427,
+      "step": 382
+    },
+    {
+      "epoch": 0.40831556503198296,
+      "grad_norm": 0.8602508992172722,
+      "learning_rate": 6.533049040511728e-05,
+      "loss": 0.4417,
+      "step": 383
+    },
+    {
+      "epoch": 0.4093816631130064,
+      "grad_norm": 1.0830008793393127,
+      "learning_rate": 6.550106609808102e-05,
+      "loss": 0.4438,
+      "step": 384
+    },
+    {
+      "epoch": 0.41044776119402987,
+      "grad_norm": 1.556613073128131,
+      "learning_rate": 6.567164179104479e-05,
+      "loss": 0.4564,
+      "step": 385
+    },
+    {
+      "epoch": 0.4115138592750533,
+      "grad_norm": 0.6513365090394059,
+      "learning_rate": 6.584221748400853e-05,
+      "loss": 0.4549,
+      "step": 386
+    },
+    {
+      "epoch": 0.4125799573560768,
+      "grad_norm": 1.7019947089510525,
+      "learning_rate": 6.60127931769723e-05,
+      "loss": 0.4577,
+      "step": 387
+    },
+    {
+      "epoch": 0.4136460554371002,
+      "grad_norm": 0.8602634122130145,
+      "learning_rate": 6.618336886993603e-05,
+      "loss": 0.4453,
+      "step": 388
+    },
+    {
+      "epoch": 0.4147121535181237,
+      "grad_norm": 1.4701074651429544,
+      "learning_rate": 6.63539445628998e-05,
+      "loss": 0.4553,
+      "step": 389
+    },
+    {
+      "epoch": 0.4157782515991471,
+      "grad_norm": 1.0042942676179272,
+      "learning_rate": 6.652452025586354e-05,
+      "loss": 0.4517,
+      "step": 390
+    },
+    {
+      "epoch": 0.4168443496801706,
+      "grad_norm": 1.2563707215527287,
+      "learning_rate": 6.66950959488273e-05,
+      "loss": 0.4583,
+      "step": 391
+    },
+    {
+      "epoch": 0.417910447761194,
+      "grad_norm": 0.8702664926874648,
+      "learning_rate": 6.686567164179104e-05,
+      "loss": 0.4447,
+      "step": 392
+    },
+    {
+      "epoch": 0.4189765458422175,
+      "grad_norm": 1.2431851878497833,
+      "learning_rate": 6.703624733475481e-05,
+      "loss": 0.4561,
+      "step": 393
+    },
+    {
+      "epoch": 0.4200426439232409,
+      "grad_norm": 0.74543869723929,
+      "learning_rate": 6.720682302771855e-05,
+      "loss": 0.4433,
+      "step": 394
+    },
+    {
+      "epoch": 0.4211087420042644,
+      "grad_norm": 1.0726381467701929,
+      "learning_rate": 6.737739872068231e-05,
+      "loss": 0.4435,
+      "step": 395
+    },
+    {
+      "epoch": 0.42217484008528783,
+      "grad_norm": 0.8897029348079197,
+      "learning_rate": 6.754797441364605e-05,
+      "loss": 0.4432,
+      "step": 396
+    },
+    {
+      "epoch": 0.4232409381663113,
+      "grad_norm": 0.9052325689021318,
+      "learning_rate": 6.771855010660982e-05,
+      "loss": 0.4468,
+      "step": 397
+    },
+    {
+      "epoch": 0.42430703624733473,
+      "grad_norm": 0.9903919600049477,
+      "learning_rate": 6.788912579957356e-05,
+      "loss": 0.4421,
+      "step": 398
+    },
+    {
+      "epoch": 0.4253731343283582,
+      "grad_norm": 1.2565089097200555,
+      "learning_rate": 6.805970149253732e-05,
+      "loss": 0.4484,
+      "step": 399
+    },
+    {
+      "epoch": 0.42643923240938164,
+      "grad_norm": 1.217499983640722,
+      "learning_rate": 6.823027718550106e-05,
+      "loss": 0.45,
+      "step": 400
+    },
+    {
+      "epoch": 0.4275053304904051,
+      "grad_norm": 0.8648975399735686,
+      "learning_rate": 6.840085287846483e-05,
+      "loss": 0.4411,
+      "step": 401
+    },
+    {
+      "epoch": 0.42857142857142855,
+      "grad_norm": 0.7803010542574316,
+      "learning_rate": 6.857142857142857e-05,
+      "loss": 0.4452,
+      "step": 402
+    },
+    {
+      "epoch": 0.42963752665245203,
+      "grad_norm": 1.070484531211215,
+      "learning_rate": 6.874200426439233e-05,
+      "loss": 0.4415,
+      "step": 403
+    },
+    {
+      "epoch": 0.43070362473347545,
+      "grad_norm": 1.0518099835702677,
+      "learning_rate": 6.891257995735607e-05,
+      "loss": 0.4483,
+      "step": 404
+    },
+    {
+      "epoch": 0.43176972281449894,
+      "grad_norm": 1.0987178262200696,
+      "learning_rate": 6.908315565031984e-05,
+      "loss": 0.4504,
+      "step": 405
+    },
+    {
+      "epoch": 0.43283582089552236,
+      "grad_norm": 1.298631454730131,
+      "learning_rate": 6.925373134328358e-05,
+      "loss": 0.4484,
+      "step": 406
+    },
+    {
+      "epoch": 0.43390191897654584,
+      "grad_norm": 0.7350226964695964,
+      "learning_rate": 6.942430703624734e-05,
+      "loss": 0.4469,
+      "step": 407
+    },
+    {
+      "epoch": 0.4349680170575693,
+      "grad_norm": 1.003473508893585,
+      "learning_rate": 6.95948827292111e-05,
+      "loss": 0.4473,
+      "step": 408
+    },
+    {
+      "epoch": 0.43603411513859275,
+      "grad_norm": 1.5274417692020188,
+      "learning_rate": 6.976545842217485e-05,
+      "loss": 0.4406,
+      "step": 409
+    },
+    {
+      "epoch": 0.43710021321961623,
+      "grad_norm": 0.795881436552545,
+      "learning_rate": 6.99360341151386e-05,
+      "loss": 0.4458,
+      "step": 410
+    },
+    {
+      "epoch": 0.43816631130063965,
+      "grad_norm": 1.8802864262284982,
+      "learning_rate": 7.010660980810235e-05,
+      "loss": 0.452,
+      "step": 411
+    },
+    {
+      "epoch": 0.43923240938166314,
+      "grad_norm": 1.0226364397036434,
+      "learning_rate": 7.02771855010661e-05,
+      "loss": 0.4439,
+      "step": 412
+    },
+    {
+      "epoch": 0.44029850746268656,
+      "grad_norm": 2.2181657239884736,
+      "learning_rate": 7.044776119402986e-05,
+      "loss": 0.4635,
+      "step": 413
+    },
+    {
+      "epoch": 0.44136460554371004,
+      "grad_norm": 1.9407812835735618,
+      "learning_rate": 7.061833688699361e-05,
+      "loss": 0.4592,
+      "step": 414
+    },
+    {
+      "epoch": 0.44243070362473347,
+      "grad_norm": 1.309070344848072,
+      "learning_rate": 7.078891257995736e-05,
+      "loss": 0.4465,
+      "step": 415
+    },
+    {
+      "epoch": 0.44349680170575695,
+      "grad_norm": 1.2883250095052161,
+      "learning_rate": 7.095948827292111e-05,
+      "loss": 0.4444,
+      "step": 416
+    },
+    {
+      "epoch": 0.4445628997867804,
+      "grad_norm": 1.0903343923529414,
+      "learning_rate": 7.113006396588487e-05,
+      "loss": 0.4496,
+      "step": 417
+    },
+    {
+      "epoch": 0.44562899786780386,
+      "grad_norm": 1.183664382780786,
+      "learning_rate": 7.130063965884862e-05,
+      "loss": 0.4438,
+      "step": 418
+    },
+    {
+      "epoch": 0.4466950959488273,
+      "grad_norm": 0.658411366668824,
+      "learning_rate": 7.147121535181237e-05,
+      "loss": 0.4487,
+      "step": 419
+    },
+    {
+      "epoch": 0.44776119402985076,
+      "grad_norm": 0.9629596016665242,
+      "learning_rate": 7.164179104477612e-05,
+      "loss": 0.4492,
+      "step": 420
+    },
+    {
+      "epoch": 0.4488272921108742,
+      "grad_norm": 0.6918711094052541,
+      "learning_rate": 7.181236673773988e-05,
+      "loss": 0.4459,
+      "step": 421
+    },
+    {
+      "epoch": 0.44989339019189767,
+      "grad_norm": 0.8563536919057261,
+      "learning_rate": 7.198294243070363e-05,
+      "loss": 0.4414,
+      "step": 422
+    },
+    {
+      "epoch": 0.4509594882729211,
+      "grad_norm": 0.6997094585258404,
+      "learning_rate": 7.215351812366738e-05,
+      "loss": 0.4371,
+      "step": 423
+    },
+    {
+      "epoch": 0.4520255863539446,
+      "grad_norm": 0.9078702301189076,
+      "learning_rate": 7.232409381663113e-05,
+      "loss": 0.4385,
+      "step": 424
+    },
+    {
+      "epoch": 0.453091684434968,
+      "grad_norm": 0.9105908002111917,
+      "learning_rate": 7.249466950959489e-05,
+      "loss": 0.4435,
+      "step": 425
+    },
+    {
+      "epoch": 0.4541577825159915,
+      "grad_norm": 1.17133908131614,
+      "learning_rate": 7.266524520255864e-05,
+      "loss": 0.4436,
+      "step": 426
+    },
+    {
+      "epoch": 0.4552238805970149,
+      "grad_norm": 0.9871418897305207,
+      "learning_rate": 7.283582089552239e-05,
+      "loss": 0.4382,
+      "step": 427
+    },
+    {
+      "epoch": 0.4562899786780384,
+      "grad_norm": 1.1416965993237809,
+      "learning_rate": 7.300639658848614e-05,
+      "loss": 0.4413,
+      "step": 428
+    },
+    {
+      "epoch": 0.4573560767590618,
+      "grad_norm": 0.8995759612900418,
+      "learning_rate": 7.31769722814499e-05,
+      "loss": 0.441,
+      "step": 429
+    },
+    {
+      "epoch": 0.4584221748400853,
+      "grad_norm": 0.9218576549939405,
+      "learning_rate": 7.334754797441365e-05,
+      "loss": 0.4384,
+      "step": 430
+    },
+    {
+      "epoch": 0.4594882729211087,
+      "grad_norm": 0.858746072858543,
+      "learning_rate": 7.35181236673774e-05,
+      "loss": 0.4431,
+      "step": 431
+    },
+    {
+      "epoch": 0.4605543710021322,
+      "grad_norm": 0.935134912271826,
+      "learning_rate": 7.368869936034115e-05,
+      "loss": 0.4419,
+      "step": 432
+    },
+    {
+      "epoch": 0.4616204690831556,
+      "grad_norm": 0.9646962439556347,
+      "learning_rate": 7.38592750533049e-05,
+      "loss": 0.4447,
+      "step": 433
+    },
+    {
+      "epoch": 0.4626865671641791,
+      "grad_norm": 1.1656061718106945,
+      "learning_rate": 7.402985074626866e-05,
+      "loss": 0.4449,
+      "step": 434
+    },
+    {
+      "epoch": 0.46375266524520253,
+      "grad_norm": 0.7326350434710983,
+      "learning_rate": 7.420042643923241e-05,
+      "loss": 0.4306,
+      "step": 435
+    },
+    {
+      "epoch": 0.464818763326226,
+      "grad_norm": 0.4426357716461027,
+      "learning_rate": 7.437100213219616e-05,
+      "loss": 0.4391,
+      "step": 436
+    },
+    {
+      "epoch": 0.46588486140724944,
+      "grad_norm": 0.6078030487605691,
+      "learning_rate": 7.454157782515992e-05,
+      "loss": 0.4347,
+      "step": 437
+    },
+    {
+      "epoch": 0.4669509594882729,
+      "grad_norm": 0.8052535102563257,
+      "learning_rate": 7.471215351812367e-05,
+      "loss": 0.4399,
+      "step": 438
+    },
+    {
+      "epoch": 0.4680170575692964,
+      "grad_norm": 0.993986917853917,
+      "learning_rate": 7.488272921108743e-05,
+      "loss": 0.4491,
+      "step": 439
+    },
+    {
+      "epoch": 0.4690831556503198,
+      "grad_norm": 1.1313255325891831,
+      "learning_rate": 7.505330490405117e-05,
+      "loss": 0.4493,
+      "step": 440
+    },
+    {
+      "epoch": 0.4701492537313433,
+      "grad_norm": 0.6842654100869138,
+      "learning_rate": 7.522388059701494e-05,
+      "loss": 0.4381,
+      "step": 441
+    },
+    {
+      "epoch": 0.47121535181236673,
+      "grad_norm": 0.914667240873549,
+      "learning_rate": 7.539445628997868e-05,
+      "loss": 0.4298,
+      "step": 442
+    },
+    {
+      "epoch": 0.4722814498933902,
+      "grad_norm": 1.079253155327782,
+      "learning_rate": 7.556503198294244e-05,
+      "loss": 0.4419,
+      "step": 443
+    },
+    {
+      "epoch": 0.47334754797441364,
+      "grad_norm": 0.7735237895647086,
+      "learning_rate": 7.573560767590618e-05,
+      "loss": 0.4363,
+      "step": 444
+    },
+    {
+      "epoch": 0.4744136460554371,
+      "grad_norm": 0.8494163950632265,
+      "learning_rate": 7.590618336886995e-05,
+      "loss": 0.4424,
+      "step": 445
+    },
+    {
+      "epoch": 0.47547974413646055,
+      "grad_norm": 0.8832423465229242,
+      "learning_rate": 7.607675906183369e-05,
+      "loss": 0.4433,
+      "step": 446
+    },
+    {
+      "epoch": 0.47654584221748403,
+      "grad_norm": 0.9230538736492173,
+      "learning_rate": 7.624733475479745e-05,
+      "loss": 0.4373,
+      "step": 447
+    },
+    {
+      "epoch": 0.47761194029850745,
+      "grad_norm": 1.0547271971615721,
+      "learning_rate": 7.641791044776119e-05,
+      "loss": 0.4383,
+      "step": 448
+    },
+    {
+      "epoch": 0.47867803837953093,
+      "grad_norm": 0.8897022777657935,
+      "learning_rate": 7.658848614072496e-05,
+      "loss": 0.442,
+      "step": 449
+    },
+    {
+      "epoch": 0.47974413646055436,
+      "grad_norm": 0.9036068567889025,
+      "learning_rate": 7.675906183368871e-05,
+      "loss": 0.4438,
+      "step": 450
+    },
+    {
+      "epoch": 0.48081023454157784,
+      "grad_norm": 0.881957048697129,
+      "learning_rate": 7.692963752665246e-05,
+      "loss": 0.4417,
+      "step": 451
+    },
+    {
+      "epoch": 0.48187633262260127,
+      "grad_norm": 0.9500253572061279,
+      "learning_rate": 7.710021321961622e-05,
+      "loss": 0.4366,
+      "step": 452
+    },
+    {
+      "epoch": 0.48294243070362475,
+      "grad_norm": 1.2095267797004439,
+      "learning_rate": 7.727078891257997e-05,
+      "loss": 0.4362,
+      "step": 453
+    },
+    {
+      "epoch": 0.4840085287846482,
+      "grad_norm": 1.0482792150572497,
+      "learning_rate": 7.744136460554372e-05,
+      "loss": 0.4322,
+      "step": 454
+    },
+    {
+      "epoch": 0.48507462686567165,
+      "grad_norm": 1.0807249109863046,
+      "learning_rate": 7.761194029850747e-05,
+      "loss": 0.4463,
+      "step": 455
+    },
+    {
+      "epoch": 0.4861407249466951,
+      "grad_norm": 1.0949378713445743,
+      "learning_rate": 7.778251599147123e-05,
+      "loss": 0.4384,
+      "step": 456
+    },
+    {
+      "epoch": 0.48720682302771856,
+      "grad_norm": 1.1230076391837633,
+      "learning_rate": 7.795309168443498e-05,
+      "loss": 0.4328,
+      "step": 457
+    },
+    {
+      "epoch": 0.488272921108742,
+      "grad_norm": 0.9284196769926588,
+      "learning_rate": 7.812366737739873e-05,
+      "loss": 0.433,
+      "step": 458
+    },
+    {
+      "epoch": 0.48933901918976547,
+      "grad_norm": 0.8307602310830237,
+      "learning_rate": 7.829424307036248e-05,
+      "loss": 0.4355,
+      "step": 459
+    },
+    {
+      "epoch": 0.4904051172707889,
+      "grad_norm": 0.6845844913178624,
+      "learning_rate": 7.846481876332623e-05,
+      "loss": 0.4324,
+      "step": 460
+    },
+    {
+      "epoch": 0.4914712153518124,
+      "grad_norm": 0.6727473022234879,
+      "learning_rate": 7.863539445628999e-05,
+      "loss": 0.4343,
+      "step": 461
+    },
+    {
+      "epoch": 0.4925373134328358,
+      "grad_norm": 0.9118111002510331,
+      "learning_rate": 7.880597014925374e-05,
+      "loss": 0.4332,
+      "step": 462
+    },
+    {
+      "epoch": 0.4936034115138593,
+      "grad_norm": 1.0468419215009597,
+      "learning_rate": 7.897654584221749e-05,
+      "loss": 0.4303,
+      "step": 463
+    },
+    {
+      "epoch": 0.4946695095948827,
+      "grad_norm": 1.0423557702670259,
+      "learning_rate": 7.914712153518124e-05,
+      "loss": 0.4468,
+      "step": 464
+    },
+    {
+      "epoch": 0.4957356076759062,
+      "grad_norm": 1.0941125408721255,
+      "learning_rate": 7.9317697228145e-05,
+      "loss": 0.437,
+      "step": 465
+    },
+    {
+      "epoch": 0.4968017057569296,
+      "grad_norm": 1.009365737040848,
+      "learning_rate": 7.948827292110875e-05,
+      "loss": 0.4344,
+      "step": 466
+    },
+    {
+      "epoch": 0.4978678038379531,
+      "grad_norm": 0.976490369885011,
+      "learning_rate": 7.96588486140725e-05,
+      "loss": 0.4405,
+      "step": 467
+    },
+    {
+      "epoch": 0.4989339019189765,
+      "grad_norm": 1.0336971441325562,
+      "learning_rate": 7.982942430703625e-05,
+      "loss": 0.4415,
+      "step": 468
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1162353727508352,
+      "learning_rate": 8e-05,
+      "loss": 0.441,
+      "step": 469
+    },
+    {
+      "epoch": 0.5010660980810234,
+      "grad_norm": 0.870590499557518,
+      "learning_rate": 7.999998892103944e-05,
+      "loss": 0.4385,
+      "step": 470
+    },
+    {
+      "epoch": 0.502132196162047,
+      "grad_norm": 1.069571130953163,
+      "learning_rate": 7.999995568416386e-05,
+      "loss": 0.436,
+      "step": 471
+    },
+    {
+      "epoch": 0.5031982942430704,
+      "grad_norm": 0.9278286961514632,
+      "learning_rate": 7.99999002893917e-05,
+      "loss": 0.439,
+      "step": 472
+    },
+    {
+      "epoch": 0.5042643923240938,
+      "grad_norm": 0.7961408606111287,
+      "learning_rate": 7.999982273675363e-05,
+      "loss": 0.4411,
+      "step": 473
+    },
+    {
+      "epoch": 0.5053304904051172,
+      "grad_norm": 0.8234137572564111,
+      "learning_rate": 7.999972302629264e-05,
+      "loss": 0.4301,
+      "step": 474
+    },
+    {
+      "epoch": 0.5063965884861408,
+      "grad_norm": 0.8624956407119708,
+      "learning_rate": 7.999960115806391e-05,
+      "loss": 0.4329,
+      "step": 475
+    },
+    {
+      "epoch": 0.5074626865671642,
+      "grad_norm": 0.8970331272898654,
+      "learning_rate": 7.999945713213499e-05,
+      "loss": 0.4326,
+      "step": 476
+    },
+    {
+      "epoch": 0.5085287846481876,
+      "grad_norm": 1.0727116898193436,
+      "learning_rate": 7.999929094858566e-05,
+      "loss": 0.4345,
+      "step": 477
+    },
+    {
+      "epoch": 0.509594882729211,
+      "grad_norm": 1.3157426521337816,
+      "learning_rate": 7.999910260750796e-05,
+      "loss": 0.4431,
+      "step": 478
+    },
+    {
+      "epoch": 0.5106609808102346,
+      "grad_norm": 0.8323751755101396,
+      "learning_rate": 7.999889210900623e-05,
+      "loss": 0.4368,
+      "step": 479
+    },
+    {
+      "epoch": 0.511727078891258,
+      "grad_norm": 0.7429493009987966,
+      "learning_rate": 7.999865945319708e-05,
+      "loss": 0.4352,
+      "step": 480
+    },
+    {
+      "epoch": 0.5127931769722814,
+      "grad_norm": 0.7605461290267354,
+      "learning_rate": 7.999840464020936e-05,
+      "loss": 0.438,
+      "step": 481
+    },
+    {
+      "epoch": 0.5138592750533049,
+      "grad_norm": 1.0898292270229712,
+      "learning_rate": 7.999812767018428e-05,
+      "loss": 0.4397,
+      "step": 482
+    },
+    {
+      "epoch": 0.5149253731343284,
+      "grad_norm": 1.1380886467744538,
+      "learning_rate": 7.999782854327523e-05,
+      "loss": 0.4367,
+      "step": 483
+    },
+    {
+      "epoch": 0.5159914712153518,
+      "grad_norm": 0.923359986404487,
+      "learning_rate": 7.99975072596479e-05,
+      "loss": 0.4384,
+      "step": 484
+    },
+    {
+      "epoch": 0.5170575692963753,
+      "grad_norm": 0.8375674330038942,
+      "learning_rate": 7.999716381948029e-05,
+      "loss": 0.436,
+      "step": 485
+    },
+    {
+      "epoch": 0.5181236673773987,
+      "grad_norm": 0.7570588135036903,
+      "learning_rate": 7.999679822296263e-05,
+      "loss": 0.4333,
+      "step": 486
+    },
+    {
+      "epoch": 0.5191897654584222,
+      "grad_norm": 0.8549659681016898,
+      "learning_rate": 7.999641047029747e-05,
+      "loss": 0.4356,
+      "step": 487
+    },
+    {
+      "epoch": 0.5202558635394456,
+      "grad_norm": 0.8852253500570002,
+      "learning_rate": 7.999600056169956e-05,
+      "loss": 0.4367,
+      "step": 488
+    },
+    {
+      "epoch": 0.5213219616204691,
+      "grad_norm": 0.9048960575709987,
+      "learning_rate": 7.9995568497396e-05,
+      "loss": 0.435,
+      "step": 489
+    },
+    {
+      "epoch": 0.5223880597014925,
+      "grad_norm": 0.88758605404917,
+      "learning_rate": 7.999511427762612e-05,
+      "loss": 0.4326,
+      "step": 490
+    },
+    {
+      "epoch": 0.523454157782516,
+      "grad_norm": 0.9051344896994227,
+      "learning_rate": 7.999463790264155e-05,
+      "loss": 0.4345,
+      "step": 491
+    },
+    {
+      "epoch": 0.5245202558635395,
+      "grad_norm": 0.9863125025310046,
+      "learning_rate": 7.999413937270616e-05,
+      "loss": 0.4302,
+      "step": 492
+    },
+    {
+      "epoch": 0.5255863539445629,
+      "grad_norm": 1.0414867859886516,
+      "learning_rate": 7.999361868809611e-05,
+      "loss": 0.4339,
+      "step": 493
+    },
+    {
+      "epoch": 0.5266524520255863,
+      "grad_norm": 0.9324905753531538,
+      "learning_rate": 7.999307584909985e-05,
+      "loss": 0.4411,
+      "step": 494
+    },
+    {
+      "epoch": 0.5277185501066098,
+      "grad_norm": 0.8218386585042996,
+      "learning_rate": 7.999251085601804e-05,
+      "loss": 0.4349,
+      "step": 495
+    },
+    {
+      "epoch": 0.5287846481876333,
+      "grad_norm": 0.6954291556313583,
+      "learning_rate": 7.999192370916371e-05,
+      "loss": 0.4406,
+      "step": 496
+    },
+    {
+      "epoch": 0.5298507462686567,
+      "grad_norm": 0.599463593583278,
+      "learning_rate": 7.999131440886208e-05,
+      "loss": 0.4294,
+      "step": 497
+    },
+    {
+      "epoch": 0.5309168443496801,
+      "grad_norm": 0.6547393051285391,
+      "learning_rate": 7.999068295545068e-05,
+      "loss": 0.4257,
+      "step": 498
+    },
+    {
+      "epoch": 0.5319829424307037,
+      "grad_norm": 0.5821330571530523,
+      "learning_rate": 7.99900293492793e-05,
+      "loss": 0.4283,
+      "step": 499
+    },
+    {
+      "epoch": 0.5330490405117271,
+      "grad_norm": 0.6804915712665963,
+      "learning_rate": 7.998935359071001e-05,
+      "loss": 0.4328,
+      "step": 500
+    },
+    {
+      "epoch": 0.5341151385927505,
+      "grad_norm": 0.8261518180898303,
+      "learning_rate": 7.998865568011713e-05,
+      "loss": 0.4272,
+      "step": 501
+    },
+    {
+      "epoch": 0.535181236673774,
+      "grad_norm": 0.984885443645453,
+      "learning_rate": 7.998793561788727e-05,
+      "loss": 0.4304,
+      "step": 502
+    },
+    {
+      "epoch": 0.5362473347547975,
+      "grad_norm": 1.0533079413867938,
+      "learning_rate": 7.998719340441933e-05,
+      "loss": 0.4325,
+      "step": 503
+    },
+    {
+      "epoch": 0.5373134328358209,
+      "grad_norm": 0.7930549662599339,
+      "learning_rate": 7.998642904012442e-05,
+      "loss": 0.4318,
+      "step": 504
+    },
+    {
+      "epoch": 0.5383795309168443,
+      "grad_norm": 0.7826200841082803,
+      "learning_rate": 7.998564252542599e-05,
+      "loss": 0.4335,
+      "step": 505
+    },
+    {
+      "epoch": 0.5394456289978679,
+      "grad_norm": 0.9783531802962122,
+      "learning_rate": 7.998483386075972e-05,
+      "loss": 0.4359,
+      "step": 506
+    },
+    {
+      "epoch": 0.5405117270788913,
+      "grad_norm": 1.0305053743252428,
+      "learning_rate": 7.998400304657356e-05,
+      "loss": 0.4295,
+      "step": 507
+    },
+    {
+      "epoch": 0.5415778251599147,
+      "grad_norm": 1.0480888416790912,
+      "learning_rate": 7.998315008332773e-05,
+      "loss": 0.4312,
+      "step": 508
+    },
+    {
+      "epoch": 0.5426439232409381,
+      "grad_norm": 1.0913103402452111,
+      "learning_rate": 7.998227497149475e-05,
+      "loss": 0.4359,
+      "step": 509
+    },
+    {
+      "epoch": 0.5437100213219617,
+      "grad_norm": 0.7828175941941907,
+      "learning_rate": 7.998137771155938e-05,
+      "loss": 0.4253,
+      "step": 510
+    },
+    {
+      "epoch": 0.5447761194029851,
+      "grad_norm": 0.753244557765304,
+      "learning_rate": 7.998045830401864e-05,
+      "loss": 0.4393,
+      "step": 511
+    },
+    {
+      "epoch": 0.5458422174840085,
+      "grad_norm": 0.7846311196148755,
+      "learning_rate": 7.997951674938185e-05,
+      "loss": 0.4381,
+      "step": 512
+    },
+    {
+      "epoch": 0.5469083155650319,
+      "grad_norm": 0.5773342532146049,
+      "learning_rate": 7.997855304817059e-05,
+      "loss": 0.4324,
+      "step": 513
+    },
+    {
+      "epoch": 0.5479744136460555,
+      "grad_norm": 0.5590862741693943,
+      "learning_rate": 7.997756720091866e-05,
+      "loss": 0.4347,
+      "step": 514
+    },
+    {
+      "epoch": 0.5490405117270789,
+      "grad_norm": 0.5750952292279558,
+      "learning_rate": 7.997655920817222e-05,
+      "loss": 0.4316,
+      "step": 515
+    },
+    {
+      "epoch": 0.5501066098081023,
+      "grad_norm": 0.6003198820780814,
+      "learning_rate": 7.997552907048961e-05,
+      "loss": 0.4315,
+      "step": 516
+    },
+    {
+      "epoch": 0.5511727078891258,
+      "grad_norm": 0.7339386833494144,
+      "learning_rate": 7.997447678844148e-05,
+      "loss": 0.4334,
+      "step": 517
+    },
+    {
+      "epoch": 0.5522388059701493,
+      "grad_norm": 0.901683791960714,
+      "learning_rate": 7.997340236261076e-05,
+      "loss": 0.4301,
+      "step": 518
+    },
+    {
+      "epoch": 0.5533049040511727,
+      "grad_norm": 1.1497802402889508,
+      "learning_rate": 7.997230579359261e-05,
+      "loss": 0.4305,
+      "step": 519
+    },
+    {
+      "epoch": 0.5543710021321961,
+      "grad_norm": 1.0986758974484423,
+      "learning_rate": 7.997118708199447e-05,
+      "loss": 0.4319,
+      "step": 520
+    },
+    {
+      "epoch": 0.5554371002132196,
+      "grad_norm": 0.9768374453514556,
+      "learning_rate": 7.997004622843603e-05,
+      "loss": 0.4298,
+      "step": 521
+    },
+    {
+      "epoch": 0.5565031982942431,
+      "grad_norm": 0.8738815441198616,
+      "learning_rate": 7.996888323354932e-05,
+      "loss": 0.4334,
+      "step": 522
+    },
+    {
+      "epoch": 0.5575692963752665,
+      "grad_norm": 0.7838623932210756,
+      "learning_rate": 7.996769809797851e-05,
+      "loss": 0.4188,
+      "step": 523
+    },
+    {
+      "epoch": 0.55863539445629,
+      "grad_norm": 0.7503706641165158,
+      "learning_rate": 7.996649082238015e-05,
+      "loss": 0.4293,
+      "step": 524
+    },
+    {
+      "epoch": 0.5597014925373134,
+      "grad_norm": 0.7732501093769599,
+      "learning_rate": 7.9965261407423e-05,
+      "loss": 0.4222,
+      "step": 525
+    },
+    {
+      "epoch": 0.5607675906183369,
+      "grad_norm": 0.7009167747253003,
+      "learning_rate": 7.996400985378807e-05,
+      "loss": 0.4208,
+      "step": 526
+    },
+    {
+      "epoch": 0.5618336886993603,
+      "grad_norm": 0.6810945772619069,
+      "learning_rate": 7.996273616216868e-05,
+      "loss": 0.4327,
+      "step": 527
+    },
+    {
+      "epoch": 0.5628997867803838,
+      "grad_norm": 0.683953064924211,
+      "learning_rate": 7.996144033327038e-05,
+      "loss": 0.432,
+      "step": 528
+    },
+    {
+      "epoch": 0.5639658848614072,
+      "grad_norm": 0.6899323492718954,
+      "learning_rate": 7.9960122367811e-05,
+      "loss": 0.4213,
+      "step": 529
+    },
+    {
+      "epoch": 0.5650319829424307,
+      "grad_norm": 0.7321597867806313,
+      "learning_rate": 7.995878226652061e-05,
+      "loss": 0.4324,
+      "step": 530
+    },
+    {
+      "epoch": 0.5660980810234542,
+      "grad_norm": 0.8962870617353619,
+      "learning_rate": 7.995742003014156e-05,
+      "loss": 0.4331,
+      "step": 531
+    },
+    {
+      "epoch": 0.5671641791044776,
+      "grad_norm": 0.9482010305629667,
+      "learning_rate": 7.995603565942846e-05,
+      "loss": 0.4305,
+      "step": 532
+    },
+    {
+      "epoch": 0.5682302771855011,
+      "grad_norm": 0.8547510485520907,
+      "learning_rate": 7.995462915514819e-05,
+      "loss": 0.428,
+      "step": 533
+    },
+    {
+      "epoch": 0.5692963752665245,
+      "grad_norm": 0.7205166965219559,
+      "learning_rate": 7.995320051807987e-05,
+      "loss": 0.4274,
+      "step": 534
+    },
+    {
+      "epoch": 0.570362473347548,
+      "grad_norm": 0.6501576811903161,
+      "learning_rate": 7.995174974901489e-05,
+      "loss": 0.4349,
+      "step": 535
+    },
+    {
+      "epoch": 0.5714285714285714,
+      "grad_norm": 0.7124693365482168,
+      "learning_rate": 7.99502768487569e-05,
+      "loss": 0.4332,
+      "step": 536
+    },
+    {
+      "epoch": 0.5724946695095949,
+      "grad_norm": 0.9137269789343048,
+      "learning_rate": 7.994878181812181e-05,
+      "loss": 0.4329,
+      "step": 537
+    },
+    {
+      "epoch": 0.5735607675906184,
+      "grad_norm": 1.042428645077687,
+      "learning_rate": 7.994726465793782e-05,
+      "loss": 0.4251,
+      "step": 538
+    },
+    {
+      "epoch": 0.5746268656716418,
+      "grad_norm": 0.9004373071226697,
+      "learning_rate": 7.994572536904529e-05,
+      "loss": 0.4285,
+      "step": 539
+    },
+    {
+      "epoch": 0.5756929637526652,
+      "grad_norm": 0.8473522603996991,
+      "learning_rate": 7.994416395229696e-05,
+      "loss": 0.4366,
+      "step": 540
+    },
+    {
+      "epoch": 0.5767590618336887,
+      "grad_norm": 0.7717855677683063,
+      "learning_rate": 7.994258040855776e-05,
+      "loss": 0.426,
+      "step": 541
+    },
+    {
+      "epoch": 0.5778251599147122,
+      "grad_norm": 0.7447673936934801,
+      "learning_rate": 7.99409747387049e-05,
+      "loss": 0.4301,
+      "step": 542
+    },
+    {
+      "epoch": 0.5788912579957356,
+      "grad_norm": 0.7651727619884234,
+      "learning_rate": 7.993934694362782e-05,
+      "loss": 0.429,
+      "step": 543
+    },
+    {
+      "epoch": 0.579957356076759,
+      "grad_norm": 0.771334240782246,
+      "learning_rate": 7.993769702422824e-05,
+      "loss": 0.4351,
+      "step": 544
+    },
+    {
+      "epoch": 0.5810234541577826,
+      "grad_norm": 0.8138260174231243,
+      "learning_rate": 7.993602498142015e-05,
+      "loss": 0.4296,
+      "step": 545
+    },
+    {
+      "epoch": 0.582089552238806,
+      "grad_norm": 0.9083617257757822,
+      "learning_rate": 7.993433081612975e-05,
+      "loss": 0.4207,
+      "step": 546
+    },
+    {
+      "epoch": 0.5831556503198294,
+      "grad_norm": 1.015947743360487,
+      "learning_rate": 7.993261452929551e-05,
+      "loss": 0.435,
+      "step": 547
+    },
+    {
+      "epoch": 0.5842217484008528,
+      "grad_norm": 0.9636126108742088,
+      "learning_rate": 7.99308761218682e-05,
+      "loss": 0.43,
+      "step": 548
+    },
+    {
+      "epoch": 0.5852878464818764,
+      "grad_norm": 0.9480436655111312,
+      "learning_rate": 7.992911559481077e-05,
+      "loss": 0.4333,
+      "step": 549
+    },
+    {
+      "epoch": 0.5863539445628998,
+      "grad_norm": 0.9004301184201612,
+      "learning_rate": 7.992733294909848e-05,
+      "loss": 0.4287,
+      "step": 550
+    },
+    {
+      "epoch": 0.5874200426439232,
+      "grad_norm": 0.8680389446297649,
+      "learning_rate": 7.992552818571883e-05,
+      "loss": 0.4286,
+      "step": 551
+    },
+    {
+      "epoch": 0.5884861407249466,
+      "grad_norm": 0.8221740761869751,
+      "learning_rate": 7.992370130567155e-05,
+      "loss": 0.4264,
+      "step": 552
+    },
+    {
+      "epoch": 0.5895522388059702,
+      "grad_norm": 0.6856491729915255,
+      "learning_rate": 7.992185230996864e-05,
+      "loss": 0.4271,
+      "step": 553
+    },
+    {
+      "epoch": 0.5906183368869936,
+      "grad_norm": 0.587285396525197,
+      "learning_rate": 7.991998119963436e-05,
+      "loss": 0.4249,
+      "step": 554
+    },
+    {
+      "epoch": 0.591684434968017,
+      "grad_norm": 0.6202171961164545,
+      "learning_rate": 7.991808797570519e-05,
+      "loss": 0.437,
+      "step": 555
+    },
+    {
+      "epoch": 0.5927505330490405,
+      "grad_norm": 0.5703467722471379,
+      "learning_rate": 7.991617263922988e-05,
+      "loss": 0.4261,
+      "step": 556
+    },
+    {
+      "epoch": 0.593816631130064,
+      "grad_norm": 0.5467752825405972,
+      "learning_rate": 7.991423519126945e-05,
+      "loss": 0.4261,
+      "step": 557
+    },
+    {
+      "epoch": 0.5948827292110874,
+      "grad_norm": 0.584586914451865,
+      "learning_rate": 7.991227563289713e-05,
+      "loss": 0.4306,
+      "step": 558
+    },
+    {
+      "epoch": 0.5959488272921108,
+      "grad_norm": 0.5995377319354194,
+      "learning_rate": 7.991029396519839e-05,
+      "loss": 0.4295,
+      "step": 559
+    },
+    {
+      "epoch": 0.5970149253731343,
+      "grad_norm": 0.5825912670190095,
+      "learning_rate": 7.9908290189271e-05,
+      "loss": 0.4246,
+      "step": 560
+    },
+    {
+      "epoch": 0.5980810234541578,
+      "grad_norm": 0.5445221492611624,
+      "learning_rate": 7.990626430622494e-05,
+      "loss": 0.4258,
+      "step": 561
+    },
+    {
+      "epoch": 0.5991471215351812,
+      "grad_norm": 0.6388885844122285,
+      "learning_rate": 7.990421631718244e-05,
+      "loss": 0.4237,
+      "step": 562
+    },
+    {
+      "epoch": 0.6002132196162047,
+      "grad_norm": 0.7985864321661731,
+      "learning_rate": 7.9902146223278e-05,
+      "loss": 0.4263,
+      "step": 563
+    },
+    {
+      "epoch": 0.6012793176972282,
+      "grad_norm": 0.9243564293353863,
+      "learning_rate": 7.990005402565831e-05,
+      "loss": 0.4319,
+      "step": 564
+    },
+    {
+      "epoch": 0.6023454157782516,
+      "grad_norm": 1.0041597610953659,
+      "learning_rate": 7.989793972548236e-05,
+      "loss": 0.4317,
+      "step": 565
+    },
+    {
+      "epoch": 0.603411513859275,
+      "grad_norm": 0.9913301041254353,
+      "learning_rate": 7.989580332392137e-05,
+      "loss": 0.4228,
+      "step": 566
+    },
+    {
+      "epoch": 0.6044776119402985,
+      "grad_norm": 0.9215600177004183,
+      "learning_rate": 7.989364482215878e-05,
+      "loss": 0.4349,
+      "step": 567
+    },
+    {
+      "epoch": 0.605543710021322,
+      "grad_norm": 0.8910563956821462,
+      "learning_rate": 7.989146422139029e-05,
+      "loss": 0.4256,
+      "step": 568
+    },
+    {
+      "epoch": 0.6066098081023454,
+      "grad_norm": 0.9133910919719787,
+      "learning_rate": 7.988926152282384e-05,
+      "loss": 0.4269,
+      "step": 569
+    },
+    {
+      "epoch": 0.6076759061833689,
+      "grad_norm": 0.964984044461557,
+      "learning_rate": 7.988703672767962e-05,
+      "loss": 0.4271,
+      "step": 570
+    },
+    {
+      "epoch": 0.6087420042643923,
+      "grad_norm": 1.0759011919787849,
+      "learning_rate": 7.988478983719003e-05,
+      "loss": 0.4252,
+      "step": 571
+    },
+    {
+      "epoch": 0.6098081023454158,
+      "grad_norm": 0.8837314544429662,
+      "learning_rate": 7.988252085259976e-05,
+      "loss": 0.4297,
+      "step": 572
+    },
+    {
+      "epoch": 0.6108742004264393,
+      "grad_norm": 0.6734363195402464,
+      "learning_rate": 7.988022977516569e-05,
+      "loss": 0.4295,
+      "step": 573
+    },
+    {
+      "epoch": 0.6119402985074627,
+      "grad_norm": 0.6034655355722486,
+      "learning_rate": 7.987791660615695e-05,
+      "loss": 0.4251,
+      "step": 574
+    },
+    {
+      "epoch": 0.6130063965884861,
+      "grad_norm": 0.5138841355005869,
+      "learning_rate": 7.987558134685494e-05,
+      "loss": 0.4257,
+      "step": 575
+    },
+    {
+      "epoch": 0.6140724946695096,
+      "grad_norm": 0.5645600852787206,
+      "learning_rate": 7.987322399855324e-05,
+      "loss": 0.4241,
+      "step": 576
+    },
+    {
+      "epoch": 0.6151385927505331,
+      "grad_norm": 0.5906802943834981,
+      "learning_rate": 7.987084456255773e-05,
+      "loss": 0.4315,
+      "step": 577
+    },
+    {
+      "epoch": 0.6162046908315565,
+      "grad_norm": 0.6711742756532632,
+      "learning_rate": 7.986844304018649e-05,
+      "loss": 0.429,
+      "step": 578
+    },
+    {
+      "epoch": 0.6172707889125799,
+      "grad_norm": 0.8364188220124186,
+      "learning_rate": 7.986601943276982e-05,
+      "loss": 0.4195,
+      "step": 579
+    },
+    {
+      "epoch": 0.6183368869936035,
+      "grad_norm": 0.9774403295859477,
+      "learning_rate": 7.986357374165028e-05,
+      "loss": 0.4332,
+      "step": 580
+    },
+    {
+      "epoch": 0.6194029850746269,
+      "grad_norm": 1.04494654978909,
+      "learning_rate": 7.986110596818265e-05,
+      "loss": 0.428,
+      "step": 581
+    },
+    {
+      "epoch": 0.6204690831556503,
+      "grad_norm": 0.9878512856031877,
+      "learning_rate": 7.985861611373397e-05,
+      "loss": 0.4248,
+      "step": 582
+    },
+    {
+      "epoch": 0.6215351812366737,
+      "grad_norm": 0.9524713066725912,
+      "learning_rate": 7.985610417968348e-05,
+      "loss": 0.4251,
+      "step": 583
+    },
+    {
+      "epoch": 0.6226012793176973,
+      "grad_norm": 0.74751402505921,
+      "learning_rate": 7.985357016742264e-05,
+      "loss": 0.4319,
+      "step": 584
+    },
+    {
+      "epoch": 0.6236673773987207,
+      "grad_norm": 0.4034413623064052,
+      "learning_rate": 7.985101407835519e-05,
+      "loss": 0.4271,
+      "step": 585
+    },
+    {
+      "epoch": 0.6247334754797441,
+      "grad_norm": 0.43538759493027,
+      "learning_rate": 7.984843591389706e-05,
+      "loss": 0.4295,
+      "step": 586
+    },
+    {
+      "epoch": 0.6257995735607675,
+      "grad_norm": 0.5770360682304801,
+      "learning_rate": 7.984583567547643e-05,
+      "loss": 0.4247,
+      "step": 587
+    },
+    {
+      "epoch": 0.6268656716417911,
+      "grad_norm": 0.601220491680326,
+      "learning_rate": 7.984321336453364e-05,
+      "loss": 0.4217,
+      "step": 588
+    },
+    {
+      "epoch": 0.6279317697228145,
+      "grad_norm": 0.5416809102850237,
+      "learning_rate": 7.984056898252141e-05,
+      "loss": 0.4178,
+      "step": 589
+    },
+    {
+      "epoch": 0.6289978678038379,
+      "grad_norm": 0.5991300962093767,
+      "learning_rate": 7.983790253090452e-05,
+      "loss": 0.4268,
+      "step": 590
+    },
+    {
+      "epoch": 0.6300639658848614,
+      "grad_norm": 0.5939600973200748,
+      "learning_rate": 7.983521401116005e-05,
+      "loss": 0.4201,
+      "step": 591
+    },
+    {
+      "epoch": 0.6311300639658849,
+      "grad_norm": 0.5745808408760033,
+      "learning_rate": 7.983250342477733e-05,
+      "loss": 0.4263,
+      "step": 592
+    },
+    {
+      "epoch": 0.6321961620469083,
+      "grad_norm": 0.6369439414081867,
+      "learning_rate": 7.982977077325788e-05,
+      "loss": 0.4221,
+      "step": 593
+    },
+    {
+      "epoch": 0.6332622601279317,
+      "grad_norm": 0.7105979728558217,
+      "learning_rate": 7.98270160581154e-05,
+      "loss": 0.4235,
+      "step": 594
+    },
+    {
+      "epoch": 0.6343283582089553,
+      "grad_norm": 0.6814030775883104,
+      "learning_rate": 7.982423928087593e-05,
+      "loss": 0.4255,
+      "step": 595
+    },
+    {
+      "epoch": 0.6353944562899787,
+      "grad_norm": 0.6048922616611062,
+      "learning_rate": 7.982144044307762e-05,
+      "loss": 0.4216,
+      "step": 596
+    },
+    {
+      "epoch": 0.6364605543710021,
+      "grad_norm": 0.6271148450150956,
+      "learning_rate": 7.981861954627088e-05,
+      "loss": 0.429,
+      "step": 597
+    },
+    {
+      "epoch": 0.6375266524520256,
+      "grad_norm": 0.7648541301808355,
+      "learning_rate": 7.981577659201833e-05,
+      "loss": 0.425,
+      "step": 598
+    },
+    {
+      "epoch": 0.6385927505330491,
+      "grad_norm": 0.7722433311546377,
+      "learning_rate": 7.981291158189486e-05,
+      "loss": 0.4266,
+      "step": 599
+    },
+    {
+      "epoch": 0.6396588486140725,
+      "grad_norm": 0.8045597536473933,
+      "learning_rate": 7.98100245174875e-05,
+      "loss": 0.4226,
+      "step": 600
+    },
+    {
+      "epoch": 0.6407249466950959,
+      "grad_norm": 0.985587897420224,
+      "learning_rate": 7.980711540039554e-05,
+      "loss": 0.422,
+      "step": 601
+    },
+    {
+      "epoch": 0.6417910447761194,
+      "grad_norm": 1.1286184314617196,
+      "learning_rate": 7.980418423223049e-05,
+      "loss": 0.4215,
+      "step": 602
+    },
+    {
+      "epoch": 0.6428571428571429,
+      "grad_norm": 0.7080020035877171,
+      "learning_rate": 7.980123101461606e-05,
+      "loss": 0.4247,
+      "step": 603
+    },
+    {
+      "epoch": 0.6439232409381663,
+      "grad_norm": 0.5626555829828844,
+      "learning_rate": 7.979825574918818e-05,
+      "loss": 0.419,
+      "step": 604
+    },
+    {
+      "epoch": 0.6449893390191898,
+      "grad_norm": 0.636169281039428,
+      "learning_rate": 7.979525843759499e-05,
+      "loss": 0.4278,
+      "step": 605
+    },
+    {
+      "epoch": 0.6460554371002132,
+      "grad_norm": 0.7198635719149019,
+      "learning_rate": 7.979223908149685e-05,
+      "loss": 0.4255,
+      "step": 606
+    },
+    {
+      "epoch": 0.6471215351812367,
+      "grad_norm": 0.8342343801146002,
+      "learning_rate": 7.978919768256631e-05,
+      "loss": 0.4227,
+      "step": 607
+    },
+    {
+      "epoch": 0.6481876332622601,
+      "grad_norm": 0.9031065878045232,
+      "learning_rate": 7.978613424248818e-05,
+      "loss": 0.4308,
+      "step": 608
+    },
+    {
+      "epoch": 0.6492537313432836,
+      "grad_norm": 0.8023633786215478,
+      "learning_rate": 7.978304876295941e-05,
+      "loss": 0.419,
+      "step": 609
+    },
+    {
+      "epoch": 0.650319829424307,
+      "grad_norm": 0.9239599844896783,
+      "learning_rate": 7.977994124568922e-05,
+      "loss": 0.424,
+      "step": 610
+    },
+    {
+      "epoch": 0.6513859275053305,
+      "grad_norm": 0.9482115919888203,
+      "learning_rate": 7.9776811692399e-05,
+      "loss": 0.4186,
+      "step": 611
+    },
+    {
+      "epoch": 0.652452025586354,
+      "grad_norm": 0.786317946938952,
+      "learning_rate": 7.977366010482236e-05,
+      "loss": 0.4141,
+      "step": 612
+    },
+    {
+      "epoch": 0.6535181236673774,
+      "grad_norm": 0.6264488608057049,
+      "learning_rate": 7.977048648470513e-05,
+      "loss": 0.4209,
+      "step": 613
+    },
+    {
+      "epoch": 0.6545842217484008,
+      "grad_norm": 0.7117064313531933,
+      "learning_rate": 7.976729083380532e-05,
+      "loss": 0.4262,
+      "step": 614
+    },
+    {
+      "epoch": 0.6556503198294243,
+      "grad_norm": 0.7175291742270824,
+      "learning_rate": 7.976407315389314e-05,
+      "loss": 0.4251,
+      "step": 615
+    },
+    {
+      "epoch": 0.6567164179104478,
+      "grad_norm": 0.6490475195317971,
+      "learning_rate": 7.976083344675105e-05,
+      "loss": 0.4137,
+      "step": 616
+    },
+    {
+      "epoch": 0.6577825159914712,
+      "grad_norm": 0.6621191439676352,
+      "learning_rate": 7.975757171417365e-05,
+      "loss": 0.4216,
+      "step": 617
+    },
+    {
+      "epoch": 0.6588486140724946,
+      "grad_norm": 0.6960151950050518,
+      "learning_rate": 7.97542879579678e-05,
+      "loss": 0.4276,
+      "step": 618
+    },
+    {
+      "epoch": 0.6599147121535182,
+      "grad_norm": 0.6648695880576607,
+      "learning_rate": 7.975098217995248e-05,
+      "loss": 0.4209,
+      "step": 619
+    },
+    {
+      "epoch": 0.6609808102345416,
+      "grad_norm": 0.6121386791069903,
+      "learning_rate": 7.974765438195897e-05,
+      "loss": 0.419,
+      "step": 620
+    },
+    {
+      "epoch": 0.662046908315565,
+      "grad_norm": 0.5748855152032502,
+      "learning_rate": 7.974430456583069e-05,
+      "loss": 0.4249,
+      "step": 621
+    },
+    {
+      "epoch": 0.6631130063965884,
+      "grad_norm": 0.5788717188362661,
+      "learning_rate": 7.974093273342325e-05,
+      "loss": 0.4204,
+      "step": 622
+    },
+    {
+      "epoch": 0.664179104477612,
+      "grad_norm": 0.499053345156499,
+      "learning_rate": 7.973753888660446e-05,
+      "loss": 0.4258,
+      "step": 623
+    },
+    {
+      "epoch": 0.6652452025586354,
+      "grad_norm": 0.6143099826003396,
+      "learning_rate": 7.973412302725435e-05,
+      "loss": 0.4233,
+      "step": 624
+    },
+    {
+      "epoch": 0.6663113006396588,
+      "grad_norm": 0.7815047847063621,
+      "learning_rate": 7.973068515726514e-05,
+      "loss": 0.4226,
+      "step": 625
+    },
+    {
+      "epoch": 0.6673773987206824,
+      "grad_norm": 0.8746470159019124,
+      "learning_rate": 7.972722527854119e-05,
+      "loss": 0.4286,
+      "step": 626
+    },
+    {
+      "epoch": 0.6684434968017058,
+      "grad_norm": 0.8412210250981513,
+      "learning_rate": 7.972374339299915e-05,
+      "loss": 0.4222,
+      "step": 627
+    },
+    {
+      "epoch": 0.6695095948827292,
+      "grad_norm": 0.8583591463497402,
+      "learning_rate": 7.972023950256775e-05,
+      "loss": 0.4248,
+      "step": 628
+    },
+    {
+      "epoch": 0.6705756929637526,
+      "grad_norm": 0.8975533200078734,
+      "learning_rate": 7.9716713609188e-05,
+      "loss": 0.4332,
+      "step": 629
+    },
+    {
+      "epoch": 0.6716417910447762,
+      "grad_norm": 0.9787622637115202,
+      "learning_rate": 7.971316571481306e-05,
+      "loss": 0.4305,
+      "step": 630
+    },
+    {
+      "epoch": 0.6727078891257996,
+      "grad_norm": 0.9814190373327007,
+      "learning_rate": 7.970959582140825e-05,
+      "loss": 0.4339,
+      "step": 631
+    },
+    {
+      "epoch": 0.673773987206823,
+      "grad_norm": 0.9082356663533697,
+      "learning_rate": 7.970600393095113e-05,
+      "loss": 0.4262,
+      "step": 632
+    },
+    {
+      "epoch": 0.6748400852878464,
+      "grad_norm": 0.7973168134733778,
+      "learning_rate": 7.970239004543141e-05,
+      "loss": 0.4263,
+      "step": 633
+    },
+    {
+      "epoch": 0.67590618336887,
+      "grad_norm": 0.7223873106620373,
+      "learning_rate": 7.969875416685101e-05,
+      "loss": 0.4174,
+      "step": 634
+    },
+    {
+      "epoch": 0.6769722814498934,
+      "grad_norm": 0.7431984730465593,
+      "learning_rate": 7.9695096297224e-05,
+      "loss": 0.4296,
+      "step": 635
+    },
+    {
+      "epoch": 0.6780383795309168,
+      "grad_norm": 0.7282657781320044,
+      "learning_rate": 7.969141643857665e-05,
+      "loss": 0.4267,
+      "step": 636
+    },
+    {
+      "epoch": 0.6791044776119403,
+      "grad_norm": 0.5869814576553907,
+      "learning_rate": 7.968771459294742e-05,
+      "loss": 0.4191,
+      "step": 637
+    },
+    {
+      "epoch": 0.6801705756929638,
+      "grad_norm": 0.5004088468685179,
+      "learning_rate": 7.968399076238694e-05,
+      "loss": 0.4198,
+      "step": 638
+    },
+    {
+      "epoch": 0.6812366737739872,
+      "grad_norm": 0.6199279637183518,
+      "learning_rate": 7.968024494895802e-05,
+      "loss": 0.423,
+      "step": 639
+    },
+    {
+      "epoch": 0.6823027718550106,
+      "grad_norm": 0.6026223540177856,
+      "learning_rate": 7.967647715473563e-05,
+      "loss": 0.4252,
+      "step": 640
+    },
+    {
+      "epoch": 0.6833688699360341,
+      "grad_norm": 0.5355415437778251,
+      "learning_rate": 7.967268738180694e-05,
+      "loss": 0.4131,
+      "step": 641
+    },
+    {
+      "epoch": 0.6844349680170576,
+      "grad_norm": 0.5663522144573513,
+      "learning_rate": 7.966887563227132e-05,
+      "loss": 0.4252,
+      "step": 642
+    },
+    {
+      "epoch": 0.685501066098081,
+      "grad_norm": 0.5690136220036935,
+      "learning_rate": 7.966504190824021e-05,
+      "loss": 0.4186,
+      "step": 643
+    },
+    {
+      "epoch": 0.6865671641791045,
+      "grad_norm": 0.6635581651982301,
+      "learning_rate": 7.966118621183735e-05,
+      "loss": 0.4161,
+      "step": 644
+    },
+    {
+      "epoch": 0.6876332622601279,
+      "grad_norm": 0.7754746339129004,
+      "learning_rate": 7.96573085451986e-05,
+      "loss": 0.4285,
+      "step": 645
+    },
+    {
+      "epoch": 0.6886993603411514,
+      "grad_norm": 0.9415320034442117,
+      "learning_rate": 7.965340891047196e-05,
+      "loss": 0.4242,
+      "step": 646
+    },
+    {
+      "epoch": 0.6897654584221748,
+      "grad_norm": 1.133439191591166,
+      "learning_rate": 7.964948730981763e-05,
+      "loss": 0.4264,
+      "step": 647
+    },
+    {
+      "epoch": 0.6908315565031983,
+      "grad_norm": 0.8248991981565036,
+      "learning_rate": 7.964554374540797e-05,
+      "loss": 0.4271,
+      "step": 648
+    },
+    {
+      "epoch": 0.6918976545842217,
+      "grad_norm": 0.49941087039661575,
+      "learning_rate": 7.964157821942752e-05,
+      "loss": 0.4185,
+      "step": 649
+    },
+    {
+      "epoch": 0.6929637526652452,
+      "grad_norm": 0.5088744299454488,
+      "learning_rate": 7.963759073407297e-05,
+      "loss": 0.4228,
+      "step": 650
+    },
+    {
+      "epoch": 0.6940298507462687,
+      "grad_norm": 0.5561088080079999,
+      "learning_rate": 7.963358129155318e-05,
+      "loss": 0.4199,
+      "step": 651
+    },
+    {
+      "epoch": 0.6950959488272921,
+      "grad_norm": 0.5380479849912571,
+      "learning_rate": 7.962954989408916e-05,
+      "loss": 0.4251,
+      "step": 652
+    },
+    {
+      "epoch": 0.6961620469083155,
+      "grad_norm": 0.5750528453656307,
+      "learning_rate": 7.962549654391412e-05,
+      "loss": 0.4199,
+      "step": 653
+    },
+    {
+      "epoch": 0.697228144989339,
+      "grad_norm": 0.6325892582842803,
+      "learning_rate": 7.962142124327338e-05,
+      "loss": 0.4206,
+      "step": 654
+    },
+    {
+      "epoch": 0.6982942430703625,
+      "grad_norm": 0.7898343271492005,
+      "learning_rate": 7.961732399442448e-05,
+      "loss": 0.422,
+      "step": 655
+    },
+    {
+      "epoch": 0.6993603411513859,
+      "grad_norm": 0.9578185478631844,
+      "learning_rate": 7.961320479963703e-05,
+      "loss": 0.4158,
+      "step": 656
+    },
+    {
+      "epoch": 0.7004264392324094,
+      "grad_norm": 1.064875777304307,
+      "learning_rate": 7.96090636611929e-05,
+      "loss": 0.4212,
+      "step": 657
+    },
+    {
+      "epoch": 0.7014925373134329,
+      "grad_norm": 0.870822886466798,
+      "learning_rate": 7.960490058138604e-05,
+      "loss": 0.4294,
+      "step": 658
+    },
+    {
+      "epoch": 0.7025586353944563,
+      "grad_norm": 0.7665000673413538,
+      "learning_rate": 7.960071556252259e-05,
+      "loss": 0.4293,
+      "step": 659
+    },
+    {
+      "epoch": 0.7036247334754797,
+      "grad_norm": 0.7943565964649919,
+      "learning_rate": 7.959650860692082e-05,
+      "loss": 0.4153,
+      "step": 660
+    },
+    {
+      "epoch": 0.7046908315565032,
+      "grad_norm": 0.820866857806031,
+      "learning_rate": 7.959227971691118e-05,
+      "loss": 0.4218,
+      "step": 661
+    },
+    {
+      "epoch": 0.7057569296375267,
+      "grad_norm": 0.6751476503197974,
+      "learning_rate": 7.958802889483626e-05,
+      "loss": 0.4193,
+      "step": 662
+    },
+    {
+      "epoch": 0.7068230277185501,
+      "grad_norm": 0.3829844790409102,
+      "learning_rate": 7.958375614305076e-05,
+      "loss": 0.42,
+      "step": 663
+    },
+    {
+      "epoch": 0.7078891257995735,
+      "grad_norm": 0.5522600827629381,
+      "learning_rate": 7.957946146392159e-05,
+      "loss": 0.4224,
+      "step": 664
+    },
+    {
+      "epoch": 0.7089552238805971,
+      "grad_norm": 0.7878606313595656,
+      "learning_rate": 7.957514485982778e-05,
+      "loss": 0.4144,
+      "step": 665
+    },
+    {
+      "epoch": 0.7100213219616205,
+      "grad_norm": 0.7385093188465528,
+      "learning_rate": 7.95708063331605e-05,
+      "loss": 0.4259,
+      "step": 666
+    },
+    {
+      "epoch": 0.7110874200426439,
+      "grad_norm": 0.58178433551703,
+      "learning_rate": 7.956644588632307e-05,
+      "loss": 0.422,
+      "step": 667
+    },
+    {
+      "epoch": 0.7121535181236673,
+      "grad_norm": 0.6284258703892865,
+      "learning_rate": 7.956206352173093e-05,
+      "loss": 0.421,
+      "step": 668
+    },
+    {
+      "epoch": 0.7132196162046909,
+      "grad_norm": 0.7936414413541673,
+      "learning_rate": 7.95576592418117e-05,
+      "loss": 0.4259,
+      "step": 669
+    },
+    {
+      "epoch": 0.7142857142857143,
+      "grad_norm": 0.8521993846503206,
+      "learning_rate": 7.955323304900514e-05,
+      "loss": 0.4228,
+      "step": 670
+    },
+    {
+      "epoch": 0.7153518123667377,
+      "grad_norm": 0.7127889848840783,
+      "learning_rate": 7.954878494576312e-05,
+      "loss": 0.4201,
+      "step": 671
+    },
+    {
+      "epoch": 0.7164179104477612,
+      "grad_norm": 0.6074747505915871,
+      "learning_rate": 7.954431493454964e-05,
+      "loss": 0.4213,
+      "step": 672
+    },
+    {
+      "epoch": 0.7174840085287847,
+      "grad_norm": 0.7129990882836668,
+      "learning_rate": 7.953982301784085e-05,
+      "loss": 0.4234,
+      "step": 673
+    },
+    {
+      "epoch": 0.7185501066098081,
+      "grad_norm": 0.7722302015202298,
+      "learning_rate": 7.953530919812506e-05,
+      "loss": 0.4191,
+      "step": 674
+    },
+    {
+      "epoch": 0.7196162046908315,
+      "grad_norm": 0.7433421116192664,
+      "learning_rate": 7.95307734779027e-05,
+      "loss": 0.4211,
+      "step": 675
+    },
+    {
+      "epoch": 0.720682302771855,
+      "grad_norm": 0.8264421753718556,
+      "learning_rate": 7.95262158596863e-05,
+      "loss": 0.4218,
+      "step": 676
+    },
+    {
+      "epoch": 0.7217484008528785,
+      "grad_norm": 1.0203826309064437,
+      "learning_rate": 7.952163634600055e-05,
+      "loss": 0.4171,
+      "step": 677
+    },
+    {
+      "epoch": 0.7228144989339019,
+      "grad_norm": 0.8482809756462577,
+      "learning_rate": 7.951703493938226e-05,
+      "loss": 0.4194,
+      "step": 678
+    },
+    {
+      "epoch": 0.7238805970149254,
+      "grad_norm": 0.5426953720506598,
+      "learning_rate": 7.951241164238039e-05,
+      "loss": 0.4191,
+      "step": 679
+    },
+    {
+      "epoch": 0.7249466950959488,
+      "grad_norm": 0.7193984844467518,
+      "learning_rate": 7.950776645755596e-05,
+      "loss": 0.4213,
+      "step": 680
+    },
+    {
+      "epoch": 0.7260127931769723,
+      "grad_norm": 0.7142598516288384,
+      "learning_rate": 7.950309938748221e-05,
+      "loss": 0.4205,
+      "step": 681
+    },
+    {
+      "epoch": 0.7270788912579957,
+      "grad_norm": 0.39915429482120396,
+      "learning_rate": 7.949841043474445e-05,
+      "loss": 0.4147,
+      "step": 682
+    },
+    {
+      "epoch": 0.7281449893390192,
+      "grad_norm": 0.5327714348171071,
+      "learning_rate": 7.949369960194009e-05,
+      "loss": 0.4183,
+      "step": 683
+    },
+    {
+      "epoch": 0.7292110874200426,
+      "grad_norm": 0.5612054527272023,
+      "learning_rate": 7.94889668916787e-05,
+      "loss": 0.4185,
+      "step": 684
+    },
+    {
+      "epoch": 0.7302771855010661,
+      "grad_norm": 0.4666415892203282,
+      "learning_rate": 7.948421230658196e-05,
+      "loss": 0.4182,
+      "step": 685
+    },
+    {
+      "epoch": 0.7313432835820896,
+      "grad_norm": 0.520314689388758,
+      "learning_rate": 7.947943584928364e-05,
+      "loss": 0.4237,
+      "step": 686
+    },
+    {
+      "epoch": 0.732409381663113,
+      "grad_norm": 0.41852601792232424,
+      "learning_rate": 7.947463752242968e-05,
+      "loss": 0.4136,
+      "step": 687
+    },
+    {
+      "epoch": 0.7334754797441365,
+      "grad_norm": 0.44187419017345353,
+      "learning_rate": 7.946981732867808e-05,
+      "loss": 0.4188,
+      "step": 688
+    },
+    {
+      "epoch": 0.7345415778251599,
+      "grad_norm": 0.5519060149192045,
+      "learning_rate": 7.946497527069898e-05,
+      "loss": 0.4178,
+      "step": 689
+    },
+    {
+      "epoch": 0.7356076759061834,
+      "grad_norm": 0.5189998274103562,
+      "learning_rate": 7.946011135117466e-05,
+      "loss": 0.4156,
+      "step": 690
+    },
+    {
+      "epoch": 0.7366737739872068,
+      "grad_norm": 0.5845444307770654,
+      "learning_rate": 7.945522557279944e-05,
+      "loss": 0.413,
+      "step": 691
+    },
+    {
+      "epoch": 0.7377398720682303,
+      "grad_norm": 0.7034648713070897,
+      "learning_rate": 7.94503179382798e-05,
+      "loss": 0.4234,
+      "step": 692
+    },
+    {
+      "epoch": 0.7388059701492538,
+      "grad_norm": 0.8180017083563943,
+      "learning_rate": 7.944538845033431e-05,
+      "loss": 0.4181,
+      "step": 693
+    },
+    {
+      "epoch": 0.7398720682302772,
+      "grad_norm": 1.0411030646276598,
+      "learning_rate": 7.944043711169367e-05,
+      "loss": 0.4177,
+      "step": 694
+    },
+    {
+      "epoch": 0.7409381663113006,
+      "grad_norm": 1.1183438708338787,
+      "learning_rate": 7.943546392510065e-05,
+      "loss": 0.4245,
+      "step": 695
+    },
+    {
+      "epoch": 0.7420042643923241,
+      "grad_norm": 0.7568820848776477,
+      "learning_rate": 7.943046889331013e-05,
+      "loss": 0.4223,
+      "step": 696
+    },
+    {
+      "epoch": 0.7430703624733476,
+      "grad_norm": 0.4694996552198355,
+      "learning_rate": 7.94254520190891e-05,
+      "loss": 0.4204,
+      "step": 697
+    },
+    {
+      "epoch": 0.744136460554371,
+      "grad_norm": 0.42506834068311544,
+      "learning_rate": 7.942041330521665e-05,
+      "loss": 0.4133,
+      "step": 698
+    },
+    {
+      "epoch": 0.7452025586353944,
+      "grad_norm": 0.7432323993182501,
+      "learning_rate": 7.941535275448399e-05,
+      "loss": 0.4204,
+      "step": 699
+    },
+    {
+      "epoch": 0.746268656716418,
+      "grad_norm": 1.017099030937784,
+      "learning_rate": 7.941027036969437e-05,
+      "loss": 0.4151,
+      "step": 700
+    },
+    {
+      "epoch": 0.7473347547974414,
+      "grad_norm": 1.042873613244398,
+      "learning_rate": 7.940516615366318e-05,
+      "loss": 0.4225,
+      "step": 701
+    },
+    {
+      "epoch": 0.7484008528784648,
+      "grad_norm": 0.672037071535835,
+      "learning_rate": 7.940004010921787e-05,
+      "loss": 0.4155,
+      "step": 702
+    },
+    {
+      "epoch": 0.7494669509594882,
+      "grad_norm": 0.47977822737782094,
+      "learning_rate": 7.939489223919803e-05,
+      "loss": 0.4239,
+      "step": 703
+    },
+    {
+      "epoch": 0.7505330490405118,
+      "grad_norm": 0.5839565536019953,
+      "learning_rate": 7.938972254645529e-05,
+      "loss": 0.4157,
+      "step": 704
+    },
+    {
+      "epoch": 0.7515991471215352,
+      "grad_norm": 0.659589798795434,
+      "learning_rate": 7.938453103385343e-05,
+      "loss": 0.4297,
+      "step": 705
+    },
+    {
+      "epoch": 0.7526652452025586,
+      "grad_norm": 0.6224572447097061,
+      "learning_rate": 7.937931770426825e-05,
+      "loss": 0.4178,
+      "step": 706
+    },
+    {
+      "epoch": 0.753731343283582,
+      "grad_norm": 0.5223926429775323,
+      "learning_rate": 7.937408256058764e-05,
+      "loss": 0.4195,
+      "step": 707
+    },
+    {
+      "epoch": 0.7547974413646056,
+      "grad_norm": 0.5263236599515704,
+      "learning_rate": 7.936882560571165e-05,
+      "loss": 0.4225,
+      "step": 708
+    },
+    {
+      "epoch": 0.755863539445629,
+      "grad_norm": 0.6701920085872527,
+      "learning_rate": 7.936354684255231e-05,
+      "loss": 0.423,
+      "step": 709
+    },
+    {
+      "epoch": 0.7569296375266524,
+      "grad_norm": 0.8271768873467236,
+      "learning_rate": 7.935824627403382e-05,
+      "loss": 0.4197,
+      "step": 710
+    },
+    {
+      "epoch": 0.7579957356076759,
+      "grad_norm": 0.8903386742919085,
+      "learning_rate": 7.935292390309239e-05,
+      "loss": 0.4161,
+      "step": 711
+    },
+    {
+      "epoch": 0.7590618336886994,
+      "grad_norm": 0.8971908647303243,
+      "learning_rate": 7.934757973267636e-05,
+      "loss": 0.4212,
+      "step": 712
+    },
+    {
+      "epoch": 0.7601279317697228,
+      "grad_norm": 0.8738601854484014,
+      "learning_rate": 7.93422137657461e-05,
+      "loss": 0.4145,
+      "step": 713
+    },
+    {
+      "epoch": 0.7611940298507462,
+      "grad_norm": 0.8238424939346762,
+      "learning_rate": 7.93368260052741e-05,
+      "loss": 0.4231,
+      "step": 714
+    },
+    {
+      "epoch": 0.7622601279317697,
+      "grad_norm": 0.8322879935982136,
+      "learning_rate": 7.933141645424489e-05,
+      "loss": 0.4123,
+      "step": 715
+    },
+    {
+      "epoch": 0.7633262260127932,
+      "grad_norm": 0.7762441733310584,
+      "learning_rate": 7.932598511565506e-05,
+      "loss": 0.4139,
+      "step": 716
+    },
+    {
+      "epoch": 0.7643923240938166,
+      "grad_norm": 0.7714495728551265,
+      "learning_rate": 7.932053199251332e-05,
+      "loss": 0.4172,
+      "step": 717
+    },
+    {
+      "epoch": 0.7654584221748401,
+      "grad_norm": 0.6994068976148212,
+      "learning_rate": 7.931505708784042e-05,
+      "loss": 0.4209,
+      "step": 718
+    },
+    {
+      "epoch": 0.7665245202558635,
+      "grad_norm": 0.4918911877930488,
+      "learning_rate": 7.930956040466912e-05,
+      "loss": 0.4187,
+      "step": 719
+    },
+    {
+      "epoch": 0.767590618336887,
+      "grad_norm": 0.4206999757056118,
+      "learning_rate": 7.930404194604436e-05,
+      "loss": 0.4197,
+      "step": 720
+    },
+    {
+      "epoch": 0.7686567164179104,
+      "grad_norm": 0.491102787820834,
+      "learning_rate": 7.929850171502304e-05,
+      "loss": 0.4223,
+      "step": 721
+    },
+    {
+      "epoch": 0.7697228144989339,
+      "grad_norm": 0.5414023421551015,
+      "learning_rate": 7.92929397146742e-05,
+      "loss": 0.4158,
+      "step": 722
+    },
+    {
+      "epoch": 0.7707889125799574,
+      "grad_norm": 0.43007919824776014,
+      "learning_rate": 7.928735594807885e-05,
+      "loss": 0.4197,
+      "step": 723
+    },
+    {
+      "epoch": 0.7718550106609808,
+      "grad_norm": 0.357619381539661,
+      "learning_rate": 7.928175041833014e-05,
+      "loss": 0.4111,
+      "step": 724
+    },
+    {
+      "epoch": 0.7729211087420043,
+      "grad_norm": 0.44213212514382777,
+      "learning_rate": 7.927612312853321e-05,
+      "loss": 0.4233,
+      "step": 725
+    },
+    {
+      "epoch": 0.7739872068230277,
+      "grad_norm": 0.4580051310888454,
+      "learning_rate": 7.927047408180533e-05,
+      "loss": 0.409,
+      "step": 726
+    },
+    {
+      "epoch": 0.7750533049040512,
+      "grad_norm": 0.5077929282870794,
+      "learning_rate": 7.926480328127573e-05,
+      "loss": 0.4198,
+      "step": 727
+    },
+    {
+      "epoch": 0.7761194029850746,
+      "grad_norm": 0.6011826055428627,
+      "learning_rate": 7.925911073008578e-05,
+      "loss": 0.4217,
+      "step": 728
+    },
+    {
+      "epoch": 0.7771855010660981,
+      "grad_norm": 0.9129504733266989,
+      "learning_rate": 7.925339643138885e-05,
+      "loss": 0.4231,
+      "step": 729
+    },
+    {
+      "epoch": 0.7782515991471215,
+      "grad_norm": 1.2218378663823979,
+      "learning_rate": 7.924766038835035e-05,
+      "loss": 0.4243,
+      "step": 730
+    },
+    {
+      "epoch": 0.779317697228145,
+      "grad_norm": 0.7903329161767807,
+      "learning_rate": 7.924190260414776e-05,
+      "loss": 0.4208,
+      "step": 731
+    },
+    {
+      "epoch": 0.7803837953091685,
+      "grad_norm": 0.6805821512299308,
+      "learning_rate": 7.923612308197058e-05,
+      "loss": 0.42,
+      "step": 732
+    },
+    {
+      "epoch": 0.7814498933901919,
+      "grad_norm": 0.6416299900106649,
+      "learning_rate": 7.923032182502037e-05,
+      "loss": 0.4201,
+      "step": 733
+    },
+    {
+      "epoch": 0.7825159914712153,
+      "grad_norm": 0.6659708417529658,
+      "learning_rate": 7.922449883651074e-05,
+      "loss": 0.4144,
+      "step": 734
+    },
+    {
+      "epoch": 0.7835820895522388,
+      "grad_norm": 0.5293729604666099,
+      "learning_rate": 7.92186541196673e-05,
+      "loss": 0.4219,
+      "step": 735
+    },
+    {
+      "epoch": 0.7846481876332623,
+      "grad_norm": 0.4286984074785958,
+      "learning_rate": 7.921278767772774e-05,
+      "loss": 0.4141,
+      "step": 736
+    },
+    {
+      "epoch": 0.7857142857142857,
+      "grad_norm": 0.49672004967076516,
+      "learning_rate": 7.920689951394175e-05,
+      "loss": 0.4222,
+      "step": 737
+    },
+    {
+      "epoch": 0.7867803837953091,
+      "grad_norm": 0.6284915886717239,
+      "learning_rate": 7.920098963157108e-05,
+      "loss": 0.4252,
+      "step": 738
+    },
+    {
+      "epoch": 0.7878464818763327,
+      "grad_norm": 0.787902960458551,
+      "learning_rate": 7.919505803388949e-05,
+      "loss": 0.4248,
+      "step": 739
+    },
+    {
+      "epoch": 0.7889125799573561,
+      "grad_norm": 0.8686899347398345,
+      "learning_rate": 7.918910472418278e-05,
+      "loss": 0.415,
+      "step": 740
+    },
+    {
+      "epoch": 0.7899786780383795,
+      "grad_norm": 0.8437670042808346,
+      "learning_rate": 7.918312970574875e-05,
+      "loss": 0.4164,
+      "step": 741
+    },
+    {
+      "epoch": 0.7910447761194029,
+      "grad_norm": 0.8540882179405255,
+      "learning_rate": 7.917713298189728e-05,
+      "loss": 0.4253,
+      "step": 742
+    },
+    {
+      "epoch": 0.7921108742004265,
+      "grad_norm": 0.9478085360927803,
+      "learning_rate": 7.917111455595023e-05,
+      "loss": 0.4196,
+      "step": 743
+    },
+    {
+      "epoch": 0.7931769722814499,
+      "grad_norm": 0.9327562081026817,
+      "learning_rate": 7.916507443124153e-05,
+      "loss": 0.4242,
+      "step": 744
+    },
+    {
+      "epoch": 0.7942430703624733,
+      "grad_norm": 0.8551898160448683,
+      "learning_rate": 7.915901261111703e-05,
+      "loss": 0.4178,
+      "step": 745
+    },
+    {
+      "epoch": 0.7953091684434968,
+      "grad_norm": 0.6425189560126273,
+      "learning_rate": 7.91529290989347e-05,
+      "loss": 0.4213,
+      "step": 746
+    },
+    {
+      "epoch": 0.7963752665245203,
+      "grad_norm": 0.5781816257700045,
+      "learning_rate": 7.91468238980645e-05,
+      "loss": 0.4291,
+      "step": 747
+    },
+    {
+      "epoch": 0.7974413646055437,
+      "grad_norm": 0.6921643667155062,
+      "learning_rate": 7.914069701188837e-05,
+      "loss": 0.4197,
+      "step": 748
+    },
+    {
+      "epoch": 0.7985074626865671,
+      "grad_norm": 0.7614603203693651,
+      "learning_rate": 7.913454844380031e-05,
+      "loss": 0.4156,
+      "step": 749
+    },
+    {
+      "epoch": 0.7995735607675906,
+      "grad_norm": 0.7095354236567498,
+      "learning_rate": 7.912837819720628e-05,
+      "loss": 0.4127,
+      "step": 750
+    },
+    {
+      "epoch": 0.8006396588486141,
+      "grad_norm": 0.6873164412119117,
+      "learning_rate": 7.91221862755243e-05,
+      "loss": 0.417,
+      "step": 751
+    },
+    {
+      "epoch": 0.8017057569296375,
+      "grad_norm": 0.718103908521537,
+      "learning_rate": 7.911597268218435e-05,
+      "loss": 0.4234,
+      "step": 752
+    },
+    {
+      "epoch": 0.802771855010661,
+      "grad_norm": 0.6911031732156341,
+      "learning_rate": 7.910973742062847e-05,
+      "loss": 0.4142,
+      "step": 753
+    },
+    {
+      "epoch": 0.8038379530916845,
+      "grad_norm": 0.7537929051659943,
+      "learning_rate": 7.910348049431064e-05,
+      "loss": 0.4161,
+      "step": 754
+    },
+    {
+      "epoch": 0.8049040511727079,
+      "grad_norm": 0.845377153307119,
+      "learning_rate": 7.909720190669689e-05,
+      "loss": 0.4135,
+      "step": 755
+    },
+    {
+      "epoch": 0.8059701492537313,
+      "grad_norm": 0.9386808147663165,
+      "learning_rate": 7.909090166126523e-05,
+      "loss": 0.4175,
+      "step": 756
+    },
+    {
+      "epoch": 0.8070362473347548,
+      "grad_norm": 0.874198077040556,
+      "learning_rate": 7.908457976150565e-05,
+      "loss": 0.4168,
+      "step": 757
+    },
+    {
+      "epoch": 0.8081023454157783,
+      "grad_norm": 0.734655468857772,
+      "learning_rate": 7.907823621092017e-05,
+      "loss": 0.4218,
+      "step": 758
+    },
+    {
+      "epoch": 0.8091684434968017,
+      "grad_norm": 0.664396615528394,
+      "learning_rate": 7.907187101302279e-05,
+      "loss": 0.4095,
+      "step": 759
+    },
+    {
+      "epoch": 0.8102345415778252,
+      "grad_norm": 0.5039424928453715,
+      "learning_rate": 7.90654841713395e-05,
+      "loss": 0.4102,
+      "step": 760
+    },
+    {
+      "epoch": 0.8113006396588486,
+      "grad_norm": 0.38736014140762187,
+      "learning_rate": 7.905907568940825e-05,
+      "loss": 0.4075,
+      "step": 761
+    },
+    {
+      "epoch": 0.8123667377398721,
+      "grad_norm": 0.5398794260824558,
+      "learning_rate": 7.905264557077905e-05,
+      "loss": 0.4185,
+      "step": 762
+    },
+    {
+      "epoch": 0.8134328358208955,
+      "grad_norm": 0.6961113732246285,
+      "learning_rate": 7.904619381901382e-05,
+      "loss": 0.419,
+      "step": 763
+    },
+    {
+      "epoch": 0.814498933901919,
+      "grad_norm": 0.7099747167015686,
+      "learning_rate": 7.903972043768652e-05,
+      "loss": 0.4192,
+      "step": 764
+    },
+    {
+      "epoch": 0.8155650319829424,
+      "grad_norm": 0.7111718390365964,
+      "learning_rate": 7.903322543038302e-05,
+      "loss": 0.4231,
+      "step": 765
+    },
+    {
+      "epoch": 0.8166311300639659,
+      "grad_norm": 0.7243885981248829,
+      "learning_rate": 7.902670880070126e-05,
+      "loss": 0.4213,
+      "step": 766
+    },
+    {
+      "epoch": 0.8176972281449894,
+      "grad_norm": 0.8443210902757722,
+      "learning_rate": 7.902017055225111e-05,
+      "loss": 0.4173,
+      "step": 767
+    },
+    {
+      "epoch": 0.8187633262260128,
+      "grad_norm": 0.9889792776736421,
+      "learning_rate": 7.901361068865441e-05,
+      "loss": 0.4188,
+      "step": 768
+    },
+    {
+      "epoch": 0.8198294243070362,
+      "grad_norm": 1.1016684504340584,
+      "learning_rate": 7.9007029213545e-05,
+      "loss": 0.4206,
+      "step": 769
+    },
+    {
+      "epoch": 0.8208955223880597,
+      "grad_norm": 0.6894548537797883,
+      "learning_rate": 7.900042613056864e-05,
+      "loss": 0.4147,
+      "step": 770
+    },
+    {
+      "epoch": 0.8219616204690832,
+      "grad_norm": 0.4780239669310881,
+      "learning_rate": 7.899380144338313e-05,
+      "loss": 0.4098,
+      "step": 771
+    },
+    {
+      "epoch": 0.8230277185501066,
+      "grad_norm": 0.7367707580027851,
+      "learning_rate": 7.898715515565817e-05,
+      "loss": 0.425,
+      "step": 772
+    },
+    {
+      "epoch": 0.82409381663113,
+      "grad_norm": 0.9588196347102074,
+      "learning_rate": 7.898048727107549e-05,
+      "loss": 0.4201,
+      "step": 773
+    },
+    {
+      "epoch": 0.8251599147121536,
+      "grad_norm": 0.9328164536411601,
+      "learning_rate": 7.897379779332873e-05,
+      "loss": 0.4086,
+      "step": 774
+    },
+    {
+      "epoch": 0.826226012793177,
+      "grad_norm": 0.7663371340787454,
+      "learning_rate": 7.896708672612352e-05,
+      "loss": 0.4157,
+      "step": 775
+    },
+    {
+      "epoch": 0.8272921108742004,
+      "grad_norm": 0.5139058112934316,
+      "learning_rate": 7.896035407317746e-05,
+      "loss": 0.417,
+      "step": 776
+    },
+    {
+      "epoch": 0.8283582089552238,
+      "grad_norm": 0.5600567127054864,
+      "learning_rate": 7.895359983822004e-05,
+      "loss": 0.4164,
+      "step": 777
+    },
+    {
+      "epoch": 0.8294243070362474,
+      "grad_norm": 0.7393036162635956,
+      "learning_rate": 7.894682402499283e-05,
+      "loss": 0.4254,
+      "step": 778
+    },
+    {
+      "epoch": 0.8304904051172708,
+      "grad_norm": 0.7834777517715102,
+      "learning_rate": 7.894002663724921e-05,
+      "loss": 0.4156,
+      "step": 779
+    },
+    {
+      "epoch": 0.8315565031982942,
+      "grad_norm": 0.760746777211738,
+      "learning_rate": 7.89332076787546e-05,
+      "loss": 0.4183,
+      "step": 780
+    },
+    {
+      "epoch": 0.8326226012793176,
+      "grad_norm": 0.7072386884303129,
+      "learning_rate": 7.892636715328638e-05,
+      "loss": 0.4085,
+      "step": 781
+    },
+    {
+      "epoch": 0.8336886993603412,
+      "grad_norm": 0.5891467963351983,
+      "learning_rate": 7.89195050646338e-05,
+      "loss": 0.4174,
+      "step": 782
+    },
+    {
+      "epoch": 0.8347547974413646,
+      "grad_norm": 0.4907232824657599,
+      "learning_rate": 7.891262141659812e-05,
+      "loss": 0.4172,
+      "step": 783
+    },
+    {
+      "epoch": 0.835820895522388,
+      "grad_norm": 0.46554454666441136,
+      "learning_rate": 7.890571621299252e-05,
+      "loss": 0.4135,
+      "step": 784
+    },
+    {
+      "epoch": 0.8368869936034116,
+      "grad_norm": 0.448212012101914,
+      "learning_rate": 7.889878945764215e-05,
+      "loss": 0.4187,
+      "step": 785
+    },
+    {
+      "epoch": 0.837953091684435,
+      "grad_norm": 0.6062486130131177,
+      "learning_rate": 7.889184115438403e-05,
+      "loss": 0.4184,
+      "step": 786
+    },
+    {
+      "epoch": 0.8390191897654584,
+      "grad_norm": 0.6256523573917208,
+      "learning_rate": 7.888487130706719e-05,
+      "loss": 0.4108,
+      "step": 787
+    },
+    {
+      "epoch": 0.8400852878464818,
+      "grad_norm": 0.5794981444661234,
+      "learning_rate": 7.887787991955254e-05,
+      "loss": 0.4156,
+      "step": 788
+    },
+    {
+      "epoch": 0.8411513859275054,
+      "grad_norm": 0.5092450399277311,
+      "learning_rate": 7.887086699571297e-05,
+      "loss": 0.4138,
+      "step": 789
+    },
+    {
+      "epoch": 0.8422174840085288,
+      "grad_norm": 0.3853513561237604,
+      "learning_rate": 7.886383253943326e-05,
+      "loss": 0.4101,
+      "step": 790
+    },
+    {
+      "epoch": 0.8432835820895522,
+      "grad_norm": 0.5174800575237039,
+      "learning_rate": 7.885677655461013e-05,
+      "loss": 0.415,
+      "step": 791
+    },
+    {
+      "epoch": 0.8443496801705757,
+      "grad_norm": 0.7565196574017777,
+      "learning_rate": 7.884969904515224e-05,
+      "loss": 0.4163,
+      "step": 792
+    },
+    {
+      "epoch": 0.8454157782515992,
+      "grad_norm": 0.9209372045007519,
+      "learning_rate": 7.884260001498015e-05,
+      "loss": 0.4153,
+      "step": 793
+    },
+    {
+      "epoch": 0.8464818763326226,
+      "grad_norm": 1.0713514936667083,
+      "learning_rate": 7.883547946802637e-05,
+      "loss": 0.4158,
+      "step": 794
+    },
+    {
+      "epoch": 0.847547974413646,
+      "grad_norm": 0.8424265293921103,
+      "learning_rate": 7.882833740823531e-05,
+      "loss": 0.4085,
+      "step": 795
+    },
+    {
+      "epoch": 0.8486140724946695,
+      "grad_norm": 0.4698049947846306,
+      "learning_rate": 7.882117383956328e-05,
+      "loss": 0.4176,
+      "step": 796
+    },
+    {
+      "epoch": 0.849680170575693,
+      "grad_norm": 0.45769071058539357,
+      "learning_rate": 7.881398876597855e-05,
+      "loss": 0.4122,
+      "step": 797
+    },
+    {
+      "epoch": 0.8507462686567164,
+      "grad_norm": 0.5978831801648813,
+      "learning_rate": 7.880678219146125e-05,
+      "loss": 0.4134,
+      "step": 798
+    },
+    {
+      "epoch": 0.8518123667377399,
+      "grad_norm": 0.6151035364335734,
+      "learning_rate": 7.879955412000348e-05,
+      "loss": 0.413,
+      "step": 799
+    },
+    {
+      "epoch": 0.8528784648187633,
+      "grad_norm": 0.5747882183128944,
+      "learning_rate": 7.87923045556092e-05,
+      "loss": 0.4119,
+      "step": 800
+    },
+    {
+      "epoch": 0.8539445628997868,
+      "grad_norm": 0.5283977724250003,
+      "learning_rate": 7.878503350229428e-05,
+      "loss": 0.4123,
+      "step": 801
+    },
+    {
+      "epoch": 0.8550106609808102,
+      "grad_norm": 0.5661813646706929,
+      "learning_rate": 7.877774096408652e-05,
+      "loss": 0.4173,
+      "step": 802
+    },
+    {
+      "epoch": 0.8560767590618337,
+      "grad_norm": 0.7555232731803762,
+      "learning_rate": 7.87704269450256e-05,
+      "loss": 0.414,
+      "step": 803
+    },
+    {
+      "epoch": 0.8571428571428571,
+      "grad_norm": 0.9330140626616096,
+      "learning_rate": 7.876309144916312e-05,
+      "loss": 0.4174,
+      "step": 804
+    },
+    {
+      "epoch": 0.8582089552238806,
+      "grad_norm": 1.0609474276194846,
+      "learning_rate": 7.875573448056255e-05,
+      "loss": 0.4184,
+      "step": 805
+    },
+    {
+      "epoch": 0.8592750533049041,
+      "grad_norm": 0.9156452459182889,
+      "learning_rate": 7.874835604329928e-05,
+      "loss": 0.4129,
+      "step": 806
+    },
+    {
+      "epoch": 0.8603411513859275,
+      "grad_norm": 0.7492113965064349,
+      "learning_rate": 7.874095614146057e-05,
+      "loss": 0.4134,
+      "step": 807
+    },
+    {
+      "epoch": 0.8614072494669509,
+      "grad_norm": 0.5003348192655799,
+      "learning_rate": 7.873353477914559e-05,
+      "loss": 0.4194,
+      "step": 808
+    },
+    {
+      "epoch": 0.8624733475479744,
+      "grad_norm": 0.5127452250412682,
+      "learning_rate": 7.872609196046537e-05,
+      "loss": 0.4098,
+      "step": 809
+    },
+    {
+      "epoch": 0.8635394456289979,
+      "grad_norm": 0.6332821308118973,
+      "learning_rate": 7.871862768954285e-05,
+      "loss": 0.4158,
+      "step": 810
+    },
+    {
+      "epoch": 0.8646055437100213,
+      "grad_norm": 0.7546855273362159,
+      "learning_rate": 7.871114197051289e-05,
+      "loss": 0.4182,
+      "step": 811
+    },
+    {
+      "epoch": 0.8656716417910447,
+      "grad_norm": 0.7220245458809873,
+      "learning_rate": 7.870363480752214e-05,
+      "loss": 0.4088,
+      "step": 812
+    },
+    {
+      "epoch": 0.8667377398720683,
+      "grad_norm": 0.6636256075414101,
+      "learning_rate": 7.869610620472918e-05,
+      "loss": 0.4169,
+      "step": 813
+    },
+    {
+      "epoch": 0.8678038379530917,
+      "grad_norm": 0.5862374738875737,
+      "learning_rate": 7.86885561663045e-05,
+      "loss": 0.4182,
+      "step": 814
+    },
+    {
+      "epoch": 0.8688699360341151,
+      "grad_norm": 0.47483648705484865,
+      "learning_rate": 7.868098469643039e-05,
+      "loss": 0.4151,
+      "step": 815
+    },
+    {
+      "epoch": 0.8699360341151386,
+      "grad_norm": 0.45874603100581357,
+      "learning_rate": 7.867339179930108e-05,
+      "loss": 0.4139,
+      "step": 816
+    },
+    {
+      "epoch": 0.8710021321961621,
+      "grad_norm": 0.5617556695876103,
+      "learning_rate": 7.866577747912262e-05,
+      "loss": 0.4123,
+      "step": 817
+    },
+    {
+      "epoch": 0.8720682302771855,
+      "grad_norm": 0.6132829210691816,
+      "learning_rate": 7.865814174011295e-05,
+      "loss": 0.4102,
+      "step": 818
+    },
+    {
+      "epoch": 0.8731343283582089,
+      "grad_norm": 0.6010015927279606,
+      "learning_rate": 7.86504845865019e-05,
+      "loss": 0.419,
+      "step": 819
+    },
+    {
+      "epoch": 0.8742004264392325,
+      "grad_norm": 0.7239480812231027,
+      "learning_rate": 7.864280602253109e-05,
+      "loss": 0.4208,
+      "step": 820
+    },
+    {
+      "epoch": 0.8752665245202559,
+      "grad_norm": 0.8304065457671317,
+      "learning_rate": 7.863510605245409e-05,
+      "loss": 0.4241,
+      "step": 821
+    },
+    {
+      "epoch": 0.8763326226012793,
+      "grad_norm": 0.9073014145416783,
+      "learning_rate": 7.862738468053625e-05,
+      "loss": 0.4175,
+      "step": 822
+    },
+    {
+      "epoch": 0.8773987206823027,
+      "grad_norm": 0.8823598817021182,
+      "learning_rate": 7.861964191105483e-05,
+      "loss": 0.4191,
+      "step": 823
+    },
+    {
+      "epoch": 0.8784648187633263,
+      "grad_norm": 0.7779060734831246,
+      "learning_rate": 7.861187774829891e-05,
+      "loss": 0.4101,
+      "step": 824
+    },
+    {
+      "epoch": 0.8795309168443497,
+      "grad_norm": 0.7580593362013702,
+      "learning_rate": 7.860409219656942e-05,
+      "loss": 0.4122,
+      "step": 825
+    },
+    {
+      "epoch": 0.8805970149253731,
+      "grad_norm": 0.877095795887976,
+      "learning_rate": 7.85962852601792e-05,
+      "loss": 0.4186,
+      "step": 826
+    },
+    {
+      "epoch": 0.8816631130063965,
+      "grad_norm": 0.965490750532556,
+      "learning_rate": 7.858845694345283e-05,
+      "loss": 0.4114,
+      "step": 827
+    },
+    {
+      "epoch": 0.8827292110874201,
+      "grad_norm": 0.9339934525386837,
+      "learning_rate": 7.858060725072682e-05,
+      "loss": 0.4119,
+      "step": 828
+    },
+    {
+      "epoch": 0.8837953091684435,
+      "grad_norm": 0.7470003853390679,
+      "learning_rate": 7.857273618634949e-05,
+      "loss": 0.411,
+      "step": 829
+    },
+    {
+      "epoch": 0.8848614072494669,
+      "grad_norm": 0.5531734893024891,
+      "learning_rate": 7.8564843754681e-05,
+      "loss": 0.4061,
+      "step": 830
+    },
+    {
+      "epoch": 0.8859275053304904,
+      "grad_norm": 0.569407252019787,
+      "learning_rate": 7.855692996009332e-05,
+      "loss": 0.409,
+      "step": 831
+    },
+    {
+      "epoch": 0.8869936034115139,
+      "grad_norm": 0.6880810896792248,
+      "learning_rate": 7.854899480697033e-05,
+      "loss": 0.4112,
+      "step": 832
+    },
+    {
+      "epoch": 0.8880597014925373,
+      "grad_norm": 0.7188833271346624,
+      "learning_rate": 7.854103829970765e-05,
+      "loss": 0.4159,
+      "step": 833
+    },
+    {
+      "epoch": 0.8891257995735607,
+      "grad_norm": 0.6200962748194073,
+      "learning_rate": 7.853306044271281e-05,
+      "loss": 0.4104,
+      "step": 834
+    },
+    {
+      "epoch": 0.8901918976545842,
+      "grad_norm": 0.6029255662836998,
+      "learning_rate": 7.852506124040509e-05,
+      "loss": 0.41,
+      "step": 835
+    },
+    {
+      "epoch": 0.8912579957356077,
+      "grad_norm": 0.5455766236674676,
+      "learning_rate": 7.851704069721567e-05,
+      "loss": 0.4155,
+      "step": 836
+    },
+    {
+      "epoch": 0.8923240938166311,
+      "grad_norm": 0.59815834722666,
+      "learning_rate": 7.850899881758746e-05,
+      "loss": 0.4099,
+      "step": 837
+    },
+    {
+      "epoch": 0.8933901918976546,
+      "grad_norm": 0.6269668637980782,
+      "learning_rate": 7.850093560597529e-05,
+      "loss": 0.4086,
+      "step": 838
+    },
+    {
+      "epoch": 0.894456289978678,
+      "grad_norm": 0.5257832120028597,
+      "learning_rate": 7.849285106684576e-05,
+      "loss": 0.41,
+      "step": 839
+    },
+    {
+      "epoch": 0.8955223880597015,
+      "grad_norm": 0.6116996403200065,
+      "learning_rate": 7.848474520467727e-05,
+      "loss": 0.4201,
+      "step": 840
+    },
+    {
+      "epoch": 0.896588486140725,
+      "grad_norm": 0.6371232956894439,
+      "learning_rate": 7.847661802396004e-05,
+      "loss": 0.412,
+      "step": 841
+    },
+    {
+      "epoch": 0.8976545842217484,
+      "grad_norm": 0.5766017052591853,
+      "learning_rate": 7.84684695291961e-05,
+      "loss": 0.4166,
+      "step": 842
+    },
+    {
+      "epoch": 0.8987206823027718,
+      "grad_norm": 0.6967805160059684,
+      "learning_rate": 7.846029972489932e-05,
+      "loss": 0.4134,
+      "step": 843
+    },
+    {
+      "epoch": 0.8997867803837953,
+      "grad_norm": 0.7884289858058776,
+      "learning_rate": 7.845210861559533e-05,
+      "loss": 0.4128,
+      "step": 844
+    },
+    {
+      "epoch": 0.9008528784648188,
+      "grad_norm": 0.7180820620379316,
+      "learning_rate": 7.84438962058216e-05,
+      "loss": 0.4151,
+      "step": 845
+    },
+    {
+      "epoch": 0.9019189765458422,
+      "grad_norm": 0.595379884431672,
+      "learning_rate": 7.843566250012734e-05,
+      "loss": 0.4197,
+      "step": 846
+    },
+    {
+      "epoch": 0.9029850746268657,
+      "grad_norm": 0.44736210153445316,
+      "learning_rate": 7.842740750307362e-05,
+      "loss": 0.4096,
+      "step": 847
+    },
+    {
+      "epoch": 0.9040511727078892,
+      "grad_norm": 0.41043276693031083,
+      "learning_rate": 7.841913121923327e-05,
+      "loss": 0.4179,
+      "step": 848
+    },
+    {
+      "epoch": 0.9051172707889126,
+      "grad_norm": 0.3848671265951802,
+      "learning_rate": 7.841083365319093e-05,
+      "loss": 0.4119,
+      "step": 849
+    },
+    {
+      "epoch": 0.906183368869936,
+      "grad_norm": 0.3978522634845732,
+      "learning_rate": 7.840251480954302e-05,
+      "loss": 0.4109,
+      "step": 850
+    },
+    {
+      "epoch": 0.9072494669509595,
+      "grad_norm": 0.5471539558932556,
+      "learning_rate": 7.839417469289773e-05,
+      "loss": 0.4137,
+      "step": 851
+    },
+    {
+      "epoch": 0.908315565031983,
+      "grad_norm": 0.6676260565465565,
+      "learning_rate": 7.838581330787508e-05,
+      "loss": 0.4132,
+      "step": 852
+    },
+    {
+      "epoch": 0.9093816631130064,
+      "grad_norm": 0.5826596489602703,
+      "learning_rate": 7.837743065910682e-05,
+      "loss": 0.4066,
+      "step": 853
+    },
+    {
+      "epoch": 0.9104477611940298,
+      "grad_norm": 0.5230535811454794,
+      "learning_rate": 7.83690267512365e-05,
+      "loss": 0.405,
+      "step": 854
+    },
+    {
+      "epoch": 0.9115138592750534,
+      "grad_norm": 0.5626160739139322,
+      "learning_rate": 7.836060158891947e-05,
+      "loss": 0.4089,
+      "step": 855
+    },
+    {
+      "epoch": 0.9125799573560768,
+      "grad_norm": 0.4697100322492265,
+      "learning_rate": 7.835215517682282e-05,
+      "loss": 0.4143,
+      "step": 856
+    },
+    {
+      "epoch": 0.9136460554371002,
+      "grad_norm": 0.36866479086378523,
+      "learning_rate": 7.834368751962542e-05,
+      "loss": 0.405,
+      "step": 857
+    },
+    {
+      "epoch": 0.9147121535181236,
+      "grad_norm": 0.4787182490306844,
+      "learning_rate": 7.833519862201791e-05,
+      "loss": 0.4144,
+      "step": 858
+    },
+    {
+      "epoch": 0.9157782515991472,
+      "grad_norm": 0.5878869019859624,
+      "learning_rate": 7.83266884887027e-05,
+      "loss": 0.4183,
+      "step": 859
+    },
+    {
+      "epoch": 0.9168443496801706,
+      "grad_norm": 0.7152505035199795,
+      "learning_rate": 7.831815712439397e-05,
+      "loss": 0.4095,
+      "step": 860
+    },
+    {
+      "epoch": 0.917910447761194,
+      "grad_norm": 0.7805590910529553,
+      "learning_rate": 7.830960453381764e-05,
+      "loss": 0.4106,
+      "step": 861
+    },
+    {
+      "epoch": 0.9189765458422174,
+      "grad_norm": 0.7572280791863363,
+      "learning_rate": 7.830103072171142e-05,
+      "loss": 0.4102,
+      "step": 862
+    },
+    {
+      "epoch": 0.920042643923241,
+      "grad_norm": 0.7493776976712094,
+      "learning_rate": 7.829243569282473e-05,
+      "loss": 0.4088,
+      "step": 863
+    },
+    {
+      "epoch": 0.9211087420042644,
+      "grad_norm": 0.7609395686522328,
+      "learning_rate": 7.828381945191879e-05,
+      "loss": 0.4111,
+      "step": 864
+    },
+    {
+      "epoch": 0.9221748400852878,
+      "grad_norm": 0.8239887481859134,
+      "learning_rate": 7.827518200376654e-05,
+      "loss": 0.4195,
+      "step": 865
+    },
+    {
+      "epoch": 0.9232409381663113,
+      "grad_norm": 0.8016884574752428,
+      "learning_rate": 7.826652335315268e-05,
+      "loss": 0.4056,
+      "step": 866
+    },
+    {
+      "epoch": 0.9243070362473348,
+      "grad_norm": 0.7736961735336623,
+      "learning_rate": 7.825784350487365e-05,
+      "loss": 0.4184,
+      "step": 867
+    },
+    {
+      "epoch": 0.9253731343283582,
+      "grad_norm": 0.8458442232897897,
+      "learning_rate": 7.824914246373764e-05,
+      "loss": 0.4159,
+      "step": 868
+    },
+    {
+      "epoch": 0.9264392324093816,
+      "grad_norm": 0.9770784658028318,
+      "learning_rate": 7.824042023456458e-05,
+      "loss": 0.414,
+      "step": 869
+    },
+    {
+      "epoch": 0.9275053304904051,
+      "grad_norm": 0.9265743591320977,
+      "learning_rate": 7.823167682218611e-05,
+      "loss": 0.4145,
+      "step": 870
+    },
+    {
+      "epoch": 0.9285714285714286,
+      "grad_norm": 0.7609304260280078,
+      "learning_rate": 7.822291223144564e-05,
+      "loss": 0.4133,
+      "step": 871
+    },
+    {
+      "epoch": 0.929637526652452,
+      "grad_norm": 0.6283447272354892,
+      "learning_rate": 7.821412646719829e-05,
+      "loss": 0.412,
+      "step": 872
+    },
+    {
+      "epoch": 0.9307036247334755,
+      "grad_norm": 0.6090160541584229,
+      "learning_rate": 7.820531953431093e-05,
+      "loss": 0.4083,
+      "step": 873
+    },
+    {
+      "epoch": 0.9317697228144989,
+      "grad_norm": 0.7086782043281862,
+      "learning_rate": 7.819649143766215e-05,
+      "loss": 0.4121,
+      "step": 874
+    },
+    {
+      "epoch": 0.9328358208955224,
+      "grad_norm": 0.6602890170546788,
+      "learning_rate": 7.818764218214224e-05,
+      "loss": 0.403,
+      "step": 875
+    },
+    {
+      "epoch": 0.9339019189765458,
+      "grad_norm": 0.5431433609726145,
+      "learning_rate": 7.817877177265323e-05,
+      "loss": 0.4145,
+      "step": 876
+    },
+    {
+      "epoch": 0.9349680170575693,
+      "grad_norm": 0.5717886780149007,
+      "learning_rate": 7.816988021410885e-05,
+      "loss": 0.4162,
+      "step": 877
+    },
+    {
+      "epoch": 0.9360341151385928,
+      "grad_norm": 0.518417714023232,
+      "learning_rate": 7.816096751143459e-05,
+      "loss": 0.4058,
+      "step": 878
+    },
+    {
+      "epoch": 0.9371002132196162,
+      "grad_norm": 0.43803136693398226,
+      "learning_rate": 7.815203366956762e-05,
+      "loss": 0.4071,
+      "step": 879
+    },
+    {
+      "epoch": 0.9381663113006397,
+      "grad_norm": 0.5244904028290258,
+      "learning_rate": 7.814307869345682e-05,
+      "loss": 0.4214,
+      "step": 880
+    },
+    {
+      "epoch": 0.9392324093816631,
+      "grad_norm": 0.637533228472145,
+      "learning_rate": 7.813410258806275e-05,
+      "loss": 0.4178,
+      "step": 881
+    },
+    {
+      "epoch": 0.9402985074626866,
+      "grad_norm": 0.7047581414477792,
+      "learning_rate": 7.812510535835775e-05,
+      "loss": 0.4073,
+      "step": 882
+    },
+    {
+      "epoch": 0.94136460554371,
+      "grad_norm": 0.641172550785535,
+      "learning_rate": 7.811608700932582e-05,
+      "loss": 0.4103,
+      "step": 883
+    },
+    {
+      "epoch": 0.9424307036247335,
+      "grad_norm": 0.5994428458639327,
+      "learning_rate": 7.810704754596263e-05,
+      "loss": 0.4135,
+      "step": 884
+    },
+    {
+      "epoch": 0.9434968017057569,
+      "grad_norm": 0.6070504903362325,
+      "learning_rate": 7.809798697327558e-05,
+      "loss": 0.4168,
+      "step": 885
+    },
+    {
+      "epoch": 0.9445628997867804,
+      "grad_norm": 0.6389004815644979,
+      "learning_rate": 7.808890529628374e-05,
+      "loss": 0.4126,
+      "step": 886
+    },
+    {
+      "epoch": 0.9456289978678039,
+      "grad_norm": 0.5841071688845674,
+      "learning_rate": 7.807980252001791e-05,
+      "loss": 0.4163,
+      "step": 887
+    },
+    {
+      "epoch": 0.9466950959488273,
+      "grad_norm": 0.5441565394837324,
+      "learning_rate": 7.807067864952055e-05,
+      "loss": 0.4073,
+      "step": 888
+    },
+    {
+      "epoch": 0.9477611940298507,
+      "grad_norm": 0.5817224526693675,
+      "learning_rate": 7.806153368984583e-05,
+      "loss": 0.4095,
+      "step": 889
+    },
+    {
+      "epoch": 0.9488272921108742,
+      "grad_norm": 0.6150981739518016,
+      "learning_rate": 7.805236764605954e-05,
+      "loss": 0.4111,
+      "step": 890
+    },
+    {
+      "epoch": 0.9498933901918977,
+      "grad_norm": 0.6063228912727467,
+      "learning_rate": 7.804318052323922e-05,
+      "loss": 0.4118,
+      "step": 891
+    },
+    {
+      "epoch": 0.9509594882729211,
+      "grad_norm": 0.5902912758191065,
+      "learning_rate": 7.803397232647406e-05,
+      "loss": 0.4145,
+      "step": 892
+    },
+    {
+      "epoch": 0.9520255863539445,
+      "grad_norm": 0.5500221760128661,
+      "learning_rate": 7.80247430608649e-05,
+      "loss": 0.4173,
+      "step": 893
+    },
+    {
+      "epoch": 0.9530916844349681,
+      "grad_norm": 0.5579754203833759,
+      "learning_rate": 7.80154927315243e-05,
+      "loss": 0.4088,
+      "step": 894
+    },
+    {
+      "epoch": 0.9541577825159915,
+      "grad_norm": 0.6650648355002524,
+      "learning_rate": 7.800622134357644e-05,
+      "loss": 0.405,
+      "step": 895
+    },
+    {
+      "epoch": 0.9552238805970149,
+      "grad_norm": 0.7556125255322795,
+      "learning_rate": 7.799692890215721e-05,
+      "loss": 0.4125,
+      "step": 896
+    },
+    {
+      "epoch": 0.9562899786780383,
+      "grad_norm": 0.7165371046432817,
+      "learning_rate": 7.798761541241413e-05,
+      "loss": 0.4086,
+      "step": 897
+    },
+    {
+      "epoch": 0.9573560767590619,
+      "grad_norm": 0.6697196433407722,
+      "learning_rate": 7.797828087950637e-05,
+      "loss": 0.4125,
+      "step": 898
+    },
+    {
+      "epoch": 0.9584221748400853,
+      "grad_norm": 0.6629917240602252,
+      "learning_rate": 7.79689253086048e-05,
+      "loss": 0.4172,
+      "step": 899
+    },
+    {
+      "epoch": 0.9594882729211087,
+      "grad_norm": 0.6424942713817352,
+      "learning_rate": 7.795954870489191e-05,
+      "loss": 0.4131,
+      "step": 900
+    },
+    {
+      "epoch": 0.9605543710021321,
+      "grad_norm": 0.4594700765306279,
+      "learning_rate": 7.795015107356186e-05,
+      "loss": 0.4122,
+      "step": 901
+    },
+    {
+      "epoch": 0.9616204690831557,
+      "grad_norm": 0.40226566645245615,
+      "learning_rate": 7.794073241982043e-05,
+      "loss": 0.4141,
+      "step": 902
+    },
+    {
+      "epoch": 0.9626865671641791,
+      "grad_norm": 0.5260998304778581,
+      "learning_rate": 7.793129274888508e-05,
+      "loss": 0.4119,
+      "step": 903
+    },
+    {
+      "epoch": 0.9637526652452025,
+      "grad_norm": 0.628398237724531,
+      "learning_rate": 7.792183206598491e-05,
+      "loss": 0.4088,
+      "step": 904
+    },
+    {
+      "epoch": 0.964818763326226,
+      "grad_norm": 0.6607508208979607,
+      "learning_rate": 7.791235037636062e-05,
+      "loss": 0.4126,
+      "step": 905
+    },
+    {
+      "epoch": 0.9658848614072495,
+      "grad_norm": 0.7250208135809759,
+      "learning_rate": 7.79028476852646e-05,
+      "loss": 0.4159,
+      "step": 906
+    },
+    {
+      "epoch": 0.9669509594882729,
+      "grad_norm": 0.7410413741786345,
+      "learning_rate": 7.789332399796079e-05,
+      "loss": 0.4099,
+      "step": 907
+    },
+    {
+      "epoch": 0.9680170575692963,
+      "grad_norm": 0.6585611107689859,
+      "learning_rate": 7.78837793197249e-05,
+      "loss": 0.4099,
+      "step": 908
+    },
+    {
+      "epoch": 0.9690831556503199,
+      "grad_norm": 0.6090162859858071,
+      "learning_rate": 7.787421365584414e-05,
+      "loss": 0.4113,
+      "step": 909
+    },
+    {
+      "epoch": 0.9701492537313433,
+      "grad_norm": 0.6733319020369772,
+      "learning_rate": 7.786462701161738e-05,
+      "loss": 0.4093,
+      "step": 910
+    },
+    {
+      "epoch": 0.9712153518123667,
+      "grad_norm": 0.7485470209020791,
+      "learning_rate": 7.785501939235513e-05,
+      "loss": 0.4141,
+      "step": 911
+    },
+    {
+      "epoch": 0.9722814498933902,
+      "grad_norm": 0.6601939210618039,
+      "learning_rate": 7.784539080337955e-05,
+      "loss": 0.4079,
+      "step": 912
+    },
+    {
+      "epoch": 0.9733475479744137,
+      "grad_norm": 0.5436166438340034,
+      "learning_rate": 7.783574125002432e-05,
+      "loss": 0.4041,
+      "step": 913
+    },
+    {
+      "epoch": 0.9744136460554371,
+      "grad_norm": 0.617167761753248,
+      "learning_rate": 7.782607073763484e-05,
+      "loss": 0.4086,
+      "step": 914
+    },
+    {
+      "epoch": 0.9754797441364605,
+      "grad_norm": 0.6443414561615247,
+      "learning_rate": 7.781637927156804e-05,
+      "loss": 0.4104,
+      "step": 915
+    },
+    {
+      "epoch": 0.976545842217484,
+      "grad_norm": 0.5741265835867478,
+      "learning_rate": 7.780666685719249e-05,
+      "loss": 0.4148,
+      "step": 916
+    },
+    {
+      "epoch": 0.9776119402985075,
+      "grad_norm": 0.42635079769878254,
+      "learning_rate": 7.779693349988839e-05,
+      "loss": 0.4048,
+      "step": 917
+    },
+    {
+      "epoch": 0.9786780383795309,
+      "grad_norm": 0.47310627157488677,
+      "learning_rate": 7.77871792050475e-05,
+      "loss": 0.409,
+      "step": 918
+    },
+    {
+      "epoch": 0.9797441364605544,
+      "grad_norm": 0.5694527852620955,
+      "learning_rate": 7.777740397807316e-05,
+      "loss": 0.4093,
+      "step": 919
+    },
+    {
+      "epoch": 0.9808102345415778,
+      "grad_norm": 0.5924875514072048,
+      "learning_rate": 7.776760782438038e-05,
+      "loss": 0.4111,
+      "step": 920
+    },
+    {
+      "epoch": 0.9818763326226013,
+      "grad_norm": 0.49411522824636867,
+      "learning_rate": 7.775779074939571e-05,
+      "loss": 0.4154,
+      "step": 921
+    },
+    {
+      "epoch": 0.9829424307036247,
+      "grad_norm": 0.5511529858288959,
+      "learning_rate": 7.77479527585573e-05,
+      "loss": 0.4132,
+      "step": 922
+    },
+    {
+      "epoch": 0.9840085287846482,
+      "grad_norm": 0.6881028326855904,
+      "learning_rate": 7.773809385731487e-05,
+      "loss": 0.4013,
+      "step": 923
+    },
+    {
+      "epoch": 0.9850746268656716,
+      "grad_norm": 0.8172075890856048,
+      "learning_rate": 7.772821405112974e-05,
+      "loss": 0.417,
+      "step": 924
+    },
+    {
+      "epoch": 0.9861407249466951,
+      "grad_norm": 0.9223848954361227,
+      "learning_rate": 7.771831334547483e-05,
+      "loss": 0.4071,
+      "step": 925
+    },
+    {
+      "epoch": 0.9872068230277186,
+      "grad_norm": 0.9603810119097379,
+      "learning_rate": 7.77083917458346e-05,
+      "loss": 0.4114,
+      "step": 926
+    },
+    {
+      "epoch": 0.988272921108742,
+      "grad_norm": 1.0088704018315873,
+      "learning_rate": 7.769844925770512e-05,
+      "loss": 0.4168,
+      "step": 927
+    },
+    {
+      "epoch": 0.9893390191897654,
+      "grad_norm": 0.9263247355978909,
+      "learning_rate": 7.7688485886594e-05,
+      "loss": 0.4068,
+      "step": 928
+    },
+    {
+      "epoch": 0.990405117270789,
+      "grad_norm": 0.6336714165848655,
+      "learning_rate": 7.76785016380204e-05,
+      "loss": 0.4039,
+      "step": 929
+    },
+    {
+      "epoch": 0.9914712153518124,
+      "grad_norm": 0.3609254347454348,
+      "learning_rate": 7.766849651751512e-05,
+      "loss": 0.4094,
+      "step": 930
+    },
+    {
+      "epoch": 0.9925373134328358,
+      "grad_norm": 0.37973843808767777,
+      "learning_rate": 7.765847053062046e-05,
+      "loss": 0.4124,
+      "step": 931
+    },
+    {
+      "epoch": 0.9936034115138592,
+      "grad_norm": 0.6199111977710898,
+      "learning_rate": 7.764842368289028e-05,
+      "loss": 0.4065,
+      "step": 932
+    },
+    {
+      "epoch": 0.9946695095948828,
+      "grad_norm": 0.7388812044022451,
+      "learning_rate": 7.763835597989005e-05,
+      "loss": 0.4126,
+      "step": 933
+    },
+    {
+      "epoch": 0.9957356076759062,
+      "grad_norm": 0.6631941161213446,
+      "learning_rate": 7.762826742719672e-05,
+      "loss": 0.4077,
+      "step": 934
+    },
+    {
+      "epoch": 0.9968017057569296,
+      "grad_norm": 0.5013692544152526,
+      "learning_rate": 7.761815803039883e-05,
+      "loss": 0.4088,
+      "step": 935
+    },
+    {
+      "epoch": 0.997867803837953,
+      "grad_norm": 0.5138742991856071,
+      "learning_rate": 7.760802779509647e-05,
+      "loss": 0.4109,
+      "step": 936
+    },
+    {
+      "epoch": 0.9989339019189766,
+      "grad_norm": 0.5647725207929322,
+      "learning_rate": 7.759787672690124e-05,
+      "loss": 0.4136,
+      "step": 937
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.49646274447841793,
+      "learning_rate": 7.758770483143634e-05,
+      "loss": 0.4046,
+      "step": 938
+    },
+    {
+      "epoch": 1.0010660980810235,
+      "grad_norm": 0.41415428574276214,
+      "learning_rate": 7.757751211433646e-05,
+      "loss": 0.3978,
+      "step": 939
+    },
+    {
+      "epoch": 1.0021321961620469,
+      "grad_norm": 0.4418990401143746,
+      "learning_rate": 7.75672985812478e-05,
+      "loss": 0.4007,
+      "step": 940
+    },
+    {
+      "epoch": 1.0031982942430704,
+      "grad_norm": 0.5702542511470434,
+      "learning_rate": 7.75570642378282e-05,
+      "loss": 0.4008,
+      "step": 941
+    },
+    {
+      "epoch": 1.004264392324094,
+      "grad_norm": 0.6510908049787664,
+      "learning_rate": 7.754680908974687e-05,
+      "loss": 0.4006,
+      "step": 942
+    },
+    {
+      "epoch": 1.0053304904051172,
+      "grad_norm": 0.7117289661008723,
+      "learning_rate": 7.75365331426847e-05,
+      "loss": 0.4029,
+      "step": 943
+    },
+    {
+      "epoch": 1.0063965884861408,
+      "grad_norm": 0.8003621078354415,
+      "learning_rate": 7.752623640233398e-05,
+      "loss": 0.4028,
+      "step": 944
+    },
+    {
+      "epoch": 1.007462686567164,
+      "grad_norm": 0.7457196896824791,
+      "learning_rate": 7.751591887439859e-05,
+      "loss": 0.4077,
+      "step": 945
+    },
+    {
+      "epoch": 1.0085287846481876,
+      "grad_norm": 0.650649322119117,
+      "learning_rate": 7.75055805645939e-05,
+      "loss": 0.4018,
+      "step": 946
+    },
+    {
+      "epoch": 1.0095948827292112,
+      "grad_norm": 0.7354343157171254,
+      "learning_rate": 7.749522147864681e-05,
+      "loss": 0.397,
+      "step": 947
+    },
+    {
+      "epoch": 1.0106609808102345,
+      "grad_norm": 0.8392238195122467,
+      "learning_rate": 7.748484162229572e-05,
+      "loss": 0.3956,
+      "step": 948
+    },
+    {
+      "epoch": 1.011727078891258,
+      "grad_norm": 0.6811125765381959,
+      "learning_rate": 7.747444100129048e-05,
+      "loss": 0.4008,
+      "step": 949
+    },
+    {
+      "epoch": 1.0127931769722816,
+      "grad_norm": 0.5072610455516853,
+      "learning_rate": 7.746401962139255e-05,
+      "loss": 0.3975,
+      "step": 950
+    },
+    {
+      "epoch": 1.0138592750533049,
+      "grad_norm": 0.475527294382562,
+      "learning_rate": 7.745357748837482e-05,
+      "loss": 0.3977,
+      "step": 951
+    },
+    {
+      "epoch": 1.0149253731343284,
+      "grad_norm": 0.6012477231886632,
+      "learning_rate": 7.744311460802166e-05,
+      "loss": 0.4016,
+      "step": 952
+    },
+    {
+      "epoch": 1.0159914712153517,
+      "grad_norm": 0.6774620182327961,
+      "learning_rate": 7.7432630986129e-05,
+      "loss": 0.395,
+      "step": 953
+    },
+    {
+      "epoch": 1.0170575692963753,
+      "grad_norm": 0.6361012701129908,
+      "learning_rate": 7.742212662850421e-05,
+      "loss": 0.3961,
+      "step": 954
+    },
+    {
+      "epoch": 1.0181236673773988,
+      "grad_norm": 0.4593566880462273,
+      "learning_rate": 7.741160154096614e-05,
+      "loss": 0.4015,
+      "step": 955
+    },
+    {
+      "epoch": 1.019189765458422,
+      "grad_norm": 0.4253384790692246,
+      "learning_rate": 7.740105572934516e-05,
+      "loss": 0.3969,
+      "step": 956
+    },
+    {
+      "epoch": 1.0202558635394456,
+      "grad_norm": 0.5377056528363119,
+      "learning_rate": 7.739048919948309e-05,
+      "loss": 0.3925,
+      "step": 957
+    },
+    {
+      "epoch": 1.0213219616204692,
+      "grad_norm": 0.4792515921280752,
+      "learning_rate": 7.737990195723325e-05,
+      "loss": 0.3954,
+      "step": 958
+    },
+    {
+      "epoch": 1.0223880597014925,
+      "grad_norm": 0.37572937988220495,
+      "learning_rate": 7.736929400846041e-05,
+      "loss": 0.3946,
+      "step": 959
+    },
+    {
+      "epoch": 1.023454157782516,
+      "grad_norm": 0.4748049142013816,
+      "learning_rate": 7.735866535904083e-05,
+      "loss": 0.3938,
+      "step": 960
+    },
+    {
+      "epoch": 1.0245202558635393,
+      "grad_norm": 0.5674490975829488,
+      "learning_rate": 7.734801601486224e-05,
+      "loss": 0.4053,
+      "step": 961
+    },
+    {
+      "epoch": 1.0255863539445629,
+      "grad_norm": 0.49630911394420096,
+      "learning_rate": 7.733734598182379e-05,
+      "loss": 0.391,
+      "step": 962
+    },
+    {
+      "epoch": 1.0266524520255864,
+      "grad_norm": 0.41816138182017965,
+      "learning_rate": 7.732665526583616e-05,
+      "loss": 0.3974,
+      "step": 963
+    },
+    {
+      "epoch": 1.0277185501066097,
+      "grad_norm": 0.48628841101926257,
+      "learning_rate": 7.731594387282144e-05,
+      "loss": 0.3897,
+      "step": 964
+    },
+    {
+      "epoch": 1.0287846481876333,
+      "grad_norm": 0.5327676523101473,
+      "learning_rate": 7.730521180871317e-05,
+      "loss": 0.3958,
+      "step": 965
+    },
+    {
+      "epoch": 1.0298507462686568,
+      "grad_norm": 0.5698192622774071,
+      "learning_rate": 7.729445907945637e-05,
+      "loss": 0.3996,
+      "step": 966
+    },
+    {
+      "epoch": 1.0309168443496801,
+      "grad_norm": 0.6600227288417884,
+      "learning_rate": 7.728368569100749e-05,
+      "loss": 0.398,
+      "step": 967
+    },
+    {
+      "epoch": 1.0319829424307037,
+      "grad_norm": 0.7199584858108914,
+      "learning_rate": 7.727289164933443e-05,
+      "loss": 0.3952,
+      "step": 968
+    },
+    {
+      "epoch": 1.033049040511727,
+      "grad_norm": 0.6780630536094696,
+      "learning_rate": 7.726207696041653e-05,
+      "loss": 0.3901,
+      "step": 969
+    },
+    {
+      "epoch": 1.0341151385927505,
+      "grad_norm": 0.6051350564389102,
+      "learning_rate": 7.725124163024456e-05,
+      "loss": 0.4027,
+      "step": 970
+    },
+    {
+      "epoch": 1.035181236673774,
+      "grad_norm": 0.6703569464507834,
+      "learning_rate": 7.724038566482073e-05,
+      "loss": 0.3982,
+      "step": 971
+    },
+    {
+      "epoch": 1.0362473347547974,
+      "grad_norm": 0.701997393773405,
+      "learning_rate": 7.722950907015867e-05,
+      "loss": 0.4064,
+      "step": 972
+    },
+    {
+      "epoch": 1.037313432835821,
+      "grad_norm": 0.7033610282992381,
+      "learning_rate": 7.721861185228347e-05,
+      "loss": 0.4024,
+      "step": 973
+    },
+    {
+      "epoch": 1.0383795309168444,
+      "grad_norm": 0.7094358785590401,
+      "learning_rate": 7.72076940172316e-05,
+      "loss": 0.4075,
+      "step": 974
+    },
+    {
+      "epoch": 1.0394456289978677,
+      "grad_norm": 0.5778245595035633,
+      "learning_rate": 7.719675557105101e-05,
+      "loss": 0.399,
+      "step": 975
+    },
+    {
+      "epoch": 1.0405117270788913,
+      "grad_norm": 0.495341844608196,
+      "learning_rate": 7.718579651980099e-05,
+      "loss": 0.3987,
+      "step": 976
+    },
+    {
+      "epoch": 1.0415778251599148,
+      "grad_norm": 0.5705131082078807,
+      "learning_rate": 7.717481686955231e-05,
+      "loss": 0.4012,
+      "step": 977
+    },
+    {
+      "epoch": 1.0426439232409381,
+      "grad_norm": 0.7460655829453172,
+      "learning_rate": 7.71638166263871e-05,
+      "loss": 0.3996,
+      "step": 978
+    },
+    {
+      "epoch": 1.0437100213219617,
+      "grad_norm": 0.8323692131007797,
+      "learning_rate": 7.715279579639895e-05,
+      "loss": 0.3893,
+      "step": 979
+    },
+    {
+      "epoch": 1.044776119402985,
+      "grad_norm": 0.7432520305478516,
+      "learning_rate": 7.714175438569282e-05,
+      "loss": 0.398,
+      "step": 980
+    },
+    {
+      "epoch": 1.0458422174840085,
+      "grad_norm": 0.6728948975330855,
+      "learning_rate": 7.713069240038506e-05,
+      "loss": 0.4017,
+      "step": 981
+    },
+    {
+      "epoch": 1.046908315565032,
+      "grad_norm": 0.4983139656283151,
+      "learning_rate": 7.711960984660346e-05,
+      "loss": 0.394,
+      "step": 982
+    },
+    {
+      "epoch": 1.0479744136460554,
+      "grad_norm": 0.33754531605342536,
+      "learning_rate": 7.710850673048717e-05,
+      "loss": 0.3944,
+      "step": 983
+    },
+    {
+      "epoch": 1.049040511727079,
+      "grad_norm": 0.3725573631734465,
+      "learning_rate": 7.709738305818674e-05,
+      "loss": 0.4005,
+      "step": 984
+    },
+    {
+      "epoch": 1.0501066098081024,
+      "grad_norm": 0.4551248038249476,
+      "learning_rate": 7.708623883586409e-05,
+      "loss": 0.3971,
+      "step": 985
+    },
+    {
+      "epoch": 1.0511727078891258,
+      "grad_norm": 0.41338624681327313,
+      "learning_rate": 7.707507406969256e-05,
+      "loss": 0.4075,
+      "step": 986
+    },
+    {
+      "epoch": 1.0522388059701493,
+      "grad_norm": 0.3693936866798565,
+      "learning_rate": 7.706388876585685e-05,
+      "loss": 0.4035,
+      "step": 987
+    },
+    {
+      "epoch": 1.0533049040511726,
+      "grad_norm": 0.3457954659560392,
+      "learning_rate": 7.705268293055302e-05,
+      "loss": 0.3988,
+      "step": 988
+    },
+    {
+      "epoch": 1.0543710021321961,
+      "grad_norm": 0.3988913386622119,
+      "learning_rate": 7.704145656998853e-05,
+      "loss": 0.3896,
+      "step": 989
+    },
+    {
+      "epoch": 1.0554371002132197,
+      "grad_norm": 0.48450611938317545,
+      "learning_rate": 7.703020969038222e-05,
+      "loss": 0.392,
+      "step": 990
+    },
+    {
+      "epoch": 1.056503198294243,
+      "grad_norm": 0.5311541916427419,
+      "learning_rate": 7.701894229796424e-05,
+      "loss": 0.3952,
+      "step": 991
+    },
+    {
+      "epoch": 1.0575692963752665,
+      "grad_norm": 0.5299103995739938,
+      "learning_rate": 7.700765439897616e-05,
+      "loss": 0.3959,
+      "step": 992
+    },
+    {
+      "epoch": 1.05863539445629,
+      "grad_norm": 0.5180100310098522,
+      "learning_rate": 7.69963459996709e-05,
+      "loss": 0.3953,
+      "step": 993
+    },
+    {
+      "epoch": 1.0597014925373134,
+      "grad_norm": 0.6335245544459096,
+      "learning_rate": 7.69850171063127e-05,
+      "loss": 0.3963,
+      "step": 994
+    },
+    {
+      "epoch": 1.060767590618337,
+      "grad_norm": 0.7565263057674985,
+      "learning_rate": 7.697366772517719e-05,
+      "loss": 0.4011,
+      "step": 995
+    },
+    {
+      "epoch": 1.0618336886993602,
+      "grad_norm": 0.7626681326172062,
+      "learning_rate": 7.696229786255136e-05,
+      "loss": 0.3973,
+      "step": 996
+    },
+    {
+      "epoch": 1.0628997867803838,
+      "grad_norm": 0.7180860950610816,
+      "learning_rate": 7.695090752473348e-05,
+      "loss": 0.4045,
+      "step": 997
+    },
+    {
+      "epoch": 1.0639658848614073,
+      "grad_norm": 0.6227425406154337,
+      "learning_rate": 7.693949671803323e-05,
+      "loss": 0.3976,
+      "step": 998
+    },
+    {
+      "epoch": 1.0650319829424306,
+      "grad_norm": 0.7201441487555086,
+      "learning_rate": 7.69280654487716e-05,
+      "loss": 0.3985,
+      "step": 999
+    },
+    {
+      "epoch": 1.0660980810234542,
+      "grad_norm": 0.8997609053026269,
+      "learning_rate": 7.691661372328093e-05,
+      "loss": 0.3986,
+      "step": 1000
+    },
+    {
+      "epoch": 1.0671641791044777,
+      "grad_norm": 1.000428399051643,
+      "learning_rate": 7.690514154790485e-05,
+      "loss": 0.3952,
+      "step": 1001
+    },
+    {
+      "epoch": 1.068230277185501,
+      "grad_norm": 0.8713614689438668,
+      "learning_rate": 7.689364892899838e-05,
+      "loss": 0.397,
+      "step": 1002
+    },
+    {
+      "epoch": 1.0692963752665245,
+      "grad_norm": 0.7060560170201216,
+      "learning_rate": 7.688213587292783e-05,
+      "loss": 0.398,
+      "step": 1003
+    },
+    {
+      "epoch": 1.070362473347548,
+      "grad_norm": 0.5741051379265467,
+      "learning_rate": 7.687060238607082e-05,
+      "loss": 0.3979,
+      "step": 1004
+    },
+    {
+      "epoch": 1.0714285714285714,
+      "grad_norm": 0.46041215316462103,
+      "learning_rate": 7.685904847481631e-05,
+      "loss": 0.3959,
+      "step": 1005
+    },
+    {
+      "epoch": 1.072494669509595,
+      "grad_norm": 0.3666940747393567,
+      "learning_rate": 7.684747414556457e-05,
+      "loss": 0.3933,
+      "step": 1006
+    },
+    {
+      "epoch": 1.0735607675906182,
+      "grad_norm": 0.3941609183280189,
+      "learning_rate": 7.683587940472716e-05,
+      "loss": 0.4004,
+      "step": 1007
+    },
+    {
+      "epoch": 1.0746268656716418,
+      "grad_norm": 0.46186462913236076,
+      "learning_rate": 7.6824264258727e-05,
+      "loss": 0.4004,
+      "step": 1008
+    },
+    {
+      "epoch": 1.0756929637526653,
+      "grad_norm": 0.5333763291785089,
+      "learning_rate": 7.681262871399824e-05,
+      "loss": 0.3989,
+      "step": 1009
+    },
+    {
+      "epoch": 1.0767590618336886,
+      "grad_norm": 0.4861997016882574,
+      "learning_rate": 7.680097277698637e-05,
+      "loss": 0.3974,
+      "step": 1010
+    },
+    {
+      "epoch": 1.0778251599147122,
+      "grad_norm": 0.420953916020597,
+      "learning_rate": 7.678929645414822e-05,
+      "loss": 0.4036,
+      "step": 1011
+    },
+    {
+      "epoch": 1.0788912579957357,
+      "grad_norm": 0.41234855570784623,
+      "learning_rate": 7.67775997519518e-05,
+      "loss": 0.3946,
+      "step": 1012
+    },
+    {
+      "epoch": 1.079957356076759,
+      "grad_norm": 0.542278036057536,
+      "learning_rate": 7.676588267687651e-05,
+      "loss": 0.3957,
+      "step": 1013
+    },
+    {
+      "epoch": 1.0810234541577826,
+      "grad_norm": 0.6900107120803527,
+      "learning_rate": 7.6754145235413e-05,
+      "loss": 0.3974,
+      "step": 1014
+    },
+    {
+      "epoch": 1.0820895522388059,
+      "grad_norm": 0.7866757471169271,
+      "learning_rate": 7.674238743406319e-05,
+      "loss": 0.4015,
+      "step": 1015
+    },
+    {
+      "epoch": 1.0831556503198294,
+      "grad_norm": 0.8352910554164142,
+      "learning_rate": 7.673060927934032e-05,
+      "loss": 0.404,
+      "step": 1016
+    },
+    {
+      "epoch": 1.084221748400853,
+      "grad_norm": 0.7695238174542417,
+      "learning_rate": 7.671881077776884e-05,
+      "loss": 0.3926,
+      "step": 1017
+    },
+    {
+      "epoch": 1.0852878464818763,
+      "grad_norm": 0.5536485102257792,
+      "learning_rate": 7.670699193588453e-05,
+      "loss": 0.3985,
+      "step": 1018
+    },
+    {
+      "epoch": 1.0863539445628998,
+      "grad_norm": 0.4228656833300754,
+      "learning_rate": 7.66951527602344e-05,
+      "loss": 0.3996,
+      "step": 1019
+    },
+    {
+      "epoch": 1.0874200426439233,
+      "grad_norm": 0.46124002487349797,
+      "learning_rate": 7.668329325737674e-05,
+      "loss": 0.3891,
+      "step": 1020
+    },
+    {
+      "epoch": 1.0884861407249466,
+      "grad_norm": 0.4882167434505241,
+      "learning_rate": 7.667141343388111e-05,
+      "loss": 0.4026,
+      "step": 1021
+    },
+    {
+      "epoch": 1.0895522388059702,
+      "grad_norm": 0.44551496006705965,
+      "learning_rate": 7.665951329632829e-05,
+      "loss": 0.4028,
+      "step": 1022
+    },
+    {
+      "epoch": 1.0906183368869935,
+      "grad_norm": 0.3735062426653466,
+      "learning_rate": 7.664759285131039e-05,
+      "loss": 0.3934,
+      "step": 1023
+    },
+    {
+      "epoch": 1.091684434968017,
+      "grad_norm": 0.3517866272882918,
+      "learning_rate": 7.663565210543065e-05,
+      "loss": 0.3908,
+      "step": 1024
+    },
+    {
+      "epoch": 1.0927505330490406,
+      "grad_norm": 0.36578234461021986,
+      "learning_rate": 7.662369106530367e-05,
+      "loss": 0.401,
+      "step": 1025
+    },
+    {
+      "epoch": 1.0938166311300639,
+      "grad_norm": 0.37034761078040107,
+      "learning_rate": 7.661170973755523e-05,
+      "loss": 0.3963,
+      "step": 1026
+    },
+    {
+      "epoch": 1.0948827292110874,
+      "grad_norm": 0.44787182009231896,
+      "learning_rate": 7.659970812882236e-05,
+      "loss": 0.396,
+      "step": 1027
+    },
+    {
+      "epoch": 1.095948827292111,
+      "grad_norm": 0.46802593434023804,
+      "learning_rate": 7.658768624575331e-05,
+      "loss": 0.3895,
+      "step": 1028
+    },
+    {
+      "epoch": 1.0970149253731343,
+      "grad_norm": 0.46152503018911306,
+      "learning_rate": 7.657564409500763e-05,
+      "loss": 0.4017,
+      "step": 1029
+    },
+    {
+      "epoch": 1.0980810234541578,
+      "grad_norm": 0.46014261998376804,
+      "learning_rate": 7.6563581683256e-05,
+      "loss": 0.3986,
+      "step": 1030
+    },
+    {
+      "epoch": 1.0991471215351813,
+      "grad_norm": 0.5602013316636195,
+      "learning_rate": 7.655149901718038e-05,
+      "loss": 0.3998,
+      "step": 1031
+    },
+    {
+      "epoch": 1.1002132196162047,
+      "grad_norm": 0.6361354791978112,
+      "learning_rate": 7.653939610347393e-05,
+      "loss": 0.3931,
+      "step": 1032
+    },
+    {
+      "epoch": 1.1012793176972282,
+      "grad_norm": 0.6096074276843847,
+      "learning_rate": 7.652727294884107e-05,
+      "loss": 0.403,
+      "step": 1033
+    },
+    {
+      "epoch": 1.1023454157782515,
+      "grad_norm": 0.6436284856894914,
+      "learning_rate": 7.651512955999737e-05,
+      "loss": 0.4062,
+      "step": 1034
+    },
+    {
+      "epoch": 1.103411513859275,
+      "grad_norm": 0.7789944154844247,
+      "learning_rate": 7.650296594366962e-05,
+      "loss": 0.4001,
+      "step": 1035
+    },
+    {
+      "epoch": 1.1044776119402986,
+      "grad_norm": 0.7656806635135831,
+      "learning_rate": 7.649078210659587e-05,
+      "loss": 0.3941,
+      "step": 1036
+    },
+    {
+      "epoch": 1.105543710021322,
+      "grad_norm": 0.6743206765604671,
+      "learning_rate": 7.647857805552532e-05,
+      "loss": 0.3982,
+      "step": 1037
+    },
+    {
+      "epoch": 1.1066098081023454,
+      "grad_norm": 0.6738594516444768,
+      "learning_rate": 7.646635379721837e-05,
+      "loss": 0.3995,
+      "step": 1038
+    },
+    {
+      "epoch": 1.1076759061833688,
+      "grad_norm": 0.5665383403100596,
+      "learning_rate": 7.645410933844663e-05,
+      "loss": 0.3996,
+      "step": 1039
+    },
+    {
+      "epoch": 1.1087420042643923,
+      "grad_norm": 0.4665621519906685,
+      "learning_rate": 7.644184468599289e-05,
+      "loss": 0.3986,
+      "step": 1040
+    },
+    {
+      "epoch": 1.1098081023454158,
+      "grad_norm": 0.39730689152255527,
+      "learning_rate": 7.642955984665113e-05,
+      "loss": 0.4026,
+      "step": 1041
+    },
+    {
+      "epoch": 1.1108742004264391,
+      "grad_norm": 0.4095924968703311,
+      "learning_rate": 7.641725482722651e-05,
+      "loss": 0.3944,
+      "step": 1042
+    },
+    {
+      "epoch": 1.1119402985074627,
+      "grad_norm": 0.42742701569060193,
+      "learning_rate": 7.640492963453538e-05,
+      "loss": 0.4004,
+      "step": 1043
+    },
+    {
+      "epoch": 1.1130063965884862,
+      "grad_norm": 0.4714508774381032,
+      "learning_rate": 7.639258427540526e-05,
+      "loss": 0.3923,
+      "step": 1044
+    },
+    {
+      "epoch": 1.1140724946695095,
+      "grad_norm": 0.43458932184535304,
+      "learning_rate": 7.638021875667483e-05,
+      "loss": 0.4013,
+      "step": 1045
+    },
+    {
+      "epoch": 1.115138592750533,
+      "grad_norm": 0.4337664712771011,
+      "learning_rate": 7.636783308519394e-05,
+      "loss": 0.3971,
+      "step": 1046
+    },
+    {
+      "epoch": 1.1162046908315566,
+      "grad_norm": 0.5374420364179895,
+      "learning_rate": 7.63554272678236e-05,
+      "loss": 0.3987,
+      "step": 1047
+    },
+    {
+      "epoch": 1.11727078891258,
+      "grad_norm": 0.599484068723148,
+      "learning_rate": 7.634300131143601e-05,
+      "loss": 0.4031,
+      "step": 1048
+    },
+    {
+      "epoch": 1.1183368869936035,
+      "grad_norm": 0.5397453012699273,
+      "learning_rate": 7.63305552229145e-05,
+      "loss": 0.3929,
+      "step": 1049
+    },
+    {
+      "epoch": 1.1194029850746268,
+      "grad_norm": 0.49396537617545705,
+      "learning_rate": 7.631808900915355e-05,
+      "loss": 0.4011,
+      "step": 1050
+    },
+    {
+      "epoch": 1.1204690831556503,
+      "grad_norm": 0.5246477358086288,
+      "learning_rate": 7.63056026770588e-05,
+      "loss": 0.395,
+      "step": 1051
+    },
+    {
+      "epoch": 1.1215351812366738,
+      "grad_norm": 0.4694349641844175,
+      "learning_rate": 7.6293096233547e-05,
+      "loss": 0.4,
+      "step": 1052
+    },
+    {
+      "epoch": 1.1226012793176972,
+      "grad_norm": 0.49101259479940984,
+      "learning_rate": 7.628056968554613e-05,
+      "loss": 0.4023,
+      "step": 1053
+    },
+    {
+      "epoch": 1.1236673773987207,
+      "grad_norm": 0.5743556431588469,
+      "learning_rate": 7.626802303999519e-05,
+      "loss": 0.3978,
+      "step": 1054
+    },
+    {
+      "epoch": 1.1247334754797442,
+      "grad_norm": 0.5806632736863245,
+      "learning_rate": 7.62554563038444e-05,
+      "loss": 0.3934,
+      "step": 1055
+    },
+    {
+      "epoch": 1.1257995735607675,
+      "grad_norm": 0.5458834450664017,
+      "learning_rate": 7.624286948405506e-05,
+      "loss": 0.4,
+      "step": 1056
+    },
+    {
+      "epoch": 1.126865671641791,
+      "grad_norm": 0.5294205958455052,
+      "learning_rate": 7.623026258759963e-05,
+      "loss": 0.4058,
+      "step": 1057
+    },
+    {
+      "epoch": 1.1279317697228146,
+      "grad_norm": 0.517163593116652,
+      "learning_rate": 7.621763562146167e-05,
+      "loss": 0.3987,
+      "step": 1058
+    },
+    {
+      "epoch": 1.128997867803838,
+      "grad_norm": 0.4383142153955423,
+      "learning_rate": 7.620498859263584e-05,
+      "loss": 0.3953,
+      "step": 1059
+    },
+    {
+      "epoch": 1.1300639658848615,
+      "grad_norm": 0.48624167313347716,
+      "learning_rate": 7.619232150812799e-05,
+      "loss": 0.4012,
+      "step": 1060
+    },
+    {
+      "epoch": 1.1311300639658848,
+      "grad_norm": 0.5650158980091758,
+      "learning_rate": 7.617963437495498e-05,
+      "loss": 0.3989,
+      "step": 1061
+    },
+    {
+      "epoch": 1.1321961620469083,
+      "grad_norm": 0.6366745345835356,
+      "learning_rate": 7.616692720014484e-05,
+      "loss": 0.3948,
+      "step": 1062
+    },
+    {
+      "epoch": 1.1332622601279319,
+      "grad_norm": 0.7232155780875018,
+      "learning_rate": 7.615419999073667e-05,
+      "loss": 0.3894,
+      "step": 1063
+    },
+    {
+      "epoch": 1.1343283582089552,
+      "grad_norm": 0.8621803906529987,
+      "learning_rate": 7.614145275378072e-05,
+      "loss": 0.4004,
+      "step": 1064
+    },
+    {
+      "epoch": 1.1353944562899787,
+      "grad_norm": 1.031817218923734,
+      "learning_rate": 7.612868549633825e-05,
+      "loss": 0.3913,
+      "step": 1065
+    },
+    {
+      "epoch": 1.136460554371002,
+      "grad_norm": 0.995716991471701,
+      "learning_rate": 7.611589822548168e-05,
+      "loss": 0.3989,
+      "step": 1066
+    },
+    {
+      "epoch": 1.1375266524520256,
+      "grad_norm": 0.9135720092202024,
+      "learning_rate": 7.61030909482945e-05,
+      "loss": 0.4013,
+      "step": 1067
+    },
+    {
+      "epoch": 1.138592750533049,
+      "grad_norm": 0.7949810308021543,
+      "learning_rate": 7.609026367187125e-05,
+      "loss": 0.3977,
+      "step": 1068
+    },
+    {
+      "epoch": 1.1396588486140724,
+      "grad_norm": 0.7103297637427972,
+      "learning_rate": 7.607741640331761e-05,
+      "loss": 0.3925,
+      "step": 1069
+    },
+    {
+      "epoch": 1.140724946695096,
+      "grad_norm": 0.6765477797632053,
+      "learning_rate": 7.606454914975029e-05,
+      "loss": 0.3957,
+      "step": 1070
+    },
+    {
+      "epoch": 1.1417910447761195,
+      "grad_norm": 0.6061563911680509,
+      "learning_rate": 7.605166191829705e-05,
+      "loss": 0.3952,
+      "step": 1071
+    },
+    {
+      "epoch": 1.1428571428571428,
+      "grad_norm": 0.6115108481298734,
+      "learning_rate": 7.603875471609677e-05,
+      "loss": 0.4006,
+      "step": 1072
+    },
+    {
+      "epoch": 1.1439232409381663,
+      "grad_norm": 0.7609527472496548,
+      "learning_rate": 7.602582755029938e-05,
+      "loss": 0.3987,
+      "step": 1073
+    },
+    {
+      "epoch": 1.1449893390191899,
+      "grad_norm": 0.7173940431562806,
+      "learning_rate": 7.601288042806582e-05,
+      "loss": 0.3995,
+      "step": 1074
+    },
+    {
+      "epoch": 1.1460554371002132,
+      "grad_norm": 0.5756472561097495,
+      "learning_rate": 7.599991335656817e-05,
+      "loss": 0.3929,
+      "step": 1075
+    },
+    {
+      "epoch": 1.1471215351812367,
+      "grad_norm": 0.5051191561774657,
+      "learning_rate": 7.598692634298949e-05,
+      "loss": 0.3974,
+      "step": 1076
+    },
+    {
+      "epoch": 1.14818763326226,
+      "grad_norm": 0.506878322162707,
+      "learning_rate": 7.59739193945239e-05,
+      "loss": 0.3988,
+      "step": 1077
+    },
+    {
+      "epoch": 1.1492537313432836,
+      "grad_norm": 0.4945186749103627,
+      "learning_rate": 7.596089251837659e-05,
+      "loss": 0.3967,
+      "step": 1078
+    },
+    {
+      "epoch": 1.150319829424307,
+      "grad_norm": 0.5094561991761961,
+      "learning_rate": 7.594784572176378e-05,
+      "loss": 0.3945,
+      "step": 1079
+    },
+    {
+      "epoch": 1.1513859275053304,
+      "grad_norm": 0.5005613849300656,
+      "learning_rate": 7.593477901191268e-05,
+      "loss": 0.3962,
+      "step": 1080
+    },
+    {
+      "epoch": 1.152452025586354,
+      "grad_norm": 0.39722111509884156,
+      "learning_rate": 7.592169239606161e-05,
+      "loss": 0.407,
+      "step": 1081
+    },
+    {
+      "epoch": 1.1535181236673775,
+      "grad_norm": 0.3086629383868203,
+      "learning_rate": 7.590858588145985e-05,
+      "loss": 0.3968,
+      "step": 1082
+    },
+    {
+      "epoch": 1.1545842217484008,
+      "grad_norm": 0.32105906743248047,
+      "learning_rate": 7.589545947536774e-05,
+      "loss": 0.4023,
+      "step": 1083
+    },
+    {
+      "epoch": 1.1556503198294243,
+      "grad_norm": 0.393821577059538,
+      "learning_rate": 7.588231318505661e-05,
+      "loss": 0.4,
+      "step": 1084
+    },
+    {
+      "epoch": 1.1567164179104479,
+      "grad_norm": 0.44231624171196116,
+      "learning_rate": 7.586914701780886e-05,
+      "loss": 0.4008,
+      "step": 1085
+    },
+    {
+      "epoch": 1.1577825159914712,
+      "grad_norm": 0.4026841091750872,
+      "learning_rate": 7.585596098091782e-05,
+      "loss": 0.3966,
+      "step": 1086
+    },
+    {
+      "epoch": 1.1588486140724947,
+      "grad_norm": 0.4378761127318457,
+      "learning_rate": 7.58427550816879e-05,
+      "loss": 0.3967,
+      "step": 1087
+    },
+    {
+      "epoch": 1.159914712153518,
+      "grad_norm": 0.5373953761219459,
+      "learning_rate": 7.582952932743445e-05,
+      "loss": 0.4052,
+      "step": 1088
+    },
+    {
+      "epoch": 1.1609808102345416,
+      "grad_norm": 0.6376217391121629,
+      "learning_rate": 7.581628372548388e-05,
+      "loss": 0.3968,
+      "step": 1089
+    },
+    {
+      "epoch": 1.1620469083155651,
+      "grad_norm": 0.7124571886953409,
+      "learning_rate": 7.580301828317354e-05,
+      "loss": 0.3967,
+      "step": 1090
+    },
+    {
+      "epoch": 1.1631130063965884,
+      "grad_norm": 0.710369838726693,
+      "learning_rate": 7.578973300785182e-05,
+      "loss": 0.3926,
+      "step": 1091
+    },
+    {
+      "epoch": 1.164179104477612,
+      "grad_norm": 0.601219772248201,
+      "learning_rate": 7.577642790687805e-05,
+      "loss": 0.3977,
+      "step": 1092
+    },
+    {
+      "epoch": 1.1652452025586353,
+      "grad_norm": 0.4939938541558801,
+      "learning_rate": 7.57631029876226e-05,
+      "loss": 0.3976,
+      "step": 1093
+    },
+    {
+      "epoch": 1.1663113006396588,
+      "grad_norm": 0.5843429305407729,
+      "learning_rate": 7.574975825746673e-05,
+      "loss": 0.3985,
+      "step": 1094
+    },
+    {
+      "epoch": 1.1673773987206824,
+      "grad_norm": 0.7423520288508058,
+      "learning_rate": 7.573639372380277e-05,
+      "loss": 0.3937,
+      "step": 1095
+    },
+    {
+      "epoch": 1.1684434968017057,
+      "grad_norm": 0.6911804134372755,
+      "learning_rate": 7.572300939403395e-05,
+      "loss": 0.3998,
+      "step": 1096
+    },
+    {
+      "epoch": 1.1695095948827292,
+      "grad_norm": 0.5980230638871175,
+      "learning_rate": 7.570960527557452e-05,
+      "loss": 0.3927,
+      "step": 1097
+    },
+    {
+      "epoch": 1.1705756929637527,
+      "grad_norm": 0.5740905237257593,
+      "learning_rate": 7.569618137584964e-05,
+      "loss": 0.3966,
+      "step": 1098
+    },
+    {
+      "epoch": 1.171641791044776,
+      "grad_norm": 0.5934276451380195,
+      "learning_rate": 7.568273770229546e-05,
+      "loss": 0.3958,
+      "step": 1099
+    },
+    {
+      "epoch": 1.1727078891257996,
+      "grad_norm": 0.6018407068991795,
+      "learning_rate": 7.566927426235909e-05,
+      "loss": 0.392,
+      "step": 1100
+    },
+    {
+      "epoch": 1.1737739872068231,
+      "grad_norm": 0.5584205973152891,
+      "learning_rate": 7.565579106349857e-05,
+      "loss": 0.403,
+      "step": 1101
+    },
+    {
+      "epoch": 1.1748400852878464,
+      "grad_norm": 0.5611127016532973,
+      "learning_rate": 7.564228811318288e-05,
+      "loss": 0.3946,
+      "step": 1102
+    },
+    {
+      "epoch": 1.17590618336887,
+      "grad_norm": 0.6160481911439404,
+      "learning_rate": 7.562876541889195e-05,
+      "loss": 0.3942,
+      "step": 1103
+    },
+    {
+      "epoch": 1.1769722814498933,
+      "grad_norm": 0.5055559528539143,
+      "learning_rate": 7.561522298811667e-05,
+      "loss": 0.3928,
+      "step": 1104
+    },
+    {
+      "epoch": 1.1780383795309168,
+      "grad_norm": 0.4031249818431398,
+      "learning_rate": 7.560166082835883e-05,
+      "loss": 0.3947,
+      "step": 1105
+    },
+    {
+      "epoch": 1.1791044776119404,
+      "grad_norm": 0.4844578748616443,
+      "learning_rate": 7.558807894713116e-05,
+      "loss": 0.4,
+      "step": 1106
+    },
+    {
+      "epoch": 1.1801705756929637,
+      "grad_norm": 0.5923421522905422,
+      "learning_rate": 7.557447735195732e-05,
+      "loss": 0.3932,
+      "step": 1107
+    },
+    {
+      "epoch": 1.1812366737739872,
+      "grad_norm": 0.5732059132565734,
+      "learning_rate": 7.556085605037191e-05,
+      "loss": 0.396,
+      "step": 1108
+    },
+    {
+      "epoch": 1.1823027718550105,
+      "grad_norm": 0.5727038238228728,
+      "learning_rate": 7.554721504992038e-05,
+      "loss": 0.3942,
+      "step": 1109
+    },
+    {
+      "epoch": 1.183368869936034,
+      "grad_norm": 0.5941550331742329,
+      "learning_rate": 7.553355435815915e-05,
+      "loss": 0.3921,
+      "step": 1110
+    },
+    {
+      "epoch": 1.1844349680170576,
+      "grad_norm": 0.6162721480823822,
+      "learning_rate": 7.551987398265554e-05,
+      "loss": 0.3973,
+      "step": 1111
+    },
+    {
+      "epoch": 1.1855010660980811,
+      "grad_norm": 0.6503625421341107,
+      "learning_rate": 7.550617393098777e-05,
+      "loss": 0.4,
+      "step": 1112
+    },
+    {
+      "epoch": 1.1865671641791045,
+      "grad_norm": 0.6428847327632391,
+      "learning_rate": 7.549245421074496e-05,
+      "loss": 0.4032,
+      "step": 1113
+    },
+    {
+      "epoch": 1.187633262260128,
+      "grad_norm": 0.6516650235926786,
+      "learning_rate": 7.54787148295271e-05,
+      "loss": 0.3972,
+      "step": 1114
+    },
+    {
+      "epoch": 1.1886993603411513,
+      "grad_norm": 0.6996227895669477,
+      "learning_rate": 7.546495579494512e-05,
+      "loss": 0.3977,
+      "step": 1115
+    },
+    {
+      "epoch": 1.1897654584221748,
+      "grad_norm": 0.6011748919765059,
+      "learning_rate": 7.54511771146208e-05,
+      "loss": 0.3937,
+      "step": 1116
+    },
+    {
+      "epoch": 1.1908315565031984,
+      "grad_norm": 0.4608169210715356,
+      "learning_rate": 7.54373787961868e-05,
+      "loss": 0.4012,
+      "step": 1117
+    },
+    {
+      "epoch": 1.1918976545842217,
+      "grad_norm": 0.41047190471676254,
+      "learning_rate": 7.542356084728669e-05,
+      "loss": 0.3999,
+      "step": 1118
+    },
+    {
+      "epoch": 1.1929637526652452,
+      "grad_norm": 0.41683875837640383,
+      "learning_rate": 7.540972327557487e-05,
+      "loss": 0.3962,
+      "step": 1119
+    },
+    {
+      "epoch": 1.1940298507462686,
+      "grad_norm": 0.4017382310711894,
+      "learning_rate": 7.539586608871667e-05,
+      "loss": 0.3971,
+      "step": 1120
+    },
+    {
+      "epoch": 1.195095948827292,
+      "grad_norm": 0.3327575746497003,
+      "learning_rate": 7.538198929438823e-05,
+      "loss": 0.3998,
+      "step": 1121
+    },
+    {
+      "epoch": 1.1961620469083156,
+      "grad_norm": 0.36060866682655995,
+      "learning_rate": 7.536809290027657e-05,
+      "loss": 0.3892,
+      "step": 1122
+    },
+    {
+      "epoch": 1.197228144989339,
+      "grad_norm": 0.4016423989687354,
+      "learning_rate": 7.53541769140796e-05,
+      "loss": 0.3928,
+      "step": 1123
+    },
+    {
+      "epoch": 1.1982942430703625,
+      "grad_norm": 0.4667699426682389,
+      "learning_rate": 7.5340241343506e-05,
+      "loss": 0.3927,
+      "step": 1124
+    },
+    {
+      "epoch": 1.199360341151386,
+      "grad_norm": 0.52739114765906,
+      "learning_rate": 7.532628619627541e-05,
+      "loss": 0.3965,
+      "step": 1125
+    },
+    {
+      "epoch": 1.2004264392324093,
+      "grad_norm": 0.5145519240447523,
+      "learning_rate": 7.531231148011821e-05,
+      "loss": 0.3934,
+      "step": 1126
+    },
+    {
+      "epoch": 1.2014925373134329,
+      "grad_norm": 0.46867941356549736,
+      "learning_rate": 7.529831720277569e-05,
+      "loss": 0.4006,
+      "step": 1127
+    },
+    {
+      "epoch": 1.2025586353944564,
+      "grad_norm": 0.394042649849781,
+      "learning_rate": 7.528430337199995e-05,
+      "loss": 0.4003,
+      "step": 1128
+    },
+    {
+      "epoch": 1.2036247334754797,
+      "grad_norm": 0.449136494508411,
+      "learning_rate": 7.527026999555393e-05,
+      "loss": 0.3978,
+      "step": 1129
+    },
+    {
+      "epoch": 1.2046908315565032,
+      "grad_norm": 0.5439656094014544,
+      "learning_rate": 7.525621708121136e-05,
+      "loss": 0.3974,
+      "step": 1130
+    },
+    {
+      "epoch": 1.2057569296375266,
+      "grad_norm": 0.6999516585400055,
+      "learning_rate": 7.524214463675686e-05,
+      "loss": 0.4066,
+      "step": 1131
+    },
+    {
+      "epoch": 1.20682302771855,
+      "grad_norm": 0.9223862712790244,
+      "learning_rate": 7.522805266998582e-05,
+      "loss": 0.3935,
+      "step": 1132
+    },
+    {
+      "epoch": 1.2078891257995736,
+      "grad_norm": 1.005522314616768,
+      "learning_rate": 7.521394118870446e-05,
+      "loss": 0.3996,
+      "step": 1133
+    },
+    {
+      "epoch": 1.208955223880597,
+      "grad_norm": 0.9189093535473554,
+      "learning_rate": 7.519981020072979e-05,
+      "loss": 0.3986,
+      "step": 1134
+    },
+    {
+      "epoch": 1.2100213219616205,
+      "grad_norm": 0.7474366825523275,
+      "learning_rate": 7.518565971388967e-05,
+      "loss": 0.3975,
+      "step": 1135
+    },
+    {
+      "epoch": 1.2110874200426438,
+      "grad_norm": 0.517722383461889,
+      "learning_rate": 7.51714897360227e-05,
+      "loss": 0.3915,
+      "step": 1136
+    },
+    {
+      "epoch": 1.2121535181236673,
+      "grad_norm": 0.460287240625528,
+      "learning_rate": 7.515730027497836e-05,
+      "loss": 0.3991,
+      "step": 1137
+    },
+    {
+      "epoch": 1.2132196162046909,
+      "grad_norm": 0.5513920397892742,
+      "learning_rate": 7.514309133861684e-05,
+      "loss": 0.3876,
+      "step": 1138
+    },
+    {
+      "epoch": 1.2142857142857142,
+      "grad_norm": 0.5926185140798839,
+      "learning_rate": 7.512886293480914e-05,
+      "loss": 0.3987,
+      "step": 1139
+    },
+    {
+      "epoch": 1.2153518123667377,
+      "grad_norm": 0.6015497667677062,
+      "learning_rate": 7.51146150714371e-05,
+      "loss": 0.3942,
+      "step": 1140
+    },
+    {
+      "epoch": 1.2164179104477613,
+      "grad_norm": 0.4979357814188255,
+      "learning_rate": 7.510034775639324e-05,
+      "loss": 0.3962,
+      "step": 1141
+    },
+    {
+      "epoch": 1.2174840085287846,
+      "grad_norm": 0.3795778689452729,
+      "learning_rate": 7.508606099758097e-05,
+      "loss": 0.4005,
+      "step": 1142
+    },
+    {
+      "epoch": 1.2185501066098081,
+      "grad_norm": 0.3928656441581091,
+      "learning_rate": 7.507175480291437e-05,
+      "loss": 0.3984,
+      "step": 1143
+    },
+    {
+      "epoch": 1.2196162046908317,
+      "grad_norm": 0.3930681703186637,
+      "learning_rate": 7.505742918031836e-05,
+      "loss": 0.3984,
+      "step": 1144
+    },
+    {
+      "epoch": 1.220682302771855,
+      "grad_norm": 0.4091879483768399,
+      "learning_rate": 7.504308413772856e-05,
+      "loss": 0.3933,
+      "step": 1145
+    },
+    {
+      "epoch": 1.2217484008528785,
+      "grad_norm": 0.46891356577284954,
+      "learning_rate": 7.502871968309139e-05,
+      "loss": 0.3988,
+      "step": 1146
+    },
+    {
+      "epoch": 1.2228144989339018,
+      "grad_norm": 0.5103860857976652,
+      "learning_rate": 7.5014335824364e-05,
+      "loss": 0.3935,
+      "step": 1147
+    },
+    {
+      "epoch": 1.2238805970149254,
+      "grad_norm": 0.5082917446932953,
+      "learning_rate": 7.499993256951433e-05,
+      "loss": 0.3927,
+      "step": 1148
+    },
+    {
+      "epoch": 1.224946695095949,
+      "grad_norm": 0.49379137994921035,
+      "learning_rate": 7.498550992652101e-05,
+      "loss": 0.4001,
+      "step": 1149
+    },
+    {
+      "epoch": 1.2260127931769722,
+      "grad_norm": 0.5353013303840448,
+      "learning_rate": 7.497106790337345e-05,
+      "loss": 0.3929,
+      "step": 1150
+    },
+    {
+      "epoch": 1.2270788912579957,
+      "grad_norm": 0.5876139459836854,
+      "learning_rate": 7.495660650807174e-05,
+      "loss": 0.3994,
+      "step": 1151
+    },
+    {
+      "epoch": 1.2281449893390193,
+      "grad_norm": 0.734210823795574,
+      "learning_rate": 7.494212574862682e-05,
+      "loss": 0.3967,
+      "step": 1152
+    },
+    {
+      "epoch": 1.2292110874200426,
+      "grad_norm": 0.7913917318938646,
+      "learning_rate": 7.49276256330602e-05,
+      "loss": 0.3936,
+      "step": 1153
+    },
+    {
+      "epoch": 1.2302771855010661,
+      "grad_norm": 0.7405044856069717,
+      "learning_rate": 7.491310616940422e-05,
+      "loss": 0.4,
+      "step": 1154
+    },
+    {
+      "epoch": 1.2313432835820897,
+      "grad_norm": 0.6663912293042208,
+      "learning_rate": 7.489856736570192e-05,
+      "loss": 0.4047,
+      "step": 1155
+    },
+    {
+      "epoch": 1.232409381663113,
+      "grad_norm": 0.6021454402370039,
+      "learning_rate": 7.488400923000703e-05,
+      "loss": 0.3907,
+      "step": 1156
+    },
+    {
+      "epoch": 1.2334754797441365,
+      "grad_norm": 0.5514602002833817,
+      "learning_rate": 7.4869431770384e-05,
+      "loss": 0.3978,
+      "step": 1157
+    },
+    {
+      "epoch": 1.2345415778251598,
+      "grad_norm": 0.48903597186198305,
+      "learning_rate": 7.485483499490799e-05,
+      "loss": 0.4005,
+      "step": 1158
+    },
+    {
+      "epoch": 1.2356076759061834,
+      "grad_norm": 0.5332694417143288,
+      "learning_rate": 7.484021891166486e-05,
+      "loss": 0.3952,
+      "step": 1159
+    },
+    {
+      "epoch": 1.236673773987207,
+      "grad_norm": 0.49543835364305194,
+      "learning_rate": 7.482558352875113e-05,
+      "loss": 0.3986,
+      "step": 1160
+    },
+    {
+      "epoch": 1.2377398720682302,
+      "grad_norm": 0.4261853368714419,
+      "learning_rate": 7.481092885427408e-05,
+      "loss": 0.39,
+      "step": 1161
+    },
+    {
+      "epoch": 1.2388059701492538,
+      "grad_norm": 0.4265314566032864,
+      "learning_rate": 7.479625489635162e-05,
+      "loss": 0.3911,
+      "step": 1162
+    },
+    {
+      "epoch": 1.239872068230277,
+      "grad_norm": 0.5029650666087466,
+      "learning_rate": 7.478156166311236e-05,
+      "loss": 0.3994,
+      "step": 1163
+    },
+    {
+      "epoch": 1.2409381663113006,
+      "grad_norm": 0.4455729452597588,
+      "learning_rate": 7.476684916269559e-05,
+      "loss": 0.3952,
+      "step": 1164
+    },
+    {
+      "epoch": 1.2420042643923241,
+      "grad_norm": 0.47338288349681196,
+      "learning_rate": 7.475211740325127e-05,
+      "loss": 0.3969,
+      "step": 1165
+    },
+    {
+      "epoch": 1.2430703624733475,
+      "grad_norm": 0.4305124151121982,
+      "learning_rate": 7.473736639294004e-05,
+      "loss": 0.3967,
+      "step": 1166
+    },
+    {
+      "epoch": 1.244136460554371,
+      "grad_norm": 0.40696000141373095,
+      "learning_rate": 7.472259613993316e-05,
+      "loss": 0.393,
+      "step": 1167
+    },
+    {
+      "epoch": 1.2452025586353945,
+      "grad_norm": 0.45477308879606393,
+      "learning_rate": 7.470780665241262e-05,
+      "loss": 0.3958,
+      "step": 1168
+    },
+    {
+      "epoch": 1.2462686567164178,
+      "grad_norm": 0.5704960820413824,
+      "learning_rate": 7.469299793857101e-05,
+      "loss": 0.3989,
+      "step": 1169
+    },
+    {
+      "epoch": 1.2473347547974414,
+      "grad_norm": 0.6135689230632373,
+      "learning_rate": 7.467817000661159e-05,
+      "loss": 0.3921,
+      "step": 1170
+    },
+    {
+      "epoch": 1.248400852878465,
+      "grad_norm": 0.6777534760469583,
+      "learning_rate": 7.466332286474826e-05,
+      "loss": 0.3962,
+      "step": 1171
+    },
+    {
+      "epoch": 1.2494669509594882,
+      "grad_norm": 0.6898047595171254,
+      "learning_rate": 7.464845652120557e-05,
+      "loss": 0.3915,
+      "step": 1172
+    },
+    {
+      "epoch": 1.2505330490405118,
+      "grad_norm": 0.7123893467359865,
+      "learning_rate": 7.46335709842187e-05,
+      "loss": 0.4025,
+      "step": 1173
+    },
+    {
+      "epoch": 1.251599147121535,
+      "grad_norm": 0.6820123806207717,
+      "learning_rate": 7.461866626203348e-05,
+      "loss": 0.3941,
+      "step": 1174
+    },
+    {
+      "epoch": 1.2526652452025586,
+      "grad_norm": 0.6533571683107087,
+      "learning_rate": 7.460374236290631e-05,
+      "loss": 0.4025,
+      "step": 1175
+    },
+    {
+      "epoch": 1.2537313432835822,
+      "grad_norm": 0.6571872810240704,
+      "learning_rate": 7.45887992951043e-05,
+      "loss": 0.3976,
+      "step": 1176
+    },
+    {
+      "epoch": 1.2547974413646055,
+      "grad_norm": 0.6849183239389355,
+      "learning_rate": 7.457383706690511e-05,
+      "loss": 0.3976,
+      "step": 1177
+    },
+    {
+      "epoch": 1.255863539445629,
+      "grad_norm": 0.6130858125092414,
+      "learning_rate": 7.455885568659705e-05,
+      "loss": 0.3917,
+      "step": 1178
+    },
+    {
+      "epoch": 1.2569296375266523,
+      "grad_norm": 0.5653602132939001,
+      "learning_rate": 7.454385516247899e-05,
+      "loss": 0.3898,
+      "step": 1179
+    },
+    {
+      "epoch": 1.2579957356076759,
+      "grad_norm": 0.6091279178319541,
+      "learning_rate": 7.452883550286049e-05,
+      "loss": 0.3928,
+      "step": 1180
+    },
+    {
+      "epoch": 1.2590618336886994,
+      "grad_norm": 0.5595753993224933,
+      "learning_rate": 7.451379671606162e-05,
+      "loss": 0.3957,
+      "step": 1181
+    },
+    {
+      "epoch": 1.260127931769723,
+      "grad_norm": 0.5079986278587565,
+      "learning_rate": 7.449873881041312e-05,
+      "loss": 0.3893,
+      "step": 1182
+    },
+    {
+      "epoch": 1.2611940298507462,
+      "grad_norm": 0.464090085159201,
+      "learning_rate": 7.448366179425628e-05,
+      "loss": 0.3926,
+      "step": 1183
+    },
+    {
+      "epoch": 1.2622601279317698,
+      "grad_norm": 0.510243217894761,
+      "learning_rate": 7.446856567594294e-05,
+      "loss": 0.3937,
+      "step": 1184
+    },
+    {
+      "epoch": 1.263326226012793,
+      "grad_norm": 0.46150844552833925,
+      "learning_rate": 7.445345046383563e-05,
+      "loss": 0.3947,
+      "step": 1185
+    },
+    {
+      "epoch": 1.2643923240938166,
+      "grad_norm": 0.3949887975107038,
+      "learning_rate": 7.443831616630734e-05,
+      "loss": 0.3884,
+      "step": 1186
+    },
+    {
+      "epoch": 1.2654584221748402,
+      "grad_norm": 0.5026720516501154,
+      "learning_rate": 7.442316279174172e-05,
+      "loss": 0.3906,
+      "step": 1187
+    },
+    {
+      "epoch": 1.2665245202558635,
+      "grad_norm": 0.502955271204735,
+      "learning_rate": 7.440799034853294e-05,
+      "loss": 0.3906,
+      "step": 1188
+    },
+    {
+      "epoch": 1.267590618336887,
+      "grad_norm": 0.43070711981801013,
+      "learning_rate": 7.439279884508573e-05,
+      "loss": 0.3959,
+      "step": 1189
+    },
+    {
+      "epoch": 1.2686567164179103,
+      "grad_norm": 0.3992942313925878,
+      "learning_rate": 7.437758828981542e-05,
+      "loss": 0.3979,
+      "step": 1190
+    },
+    {
+      "epoch": 1.2697228144989339,
+      "grad_norm": 0.41776106633422166,
+      "learning_rate": 7.436235869114785e-05,
+      "loss": 0.3888,
+      "step": 1191
+    },
+    {
+      "epoch": 1.2707889125799574,
+      "grad_norm": 0.3678243374968389,
+      "learning_rate": 7.434711005751942e-05,
+      "loss": 0.3929,
+      "step": 1192
+    },
+    {
+      "epoch": 1.271855010660981,
+      "grad_norm": 0.4462778916558436,
+      "learning_rate": 7.43318423973771e-05,
+      "loss": 0.4052,
+      "step": 1193
+    },
+    {
+      "epoch": 1.2729211087420043,
+      "grad_norm": 0.4072295381721561,
+      "learning_rate": 7.431655571917835e-05,
+      "loss": 0.3901,
+      "step": 1194
+    },
+    {
+      "epoch": 1.2739872068230278,
+      "grad_norm": 0.33668391253724705,
+      "learning_rate": 7.430125003139124e-05,
+      "loss": 0.395,
+      "step": 1195
+    },
+    {
+      "epoch": 1.275053304904051,
+      "grad_norm": 0.49984395156609207,
+      "learning_rate": 7.428592534249429e-05,
+      "loss": 0.3921,
+      "step": 1196
+    },
+    {
+      "epoch": 1.2761194029850746,
+      "grad_norm": 0.4961401301018006,
+      "learning_rate": 7.42705816609766e-05,
+      "loss": 0.3986,
+      "step": 1197
+    },
+    {
+      "epoch": 1.2771855010660982,
+      "grad_norm": 0.36987736865515364,
+      "learning_rate": 7.425521899533776e-05,
+      "loss": 0.3863,
+      "step": 1198
+    },
+    {
+      "epoch": 1.2782515991471215,
+      "grad_norm": 0.37433483688551117,
+      "learning_rate": 7.42398373540879e-05,
+      "loss": 0.4008,
+      "step": 1199
+    },
+    {
+      "epoch": 1.279317697228145,
+      "grad_norm": 0.3960664301566479,
+      "learning_rate": 7.422443674574764e-05,
+      "loss": 0.3936,
+      "step": 1200
+    },
+    {
+      "epoch": 1.2803837953091683,
+      "grad_norm": 0.42962763499848106,
+      "learning_rate": 7.42090171788481e-05,
+      "loss": 0.3968,
+      "step": 1201
+    },
+    {
+      "epoch": 1.2814498933901919,
+      "grad_norm": 0.4947438283060391,
+      "learning_rate": 7.419357866193097e-05,
+      "loss": 0.3974,
+      "step": 1202
+    },
+    {
+      "epoch": 1.2825159914712154,
+      "grad_norm": 0.7638002002790177,
+      "learning_rate": 7.417812120354833e-05,
+      "loss": 0.3987,
+      "step": 1203
+    },
+    {
+      "epoch": 1.2835820895522387,
+      "grad_norm": 0.6618293847952098,
+      "learning_rate": 7.416264481226284e-05,
+      "loss": 0.3954,
+      "step": 1204
+    },
+    {
+      "epoch": 1.2846481876332623,
+      "grad_norm": 0.7564644303522643,
+      "learning_rate": 7.414714949664761e-05,
+      "loss": 0.4025,
+      "step": 1205
+    },
+    {
+      "epoch": 1.2857142857142856,
+      "grad_norm": 0.7341957818072022,
+      "learning_rate": 7.413163526528623e-05,
+      "loss": 0.3972,
+      "step": 1206
+    },
+    {
+      "epoch": 1.2867803837953091,
+      "grad_norm": 0.7172296615238862,
+      "learning_rate": 7.41161021267728e-05,
+      "loss": 0.4024,
+      "step": 1207
+    },
+    {
+      "epoch": 1.2878464818763327,
+      "grad_norm": 0.6773018518559085,
+      "learning_rate": 7.410055008971186e-05,
+      "loss": 0.3986,
+      "step": 1208
+    },
+    {
+      "epoch": 1.2889125799573562,
+      "grad_norm": 0.7499881724261244,
+      "learning_rate": 7.40849791627184e-05,
+      "loss": 0.401,
+      "step": 1209
+    },
+    {
+      "epoch": 1.2899786780383795,
+      "grad_norm": 0.6552329889222517,
+      "learning_rate": 7.406938935441795e-05,
+      "loss": 0.399,
+      "step": 1210
+    },
+    {
+      "epoch": 1.291044776119403,
+      "grad_norm": 0.6452927503574722,
+      "learning_rate": 7.405378067344645e-05,
+      "loss": 0.4023,
+      "step": 1211
+    },
+    {
+      "epoch": 1.2921108742004264,
+      "grad_norm": 0.6122823637184173,
+      "learning_rate": 7.403815312845027e-05,
+      "loss": 0.4003,
+      "step": 1212
+    },
+    {
+      "epoch": 1.29317697228145,
+      "grad_norm": 0.4885206292712561,
+      "learning_rate": 7.402250672808627e-05,
+      "loss": 0.3948,
+      "step": 1213
+    },
+    {
+      "epoch": 1.2942430703624734,
+      "grad_norm": 0.36971065365308836,
+      "learning_rate": 7.400684148102175e-05,
+      "loss": 0.3983,
+      "step": 1214
+    },
+    {
+      "epoch": 1.2953091684434968,
+      "grad_norm": 0.5341249123573403,
+      "learning_rate": 7.399115739593444e-05,
+      "loss": 0.3998,
+      "step": 1215
+    },
+    {
+      "epoch": 1.2963752665245203,
+      "grad_norm": 0.6951670150170282,
+      "learning_rate": 7.397545448151249e-05,
+      "loss": 0.3917,
+      "step": 1216
+    },
+    {
+      "epoch": 1.2974413646055436,
+      "grad_norm": 0.6307408026599629,
+      "learning_rate": 7.395973274645452e-05,
+      "loss": 0.3966,
+      "step": 1217
+    },
+    {
+      "epoch": 1.2985074626865671,
+      "grad_norm": 0.41244623390075813,
+      "learning_rate": 7.394399219946955e-05,
+      "loss": 0.3943,
+      "step": 1218
+    },
+    {
+      "epoch": 1.2995735607675907,
+      "grad_norm": 0.4959351860092428,
+      "learning_rate": 7.392823284927704e-05,
+      "loss": 0.3969,
+      "step": 1219
+    },
+    {
+      "epoch": 1.3006396588486142,
+      "grad_norm": 0.6354141360008508,
+      "learning_rate": 7.391245470460682e-05,
+      "loss": 0.3912,
+      "step": 1220
+    },
+    {
+      "epoch": 1.3017057569296375,
+      "grad_norm": 0.5545244238798931,
+      "learning_rate": 7.389665777419916e-05,
+      "loss": 0.3959,
+      "step": 1221
+    },
+    {
+      "epoch": 1.302771855010661,
+      "grad_norm": 0.4981798925494521,
+      "learning_rate": 7.388084206680477e-05,
+      "loss": 0.3975,
+      "step": 1222
+    },
+    {
+      "epoch": 1.3038379530916844,
+      "grad_norm": 0.6996743346801269,
+      "learning_rate": 7.386500759118472e-05,
+      "loss": 0.3969,
+      "step": 1223
+    },
+    {
+      "epoch": 1.304904051172708,
+      "grad_norm": 0.8545480674302531,
+      "learning_rate": 7.384915435611047e-05,
+      "loss": 0.3911,
+      "step": 1224
+    },
+    {
+      "epoch": 1.3059701492537314,
+      "grad_norm": 0.9034425556959509,
+      "learning_rate": 7.38332823703639e-05,
+      "loss": 0.3984,
+      "step": 1225
+    },
+    {
+      "epoch": 1.3070362473347548,
+      "grad_norm": 0.8405317648162198,
+      "learning_rate": 7.381739164273727e-05,
+      "loss": 0.3919,
+      "step": 1226
+    },
+    {
+      "epoch": 1.3081023454157783,
+      "grad_norm": 0.7611193691575167,
+      "learning_rate": 7.380148218203321e-05,
+      "loss": 0.3936,
+      "step": 1227
+    },
+    {
+      "epoch": 1.3091684434968016,
+      "grad_norm": 0.7529452227808444,
+      "learning_rate": 7.378555399706473e-05,
+      "loss": 0.3976,
+      "step": 1228
+    },
+    {
+      "epoch": 1.3102345415778252,
+      "grad_norm": 0.7498395498416355,
+      "learning_rate": 7.376960709665522e-05,
+      "loss": 0.3938,
+      "step": 1229
+    },
+    {
+      "epoch": 1.3113006396588487,
+      "grad_norm": 0.617409199350578,
+      "learning_rate": 7.375364148963845e-05,
+      "loss": 0.4,
+      "step": 1230
+    },
+    {
+      "epoch": 1.312366737739872,
+      "grad_norm": 0.5012176007830088,
+      "learning_rate": 7.373765718485851e-05,
+      "loss": 0.3898,
+      "step": 1231
+    },
+    {
+      "epoch": 1.3134328358208955,
+      "grad_norm": 0.43946542231558117,
+      "learning_rate": 7.37216541911699e-05,
+      "loss": 0.4011,
+      "step": 1232
+    },
+    {
+      "epoch": 1.3144989339019189,
+      "grad_norm": 0.37173647672592797,
+      "learning_rate": 7.370563251743743e-05,
+      "loss": 0.4035,
+      "step": 1233
+    },
+    {
+      "epoch": 1.3155650319829424,
+      "grad_norm": 0.3453534441885958,
+      "learning_rate": 7.368959217253627e-05,
+      "loss": 0.3997,
+      "step": 1234
+    },
+    {
+      "epoch": 1.316631130063966,
+      "grad_norm": 0.35475360394383565,
+      "learning_rate": 7.367353316535195e-05,
+      "loss": 0.396,
+      "step": 1235
+    },
+    {
+      "epoch": 1.3176972281449895,
+      "grad_norm": 0.4807258068112041,
+      "learning_rate": 7.365745550478034e-05,
+      "loss": 0.3981,
+      "step": 1236
+    },
+    {
+      "epoch": 1.3187633262260128,
+      "grad_norm": 0.5770643962715295,
+      "learning_rate": 7.364135919972759e-05,
+      "loss": 0.397,
+      "step": 1237
+    },
+    {
+      "epoch": 1.3198294243070363,
+      "grad_norm": 0.5494777777905933,
+      "learning_rate": 7.362524425911024e-05,
+      "loss": 0.3937,
+      "step": 1238
+    },
+    {
+      "epoch": 1.3208955223880596,
+      "grad_norm": 0.4823030916464439,
+      "learning_rate": 7.360911069185513e-05,
+      "loss": 0.3946,
+      "step": 1239
+    },
+    {
+      "epoch": 1.3219616204690832,
+      "grad_norm": 0.44738974205793797,
+      "learning_rate": 7.35929585068994e-05,
+      "loss": 0.3987,
+      "step": 1240
+    },
+    {
+      "epoch": 1.3230277185501067,
+      "grad_norm": 0.4647985462455625,
+      "learning_rate": 7.357678771319055e-05,
+      "loss": 0.3989,
+      "step": 1241
+    },
+    {
+      "epoch": 1.32409381663113,
+      "grad_norm": 0.4907957965266697,
+      "learning_rate": 7.356059831968634e-05,
+      "loss": 0.3928,
+      "step": 1242
+    },
+    {
+      "epoch": 1.3251599147121536,
+      "grad_norm": 0.4834632280662174,
+      "learning_rate": 7.354439033535486e-05,
+      "loss": 0.3925,
+      "step": 1243
+    },
+    {
+      "epoch": 1.3262260127931769,
+      "grad_norm": 0.5312494947802512,
+      "learning_rate": 7.352816376917448e-05,
+      "loss": 0.395,
+      "step": 1244
+    },
+    {
+      "epoch": 1.3272921108742004,
+      "grad_norm": 0.6134924022779664,
+      "learning_rate": 7.351191863013387e-05,
+      "loss": 0.3952,
+      "step": 1245
+    },
+    {
+      "epoch": 1.328358208955224,
+      "grad_norm": 0.6854173044473056,
+      "learning_rate": 7.349565492723204e-05,
+      "loss": 0.3993,
+      "step": 1246
+    },
+    {
+      "epoch": 1.3294243070362473,
+      "grad_norm": 0.6787952149521692,
+      "learning_rate": 7.347937266947817e-05,
+      "loss": 0.388,
+      "step": 1247
+    },
+    {
+      "epoch": 1.3304904051172708,
+      "grad_norm": 0.5985162637767774,
+      "learning_rate": 7.346307186589183e-05,
+      "loss": 0.3964,
+      "step": 1248
+    },
+    {
+      "epoch": 1.331556503198294,
+      "grad_norm": 0.5267115292559772,
+      "learning_rate": 7.344675252550278e-05,
+      "loss": 0.397,
+      "step": 1249
+    },
+    {
+      "epoch": 1.3326226012793176,
+      "grad_norm": 0.4627942903946516,
+      "learning_rate": 7.343041465735115e-05,
+      "loss": 0.398,
+      "step": 1250
+    },
+    {
+      "epoch": 1.3336886993603412,
+      "grad_norm": 0.38809726070928446,
+      "learning_rate": 7.34140582704872e-05,
+      "loss": 0.3912,
+      "step": 1251
+    },
+    {
+      "epoch": 1.3347547974413647,
+      "grad_norm": 0.38768792360594717,
+      "learning_rate": 7.339768337397156e-05,
+      "loss": 0.3976,
+      "step": 1252
+    },
+    {
+      "epoch": 1.335820895522388,
+      "grad_norm": 0.8280978007735911,
+      "learning_rate": 7.338128997687505e-05,
+      "loss": 0.3916,
+      "step": 1253
+    },
+    {
+      "epoch": 1.3368869936034116,
+      "grad_norm": 0.5768542990929353,
+      "learning_rate": 7.336487808827878e-05,
+      "loss": 0.4006,
+      "step": 1254
+    },
+    {
+      "epoch": 1.3379530916844349,
+      "grad_norm": 0.5660915592706494,
+      "learning_rate": 7.334844771727407e-05,
+      "loss": 0.3889,
+      "step": 1255
+    },
+    {
+      "epoch": 1.3390191897654584,
+      "grad_norm": 0.5472277393608467,
+      "learning_rate": 7.333199887296249e-05,
+      "loss": 0.3966,
+      "step": 1256
+    },
+    {
+      "epoch": 1.340085287846482,
+      "grad_norm": 0.591414528568585,
+      "learning_rate": 7.331553156445585e-05,
+      "loss": 0.3942,
+      "step": 1257
+    },
+    {
+      "epoch": 1.3411513859275053,
+      "grad_norm": 0.551915916206759,
+      "learning_rate": 7.329904580087618e-05,
+      "loss": 0.3981,
+      "step": 1258
+    },
+    {
+      "epoch": 1.3422174840085288,
+      "grad_norm": 0.5979571239616868,
+      "learning_rate": 7.328254159135575e-05,
+      "loss": 0.391,
+      "step": 1259
+    },
+    {
+      "epoch": 1.3432835820895521,
+      "grad_norm": 0.6293378397487646,
+      "learning_rate": 7.326601894503703e-05,
+      "loss": 0.39,
+      "step": 1260
+    },
+    {
+      "epoch": 1.3443496801705757,
+      "grad_norm": 0.6163722703678122,
+      "learning_rate": 7.324947787107267e-05,
+      "loss": 0.3967,
+      "step": 1261
+    },
+    {
+      "epoch": 1.3454157782515992,
+      "grad_norm": 0.5746957535175811,
+      "learning_rate": 7.323291837862561e-05,
+      "loss": 0.3948,
+      "step": 1262
+    },
+    {
+      "epoch": 1.3464818763326227,
+      "grad_norm": 0.5421145237930761,
+      "learning_rate": 7.321634047686895e-05,
+      "loss": 0.3914,
+      "step": 1263
+    },
+    {
+      "epoch": 1.347547974413646,
+      "grad_norm": 0.5487443991159977,
+      "learning_rate": 7.319974417498594e-05,
+      "loss": 0.3895,
+      "step": 1264
+    },
+    {
+      "epoch": 1.3486140724946696,
+      "grad_norm": 0.5644289944018167,
+      "learning_rate": 7.318312948217012e-05,
+      "loss": 0.3928,
+      "step": 1265
+    },
+    {
+      "epoch": 1.349680170575693,
+      "grad_norm": 0.6211957603031802,
+      "learning_rate": 7.316649640762515e-05,
+      "loss": 0.3914,
+      "step": 1266
+    },
+    {
+      "epoch": 1.3507462686567164,
+      "grad_norm": 0.6243920858668569,
+      "learning_rate": 7.314984496056487e-05,
+      "loss": 0.3918,
+      "step": 1267
+    },
+    {
+      "epoch": 1.35181236673774,
+      "grad_norm": 0.48121327241122974,
+      "learning_rate": 7.313317515021334e-05,
+      "loss": 0.3949,
+      "step": 1268
+    },
+    {
+      "epoch": 1.3528784648187633,
+      "grad_norm": 0.42705279500592713,
+      "learning_rate": 7.311648698580475e-05,
+      "loss": 0.3914,
+      "step": 1269
+    },
+    {
+      "epoch": 1.3539445628997868,
+      "grad_norm": 0.4868001076889179,
+      "learning_rate": 7.309978047658348e-05,
+      "loss": 0.3969,
+      "step": 1270
+    },
+    {
+      "epoch": 1.3550106609808101,
+      "grad_norm": 0.5141089848205457,
+      "learning_rate": 7.308305563180409e-05,
+      "loss": 0.3947,
+      "step": 1271
+    },
+    {
+      "epoch": 1.3560767590618337,
+      "grad_norm": 0.5120767848632203,
+      "learning_rate": 7.306631246073125e-05,
+      "loss": 0.393,
+      "step": 1272
+    },
+    {
+      "epoch": 1.3571428571428572,
+      "grad_norm": 0.5231030391185801,
+      "learning_rate": 7.30495509726398e-05,
+      "loss": 0.3897,
+      "step": 1273
+    },
+    {
+      "epoch": 1.3582089552238805,
+      "grad_norm": 0.43134602113074105,
+      "learning_rate": 7.303277117681475e-05,
+      "loss": 0.3955,
+      "step": 1274
+    },
+    {
+      "epoch": 1.359275053304904,
+      "grad_norm": 0.2918563082914735,
+      "learning_rate": 7.301597308255124e-05,
+      "loss": 0.3933,
+      "step": 1275
+    },
+    {
+      "epoch": 1.3603411513859274,
+      "grad_norm": 0.2886464392254564,
+      "learning_rate": 7.299915669915454e-05,
+      "loss": 0.3947,
+      "step": 1276
+    },
+    {
+      "epoch": 1.361407249466951,
+      "grad_norm": 0.3719373094507802,
+      "learning_rate": 7.298232203594003e-05,
+      "loss": 0.3972,
+      "step": 1277
+    },
+    {
+      "epoch": 1.3624733475479744,
+      "grad_norm": 0.3893538092048672,
+      "learning_rate": 7.296546910223327e-05,
+      "loss": 0.3976,
+      "step": 1278
+    },
+    {
+      "epoch": 1.363539445628998,
+      "grad_norm": 0.33287474885809093,
+      "learning_rate": 7.294859790736989e-05,
+      "loss": 0.3899,
+      "step": 1279
+    },
+    {
+      "epoch": 1.3646055437100213,
+      "grad_norm": 0.44445251040661327,
+      "learning_rate": 7.293170846069564e-05,
+      "loss": 0.3982,
+      "step": 1280
+    },
+    {
+      "epoch": 1.3656716417910448,
+      "grad_norm": 0.5752395269544717,
+      "learning_rate": 7.291480077156642e-05,
+      "loss": 0.3968,
+      "step": 1281
+    },
+    {
+      "epoch": 1.3667377398720681,
+      "grad_norm": 0.5884059035360013,
+      "learning_rate": 7.289787484934823e-05,
+      "loss": 0.387,
+      "step": 1282
+    },
+    {
+      "epoch": 1.3678038379530917,
+      "grad_norm": 0.6411711941306089,
+      "learning_rate": 7.288093070341709e-05,
+      "loss": 0.3979,
+      "step": 1283
+    },
+    {
+      "epoch": 1.3688699360341152,
+      "grad_norm": 0.7297554798439111,
+      "learning_rate": 7.286396834315925e-05,
+      "loss": 0.3975,
+      "step": 1284
+    },
+    {
+      "epoch": 1.3699360341151385,
+      "grad_norm": 0.8133268333459146,
+      "learning_rate": 7.284698777797091e-05,
+      "loss": 0.3958,
+      "step": 1285
+    },
+    {
+      "epoch": 1.371002132196162,
+      "grad_norm": 0.8171389709979388,
+      "learning_rate": 7.282998901725846e-05,
+      "loss": 0.3935,
+      "step": 1286
+    },
+    {
+      "epoch": 1.3720682302771854,
+      "grad_norm": 0.6715918102841708,
+      "learning_rate": 7.281297207043832e-05,
+      "loss": 0.3963,
+      "step": 1287
+    },
+    {
+      "epoch": 1.373134328358209,
+      "grad_norm": 0.5821302267603919,
+      "learning_rate": 7.279593694693698e-05,
+      "loss": 0.3957,
+      "step": 1288
+    },
+    {
+      "epoch": 1.3742004264392325,
+      "grad_norm": 0.5793056879165892,
+      "learning_rate": 7.277888365619104e-05,
+      "loss": 0.399,
+      "step": 1289
+    },
+    {
+      "epoch": 1.375266524520256,
+      "grad_norm": 0.6131326531548551,
+      "learning_rate": 7.276181220764713e-05,
+      "loss": 0.391,
+      "step": 1290
+    },
+    {
+      "epoch": 1.3763326226012793,
+      "grad_norm": 0.6655360716979257,
+      "learning_rate": 7.274472261076192e-05,
+      "loss": 0.3936,
+      "step": 1291
+    },
+    {
+      "epoch": 1.3773987206823028,
+      "grad_norm": 0.6514488453949313,
+      "learning_rate": 7.272761487500219e-05,
+      "loss": 0.389,
+      "step": 1292
+    },
+    {
+      "epoch": 1.3784648187633262,
+      "grad_norm": 0.6004752538724129,
+      "learning_rate": 7.271048900984473e-05,
+      "loss": 0.3979,
+      "step": 1293
+    },
+    {
+      "epoch": 1.3795309168443497,
+      "grad_norm": 0.5286751186792054,
+      "learning_rate": 7.269334502477636e-05,
+      "loss": 0.3975,
+      "step": 1294
+    },
+    {
+      "epoch": 1.3805970149253732,
+      "grad_norm": 0.3782372078245128,
+      "learning_rate": 7.267618292929398e-05,
+      "loss": 0.3965,
+      "step": 1295
+    },
+    {
+      "epoch": 1.3816631130063965,
+      "grad_norm": 0.4478939359799176,
+      "learning_rate": 7.265900273290448e-05,
+      "loss": 0.3882,
+      "step": 1296
+    },
+    {
+      "epoch": 1.38272921108742,
+      "grad_norm": 0.40504400102254523,
+      "learning_rate": 7.264180444512481e-05,
+      "loss": 0.3979,
+      "step": 1297
+    },
+    {
+      "epoch": 1.3837953091684434,
+      "grad_norm": 0.38764841393473026,
+      "learning_rate": 7.262458807548191e-05,
+      "loss": 0.3963,
+      "step": 1298
+    },
+    {
+      "epoch": 1.384861407249467,
+      "grad_norm": 0.3656453484587496,
+      "learning_rate": 7.260735363351278e-05,
+      "loss": 0.3938,
+      "step": 1299
+    },
+    {
+      "epoch": 1.3859275053304905,
+      "grad_norm": 0.48752680564499556,
+      "learning_rate": 7.259010112876437e-05,
+      "loss": 0.3929,
+      "step": 1300
+    },
+    {
+      "epoch": 1.3869936034115138,
+      "grad_norm": 0.5660988157671498,
+      "learning_rate": 7.257283057079371e-05,
+      "loss": 0.3965,
+      "step": 1301
+    },
+    {
+      "epoch": 1.3880597014925373,
+      "grad_norm": 0.5415116199650241,
+      "learning_rate": 7.255554196916777e-05,
+      "loss": 0.3969,
+      "step": 1302
+    },
+    {
+      "epoch": 1.3891257995735606,
+      "grad_norm": 0.48579681088844257,
+      "learning_rate": 7.253823533346353e-05,
+      "loss": 0.3963,
+      "step": 1303
+    },
+    {
+      "epoch": 1.3901918976545842,
+      "grad_norm": 0.4521402830070269,
+      "learning_rate": 7.2520910673268e-05,
+      "loss": 0.3889,
+      "step": 1304
+    },
+    {
+      "epoch": 1.3912579957356077,
+      "grad_norm": 0.3868095009917062,
+      "learning_rate": 7.250356799817811e-05,
+      "loss": 0.3847,
+      "step": 1305
+    },
+    {
+      "epoch": 1.3923240938166312,
+      "grad_norm": 0.3772982157721458,
+      "learning_rate": 7.24862073178008e-05,
+      "loss": 0.3898,
+      "step": 1306
+    },
+    {
+      "epoch": 1.3933901918976546,
+      "grad_norm": 0.3776649103582633,
+      "learning_rate": 7.2468828641753e-05,
+      "loss": 0.3917,
+      "step": 1307
+    },
+    {
+      "epoch": 1.394456289978678,
+      "grad_norm": 0.3824919655119669,
+      "learning_rate": 7.245143197966158e-05,
+      "loss": 0.3938,
+      "step": 1308
+    },
+    {
+      "epoch": 1.3955223880597014,
+      "grad_norm": 0.5015303372462967,
+      "learning_rate": 7.243401734116341e-05,
+      "loss": 0.3896,
+      "step": 1309
+    },
+    {
+      "epoch": 1.396588486140725,
+      "grad_norm": 0.594587592040505,
+      "learning_rate": 7.241658473590526e-05,
+      "loss": 0.3927,
+      "step": 1310
+    },
+    {
+      "epoch": 1.3976545842217485,
+      "grad_norm": 0.560051836440119,
+      "learning_rate": 7.239913417354393e-05,
+      "loss": 0.3956,
+      "step": 1311
+    },
+    {
+      "epoch": 1.3987206823027718,
+      "grad_norm": 0.4940699663851546,
+      "learning_rate": 7.238166566374607e-05,
+      "loss": 0.3976,
+      "step": 1312
+    },
+    {
+      "epoch": 1.3997867803837953,
+      "grad_norm": 0.4683362628695071,
+      "learning_rate": 7.236417921618839e-05,
+      "loss": 0.3934,
+      "step": 1313
+    },
+    {
+      "epoch": 1.4008528784648187,
+      "grad_norm": 0.46229731243535377,
+      "learning_rate": 7.234667484055742e-05,
+      "loss": 0.3897,
+      "step": 1314
+    },
+    {
+      "epoch": 1.4019189765458422,
+      "grad_norm": 0.4363083782167336,
+      "learning_rate": 7.232915254654968e-05,
+      "loss": 0.3925,
+      "step": 1315
+    },
+    {
+      "epoch": 1.4029850746268657,
+      "grad_norm": 0.4257678057099821,
+      "learning_rate": 7.231161234387165e-05,
+      "loss": 0.3926,
+      "step": 1316
+    },
+    {
+      "epoch": 1.4040511727078893,
+      "grad_norm": 0.4846979047099743,
+      "learning_rate": 7.229405424223967e-05,
+      "loss": 0.3913,
+      "step": 1317
+    },
+    {
+      "epoch": 1.4051172707889126,
+      "grad_norm": 0.5325656515388444,
+      "learning_rate": 7.227647825137998e-05,
+      "loss": 0.3922,
+      "step": 1318
+    },
+    {
+      "epoch": 1.4061833688699361,
+      "grad_norm": 0.544343816677594,
+      "learning_rate": 7.225888438102882e-05,
+      "loss": 0.3888,
+      "step": 1319
+    },
+    {
+      "epoch": 1.4072494669509594,
+      "grad_norm": 0.5779022828781076,
+      "learning_rate": 7.224127264093225e-05,
+      "loss": 0.4001,
+      "step": 1320
+    },
+    {
+      "epoch": 1.408315565031983,
+      "grad_norm": 0.7411524315870592,
+      "learning_rate": 7.222364304084627e-05,
+      "loss": 0.3919,
+      "step": 1321
+    },
+    {
+      "epoch": 1.4093816631130065,
+      "grad_norm": 0.9208203333396868,
+      "learning_rate": 7.220599559053676e-05,
+      "loss": 0.3985,
+      "step": 1322
+    },
+    {
+      "epoch": 1.4104477611940298,
+      "grad_norm": 0.9562178280419171,
+      "learning_rate": 7.218833029977948e-05,
+      "loss": 0.4012,
+      "step": 1323
+    },
+    {
+      "epoch": 1.4115138592750534,
+      "grad_norm": 0.8517901376976156,
+      "learning_rate": 7.217064717836009e-05,
+      "loss": 0.3941,
+      "step": 1324
+    },
+    {
+      "epoch": 1.4125799573560767,
+      "grad_norm": 0.6475331249527876,
+      "learning_rate": 7.215294623607414e-05,
+      "loss": 0.3952,
+      "step": 1325
+    },
+    {
+      "epoch": 1.4136460554371002,
+      "grad_norm": 0.4701406210795358,
+      "learning_rate": 7.213522748272699e-05,
+      "loss": 0.3912,
+      "step": 1326
+    },
+    {
+      "epoch": 1.4147121535181237,
+      "grad_norm": 0.519797694341973,
+      "learning_rate": 7.211749092813395e-05,
+      "loss": 0.3941,
+      "step": 1327
+    },
+    {
+      "epoch": 1.415778251599147,
+      "grad_norm": 0.5546483656534481,
+      "learning_rate": 7.20997365821201e-05,
+      "loss": 0.39,
+      "step": 1328
+    },
+    {
+      "epoch": 1.4168443496801706,
+      "grad_norm": 0.5322694569447141,
+      "learning_rate": 7.208196445452048e-05,
+      "loss": 0.3933,
+      "step": 1329
+    },
+    {
+      "epoch": 1.417910447761194,
+      "grad_norm": 0.5615390005781126,
+      "learning_rate": 7.20641745551799e-05,
+      "loss": 0.389,
+      "step": 1330
+    },
+    {
+      "epoch": 1.4189765458422174,
+      "grad_norm": 0.5359170050084836,
+      "learning_rate": 7.204636689395304e-05,
+      "loss": 0.3915,
+      "step": 1331
+    },
+    {
+      "epoch": 1.420042643923241,
+      "grad_norm": 0.46419011837159624,
+      "learning_rate": 7.202854148070443e-05,
+      "loss": 0.3879,
+      "step": 1332
+    },
+    {
+      "epoch": 1.4211087420042645,
+      "grad_norm": 0.40341695125405613,
+      "learning_rate": 7.201069832530838e-05,
+      "loss": 0.3926,
+      "step": 1333
+    },
+    {
+      "epoch": 1.4221748400852878,
+      "grad_norm": 0.37929724565888073,
+      "learning_rate": 7.199283743764913e-05,
+      "loss": 0.3876,
+      "step": 1334
+    },
+    {
+      "epoch": 1.4232409381663114,
+      "grad_norm": 0.3708060382902423,
+      "learning_rate": 7.197495882762065e-05,
+      "loss": 0.394,
+      "step": 1335
+    },
+    {
+      "epoch": 1.4243070362473347,
+      "grad_norm": 0.42952036732595084,
+      "learning_rate": 7.195706250512676e-05,
+      "loss": 0.3911,
+      "step": 1336
+    },
+    {
+      "epoch": 1.4253731343283582,
+      "grad_norm": 0.3633415202092657,
+      "learning_rate": 7.19391484800811e-05,
+      "loss": 0.3881,
+      "step": 1337
+    },
+    {
+      "epoch": 1.4264392324093818,
+      "grad_norm": 0.3233744886500194,
+      "learning_rate": 7.192121676240713e-05,
+      "loss": 0.3969,
+      "step": 1338
+    },
+    {
+      "epoch": 1.427505330490405,
+      "grad_norm": 0.35414304904628113,
+      "learning_rate": 7.190326736203805e-05,
+      "loss": 0.3968,
+      "step": 1339
+    },
+    {
+      "epoch": 1.4285714285714286,
+      "grad_norm": 0.36418321869973685,
+      "learning_rate": 7.188530028891691e-05,
+      "loss": 0.3932,
+      "step": 1340
+    },
+    {
+      "epoch": 1.429637526652452,
+      "grad_norm": 0.32413601902229694,
+      "learning_rate": 7.186731555299654e-05,
+      "loss": 0.3936,
+      "step": 1341
+    },
+    {
+      "epoch": 1.4307036247334755,
+      "grad_norm": 0.3875245106191369,
+      "learning_rate": 7.184931316423955e-05,
+      "loss": 0.3956,
+      "step": 1342
+    },
+    {
+      "epoch": 1.431769722814499,
+      "grad_norm": 0.40455914476109633,
+      "learning_rate": 7.183129313261833e-05,
+      "loss": 0.393,
+      "step": 1343
+    },
+    {
+      "epoch": 1.4328358208955223,
+      "grad_norm": 0.4415785397804514,
+      "learning_rate": 7.181325546811503e-05,
+      "loss": 0.3961,
+      "step": 1344
+    },
+    {
+      "epoch": 1.4339019189765458,
+      "grad_norm": 0.5500841020492822,
+      "learning_rate": 7.179520018072158e-05,
+      "loss": 0.3954,
+      "step": 1345
+    },
+    {
+      "epoch": 1.4349680170575694,
+      "grad_norm": 0.6263260139567476,
+      "learning_rate": 7.177712728043967e-05,
+      "loss": 0.3981,
+      "step": 1346
+    },
+    {
+      "epoch": 1.4360341151385927,
+      "grad_norm": 0.7401499402434726,
+      "learning_rate": 7.175903677728077e-05,
+      "loss": 0.3974,
+      "step": 1347
+    },
+    {
+      "epoch": 1.4371002132196162,
+      "grad_norm": 0.8726338431226237,
+      "learning_rate": 7.174092868126604e-05,
+      "loss": 0.3871,
+      "step": 1348
+    },
+    {
+      "epoch": 1.4381663113006398,
+      "grad_norm": 0.8511811546689546,
+      "learning_rate": 7.172280300242648e-05,
+      "loss": 0.3935,
+      "step": 1349
+    },
+    {
+      "epoch": 1.439232409381663,
+      "grad_norm": 0.8021327414666732,
+      "learning_rate": 7.17046597508027e-05,
+      "loss": 0.3989,
+      "step": 1350
+    },
+    {
+      "epoch": 1.4402985074626866,
+      "grad_norm": 0.7465635016649216,
+      "learning_rate": 7.168649893644517e-05,
+      "loss": 0.3911,
+      "step": 1351
+    },
+    {
+      "epoch": 1.44136460554371,
+      "grad_norm": 0.722112156696109,
+      "learning_rate": 7.166832056941405e-05,
+      "loss": 0.4008,
+      "step": 1352
+    },
+    {
+      "epoch": 1.4424307036247335,
+      "grad_norm": 0.6928780087951198,
+      "learning_rate": 7.165012465977916e-05,
+      "loss": 0.3944,
+      "step": 1353
+    },
+    {
+      "epoch": 1.443496801705757,
+      "grad_norm": 0.5324488681456964,
+      "learning_rate": 7.163191121762012e-05,
+      "loss": 0.3923,
+      "step": 1354
+    },
+    {
+      "epoch": 1.4445628997867803,
+      "grad_norm": 0.37600301998098723,
+      "learning_rate": 7.161368025302622e-05,
+      "loss": 0.3931,
+      "step": 1355
+    },
+    {
+      "epoch": 1.4456289978678039,
+      "grad_norm": 0.43884431846324246,
+      "learning_rate": 7.159543177609648e-05,
+      "loss": 0.4019,
+      "step": 1356
+    },
+    {
+      "epoch": 1.4466950959488272,
+      "grad_norm": 0.4943768560008293,
+      "learning_rate": 7.15771657969396e-05,
+      "loss": 0.3918,
+      "step": 1357
+    },
+    {
+      "epoch": 1.4477611940298507,
+      "grad_norm": 0.5078112364255435,
+      "learning_rate": 7.155888232567396e-05,
+      "loss": 0.3931,
+      "step": 1358
+    },
+    {
+      "epoch": 1.4488272921108742,
+      "grad_norm": 0.5262670681715143,
+      "learning_rate": 7.15405813724277e-05,
+      "loss": 0.3901,
+      "step": 1359
+    },
+    {
+      "epoch": 1.4498933901918978,
+      "grad_norm": 0.44931799250014814,
+      "learning_rate": 7.152226294733857e-05,
+      "loss": 0.3928,
+      "step": 1360
+    },
+    {
+      "epoch": 1.450959488272921,
+      "grad_norm": 0.3766185113949834,
+      "learning_rate": 7.150392706055401e-05,
+      "loss": 0.3888,
+      "step": 1361
+    },
+    {
+      "epoch": 1.4520255863539446,
+      "grad_norm": 0.31545132670855125,
+      "learning_rate": 7.148557372223118e-05,
+      "loss": 0.3917,
+      "step": 1362
+    },
+    {
+      "epoch": 1.453091684434968,
+      "grad_norm": 0.4218667315766686,
+      "learning_rate": 7.146720294253687e-05,
+      "loss": 0.3958,
+      "step": 1363
+    },
+    {
+      "epoch": 1.4541577825159915,
+      "grad_norm": 0.5982279310643812,
+      "learning_rate": 7.14488147316475e-05,
+      "loss": 0.3923,
+      "step": 1364
+    },
+    {
+      "epoch": 1.455223880597015,
+      "grad_norm": 0.6848592048351014,
+      "learning_rate": 7.143040909974923e-05,
+      "loss": 0.3984,
+      "step": 1365
+    },
+    {
+      "epoch": 1.4562899786780383,
+      "grad_norm": 0.6819349249907524,
+      "learning_rate": 7.14119860570378e-05,
+      "loss": 0.3939,
+      "step": 1366
+    },
+    {
+      "epoch": 1.4573560767590619,
+      "grad_norm": 0.6440204606115185,
+      "learning_rate": 7.139354561371863e-05,
+      "loss": 0.3997,
+      "step": 1367
+    },
+    {
+      "epoch": 1.4584221748400852,
+      "grad_norm": 0.630295208498841,
+      "learning_rate": 7.137508778000676e-05,
+      "loss": 0.3948,
+      "step": 1368
+    },
+    {
+      "epoch": 1.4594882729211087,
+      "grad_norm": 0.5314317110342841,
+      "learning_rate": 7.135661256612688e-05,
+      "loss": 0.3889,
+      "step": 1369
+    },
+    {
+      "epoch": 1.4605543710021323,
+      "grad_norm": 0.34981486673823153,
+      "learning_rate": 7.133811998231327e-05,
+      "loss": 0.3935,
+      "step": 1370
+    },
+    {
+      "epoch": 1.4616204690831556,
+      "grad_norm": 0.35709223587817546,
+      "learning_rate": 7.131961003880989e-05,
+      "loss": 0.3934,
+      "step": 1371
+    },
+    {
+      "epoch": 1.462686567164179,
+      "grad_norm": 0.45371546795069073,
+      "learning_rate": 7.130108274587027e-05,
+      "loss": 0.3968,
+      "step": 1372
+    },
+    {
+      "epoch": 1.4637526652452024,
+      "grad_norm": 0.5253418087756354,
+      "learning_rate": 7.128253811375759e-05,
+      "loss": 0.394,
+      "step": 1373
+    },
+    {
+      "epoch": 1.464818763326226,
+      "grad_norm": 0.4991498770169075,
+      "learning_rate": 7.126397615274459e-05,
+      "loss": 0.3846,
+      "step": 1374
+    },
+    {
+      "epoch": 1.4658848614072495,
+      "grad_norm": 0.49528918882508394,
+      "learning_rate": 7.124539687311362e-05,
+      "loss": 0.3945,
+      "step": 1375
+    },
+    {
+      "epoch": 1.466950959488273,
+      "grad_norm": 0.5068485592396924,
+      "learning_rate": 7.122680028515668e-05,
+      "loss": 0.3959,
+      "step": 1376
+    },
+    {
+      "epoch": 1.4680170575692963,
+      "grad_norm": 0.5465623764731636,
+      "learning_rate": 7.120818639917527e-05,
+      "loss": 0.3983,
+      "step": 1377
+    },
+    {
+      "epoch": 1.4690831556503199,
+      "grad_norm": 0.6180418289803238,
+      "learning_rate": 7.118955522548053e-05,
+      "loss": 0.3906,
+      "step": 1378
+    },
+    {
+      "epoch": 1.4701492537313432,
+      "grad_norm": 0.6374392706839199,
+      "learning_rate": 7.117090677439317e-05,
+      "loss": 0.4016,
+      "step": 1379
+    },
+    {
+      "epoch": 1.4712153518123667,
+      "grad_norm": 0.5780281794791482,
+      "learning_rate": 7.115224105624346e-05,
+      "loss": 0.3891,
+      "step": 1380
+    },
+    {
+      "epoch": 1.4722814498933903,
+      "grad_norm": 0.5215481763118731,
+      "learning_rate": 7.113355808137122e-05,
+      "loss": 0.3937,
+      "step": 1381
+    },
+    {
+      "epoch": 1.4733475479744136,
+      "grad_norm": 0.5508793693931563,
+      "learning_rate": 7.111485786012588e-05,
+      "loss": 0.3993,
+      "step": 1382
+    },
+    {
+      "epoch": 1.4744136460554371,
+      "grad_norm": 0.6235158688200941,
+      "learning_rate": 7.109614040286636e-05,
+      "loss": 0.3909,
+      "step": 1383
+    },
+    {
+      "epoch": 1.4754797441364604,
+      "grad_norm": 0.6200848574811243,
+      "learning_rate": 7.107740571996118e-05,
+      "loss": 0.4002,
+      "step": 1384
+    },
+    {
+      "epoch": 1.476545842217484,
+      "grad_norm": 0.5904012350207049,
+      "learning_rate": 7.105865382178836e-05,
+      "loss": 0.3934,
+      "step": 1385
+    },
+    {
+      "epoch": 1.4776119402985075,
+      "grad_norm": 0.5289964058379455,
+      "learning_rate": 7.10398847187355e-05,
+      "loss": 0.3914,
+      "step": 1386
+    },
+    {
+      "epoch": 1.478678038379531,
+      "grad_norm": 0.443464050960444,
+      "learning_rate": 7.10210984211997e-05,
+      "loss": 0.392,
+      "step": 1387
+    },
+    {
+      "epoch": 1.4797441364605544,
+      "grad_norm": 0.4427955685061585,
+      "learning_rate": 7.100229493958757e-05,
+      "loss": 0.3969,
+      "step": 1388
+    },
+    {
+      "epoch": 1.480810234541578,
+      "grad_norm": 0.46393678647424846,
+      "learning_rate": 7.09834742843153e-05,
+      "loss": 0.3934,
+      "step": 1389
+    },
+    {
+      "epoch": 1.4818763326226012,
+      "grad_norm": 0.39212911960500413,
+      "learning_rate": 7.096463646580853e-05,
+      "loss": 0.3969,
+      "step": 1390
+    },
+    {
+      "epoch": 1.4829424307036247,
+      "grad_norm": 0.31518826807678046,
+      "learning_rate": 7.094578149450243e-05,
+      "loss": 0.3965,
+      "step": 1391
+    },
+    {
+      "epoch": 1.4840085287846483,
+      "grad_norm": 0.35137446772893605,
+      "learning_rate": 7.092690938084168e-05,
+      "loss": 0.394,
+      "step": 1392
+    },
+    {
+      "epoch": 1.4850746268656716,
+      "grad_norm": 0.43189084089863594,
+      "learning_rate": 7.090802013528047e-05,
+      "loss": 0.3907,
+      "step": 1393
+    },
+    {
+      "epoch": 1.4861407249466951,
+      "grad_norm": 0.4502661111740289,
+      "learning_rate": 7.088911376828241e-05,
+      "loss": 0.3946,
+      "step": 1394
+    },
+    {
+      "epoch": 1.4872068230277184,
+      "grad_norm": 0.44527023350907197,
+      "learning_rate": 7.087019029032071e-05,
+      "loss": 0.3977,
+      "step": 1395
+    },
+    {
+      "epoch": 1.488272921108742,
+      "grad_norm": 0.45794515563438,
+      "learning_rate": 7.085124971187794e-05,
+      "loss": 0.3974,
+      "step": 1396
+    },
+    {
+      "epoch": 1.4893390191897655,
+      "grad_norm": 0.46227336496153126,
+      "learning_rate": 7.083229204344623e-05,
+      "loss": 0.3861,
+      "step": 1397
+    },
+    {
+      "epoch": 1.4904051172707888,
+      "grad_norm": 0.4503424372012637,
+      "learning_rate": 7.081331729552712e-05,
+      "loss": 0.3951,
+      "step": 1398
+    },
+    {
+      "epoch": 1.4914712153518124,
+      "grad_norm": 0.4826116426207179,
+      "learning_rate": 7.079432547863164e-05,
+      "loss": 0.3883,
+      "step": 1399
+    },
+    {
+      "epoch": 1.4925373134328357,
+      "grad_norm": 0.526692112887762,
+      "learning_rate": 7.077531660328028e-05,
+      "loss": 0.3908,
+      "step": 1400
+    },
+    {
+      "epoch": 1.4936034115138592,
+      "grad_norm": 0.4637243917340167,
+      "learning_rate": 7.075629068000297e-05,
+      "loss": 0.3939,
+      "step": 1401
+    },
+    {
+      "epoch": 1.4946695095948828,
+      "grad_norm": 0.4348762733252528,
+      "learning_rate": 7.073724771933906e-05,
+      "loss": 0.3859,
+      "step": 1402
+    },
+    {
+      "epoch": 1.4957356076759063,
+      "grad_norm": 0.4277816665773673,
+      "learning_rate": 7.071818773183738e-05,
+      "loss": 0.3969,
+      "step": 1403
+    },
+    {
+      "epoch": 1.4968017057569296,
+      "grad_norm": 0.4767523309718055,
+      "learning_rate": 7.069911072805618e-05,
+      "loss": 0.3918,
+      "step": 1404
+    },
+    {
+      "epoch": 1.4978678038379531,
+      "grad_norm": 0.4905037191035888,
+      "learning_rate": 7.068001671856309e-05,
+      "loss": 0.3906,
+      "step": 1405
+    },
+    {
+      "epoch": 1.4989339019189765,
+      "grad_norm": 0.4362857431277552,
+      "learning_rate": 7.066090571393524e-05,
+      "loss": 0.3909,
+      "step": 1406
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.3290996660217568,
+      "learning_rate": 7.064177772475912e-05,
+      "loss": 0.3976,
+      "step": 1407
+    },
+    {
+      "epoch": 1.5010660980810235,
+      "grad_norm": 0.3744727365069323,
+      "learning_rate": 7.062263276163064e-05,
+      "loss": 0.3962,
+      "step": 1408
+    },
+    {
+      "epoch": 1.502132196162047,
+      "grad_norm": 0.5302924684211919,
+      "learning_rate": 7.060347083515511e-05,
+      "loss": 0.3847,
+      "step": 1409
+    },
+    {
+      "epoch": 1.5031982942430704,
+      "grad_norm": 0.636023047266614,
+      "learning_rate": 7.058429195594727e-05,
+      "loss": 0.3952,
+      "step": 1410
+    },
+    {
+      "epoch": 1.5042643923240937,
+      "grad_norm": 0.5944045338529551,
+      "learning_rate": 7.056509613463118e-05,
+      "loss": 0.3901,
+      "step": 1411
+    },
+    {
+      "epoch": 1.5053304904051172,
+      "grad_norm": 0.4426101595482755,
+      "learning_rate": 7.054588338184034e-05,
+      "loss": 0.3949,
+      "step": 1412
+    },
+    {
+      "epoch": 1.5063965884861408,
+      "grad_norm": 0.3582450742335668,
+      "learning_rate": 7.052665370821764e-05,
+      "loss": 0.3882,
+      "step": 1413
+    },
+    {
+      "epoch": 1.5074626865671643,
+      "grad_norm": 0.3954777880467866,
+      "learning_rate": 7.050740712441528e-05,
+      "loss": 0.3972,
+      "step": 1414
+    },
+    {
+      "epoch": 1.5085287846481876,
+      "grad_norm": 0.4416647029712508,
+      "learning_rate": 7.04881436410949e-05,
+      "loss": 0.39,
+      "step": 1415
+    },
+    {
+      "epoch": 1.509594882729211,
+      "grad_norm": 0.44707077361093805,
+      "learning_rate": 7.046886326892747e-05,
+      "loss": 0.3855,
+      "step": 1416
+    },
+    {
+      "epoch": 1.5106609808102345,
+      "grad_norm": 0.40402387918207455,
+      "learning_rate": 7.044956601859329e-05,
+      "loss": 0.3928,
+      "step": 1417
+    },
+    {
+      "epoch": 1.511727078891258,
+      "grad_norm": 0.31989191868470707,
+      "learning_rate": 7.043025190078205e-05,
+      "loss": 0.3919,
+      "step": 1418
+    },
+    {
+      "epoch": 1.5127931769722816,
+      "grad_norm": 0.2810822260181101,
+      "learning_rate": 7.041092092619277e-05,
+      "loss": 0.3852,
+      "step": 1419
+    },
+    {
+      "epoch": 1.5138592750533049,
+      "grad_norm": 0.38427721507681256,
+      "learning_rate": 7.039157310553378e-05,
+      "loss": 0.3975,
+      "step": 1420
+    },
+    {
+      "epoch": 1.5149253731343284,
+      "grad_norm": 0.43787056804713514,
+      "learning_rate": 7.03722084495228e-05,
+      "loss": 0.3986,
+      "step": 1421
+    },
+    {
+      "epoch": 1.5159914712153517,
+      "grad_norm": 0.39271707463092426,
+      "learning_rate": 7.035282696888684e-05,
+      "loss": 0.393,
+      "step": 1422
+    },
+    {
+      "epoch": 1.5170575692963753,
+      "grad_norm": 0.41954871590765713,
+      "learning_rate": 7.033342867436221e-05,
+      "loss": 0.3979,
+      "step": 1423
+    },
+    {
+      "epoch": 1.5181236673773988,
+      "grad_norm": 0.488489761777447,
+      "learning_rate": 7.031401357669456e-05,
+      "loss": 0.3881,
+      "step": 1424
+    },
+    {
+      "epoch": 1.5191897654584223,
+      "grad_norm": 0.5515408415736078,
+      "learning_rate": 7.029458168663887e-05,
+      "loss": 0.3862,
+      "step": 1425
+    },
+    {
+      "epoch": 1.5202558635394456,
+      "grad_norm": 0.6316116354397406,
+      "learning_rate": 7.027513301495937e-05,
+      "loss": 0.3889,
+      "step": 1426
+    },
+    {
+      "epoch": 1.521321961620469,
+      "grad_norm": 0.6866508216670858,
+      "learning_rate": 7.025566757242962e-05,
+      "loss": 0.3927,
+      "step": 1427
+    },
+    {
+      "epoch": 1.5223880597014925,
+      "grad_norm": 0.6861740365624607,
+      "learning_rate": 7.023618536983249e-05,
+      "loss": 0.3976,
+      "step": 1428
+    },
+    {
+      "epoch": 1.523454157782516,
+      "grad_norm": 0.6832689017816802,
+      "learning_rate": 7.021668641796008e-05,
+      "loss": 0.3935,
+      "step": 1429
+    },
+    {
+      "epoch": 1.5245202558635396,
+      "grad_norm": 0.678454190085777,
+      "learning_rate": 7.019717072761377e-05,
+      "loss": 0.3879,
+      "step": 1430
+    },
+    {
+      "epoch": 1.5255863539445629,
+      "grad_norm": 0.6514965485012345,
+      "learning_rate": 7.01776383096043e-05,
+      "loss": 0.3917,
+      "step": 1431
+    },
+    {
+      "epoch": 1.5266524520255862,
+      "grad_norm": 0.5811169015448752,
+      "learning_rate": 7.01580891747516e-05,
+      "loss": 0.3942,
+      "step": 1432
+    },
+    {
+      "epoch": 1.5277185501066097,
+      "grad_norm": 0.569097620467461,
+      "learning_rate": 7.013852333388483e-05,
+      "loss": 0.3955,
+      "step": 1433
+    },
+    {
+      "epoch": 1.5287846481876333,
+      "grad_norm": 0.4808827995992214,
+      "learning_rate": 7.011894079784248e-05,
+      "loss": 0.3908,
+      "step": 1434
+    },
+    {
+      "epoch": 1.5298507462686568,
+      "grad_norm": 0.3656185727390282,
+      "learning_rate": 7.009934157747227e-05,
+      "loss": 0.3946,
+      "step": 1435
+    },
+    {
+      "epoch": 1.5309168443496801,
+      "grad_norm": 0.5145348764394406,
+      "learning_rate": 7.007972568363112e-05,
+      "loss": 0.3882,
+      "step": 1436
+    },
+    {
+      "epoch": 1.5319829424307037,
+      "grad_norm": 0.6632114181696835,
+      "learning_rate": 7.006009312718525e-05,
+      "loss": 0.3912,
+      "step": 1437
+    },
+    {
+      "epoch": 1.533049040511727,
+      "grad_norm": 0.6714409788868576,
+      "learning_rate": 7.004044391901005e-05,
+      "loss": 0.3946,
+      "step": 1438
+    },
+    {
+      "epoch": 1.5341151385927505,
+      "grad_norm": 0.5938831067807631,
+      "learning_rate": 7.002077806999016e-05,
+      "loss": 0.3911,
+      "step": 1439
+    },
+    {
+      "epoch": 1.535181236673774,
+      "grad_norm": 0.5658150891169751,
+      "learning_rate": 7.000109559101944e-05,
+      "loss": 0.39,
+      "step": 1440
+    },
+    {
+      "epoch": 1.5362473347547976,
+      "grad_norm": 0.5719079089059944,
+      "learning_rate": 6.998139649300097e-05,
+      "loss": 0.3963,
+      "step": 1441
+    },
+    {
+      "epoch": 1.537313432835821,
+      "grad_norm": 0.4711947234722605,
+      "learning_rate": 6.996168078684702e-05,
+      "loss": 0.3909,
+      "step": 1442
+    },
+    {
+      "epoch": 1.5383795309168442,
+      "grad_norm": 0.4358846322039788,
+      "learning_rate": 6.994194848347908e-05,
+      "loss": 0.3933,
+      "step": 1443
+    },
+    {
+      "epoch": 1.5394456289978677,
+      "grad_norm": 0.6131004479680616,
+      "learning_rate": 6.99221995938278e-05,
+      "loss": 0.3995,
+      "step": 1444
+    },
+    {
+      "epoch": 1.5405117270788913,
+      "grad_norm": 0.6962404795880364,
+      "learning_rate": 6.990243412883304e-05,
+      "loss": 0.3925,
+      "step": 1445
+    },
+    {
+      "epoch": 1.5415778251599148,
+      "grad_norm": 0.6694951018990504,
+      "learning_rate": 6.988265209944387e-05,
+      "loss": 0.3892,
+      "step": 1446
+    },
+    {
+      "epoch": 1.5426439232409381,
+      "grad_norm": 0.5156094912368259,
+      "learning_rate": 6.986285351661847e-05,
+      "loss": 0.3892,
+      "step": 1447
+    },
+    {
+      "epoch": 1.5437100213219617,
+      "grad_norm": 0.41423025778526606,
+      "learning_rate": 6.984303839132425e-05,
+      "loss": 0.3964,
+      "step": 1448
+    },
+    {
+      "epoch": 1.544776119402985,
+      "grad_norm": 0.3696131621291919,
+      "learning_rate": 6.982320673453773e-05,
+      "loss": 0.3905,
+      "step": 1449
+    },
+    {
+      "epoch": 1.5458422174840085,
+      "grad_norm": 0.37566704365434295,
+      "learning_rate": 6.980335855724465e-05,
+      "loss": 0.3898,
+      "step": 1450
+    },
+    {
+      "epoch": 1.546908315565032,
+      "grad_norm": 0.44656110415033884,
+      "learning_rate": 6.978349387043986e-05,
+      "loss": 0.3954,
+      "step": 1451
+    },
+    {
+      "epoch": 1.5479744136460556,
+      "grad_norm": 0.5097702199954142,
+      "learning_rate": 6.976361268512735e-05,
+      "loss": 0.3916,
+      "step": 1452
+    },
+    {
+      "epoch": 1.549040511727079,
+      "grad_norm": 0.6115774145379382,
+      "learning_rate": 6.974371501232027e-05,
+      "loss": 0.4001,
+      "step": 1453
+    },
+    {
+      "epoch": 1.5501066098081022,
+      "grad_norm": 0.6298592763126765,
+      "learning_rate": 6.97238008630409e-05,
+      "loss": 0.3913,
+      "step": 1454
+    },
+    {
+      "epoch": 1.5511727078891258,
+      "grad_norm": 0.5640252014215751,
+      "learning_rate": 6.970387024832066e-05,
+      "loss": 0.3972,
+      "step": 1455
+    },
+    {
+      "epoch": 1.5522388059701493,
+      "grad_norm": 0.5637434287286831,
+      "learning_rate": 6.968392317920005e-05,
+      "loss": 0.3914,
+      "step": 1456
+    },
+    {
+      "epoch": 1.5533049040511728,
+      "grad_norm": 0.5758115244979767,
+      "learning_rate": 6.96639596667287e-05,
+      "loss": 0.3888,
+      "step": 1457
+    },
+    {
+      "epoch": 1.5543710021321961,
+      "grad_norm": 0.579669625426786,
+      "learning_rate": 6.96439797219654e-05,
+      "loss": 0.3964,
+      "step": 1458
+    },
+    {
+      "epoch": 1.5554371002132195,
+      "grad_norm": 0.5774000389217503,
+      "learning_rate": 6.962398335597798e-05,
+      "loss": 0.391,
+      "step": 1459
+    },
+    {
+      "epoch": 1.556503198294243,
+      "grad_norm": 0.5665645013725628,
+      "learning_rate": 6.960397057984336e-05,
+      "loss": 0.3925,
+      "step": 1460
+    },
+    {
+      "epoch": 1.5575692963752665,
+      "grad_norm": 0.5243134790028094,
+      "learning_rate": 6.958394140464761e-05,
+      "loss": 0.3931,
+      "step": 1461
+    },
+    {
+      "epoch": 1.55863539445629,
+      "grad_norm": 0.5072252071927075,
+      "learning_rate": 6.956389584148586e-05,
+      "loss": 0.3874,
+      "step": 1462
+    },
+    {
+      "epoch": 1.5597014925373134,
+      "grad_norm": 0.42649762130370344,
+      "learning_rate": 6.954383390146228e-05,
+      "loss": 0.3902,
+      "step": 1463
+    },
+    {
+      "epoch": 1.560767590618337,
+      "grad_norm": 0.33264595497577854,
+      "learning_rate": 6.952375559569016e-05,
+      "loss": 0.3863,
+      "step": 1464
+    },
+    {
+      "epoch": 1.5618336886993602,
+      "grad_norm": 0.3759252125909897,
+      "learning_rate": 6.950366093529184e-05,
+      "loss": 0.393,
+      "step": 1465
+    },
+    {
+      "epoch": 1.5628997867803838,
+      "grad_norm": 0.4265770921364556,
+      "learning_rate": 6.94835499313987e-05,
+      "loss": 0.3852,
+      "step": 1466
+    },
+    {
+      "epoch": 1.5639658848614073,
+      "grad_norm": 0.4579903850772894,
+      "learning_rate": 6.946342259515122e-05,
+      "loss": 0.3911,
+      "step": 1467
+    },
+    {
+      "epoch": 1.5650319829424308,
+      "grad_norm": 0.5479175031792544,
+      "learning_rate": 6.944327893769887e-05,
+      "loss": 0.3962,
+      "step": 1468
+    },
+    {
+      "epoch": 1.5660980810234542,
+      "grad_norm": 0.6055176521095669,
+      "learning_rate": 6.94231189702002e-05,
+      "loss": 0.3864,
+      "step": 1469
+    },
+    {
+      "epoch": 1.5671641791044775,
+      "grad_norm": 0.6727380474108113,
+      "learning_rate": 6.940294270382278e-05,
+      "loss": 0.3921,
+      "step": 1470
+    },
+    {
+      "epoch": 1.568230277185501,
+      "grad_norm": 0.6356621826960686,
+      "learning_rate": 6.938275014974323e-05,
+      "loss": 0.3947,
+      "step": 1471
+    },
+    {
+      "epoch": 1.5692963752665245,
+      "grad_norm": 0.5243747711823524,
+      "learning_rate": 6.936254131914717e-05,
+      "loss": 0.3959,
+      "step": 1472
+    },
+    {
+      "epoch": 1.570362473347548,
+      "grad_norm": 0.46811482341260074,
+      "learning_rate": 6.934231622322923e-05,
+      "loss": 0.3904,
+      "step": 1473
+    },
+    {
+      "epoch": 1.5714285714285714,
+      "grad_norm": 0.3465924585456931,
+      "learning_rate": 6.932207487319305e-05,
+      "loss": 0.393,
+      "step": 1474
+    },
+    {
+      "epoch": 1.572494669509595,
+      "grad_norm": 0.3025626410259389,
+      "learning_rate": 6.930181728025133e-05,
+      "loss": 0.3925,
+      "step": 1475
+    },
+    {
+      "epoch": 1.5735607675906182,
+      "grad_norm": 0.391223961099086,
+      "learning_rate": 6.928154345562569e-05,
+      "loss": 0.39,
+      "step": 1476
+    },
+    {
+      "epoch": 1.5746268656716418,
+      "grad_norm": 0.41487388189624425,
+      "learning_rate": 6.926125341054676e-05,
+      "loss": 0.3924,
+      "step": 1477
+    },
+    {
+      "epoch": 1.5756929637526653,
+      "grad_norm": 0.45859115746381557,
+      "learning_rate": 6.92409471562542e-05,
+      "loss": 0.3948,
+      "step": 1478
+    },
+    {
+      "epoch": 1.5767590618336889,
+      "grad_norm": 0.5136700870287039,
+      "learning_rate": 6.922062470399663e-05,
+      "loss": 0.3861,
+      "step": 1479
+    },
+    {
+      "epoch": 1.5778251599147122,
+      "grad_norm": 0.5017752722816414,
+      "learning_rate": 6.920028606503161e-05,
+      "loss": 0.3912,
+      "step": 1480
+    },
+    {
+      "epoch": 1.5788912579957355,
+      "grad_norm": 0.46815074536424645,
+      "learning_rate": 6.91799312506257e-05,
+      "loss": 0.3923,
+      "step": 1481
+    },
+    {
+      "epoch": 1.579957356076759,
+      "grad_norm": 0.4683376068962054,
+      "learning_rate": 6.915956027205438e-05,
+      "loss": 0.3891,
+      "step": 1482
+    },
+    {
+      "epoch": 1.5810234541577826,
+      "grad_norm": 0.3838678721557934,
+      "learning_rate": 6.913917314060215e-05,
+      "loss": 0.3884,
+      "step": 1483
+    },
+    {
+      "epoch": 1.582089552238806,
+      "grad_norm": 0.28512465563571143,
+      "learning_rate": 6.911876986756241e-05,
+      "loss": 0.3946,
+      "step": 1484
+    },
+    {
+      "epoch": 1.5831556503198294,
+      "grad_norm": 0.2798136165240047,
+      "learning_rate": 6.90983504642375e-05,
+      "loss": 0.3925,
+      "step": 1485
+    },
+    {
+      "epoch": 1.5842217484008527,
+      "grad_norm": 0.32595231025368043,
+      "learning_rate": 6.907791494193873e-05,
+      "loss": 0.3933,
+      "step": 1486
+    },
+    {
+      "epoch": 1.5852878464818763,
+      "grad_norm": 0.33263930837599737,
+      "learning_rate": 6.905746331198631e-05,
+      "loss": 0.3882,
+      "step": 1487
+    },
+    {
+      "epoch": 1.5863539445628998,
+      "grad_norm": 0.3001000433354499,
+      "learning_rate": 6.903699558570935e-05,
+      "loss": 0.3944,
+      "step": 1488
+    },
+    {
+      "epoch": 1.5874200426439233,
+      "grad_norm": 0.33691283404723554,
+      "learning_rate": 6.901651177444596e-05,
+      "loss": 0.3901,
+      "step": 1489
+    },
+    {
+      "epoch": 1.5884861407249466,
+      "grad_norm": 0.36842928704561506,
+      "learning_rate": 6.899601188954306e-05,
+      "loss": 0.3891,
+      "step": 1490
+    },
+    {
+      "epoch": 1.5895522388059702,
+      "grad_norm": 0.37394858126001956,
+      "learning_rate": 6.897549594235654e-05,
+      "loss": 0.3964,
+      "step": 1491
+    },
+    {
+      "epoch": 1.5906183368869935,
+      "grad_norm": 0.3288296151355204,
+      "learning_rate": 6.895496394425118e-05,
+      "loss": 0.3894,
+      "step": 1492
+    },
+    {
+      "epoch": 1.591684434968017,
+      "grad_norm": 0.33186383529359254,
+      "learning_rate": 6.893441590660064e-05,
+      "loss": 0.3964,
+      "step": 1493
+    },
+    {
+      "epoch": 1.5927505330490406,
+      "grad_norm": 0.32170536578140646,
+      "learning_rate": 6.891385184078744e-05,
+      "loss": 0.3892,
+      "step": 1494
+    },
+    {
+      "epoch": 1.593816631130064,
+      "grad_norm": 0.34823720165895466,
+      "learning_rate": 6.889327175820302e-05,
+      "loss": 0.3912,
+      "step": 1495
+    },
+    {
+      "epoch": 1.5948827292110874,
+      "grad_norm": 0.36827171478815857,
+      "learning_rate": 6.887267567024767e-05,
+      "loss": 0.3877,
+      "step": 1496
+    },
+    {
+      "epoch": 1.5959488272921107,
+      "grad_norm": 0.324329055251669,
+      "learning_rate": 6.885206358833056e-05,
+      "loss": 0.3907,
+      "step": 1497
+    },
+    {
+      "epoch": 1.5970149253731343,
+      "grad_norm": 0.3724585692258156,
+      "learning_rate": 6.883143552386971e-05,
+      "loss": 0.3882,
+      "step": 1498
+    },
+    {
+      "epoch": 1.5980810234541578,
+      "grad_norm": 0.3726849206910898,
+      "learning_rate": 6.881079148829198e-05,
+      "loss": 0.3905,
+      "step": 1499
+    },
+    {
+      "epoch": 1.5991471215351813,
+      "grad_norm": 0.37495710080763744,
+      "learning_rate": 6.879013149303312e-05,
+      "loss": 0.3935,
+      "step": 1500
+    },
+    {
+      "epoch": 1.6002132196162047,
+      "grad_norm": 0.5021479834874241,
+      "learning_rate": 6.876945554953767e-05,
+      "loss": 0.3925,
+      "step": 1501
+    },
+    {
+      "epoch": 1.6012793176972282,
+      "grad_norm": 0.6598112081849651,
+      "learning_rate": 6.874876366925904e-05,
+      "loss": 0.3917,
+      "step": 1502
+    },
+    {
+      "epoch": 1.6023454157782515,
+      "grad_norm": 0.7759163516332811,
+      "learning_rate": 6.872805586365947e-05,
+      "loss": 0.3885,
+      "step": 1503
+    },
+    {
+      "epoch": 1.603411513859275,
+      "grad_norm": 0.8164276907570536,
+      "learning_rate": 6.870733214420998e-05,
+      "loss": 0.3928,
+      "step": 1504
+    },
+    {
+      "epoch": 1.6044776119402986,
+      "grad_norm": 0.8744103848461371,
+      "learning_rate": 6.868659252239045e-05,
+      "loss": 0.394,
+      "step": 1505
+    },
+    {
+      "epoch": 1.6055437100213221,
+      "grad_norm": 0.8615414040135513,
+      "learning_rate": 6.866583700968954e-05,
+      "loss": 0.3838,
+      "step": 1506
+    },
+    {
+      "epoch": 1.6066098081023454,
+      "grad_norm": 0.7600225280740014,
+      "learning_rate": 6.864506561760474e-05,
+      "loss": 0.3963,
+      "step": 1507
+    },
+    {
+      "epoch": 1.6076759061833688,
+      "grad_norm": 0.5613832665703143,
+      "learning_rate": 6.862427835764231e-05,
+      "loss": 0.3995,
+      "step": 1508
+    },
+    {
+      "epoch": 1.6087420042643923,
+      "grad_norm": 0.3881251295902559,
+      "learning_rate": 6.860347524131733e-05,
+      "loss": 0.3852,
+      "step": 1509
+    },
+    {
+      "epoch": 1.6098081023454158,
+      "grad_norm": 0.3100627174569291,
+      "learning_rate": 6.858265628015362e-05,
+      "loss": 0.3914,
+      "step": 1510
+    },
+    {
+      "epoch": 1.6108742004264394,
+      "grad_norm": 0.3947566702558035,
+      "learning_rate": 6.856182148568382e-05,
+      "loss": 0.3961,
+      "step": 1511
+    },
+    {
+      "epoch": 1.6119402985074627,
+      "grad_norm": 0.49178390954875206,
+      "learning_rate": 6.854097086944932e-05,
+      "loss": 0.3941,
+      "step": 1512
+    },
+    {
+      "epoch": 1.613006396588486,
+      "grad_norm": 0.5414001966948181,
+      "learning_rate": 6.852010444300028e-05,
+      "loss": 0.3926,
+      "step": 1513
+    },
+    {
+      "epoch": 1.6140724946695095,
+      "grad_norm": 0.49531130512745414,
+      "learning_rate": 6.849922221789561e-05,
+      "loss": 0.3915,
+      "step": 1514
+    },
+    {
+      "epoch": 1.615138592750533,
+      "grad_norm": 0.41543603344069213,
+      "learning_rate": 6.847832420570298e-05,
+      "loss": 0.3883,
+      "step": 1515
+    },
+    {
+      "epoch": 1.6162046908315566,
+      "grad_norm": 0.33242146656919136,
+      "learning_rate": 6.84574104179988e-05,
+      "loss": 0.3948,
+      "step": 1516
+    },
+    {
+      "epoch": 1.61727078891258,
+      "grad_norm": 0.3892193202210671,
+      "learning_rate": 6.843648086636822e-05,
+      "loss": 0.3943,
+      "step": 1517
+    },
+    {
+      "epoch": 1.6183368869936035,
+      "grad_norm": 0.5073950555317511,
+      "learning_rate": 6.841553556240514e-05,
+      "loss": 0.3899,
+      "step": 1518
+    },
+    {
+      "epoch": 1.6194029850746268,
+      "grad_norm": 0.543311094919119,
+      "learning_rate": 6.839457451771214e-05,
+      "loss": 0.3915,
+      "step": 1519
+    },
+    {
+      "epoch": 1.6204690831556503,
+      "grad_norm": 0.5283833705504892,
+      "learning_rate": 6.837359774390058e-05,
+      "loss": 0.3874,
+      "step": 1520
+    },
+    {
+      "epoch": 1.6215351812366738,
+      "grad_norm": 0.5077032134792163,
+      "learning_rate": 6.835260525259048e-05,
+      "loss": 0.3904,
+      "step": 1521
+    },
+    {
+      "epoch": 1.6226012793176974,
+      "grad_norm": 0.5106764897990909,
+      "learning_rate": 6.83315970554106e-05,
+      "loss": 0.3898,
+      "step": 1522
+    },
+    {
+      "epoch": 1.6236673773987207,
+      "grad_norm": 0.5286416559892751,
+      "learning_rate": 6.831057316399839e-05,
+      "loss": 0.3886,
+      "step": 1523
+    },
+    {
+      "epoch": 1.624733475479744,
+      "grad_norm": 0.409097737243627,
+      "learning_rate": 6.828953358999998e-05,
+      "loss": 0.3899,
+      "step": 1524
+    },
+    {
+      "epoch": 1.6257995735607675,
+      "grad_norm": 0.3213269166849657,
+      "learning_rate": 6.826847834507024e-05,
+      "loss": 0.3892,
+      "step": 1525
+    },
+    {
+      "epoch": 1.626865671641791,
+      "grad_norm": 0.3126601387070219,
+      "learning_rate": 6.824740744087262e-05,
+      "loss": 0.3861,
+      "step": 1526
+    },
+    {
+      "epoch": 1.6279317697228146,
+      "grad_norm": 0.39308792247161223,
+      "learning_rate": 6.822632088907937e-05,
+      "loss": 0.3912,
+      "step": 1527
+    },
+    {
+      "epoch": 1.628997867803838,
+      "grad_norm": 0.48501314817610786,
+      "learning_rate": 6.820521870137129e-05,
+      "loss": 0.3901,
+      "step": 1528
+    },
+    {
+      "epoch": 1.6300639658848612,
+      "grad_norm": 0.5074473205522086,
+      "learning_rate": 6.818410088943791e-05,
+      "loss": 0.391,
+      "step": 1529
+    },
+    {
+      "epoch": 1.6311300639658848,
+      "grad_norm": 0.5821772106338331,
+      "learning_rate": 6.816296746497744e-05,
+      "loss": 0.3868,
+      "step": 1530
+    },
+    {
+      "epoch": 1.6321961620469083,
+      "grad_norm": 0.6719640615154807,
+      "learning_rate": 6.814181843969664e-05,
+      "loss": 0.3974,
+      "step": 1531
+    },
+    {
+      "epoch": 1.6332622601279319,
+      "grad_norm": 0.599159712940155,
+      "learning_rate": 6.812065382531101e-05,
+      "loss": 0.391,
+      "step": 1532
+    },
+    {
+      "epoch": 1.6343283582089554,
+      "grad_norm": 0.5406155214874598,
+      "learning_rate": 6.809947363354464e-05,
+      "loss": 0.3907,
+      "step": 1533
+    },
+    {
+      "epoch": 1.6353944562899787,
+      "grad_norm": 0.5404737249091326,
+      "learning_rate": 6.807827787613024e-05,
+      "loss": 0.3933,
+      "step": 1534
+    },
+    {
+      "epoch": 1.636460554371002,
+      "grad_norm": 0.5080707037084039,
+      "learning_rate": 6.805706656480917e-05,
+      "loss": 0.3802,
+      "step": 1535
+    },
+    {
+      "epoch": 1.6375266524520256,
+      "grad_norm": 0.45278920241628184,
+      "learning_rate": 6.803583971133139e-05,
+      "loss": 0.3965,
+      "step": 1536
+    },
+    {
+      "epoch": 1.638592750533049,
+      "grad_norm": 0.46161933055133486,
+      "learning_rate": 6.801459732745547e-05,
+      "loss": 0.4023,
+      "step": 1537
+    },
+    {
+      "epoch": 1.6396588486140726,
+      "grad_norm": 0.5047213330551431,
+      "learning_rate": 6.799333942494861e-05,
+      "loss": 0.386,
+      "step": 1538
+    },
+    {
+      "epoch": 1.640724946695096,
+      "grad_norm": 0.4450247054157199,
+      "learning_rate": 6.797206601558654e-05,
+      "loss": 0.3933,
+      "step": 1539
+    },
+    {
+      "epoch": 1.6417910447761193,
+      "grad_norm": 0.3955698230219666,
+      "learning_rate": 6.795077711115368e-05,
+      "loss": 0.3941,
+      "step": 1540
+    },
+    {
+      "epoch": 1.6428571428571428,
+      "grad_norm": 0.4975244131254161,
+      "learning_rate": 6.792947272344292e-05,
+      "loss": 0.3985,
+      "step": 1541
+    },
+    {
+      "epoch": 1.6439232409381663,
+      "grad_norm": 0.4201752241177451,
+      "learning_rate": 6.790815286425581e-05,
+      "loss": 0.3899,
+      "step": 1542
+    },
+    {
+      "epoch": 1.6449893390191899,
+      "grad_norm": 0.3070985227762259,
+      "learning_rate": 6.788681754540245e-05,
+      "loss": 0.3931,
+      "step": 1543
+    },
+    {
+      "epoch": 1.6460554371002132,
+      "grad_norm": 0.40901085056872827,
+      "learning_rate": 6.78654667787015e-05,
+      "loss": 0.3911,
+      "step": 1544
+    },
+    {
+      "epoch": 1.6471215351812367,
+      "grad_norm": 0.4407517158255465,
+      "learning_rate": 6.784410057598016e-05,
+      "loss": 0.3888,
+      "step": 1545
+    },
+    {
+      "epoch": 1.64818763326226,
+      "grad_norm": 0.3680848288610597,
+      "learning_rate": 6.782271894907419e-05,
+      "loss": 0.3867,
+      "step": 1546
+    },
+    {
+      "epoch": 1.6492537313432836,
+      "grad_norm": 0.38558182531740504,
+      "learning_rate": 6.78013219098279e-05,
+      "loss": 0.3873,
+      "step": 1547
+    },
+    {
+      "epoch": 1.650319829424307,
+      "grad_norm": 0.43293933049711597,
+      "learning_rate": 6.777990947009418e-05,
+      "loss": 0.3876,
+      "step": 1548
+    },
+    {
+      "epoch": 1.6513859275053306,
+      "grad_norm": 0.5130367741977939,
+      "learning_rate": 6.775848164173436e-05,
+      "loss": 0.3971,
+      "step": 1549
+    },
+    {
+      "epoch": 1.652452025586354,
+      "grad_norm": 0.5953290932568412,
+      "learning_rate": 6.773703843661837e-05,
+      "loss": 0.3919,
+      "step": 1550
+    },
+    {
+      "epoch": 1.6535181236673773,
+      "grad_norm": 0.603488942008407,
+      "learning_rate": 6.771557986662462e-05,
+      "loss": 0.39,
+      "step": 1551
+    },
+    {
+      "epoch": 1.6545842217484008,
+      "grad_norm": 0.5050491129848195,
+      "learning_rate": 6.769410594364004e-05,
+      "loss": 0.3875,
+      "step": 1552
+    },
+    {
+      "epoch": 1.6556503198294243,
+      "grad_norm": 0.4097987145826878,
+      "learning_rate": 6.767261667956009e-05,
+      "loss": 0.3923,
+      "step": 1553
+    },
+    {
+      "epoch": 1.6567164179104479,
+      "grad_norm": 0.38259958335783656,
+      "learning_rate": 6.765111208628866e-05,
+      "loss": 0.394,
+      "step": 1554
+    },
+    {
+      "epoch": 1.6577825159914712,
+      "grad_norm": 0.36285793649423737,
+      "learning_rate": 6.762959217573823e-05,
+      "loss": 0.3902,
+      "step": 1555
+    },
+    {
+      "epoch": 1.6588486140724945,
+      "grad_norm": 0.38527741858086,
+      "learning_rate": 6.760805695982967e-05,
+      "loss": 0.3934,
+      "step": 1556
+    },
+    {
+      "epoch": 1.659914712153518,
+      "grad_norm": 0.43180049398669657,
+      "learning_rate": 6.75865064504924e-05,
+      "loss": 0.3948,
+      "step": 1557
+    },
+    {
+      "epoch": 1.6609808102345416,
+      "grad_norm": 0.4605936531276027,
+      "learning_rate": 6.756494065966426e-05,
+      "loss": 0.3853,
+      "step": 1558
+    },
+    {
+      "epoch": 1.6620469083155651,
+      "grad_norm": 0.5637754400106197,
+      "learning_rate": 6.754335959929159e-05,
+      "loss": 0.3906,
+      "step": 1559
+    },
+    {
+      "epoch": 1.6631130063965884,
+      "grad_norm": 0.6706100285105229,
+      "learning_rate": 6.752176328132918e-05,
+      "loss": 0.3896,
+      "step": 1560
+    },
+    {
+      "epoch": 1.664179104477612,
+      "grad_norm": 0.6676286093604591,
+      "learning_rate": 6.750015171774025e-05,
+      "loss": 0.3916,
+      "step": 1561
+    },
+    {
+      "epoch": 1.6652452025586353,
+      "grad_norm": 0.5819487463144472,
+      "learning_rate": 6.747852492049648e-05,
+      "loss": 0.3865,
+      "step": 1562
+    },
+    {
+      "epoch": 1.6663113006396588,
+      "grad_norm": 0.4906481414544558,
+      "learning_rate": 6.745688290157803e-05,
+      "loss": 0.3901,
+      "step": 1563
+    },
+    {
+      "epoch": 1.6673773987206824,
+      "grad_norm": 0.43576919495858,
+      "learning_rate": 6.743522567297341e-05,
+      "loss": 0.3946,
+      "step": 1564
+    },
+    {
+      "epoch": 1.668443496801706,
+      "grad_norm": 0.3626934309190754,
+      "learning_rate": 6.741355324667963e-05,
+      "loss": 0.3859,
+      "step": 1565
+    },
+    {
+      "epoch": 1.6695095948827292,
+      "grad_norm": 0.30739074025899915,
+      "learning_rate": 6.739186563470208e-05,
+      "loss": 0.3855,
+      "step": 1566
+    },
+    {
+      "epoch": 1.6705756929637525,
+      "grad_norm": 0.39101458085234175,
+      "learning_rate": 6.737016284905455e-05,
+      "loss": 0.3931,
+      "step": 1567
+    },
+    {
+      "epoch": 1.671641791044776,
+      "grad_norm": 0.5051971466609865,
+      "learning_rate": 6.734844490175929e-05,
+      "loss": 0.3881,
+      "step": 1568
+    },
+    {
+      "epoch": 1.6727078891257996,
+      "grad_norm": 0.5097545702694593,
+      "learning_rate": 6.732671180484687e-05,
+      "loss": 0.385,
+      "step": 1569
+    },
+    {
+      "epoch": 1.6737739872068231,
+      "grad_norm": 0.39153153015185177,
+      "learning_rate": 6.730496357035634e-05,
+      "loss": 0.3876,
+      "step": 1570
+    },
+    {
+      "epoch": 1.6748400852878464,
+      "grad_norm": 0.3456157078554192,
+      "learning_rate": 6.728320021033509e-05,
+      "loss": 0.3913,
+      "step": 1571
+    },
+    {
+      "epoch": 1.67590618336887,
+      "grad_norm": 0.38289300413638977,
+      "learning_rate": 6.726142173683884e-05,
+      "loss": 0.3889,
+      "step": 1572
+    },
+    {
+      "epoch": 1.6769722814498933,
+      "grad_norm": 0.36549754916811916,
+      "learning_rate": 6.723962816193178e-05,
+      "loss": 0.389,
+      "step": 1573
+    },
+    {
+      "epoch": 1.6780383795309168,
+      "grad_norm": 0.33051094759676547,
+      "learning_rate": 6.721781949768639e-05,
+      "loss": 0.3919,
+      "step": 1574
+    },
+    {
+      "epoch": 1.6791044776119404,
+      "grad_norm": 0.36416872005257267,
+      "learning_rate": 6.719599575618357e-05,
+      "loss": 0.3912,
+      "step": 1575
+    },
+    {
+      "epoch": 1.680170575692964,
+      "grad_norm": 0.34559502107142376,
+      "learning_rate": 6.717415694951251e-05,
+      "loss": 0.3851,
+      "step": 1576
+    },
+    {
+      "epoch": 1.6812366737739872,
+      "grad_norm": 0.3116613026962773,
+      "learning_rate": 6.715230308977078e-05,
+      "loss": 0.3916,
+      "step": 1577
+    },
+    {
+      "epoch": 1.6823027718550105,
+      "grad_norm": 0.30050046517095325,
+      "learning_rate": 6.713043418906428e-05,
+      "loss": 0.3962,
+      "step": 1578
+    },
+    {
+      "epoch": 1.683368869936034,
+      "grad_norm": 0.3577516859630259,
+      "learning_rate": 6.710855025950727e-05,
+      "loss": 0.3888,
+      "step": 1579
+    },
+    {
+      "epoch": 1.6844349680170576,
+      "grad_norm": 0.4364133412686821,
+      "learning_rate": 6.708665131322227e-05,
+      "loss": 0.3863,
+      "step": 1580
+    },
+    {
+      "epoch": 1.6855010660980811,
+      "grad_norm": 0.4687325703384506,
+      "learning_rate": 6.706473736234018e-05,
+      "loss": 0.3915,
+      "step": 1581
+    },
+    {
+      "epoch": 1.6865671641791045,
+      "grad_norm": 0.5090267219412645,
+      "learning_rate": 6.704280841900019e-05,
+      "loss": 0.3839,
+      "step": 1582
+    },
+    {
+      "epoch": 1.6876332622601278,
+      "grad_norm": 0.5457964796096376,
+      "learning_rate": 6.70208644953498e-05,
+      "loss": 0.3956,
+      "step": 1583
+    },
+    {
+      "epoch": 1.6886993603411513,
+      "grad_norm": 0.6368745088599855,
+      "learning_rate": 6.699890560354478e-05,
+      "loss": 0.3902,
+      "step": 1584
+    },
+    {
+      "epoch": 1.6897654584221748,
+      "grad_norm": 0.7026042269970657,
+      "learning_rate": 6.697693175574923e-05,
+      "loss": 0.3898,
+      "step": 1585
+    },
+    {
+      "epoch": 1.6908315565031984,
+      "grad_norm": 0.6804973492768571,
+      "learning_rate": 6.695494296413554e-05,
+      "loss": 0.3835,
+      "step": 1586
+    },
+    {
+      "epoch": 1.6918976545842217,
+      "grad_norm": 0.6670494458267645,
+      "learning_rate": 6.693293924088432e-05,
+      "loss": 0.3951,
+      "step": 1587
+    },
+    {
+      "epoch": 1.6929637526652452,
+      "grad_norm": 0.6433701055713933,
+      "learning_rate": 6.691092059818451e-05,
+      "loss": 0.3884,
+      "step": 1588
+    },
+    {
+      "epoch": 1.6940298507462686,
+      "grad_norm": 0.6565301872411362,
+      "learning_rate": 6.688888704823329e-05,
+      "loss": 0.3951,
+      "step": 1589
+    },
+    {
+      "epoch": 1.695095948827292,
+      "grad_norm": 0.6960842723329089,
+      "learning_rate": 6.686683860323611e-05,
+      "loss": 0.3946,
+      "step": 1590
+    },
+    {
+      "epoch": 1.6961620469083156,
+      "grad_norm": 0.6818586763465625,
+      "learning_rate": 6.684477527540664e-05,
+      "loss": 0.3946,
+      "step": 1591
+    },
+    {
+      "epoch": 1.6972281449893392,
+      "grad_norm": 0.5572883382416285,
+      "learning_rate": 6.682269707696685e-05,
+      "loss": 0.3867,
+      "step": 1592
+    },
+    {
+      "epoch": 1.6982942430703625,
+      "grad_norm": 0.3990760403818603,
+      "learning_rate": 6.680060402014689e-05,
+      "loss": 0.3854,
+      "step": 1593
+    },
+    {
+      "epoch": 1.6993603411513858,
+      "grad_norm": 0.4118819183924536,
+      "learning_rate": 6.677849611718515e-05,
+      "loss": 0.3864,
+      "step": 1594
+    },
+    {
+      "epoch": 1.7004264392324093,
+      "grad_norm": 0.49548805198701124,
+      "learning_rate": 6.67563733803283e-05,
+      "loss": 0.3974,
+      "step": 1595
+    },
+    {
+      "epoch": 1.7014925373134329,
+      "grad_norm": 0.5484992450512396,
+      "learning_rate": 6.673423582183117e-05,
+      "loss": 0.3895,
+      "step": 1596
+    },
+    {
+      "epoch": 1.7025586353944564,
+      "grad_norm": 0.5575698301270008,
+      "learning_rate": 6.67120834539568e-05,
+      "loss": 0.3877,
+      "step": 1597
+    },
+    {
+      "epoch": 1.7036247334754797,
+      "grad_norm": 0.5151550662716535,
+      "learning_rate": 6.668991628897648e-05,
+      "loss": 0.3877,
+      "step": 1598
+    },
+    {
+      "epoch": 1.7046908315565032,
+      "grad_norm": 0.4728544267085605,
+      "learning_rate": 6.666773433916965e-05,
+      "loss": 0.3869,
+      "step": 1599
+    },
+    {
+      "epoch": 1.7057569296375266,
+      "grad_norm": 0.47045760280939164,
+      "learning_rate": 6.664553761682395e-05,
+      "loss": 0.3851,
+      "step": 1600
+    },
+    {
+      "epoch": 1.70682302771855,
+      "grad_norm": 0.45385159010082227,
+      "learning_rate": 6.662332613423522e-05,
+      "loss": 0.3957,
+      "step": 1601
+    },
+    {
+      "epoch": 1.7078891257995736,
+      "grad_norm": 0.39340233247394074,
+      "learning_rate": 6.660109990370747e-05,
+      "loss": 0.3831,
+      "step": 1602
+    },
+    {
+      "epoch": 1.7089552238805972,
+      "grad_norm": 0.4438034696481542,
+      "learning_rate": 6.657885893755288e-05,
+      "loss": 0.3941,
+      "step": 1603
+    },
+    {
+      "epoch": 1.7100213219616205,
+      "grad_norm": 0.5254817458294889,
+      "learning_rate": 6.655660324809177e-05,
+      "loss": 0.387,
+      "step": 1604
+    },
+    {
+      "epoch": 1.7110874200426438,
+      "grad_norm": 0.4913954053153753,
+      "learning_rate": 6.653433284765266e-05,
+      "loss": 0.3826,
+      "step": 1605
+    },
+    {
+      "epoch": 1.7121535181236673,
+      "grad_norm": 0.36756522486239,
+      "learning_rate": 6.651204774857218e-05,
+      "loss": 0.4007,
+      "step": 1606
+    },
+    {
+      "epoch": 1.7132196162046909,
+      "grad_norm": 0.34100380243193384,
+      "learning_rate": 6.648974796319512e-05,
+      "loss": 0.3932,
+      "step": 1607
+    },
+    {
+      "epoch": 1.7142857142857144,
+      "grad_norm": 0.38180887373133354,
+      "learning_rate": 6.646743350387438e-05,
+      "loss": 0.3857,
+      "step": 1608
+    },
+    {
+      "epoch": 1.7153518123667377,
+      "grad_norm": 0.4155049597785398,
+      "learning_rate": 6.644510438297105e-05,
+      "loss": 0.3925,
+      "step": 1609
+    },
+    {
+      "epoch": 1.716417910447761,
+      "grad_norm": 0.4422223445035819,
+      "learning_rate": 6.642276061285428e-05,
+      "loss": 0.387,
+      "step": 1610
+    },
+    {
+      "epoch": 1.7174840085287846,
+      "grad_norm": 0.5794607050629841,
+      "learning_rate": 6.640040220590136e-05,
+      "loss": 0.3887,
+      "step": 1611
+    },
+    {
+      "epoch": 1.7185501066098081,
+      "grad_norm": 0.6678832046144233,
+      "learning_rate": 6.63780291744977e-05,
+      "loss": 0.3899,
+      "step": 1612
+    },
+    {
+      "epoch": 1.7196162046908317,
+      "grad_norm": 0.6122776429863264,
+      "learning_rate": 6.635564153103677e-05,
+      "loss": 0.3929,
+      "step": 1613
+    },
+    {
+      "epoch": 1.720682302771855,
+      "grad_norm": 0.5059533853234256,
+      "learning_rate": 6.633323928792018e-05,
+      "loss": 0.3915,
+      "step": 1614
+    },
+    {
+      "epoch": 1.7217484008528785,
+      "grad_norm": 0.4164548494103394,
+      "learning_rate": 6.631082245755762e-05,
+      "loss": 0.3962,
+      "step": 1615
+    },
+    {
+      "epoch": 1.7228144989339018,
+      "grad_norm": 0.39276463156642827,
+      "learning_rate": 6.628839105236681e-05,
+      "loss": 0.3904,
+      "step": 1616
+    },
+    {
+      "epoch": 1.7238805970149254,
+      "grad_norm": 0.41559718909247517,
+      "learning_rate": 6.626594508477361e-05,
+      "loss": 0.3917,
+      "step": 1617
+    },
+    {
+      "epoch": 1.724946695095949,
+      "grad_norm": 0.44665059431355064,
+      "learning_rate": 6.624348456721191e-05,
+      "loss": 0.3904,
+      "step": 1618
+    },
+    {
+      "epoch": 1.7260127931769724,
+      "grad_norm": 0.4665709219109089,
+      "learning_rate": 6.622100951212368e-05,
+      "loss": 0.3901,
+      "step": 1619
+    },
+    {
+      "epoch": 1.7270788912579957,
+      "grad_norm": 0.5150232707357006,
+      "learning_rate": 6.619851993195893e-05,
+      "loss": 0.3922,
+      "step": 1620
+    },
+    {
+      "epoch": 1.728144989339019,
+      "grad_norm": 0.6026588085567588,
+      "learning_rate": 6.61760158391757e-05,
+      "loss": 0.3972,
+      "step": 1621
+    },
+    {
+      "epoch": 1.7292110874200426,
+      "grad_norm": 0.6385069316475197,
+      "learning_rate": 6.615349724624012e-05,
+      "loss": 0.3963,
+      "step": 1622
+    },
+    {
+      "epoch": 1.7302771855010661,
+      "grad_norm": 0.6066727609006709,
+      "learning_rate": 6.61309641656263e-05,
+      "loss": 0.3916,
+      "step": 1623
+    },
+    {
+      "epoch": 1.7313432835820897,
+      "grad_norm": 0.5241084773133352,
+      "learning_rate": 6.610841660981639e-05,
+      "loss": 0.3897,
+      "step": 1624
+    },
+    {
+      "epoch": 1.732409381663113,
+      "grad_norm": 0.4853459956518966,
+      "learning_rate": 6.608585459130057e-05,
+      "loss": 0.3943,
+      "step": 1625
+    },
+    {
+      "epoch": 1.7334754797441365,
+      "grad_norm": 0.43877664392065524,
+      "learning_rate": 6.606327812257705e-05,
+      "loss": 0.3889,
+      "step": 1626
+    },
+    {
+      "epoch": 1.7345415778251598,
+      "grad_norm": 0.39226116124002375,
+      "learning_rate": 6.604068721615198e-05,
+      "loss": 0.394,
+      "step": 1627
+    },
+    {
+      "epoch": 1.7356076759061834,
+      "grad_norm": 0.34110686996523315,
+      "learning_rate": 6.601808188453957e-05,
+      "loss": 0.3912,
+      "step": 1628
+    },
+    {
+      "epoch": 1.736673773987207,
+      "grad_norm": 0.3299899773295145,
+      "learning_rate": 6.599546214026199e-05,
+      "loss": 0.396,
+      "step": 1629
+    },
+    {
+      "epoch": 1.7377398720682304,
+      "grad_norm": 0.376767657465232,
+      "learning_rate": 6.597282799584941e-05,
+      "loss": 0.3878,
+      "step": 1630
+    },
+    {
+      "epoch": 1.7388059701492538,
+      "grad_norm": 0.3870812218773304,
+      "learning_rate": 6.595017946383998e-05,
+      "loss": 0.3924,
+      "step": 1631
+    },
+    {
+      "epoch": 1.739872068230277,
+      "grad_norm": 0.32816223635729713,
+      "learning_rate": 6.59275165567798e-05,
+      "loss": 0.3891,
+      "step": 1632
+    },
+    {
+      "epoch": 1.7409381663113006,
+      "grad_norm": 0.3368154332955513,
+      "learning_rate": 6.590483928722293e-05,
+      "loss": 0.3867,
+      "step": 1633
+    },
+    {
+      "epoch": 1.7420042643923241,
+      "grad_norm": 0.32846108321184647,
+      "learning_rate": 6.58821476677314e-05,
+      "loss": 0.3899,
+      "step": 1634
+    },
+    {
+      "epoch": 1.7430703624733477,
+      "grad_norm": 0.30960636706164546,
+      "learning_rate": 6.585944171087521e-05,
+      "loss": 0.3967,
+      "step": 1635
+    },
+    {
+      "epoch": 1.744136460554371,
+      "grad_norm": 0.3431044349932443,
+      "learning_rate": 6.583672142923226e-05,
+      "loss": 0.3957,
+      "step": 1636
+    },
+    {
+      "epoch": 1.7452025586353943,
+      "grad_norm": 0.3817314249874563,
+      "learning_rate": 6.581398683538842e-05,
+      "loss": 0.3938,
+      "step": 1637
+    },
+    {
+      "epoch": 1.7462686567164178,
+      "grad_norm": 0.3846143335468199,
+      "learning_rate": 6.579123794193746e-05,
+      "loss": 0.3875,
+      "step": 1638
+    },
+    {
+      "epoch": 1.7473347547974414,
+      "grad_norm": 0.40698506124086886,
+      "learning_rate": 6.576847476148109e-05,
+      "loss": 0.3878,
+      "step": 1639
+    },
+    {
+      "epoch": 1.748400852878465,
+      "grad_norm": 0.5160395538152548,
+      "learning_rate": 6.574569730662893e-05,
+      "loss": 0.3935,
+      "step": 1640
+    },
+    {
+      "epoch": 1.7494669509594882,
+      "grad_norm": 0.6377912942089725,
+      "learning_rate": 6.57229055899985e-05,
+      "loss": 0.392,
+      "step": 1641
+    },
+    {
+      "epoch": 1.7505330490405118,
+      "grad_norm": 0.6901659496461329,
+      "learning_rate": 6.570009962421523e-05,
+      "loss": 0.3912,
+      "step": 1642
+    },
+    {
+      "epoch": 1.751599147121535,
+      "grad_norm": 0.6169852874235238,
+      "learning_rate": 6.567727942191246e-05,
+      "loss": 0.391,
+      "step": 1643
+    },
+    {
+      "epoch": 1.7526652452025586,
+      "grad_norm": 0.4948279659725487,
+      "learning_rate": 6.565444499573136e-05,
+      "loss": 0.3893,
+      "step": 1644
+    },
+    {
+      "epoch": 1.7537313432835822,
+      "grad_norm": 0.5319346837051117,
+      "learning_rate": 6.563159635832105e-05,
+      "loss": 0.396,
+      "step": 1645
+    },
+    {
+      "epoch": 1.7547974413646057,
+      "grad_norm": 0.6070963028163101,
+      "learning_rate": 6.560873352233846e-05,
+      "loss": 0.3879,
+      "step": 1646
+    },
+    {
+      "epoch": 1.755863539445629,
+      "grad_norm": 0.5500219234930699,
+      "learning_rate": 6.558585650044842e-05,
+      "loss": 0.3962,
+      "step": 1647
+    },
+    {
+      "epoch": 1.7569296375266523,
+      "grad_norm": 0.3774689595977208,
+      "learning_rate": 6.556296530532364e-05,
+      "loss": 0.3853,
+      "step": 1648
+    },
+    {
+      "epoch": 1.7579957356076759,
+      "grad_norm": 0.3097496369115005,
+      "learning_rate": 6.554005994964459e-05,
+      "loss": 0.3889,
+      "step": 1649
+    },
+    {
+      "epoch": 1.7590618336886994,
+      "grad_norm": 0.3192713728666256,
+      "learning_rate": 6.55171404460997e-05,
+      "loss": 0.3904,
+      "step": 1650
+    },
+    {
+      "epoch": 1.760127931769723,
+      "grad_norm": 0.32864817312795563,
+      "learning_rate": 6.549420680738516e-05,
+      "loss": 0.3944,
+      "step": 1651
+    },
+    {
+      "epoch": 1.7611940298507462,
+      "grad_norm": 0.32211875803439644,
+      "learning_rate": 6.547125904620504e-05,
+      "loss": 0.385,
+      "step": 1652
+    },
+    {
+      "epoch": 1.7622601279317696,
+      "grad_norm": 0.30017127275221656,
+      "learning_rate": 6.544829717527118e-05,
+      "loss": 0.3885,
+      "step": 1653
+    },
+    {
+      "epoch": 1.763326226012793,
+      "grad_norm": 0.3646601960483686,
+      "learning_rate": 6.542532120730327e-05,
+      "loss": 0.3905,
+      "step": 1654
+    },
+    {
+      "epoch": 1.7643923240938166,
+      "grad_norm": 0.4908471802914999,
+      "learning_rate": 6.540233115502881e-05,
+      "loss": 0.3889,
+      "step": 1655
+    },
+    {
+      "epoch": 1.7654584221748402,
+      "grad_norm": 0.5574952396992828,
+      "learning_rate": 6.537932703118308e-05,
+      "loss": 0.3881,
+      "step": 1656
+    },
+    {
+      "epoch": 1.7665245202558635,
+      "grad_norm": 0.5663694328254277,
+      "learning_rate": 6.535630884850917e-05,
+      "loss": 0.3862,
+      "step": 1657
+    },
+    {
+      "epoch": 1.767590618336887,
+      "grad_norm": 0.5503251345750754,
+      "learning_rate": 6.533327661975799e-05,
+      "loss": 0.3826,
+      "step": 1658
+    },
+    {
+      "epoch": 1.7686567164179103,
+      "grad_norm": 0.4691123534761767,
+      "learning_rate": 6.531023035768815e-05,
+      "loss": 0.3905,
+      "step": 1659
+    },
+    {
+      "epoch": 1.7697228144989339,
+      "grad_norm": 0.41955605601856755,
+      "learning_rate": 6.528717007506612e-05,
+      "loss": 0.396,
+      "step": 1660
+    },
+    {
+      "epoch": 1.7707889125799574,
+      "grad_norm": 0.4274791936147579,
+      "learning_rate": 6.526409578466606e-05,
+      "loss": 0.3874,
+      "step": 1661
+    },
+    {
+      "epoch": 1.771855010660981,
+      "grad_norm": 0.4609745881429261,
+      "learning_rate": 6.524100749926997e-05,
+      "loss": 0.3965,
+      "step": 1662
+    },
+    {
+      "epoch": 1.7729211087420043,
+      "grad_norm": 0.5004816396707016,
+      "learning_rate": 6.521790523166752e-05,
+      "loss": 0.391,
+      "step": 1663
+    },
+    {
+      "epoch": 1.7739872068230276,
+      "grad_norm": 0.5323254145339794,
+      "learning_rate": 6.51947889946562e-05,
+      "loss": 0.3855,
+      "step": 1664
+    },
+    {
+      "epoch": 1.775053304904051,
+      "grad_norm": 0.5367601341161515,
+      "learning_rate": 6.517165880104119e-05,
+      "loss": 0.395,
+      "step": 1665
+    },
+    {
+      "epoch": 1.7761194029850746,
+      "grad_norm": 0.5721670713464904,
+      "learning_rate": 6.51485146636354e-05,
+      "loss": 0.3831,
+      "step": 1666
+    },
+    {
+      "epoch": 1.7771855010660982,
+      "grad_norm": 0.6751411525618288,
+      "learning_rate": 6.51253565952595e-05,
+      "loss": 0.3897,
+      "step": 1667
+    },
+    {
+      "epoch": 1.7782515991471215,
+      "grad_norm": 0.7475500173594719,
+      "learning_rate": 6.510218460874186e-05,
+      "loss": 0.3918,
+      "step": 1668
+    },
+    {
+      "epoch": 1.779317697228145,
+      "grad_norm": 0.7174243154443929,
+      "learning_rate": 6.507899871691852e-05,
+      "loss": 0.3872,
+      "step": 1669
+    },
+    {
+      "epoch": 1.7803837953091683,
+      "grad_norm": 0.5785749621394422,
+      "learning_rate": 6.50557989326333e-05,
+      "loss": 0.3898,
+      "step": 1670
+    },
+    {
+      "epoch": 1.7814498933901919,
+      "grad_norm": 0.4450886636964885,
+      "learning_rate": 6.503258526873767e-05,
+      "loss": 0.3824,
+      "step": 1671
+    },
+    {
+      "epoch": 1.7825159914712154,
+      "grad_norm": 0.4254053253981925,
+      "learning_rate": 6.500935773809076e-05,
+      "loss": 0.3911,
+      "step": 1672
+    },
+    {
+      "epoch": 1.783582089552239,
+      "grad_norm": 0.4474966076928897,
+      "learning_rate": 6.498611635355947e-05,
+      "loss": 0.39,
+      "step": 1673
+    },
+    {
+      "epoch": 1.7846481876332623,
+      "grad_norm": 0.45486240848719184,
+      "learning_rate": 6.496286112801826e-05,
+      "loss": 0.3914,
+      "step": 1674
+    },
+    {
+      "epoch": 1.7857142857142856,
+      "grad_norm": 0.514667280468136,
+      "learning_rate": 6.493959207434934e-05,
+      "loss": 0.3861,
+      "step": 1675
+    },
+    {
+      "epoch": 1.7867803837953091,
+      "grad_norm": 0.5337184119841549,
+      "learning_rate": 6.491630920544257e-05,
+      "loss": 0.3871,
+      "step": 1676
+    },
+    {
+      "epoch": 1.7878464818763327,
+      "grad_norm": 0.4780735793732188,
+      "learning_rate": 6.489301253419545e-05,
+      "loss": 0.3821,
+      "step": 1677
+    },
+    {
+      "epoch": 1.7889125799573562,
+      "grad_norm": 0.4231898830746279,
+      "learning_rate": 6.48697020735131e-05,
+      "loss": 0.386,
+      "step": 1678
+    },
+    {
+      "epoch": 1.7899786780383795,
+      "grad_norm": 0.3663697733311359,
+      "learning_rate": 6.484637783630832e-05,
+      "loss": 0.385,
+      "step": 1679
+    },
+    {
+      "epoch": 1.7910447761194028,
+      "grad_norm": 0.30984371824344287,
+      "learning_rate": 6.482303983550151e-05,
+      "loss": 0.391,
+      "step": 1680
+    },
+    {
+      "epoch": 1.7921108742004264,
+      "grad_norm": 0.35751913701235005,
+      "learning_rate": 6.479968808402075e-05,
+      "loss": 0.3957,
+      "step": 1681
+    },
+    {
+      "epoch": 1.79317697228145,
+      "grad_norm": 0.4093914369810105,
+      "learning_rate": 6.477632259480165e-05,
+      "loss": 0.3935,
+      "step": 1682
+    },
+    {
+      "epoch": 1.7942430703624734,
+      "grad_norm": 0.41377357296377376,
+      "learning_rate": 6.475294338078752e-05,
+      "loss": 0.3844,
+      "step": 1683
+    },
+    {
+      "epoch": 1.7953091684434968,
+      "grad_norm": 0.39868902718714505,
+      "learning_rate": 6.472955045492918e-05,
+      "loss": 0.392,
+      "step": 1684
+    },
+    {
+      "epoch": 1.7963752665245203,
+      "grad_norm": 0.42005660101933967,
+      "learning_rate": 6.470614383018512e-05,
+      "loss": 0.3932,
+      "step": 1685
+    },
+    {
+      "epoch": 1.7974413646055436,
+      "grad_norm": 0.4135157735982948,
+      "learning_rate": 6.468272351952141e-05,
+      "loss": 0.3938,
+      "step": 1686
+    },
+    {
+      "epoch": 1.7985074626865671,
+      "grad_norm": 0.42801661015742615,
+      "learning_rate": 6.465928953591165e-05,
+      "loss": 0.3879,
+      "step": 1687
+    },
+    {
+      "epoch": 1.7995735607675907,
+      "grad_norm": 0.4816981612374122,
+      "learning_rate": 6.463584189233709e-05,
+      "loss": 0.3951,
+      "step": 1688
+    },
+    {
+      "epoch": 1.8006396588486142,
+      "grad_norm": 0.46743748656875295,
+      "learning_rate": 6.461238060178647e-05,
+      "loss": 0.3939,
+      "step": 1689
+    },
+    {
+      "epoch": 1.8017057569296375,
+      "grad_norm": 0.4222158155941226,
+      "learning_rate": 6.458890567725614e-05,
+      "loss": 0.3906,
+      "step": 1690
+    },
+    {
+      "epoch": 1.8027718550106608,
+      "grad_norm": 0.3658953583582503,
+      "learning_rate": 6.456541713174999e-05,
+      "loss": 0.3962,
+      "step": 1691
+    },
+    {
+      "epoch": 1.8038379530916844,
+      "grad_norm": 0.33668688496020444,
+      "learning_rate": 6.454191497827945e-05,
+      "loss": 0.3921,
+      "step": 1692
+    },
+    {
+      "epoch": 1.804904051172708,
+      "grad_norm": 0.3206971308888266,
+      "learning_rate": 6.451839922986349e-05,
+      "loss": 0.3883,
+      "step": 1693
+    },
+    {
+      "epoch": 1.8059701492537314,
+      "grad_norm": 0.3143057384612725,
+      "learning_rate": 6.449486989952863e-05,
+      "loss": 0.3896,
+      "step": 1694
+    },
+    {
+      "epoch": 1.8070362473347548,
+      "grad_norm": 0.34510492572302714,
+      "learning_rate": 6.447132700030887e-05,
+      "loss": 0.3826,
+      "step": 1695
+    },
+    {
+      "epoch": 1.8081023454157783,
+      "grad_norm": 0.4547140804511741,
+      "learning_rate": 6.444777054524576e-05,
+      "loss": 0.3884,
+      "step": 1696
+    },
+    {
+      "epoch": 1.8091684434968016,
+      "grad_norm": 0.47384268369658744,
+      "learning_rate": 6.442420054738837e-05,
+      "loss": 0.3834,
+      "step": 1697
+    },
+    {
+      "epoch": 1.8102345415778252,
+      "grad_norm": 0.47381168033622467,
+      "learning_rate": 6.440061701979323e-05,
+      "loss": 0.3812,
+      "step": 1698
+    },
+    {
+      "epoch": 1.8113006396588487,
+      "grad_norm": 0.4548557810738229,
+      "learning_rate": 6.43770199755244e-05,
+      "loss": 0.3879,
+      "step": 1699
+    },
+    {
+      "epoch": 1.8123667377398722,
+      "grad_norm": 0.4563303678565612,
+      "learning_rate": 6.435340942765341e-05,
+      "loss": 0.3881,
+      "step": 1700
+    },
+    {
+      "epoch": 1.8134328358208955,
+      "grad_norm": 0.4345832462667786,
+      "learning_rate": 6.432978538925928e-05,
+      "loss": 0.3875,
+      "step": 1701
+    },
+    {
+      "epoch": 1.8144989339019189,
+      "grad_norm": 0.36369291668844994,
+      "learning_rate": 6.430614787342853e-05,
+      "loss": 0.3891,
+      "step": 1702
+    },
+    {
+      "epoch": 1.8155650319829424,
+      "grad_norm": 0.351336330190162,
+      "learning_rate": 6.428249689325505e-05,
+      "loss": 0.3884,
+      "step": 1703
+    },
+    {
+      "epoch": 1.816631130063966,
+      "grad_norm": 0.3720689107883433,
+      "learning_rate": 6.425883246184031e-05,
+      "loss": 0.3917,
+      "step": 1704
+    },
+    {
+      "epoch": 1.8176972281449895,
+      "grad_norm": 0.3277156951268358,
+      "learning_rate": 6.423515459229313e-05,
+      "loss": 0.3857,
+      "step": 1705
+    },
+    {
+      "epoch": 1.8187633262260128,
+      "grad_norm": 0.3146191623979247,
+      "learning_rate": 6.421146329772988e-05,
+      "loss": 0.3881,
+      "step": 1706
+    },
+    {
+      "epoch": 1.819829424307036,
+      "grad_norm": 0.39372224035911607,
+      "learning_rate": 6.418775859127424e-05,
+      "loss": 0.3847,
+      "step": 1707
+    },
+    {
+      "epoch": 1.8208955223880596,
+      "grad_norm": 0.4723072413424627,
+      "learning_rate": 6.416404048605744e-05,
+      "loss": 0.3946,
+      "step": 1708
+    },
+    {
+      "epoch": 1.8219616204690832,
+      "grad_norm": 0.5461469639906784,
+      "learning_rate": 6.414030899521802e-05,
+      "loss": 0.3977,
+      "step": 1709
+    },
+    {
+      "epoch": 1.8230277185501067,
+      "grad_norm": 0.5644867900194028,
+      "learning_rate": 6.411656413190205e-05,
+      "loss": 0.3854,
+      "step": 1710
+    },
+    {
+      "epoch": 1.82409381663113,
+      "grad_norm": 0.48007304376329935,
+      "learning_rate": 6.409280590926292e-05,
+      "loss": 0.3947,
+      "step": 1711
+    },
+    {
+      "epoch": 1.8251599147121536,
+      "grad_norm": 0.41131804519163606,
+      "learning_rate": 6.406903434046146e-05,
+      "loss": 0.3913,
+      "step": 1712
+    },
+    {
+      "epoch": 1.8262260127931769,
+      "grad_norm": 0.37379945962827266,
+      "learning_rate": 6.404524943866588e-05,
+      "loss": 0.3807,
+      "step": 1713
+    },
+    {
+      "epoch": 1.8272921108742004,
+      "grad_norm": 0.42604642417053085,
+      "learning_rate": 6.402145121705178e-05,
+      "loss": 0.3953,
+      "step": 1714
+    },
+    {
+      "epoch": 1.828358208955224,
+      "grad_norm": 0.541443650037536,
+      "learning_rate": 6.399763968880214e-05,
+      "loss": 0.3915,
+      "step": 1715
+    },
+    {
+      "epoch": 1.8294243070362475,
+      "grad_norm": 0.5461544504898415,
+      "learning_rate": 6.397381486710728e-05,
+      "loss": 0.3917,
+      "step": 1716
+    },
+    {
+      "epoch": 1.8304904051172708,
+      "grad_norm": 0.4620349232311522,
+      "learning_rate": 6.394997676516497e-05,
+      "loss": 0.3951,
+      "step": 1717
+    },
+    {
+      "epoch": 1.831556503198294,
+      "grad_norm": 0.3606206070971171,
+      "learning_rate": 6.392612539618024e-05,
+      "loss": 0.3877,
+      "step": 1718
+    },
+    {
+      "epoch": 1.8326226012793176,
+      "grad_norm": 0.3665152949807116,
+      "learning_rate": 6.39022607733655e-05,
+      "loss": 0.3911,
+      "step": 1719
+    },
+    {
+      "epoch": 1.8336886993603412,
+      "grad_norm": 0.3617479779123929,
+      "learning_rate": 6.387838290994056e-05,
+      "loss": 0.3864,
+      "step": 1720
+    },
+    {
+      "epoch": 1.8347547974413647,
+      "grad_norm": 0.35982095746420845,
+      "learning_rate": 6.385449181913246e-05,
+      "loss": 0.387,
+      "step": 1721
+    },
+    {
+      "epoch": 1.835820895522388,
+      "grad_norm": 0.42991157325695156,
+      "learning_rate": 6.383058751417566e-05,
+      "loss": 0.3872,
+      "step": 1722
+    },
+    {
+      "epoch": 1.8368869936034116,
+      "grad_norm": 0.4025624229827069,
+      "learning_rate": 6.380667000831188e-05,
+      "loss": 0.3879,
+      "step": 1723
+    },
+    {
+      "epoch": 1.8379530916844349,
+      "grad_norm": 0.37576823662421394,
+      "learning_rate": 6.37827393147902e-05,
+      "loss": 0.3934,
+      "step": 1724
+    },
+    {
+      "epoch": 1.8390191897654584,
+      "grad_norm": 0.4724397154377103,
+      "learning_rate": 6.375879544686695e-05,
+      "loss": 0.3907,
+      "step": 1725
+    },
+    {
+      "epoch": 1.840085287846482,
+      "grad_norm": 0.49890032962764724,
+      "learning_rate": 6.37348384178058e-05,
+      "loss": 0.3905,
+      "step": 1726
+    },
+    {
+      "epoch": 1.8411513859275055,
+      "grad_norm": 0.5271786209536122,
+      "learning_rate": 6.371086824087772e-05,
+      "loss": 0.3879,
+      "step": 1727
+    },
+    {
+      "epoch": 1.8422174840085288,
+      "grad_norm": 0.5789984364979213,
+      "learning_rate": 6.368688492936091e-05,
+      "loss": 0.3968,
+      "step": 1728
+    },
+    {
+      "epoch": 1.8432835820895521,
+      "grad_norm": 0.5914257408271724,
+      "learning_rate": 6.366288849654091e-05,
+      "loss": 0.3917,
+      "step": 1729
+    },
+    {
+      "epoch": 1.8443496801705757,
+      "grad_norm": 0.4741221375005763,
+      "learning_rate": 6.363887895571045e-05,
+      "loss": 0.3845,
+      "step": 1730
+    },
+    {
+      "epoch": 1.8454157782515992,
+      "grad_norm": 0.3731893195453408,
+      "learning_rate": 6.361485632016963e-05,
+      "loss": 0.3927,
+      "step": 1731
+    },
+    {
+      "epoch": 1.8464818763326227,
+      "grad_norm": 0.36291074847148286,
+      "learning_rate": 6.359082060322569e-05,
+      "loss": 0.3834,
+      "step": 1732
+    },
+    {
+      "epoch": 1.847547974413646,
+      "grad_norm": 0.33824345650854754,
+      "learning_rate": 6.356677181819319e-05,
+      "loss": 0.3923,
+      "step": 1733
+    },
+    {
+      "epoch": 1.8486140724946694,
+      "grad_norm": 0.3590581537534159,
+      "learning_rate": 6.35427099783939e-05,
+      "loss": 0.3885,
+      "step": 1734
+    },
+    {
+      "epoch": 1.849680170575693,
+      "grad_norm": 0.393824937605615,
+      "learning_rate": 6.351863509715684e-05,
+      "loss": 0.3808,
+      "step": 1735
+    },
+    {
+      "epoch": 1.8507462686567164,
+      "grad_norm": 0.38639667752223744,
+      "learning_rate": 6.349454718781822e-05,
+      "loss": 0.3893,
+      "step": 1736
+    },
+    {
+      "epoch": 1.85181236673774,
+      "grad_norm": 0.3933756538854629,
+      "learning_rate": 6.347044626372153e-05,
+      "loss": 0.3866,
+      "step": 1737
+    },
+    {
+      "epoch": 1.8528784648187633,
+      "grad_norm": 0.4385122841126179,
+      "learning_rate": 6.34463323382174e-05,
+      "loss": 0.388,
+      "step": 1738
+    },
+    {
+      "epoch": 1.8539445628997868,
+      "grad_norm": 0.4219487489482543,
+      "learning_rate": 6.342220542466368e-05,
+      "loss": 0.3902,
+      "step": 1739
+    },
+    {
+      "epoch": 1.8550106609808101,
+      "grad_norm": 0.42067268031548977,
+      "learning_rate": 6.339806553642545e-05,
+      "loss": 0.3873,
+      "step": 1740
+    },
+    {
+      "epoch": 1.8560767590618337,
+      "grad_norm": 0.4563965192041482,
+      "learning_rate": 6.337391268687495e-05,
+      "loss": 0.3865,
+      "step": 1741
+    },
+    {
+      "epoch": 1.8571428571428572,
+      "grad_norm": 0.4746787282948449,
+      "learning_rate": 6.334974688939161e-05,
+      "loss": 0.3832,
+      "step": 1742
+    },
+    {
+      "epoch": 1.8582089552238807,
+      "grad_norm": 0.4616609792809343,
+      "learning_rate": 6.3325568157362e-05,
+      "loss": 0.3878,
+      "step": 1743
+    },
+    {
+      "epoch": 1.859275053304904,
+      "grad_norm": 0.4503741534784094,
+      "learning_rate": 6.33013765041799e-05,
+      "loss": 0.3857,
+      "step": 1744
+    },
+    {
+      "epoch": 1.8603411513859274,
+      "grad_norm": 0.3431922287079442,
+      "learning_rate": 6.327717194324622e-05,
+      "loss": 0.3903,
+      "step": 1745
+    },
+    {
+      "epoch": 1.861407249466951,
+      "grad_norm": 0.2609150587662325,
+      "learning_rate": 6.325295448796903e-05,
+      "loss": 0.391,
+      "step": 1746
+    },
+    {
+      "epoch": 1.8624733475479744,
+      "grad_norm": 0.33270455925916104,
+      "learning_rate": 6.322872415176356e-05,
+      "loss": 0.3934,
+      "step": 1747
+    },
+    {
+      "epoch": 1.863539445628998,
+      "grad_norm": 0.43954183331919916,
+      "learning_rate": 6.320448094805214e-05,
+      "loss": 0.3873,
+      "step": 1748
+    },
+    {
+      "epoch": 1.8646055437100213,
+      "grad_norm": 0.5101457999652256,
+      "learning_rate": 6.318022489026425e-05,
+      "loss": 0.3906,
+      "step": 1749
+    },
+    {
+      "epoch": 1.8656716417910446,
+      "grad_norm": 0.5395181016945636,
+      "learning_rate": 6.315595599183646e-05,
+      "loss": 0.3858,
+      "step": 1750
+    },
+    {
+      "epoch": 1.8667377398720681,
+      "grad_norm": 0.6031812678730777,
+      "learning_rate": 6.313167426621253e-05,
+      "loss": 0.3868,
+      "step": 1751
+    },
+    {
+      "epoch": 1.8678038379530917,
+      "grad_norm": 0.6041907845639172,
+      "learning_rate": 6.310737972684322e-05,
+      "loss": 0.3902,
+      "step": 1752
+    },
+    {
+      "epoch": 1.8688699360341152,
+      "grad_norm": 0.5089623629692637,
+      "learning_rate": 6.308307238718649e-05,
+      "loss": 0.385,
+      "step": 1753
+    },
+    {
+      "epoch": 1.8699360341151388,
+      "grad_norm": 0.40212189395827613,
+      "learning_rate": 6.305875226070729e-05,
+      "loss": 0.3914,
+      "step": 1754
+    },
+    {
+      "epoch": 1.871002132196162,
+      "grad_norm": 0.35938255962719357,
+      "learning_rate": 6.303441936087776e-05,
+      "loss": 0.3867,
+      "step": 1755
+    },
+    {
+      "epoch": 1.8720682302771854,
+      "grad_norm": 0.42202040294266446,
+      "learning_rate": 6.301007370117703e-05,
+      "loss": 0.3902,
+      "step": 1756
+    },
+    {
+      "epoch": 1.873134328358209,
+      "grad_norm": 0.5647279892704614,
+      "learning_rate": 6.298571529509135e-05,
+      "loss": 0.3845,
+      "step": 1757
+    },
+    {
+      "epoch": 1.8742004264392325,
+      "grad_norm": 0.6637603262956403,
+      "learning_rate": 6.296134415611399e-05,
+      "loss": 0.3854,
+      "step": 1758
+    },
+    {
+      "epoch": 1.875266524520256,
+      "grad_norm": 0.7031726822916566,
+      "learning_rate": 6.29369602977453e-05,
+      "loss": 0.3855,
+      "step": 1759
+    },
+    {
+      "epoch": 1.8763326226012793,
+      "grad_norm": 0.7834542862723698,
+      "learning_rate": 6.291256373349269e-05,
+      "loss": 0.3945,
+      "step": 1760
+    },
+    {
+      "epoch": 1.8773987206823026,
+      "grad_norm": 0.8226206202632336,
+      "learning_rate": 6.288815447687056e-05,
+      "loss": 0.3901,
+      "step": 1761
+    },
+    {
+      "epoch": 1.8784648187633262,
+      "grad_norm": 0.7994895608418134,
+      "learning_rate": 6.286373254140038e-05,
+      "loss": 0.3893,
+      "step": 1762
+    },
+    {
+      "epoch": 1.8795309168443497,
+      "grad_norm": 0.7178750160169209,
+      "learning_rate": 6.283929794061065e-05,
+      "loss": 0.3917,
+      "step": 1763
+    },
+    {
+      "epoch": 1.8805970149253732,
+      "grad_norm": 0.5231999499518026,
+      "learning_rate": 6.281485068803683e-05,
+      "loss": 0.3857,
+      "step": 1764
+    },
+    {
+      "epoch": 1.8816631130063965,
+      "grad_norm": 0.39458981834934215,
+      "learning_rate": 6.279039079722147e-05,
+      "loss": 0.3873,
+      "step": 1765
+    },
+    {
+      "epoch": 1.88272921108742,
+      "grad_norm": 0.5291211636623102,
+      "learning_rate": 6.276591828171406e-05,
+      "loss": 0.386,
+      "step": 1766
+    },
+    {
+      "epoch": 1.8837953091684434,
+      "grad_norm": 0.6451705117892833,
+      "learning_rate": 6.274143315507108e-05,
+      "loss": 0.3861,
+      "step": 1767
+    },
+    {
+      "epoch": 1.884861407249467,
+      "grad_norm": 0.6150320216897379,
+      "learning_rate": 6.271693543085607e-05,
+      "loss": 0.3891,
+      "step": 1768
+    },
+    {
+      "epoch": 1.8859275053304905,
+      "grad_norm": 0.4359432097990482,
+      "learning_rate": 6.269242512263945e-05,
+      "loss": 0.3929,
+      "step": 1769
+    },
+    {
+      "epoch": 1.886993603411514,
+      "grad_norm": 0.31560057559310783,
+      "learning_rate": 6.266790224399867e-05,
+      "loss": 0.3887,
+      "step": 1770
+    },
+    {
+      "epoch": 1.8880597014925373,
+      "grad_norm": 0.4422794491574624,
+      "learning_rate": 6.264336680851813e-05,
+      "loss": 0.3902,
+      "step": 1771
+    },
+    {
+      "epoch": 1.8891257995735606,
+      "grad_norm": 0.5106376023504305,
+      "learning_rate": 6.26188188297892e-05,
+      "loss": 0.3891,
+      "step": 1772
+    },
+    {
+      "epoch": 1.8901918976545842,
+      "grad_norm": 0.4415611667008926,
+      "learning_rate": 6.259425832141017e-05,
+      "loss": 0.3859,
+      "step": 1773
+    },
+    {
+      "epoch": 1.8912579957356077,
+      "grad_norm": 0.40641816501262484,
+      "learning_rate": 6.256968529698628e-05,
+      "loss": 0.3937,
+      "step": 1774
+    },
+    {
+      "epoch": 1.8923240938166312,
+      "grad_norm": 0.448521267716046,
+      "learning_rate": 6.254509977012972e-05,
+      "loss": 0.3793,
+      "step": 1775
+    },
+    {
+      "epoch": 1.8933901918976546,
+      "grad_norm": 0.5095261552719494,
+      "learning_rate": 6.252050175445959e-05,
+      "loss": 0.3887,
+      "step": 1776
+    },
+    {
+      "epoch": 1.8944562899786779,
+      "grad_norm": 0.5730353742123191,
+      "learning_rate": 6.249589126360192e-05,
+      "loss": 0.3876,
+      "step": 1777
+    },
+    {
+      "epoch": 1.8955223880597014,
+      "grad_norm": 0.6087785843423836,
+      "learning_rate": 6.247126831118962e-05,
+      "loss": 0.391,
+      "step": 1778
+    },
+    {
+      "epoch": 1.896588486140725,
+      "grad_norm": 0.5935074468788941,
+      "learning_rate": 6.244663291086256e-05,
+      "loss": 0.392,
+      "step": 1779
+    },
+    {
+      "epoch": 1.8976545842217485,
+      "grad_norm": 0.5590142718060106,
+      "learning_rate": 6.242198507626746e-05,
+      "loss": 0.3893,
+      "step": 1780
+    },
+    {
+      "epoch": 1.8987206823027718,
+      "grad_norm": 0.5336039934734914,
+      "learning_rate": 6.23973248210579e-05,
+      "loss": 0.3855,
+      "step": 1781
+    },
+    {
+      "epoch": 1.8997867803837953,
+      "grad_norm": 0.445389003613856,
+      "learning_rate": 6.237265215889444e-05,
+      "loss": 0.3886,
+      "step": 1782
+    },
+    {
+      "epoch": 1.9008528784648187,
+      "grad_norm": 0.3716042168347448,
+      "learning_rate": 6.234796710344441e-05,
+      "loss": 0.3798,
+      "step": 1783
+    },
+    {
+      "epoch": 1.9019189765458422,
+      "grad_norm": 0.4696131039929026,
+      "learning_rate": 6.232326966838207e-05,
+      "loss": 0.3924,
+      "step": 1784
+    },
+    {
+      "epoch": 1.9029850746268657,
+      "grad_norm": 0.5401242190818367,
+      "learning_rate": 6.229855986738851e-05,
+      "loss": 0.3868,
+      "step": 1785
+    },
+    {
+      "epoch": 1.9040511727078893,
+      "grad_norm": 0.4997589943313716,
+      "learning_rate": 6.227383771415166e-05,
+      "loss": 0.3889,
+      "step": 1786
+    },
+    {
+      "epoch": 1.9051172707889126,
+      "grad_norm": 0.4625159920961848,
+      "learning_rate": 6.224910322236634e-05,
+      "loss": 0.3902,
+      "step": 1787
+    },
+    {
+      "epoch": 1.906183368869936,
+      "grad_norm": 0.4473030136236895,
+      "learning_rate": 6.222435640573414e-05,
+      "loss": 0.3922,
+      "step": 1788
+    },
+    {
+      "epoch": 1.9072494669509594,
+      "grad_norm": 0.3688121331850802,
+      "learning_rate": 6.219959727796354e-05,
+      "loss": 0.3819,
+      "step": 1789
+    },
+    {
+      "epoch": 1.908315565031983,
+      "grad_norm": 0.27997416856472385,
+      "learning_rate": 6.217482585276979e-05,
+      "loss": 0.3935,
+      "step": 1790
+    },
+    {
+      "epoch": 1.9093816631130065,
+      "grad_norm": 0.2996081067229179,
+      "learning_rate": 6.215004214387497e-05,
+      "loss": 0.3889,
+      "step": 1791
+    },
+    {
+      "epoch": 1.9104477611940298,
+      "grad_norm": 0.4281784065305154,
+      "learning_rate": 6.212524616500798e-05,
+      "loss": 0.3865,
+      "step": 1792
+    },
+    {
+      "epoch": 1.9115138592750534,
+      "grad_norm": 0.44659189904413066,
+      "learning_rate": 6.210043792990449e-05,
+      "loss": 0.3868,
+      "step": 1793
+    },
+    {
+      "epoch": 1.9125799573560767,
+      "grad_norm": 0.3623900185596153,
+      "learning_rate": 6.2075617452307e-05,
+      "loss": 0.3823,
+      "step": 1794
+    },
+    {
+      "epoch": 1.9136460554371002,
+      "grad_norm": 0.37626885568568225,
+      "learning_rate": 6.205078474596473e-05,
+      "loss": 0.3873,
+      "step": 1795
+    },
+    {
+      "epoch": 1.9147121535181237,
+      "grad_norm": 0.3768400015936373,
+      "learning_rate": 6.202593982463373e-05,
+      "loss": 0.39,
+      "step": 1796
+    },
+    {
+      "epoch": 1.9157782515991473,
+      "grad_norm": 0.4025791301348382,
+      "learning_rate": 6.200108270207679e-05,
+      "loss": 0.3791,
+      "step": 1797
+    },
+    {
+      "epoch": 1.9168443496801706,
+      "grad_norm": 0.5071478998115798,
+      "learning_rate": 6.197621339206345e-05,
+      "loss": 0.3819,
+      "step": 1798
+    },
+    {
+      "epoch": 1.917910447761194,
+      "grad_norm": 0.5571967213799047,
+      "learning_rate": 6.195133190837004e-05,
+      "loss": 0.3892,
+      "step": 1799
+    },
+    {
+      "epoch": 1.9189765458422174,
+      "grad_norm": 0.5126160369273954,
+      "learning_rate": 6.192643826477959e-05,
+      "loss": 0.3835,
+      "step": 1800
+    },
+    {
+      "epoch": 1.920042643923241,
+      "grad_norm": 0.44604407636296517,
+      "learning_rate": 6.190153247508189e-05,
+      "loss": 0.3953,
+      "step": 1801
+    },
+    {
+      "epoch": 1.9211087420042645,
+      "grad_norm": 0.39601219897116957,
+      "learning_rate": 6.187661455307346e-05,
+      "loss": 0.3916,
+      "step": 1802
+    },
+    {
+      "epoch": 1.9221748400852878,
+      "grad_norm": 0.38498451897362057,
+      "learning_rate": 6.185168451255752e-05,
+      "loss": 0.393,
+      "step": 1803
+    },
+    {
+      "epoch": 1.9232409381663111,
+      "grad_norm": 0.36258385258732584,
+      "learning_rate": 6.182674236734404e-05,
+      "loss": 0.3858,
+      "step": 1804
+    },
+    {
+      "epoch": 1.9243070362473347,
+      "grad_norm": 0.29544150204233544,
+      "learning_rate": 6.180178813124965e-05,
+      "loss": 0.385,
+      "step": 1805
+    },
+    {
+      "epoch": 1.9253731343283582,
+      "grad_norm": 0.3264189147080742,
+      "learning_rate": 6.177682181809772e-05,
+      "loss": 0.3851,
+      "step": 1806
+    },
+    {
+      "epoch": 1.9264392324093818,
+      "grad_norm": 0.3541649704837544,
+      "learning_rate": 6.175184344171827e-05,
+      "loss": 0.3885,
+      "step": 1807
+    },
+    {
+      "epoch": 1.927505330490405,
+      "grad_norm": 0.30666928053826004,
+      "learning_rate": 6.172685301594802e-05,
+      "loss": 0.3873,
+      "step": 1808
+    },
+    {
+      "epoch": 1.9285714285714286,
+      "grad_norm": 0.2639778431840215,
+      "learning_rate": 6.170185055463039e-05,
+      "loss": 0.3854,
+      "step": 1809
+    },
+    {
+      "epoch": 1.929637526652452,
+      "grad_norm": 0.277287436322927,
+      "learning_rate": 6.167683607161542e-05,
+      "loss": 0.3865,
+      "step": 1810
+    },
+    {
+      "epoch": 1.9307036247334755,
+      "grad_norm": 0.3219135209443072,
+      "learning_rate": 6.165180958075985e-05,
+      "loss": 0.3848,
+      "step": 1811
+    },
+    {
+      "epoch": 1.931769722814499,
+      "grad_norm": 0.3303760398066151,
+      "learning_rate": 6.162677109592704e-05,
+      "loss": 0.386,
+      "step": 1812
+    },
+    {
+      "epoch": 1.9328358208955225,
+      "grad_norm": 0.3084131821261726,
+      "learning_rate": 6.160172063098703e-05,
+      "loss": 0.3884,
+      "step": 1813
+    },
+    {
+      "epoch": 1.9339019189765458,
+      "grad_norm": 0.31138352082161325,
+      "learning_rate": 6.157665819981646e-05,
+      "loss": 0.3832,
+      "step": 1814
+    },
+    {
+      "epoch": 1.9349680170575692,
+      "grad_norm": 0.3417574287817514,
+      "learning_rate": 6.155158381629863e-05,
+      "loss": 0.39,
+      "step": 1815
+    },
+    {
+      "epoch": 1.9360341151385927,
+      "grad_norm": 0.29411408973125125,
+      "learning_rate": 6.152649749432343e-05,
+      "loss": 0.3857,
+      "step": 1816
+    },
+    {
+      "epoch": 1.9371002132196162,
+      "grad_norm": 0.28770100823418326,
+      "learning_rate": 6.150139924778738e-05,
+      "loss": 0.3868,
+      "step": 1817
+    },
+    {
+      "epoch": 1.9381663113006398,
+      "grad_norm": 0.3600800861434762,
+      "learning_rate": 6.14762890905936e-05,
+      "loss": 0.3907,
+      "step": 1818
+    },
+    {
+      "epoch": 1.939232409381663,
+      "grad_norm": 0.40220665801015487,
+      "learning_rate": 6.145116703665184e-05,
+      "loss": 0.3839,
+      "step": 1819
+    },
+    {
+      "epoch": 1.9402985074626866,
+      "grad_norm": 0.4133033373544916,
+      "learning_rate": 6.142603309987838e-05,
+      "loss": 0.387,
+      "step": 1820
+    },
+    {
+      "epoch": 1.94136460554371,
+      "grad_norm": 0.47788470275197026,
+      "learning_rate": 6.140088729419613e-05,
+      "loss": 0.3903,
+      "step": 1821
+    },
+    {
+      "epoch": 1.9424307036247335,
+      "grad_norm": 0.4896797272268825,
+      "learning_rate": 6.137572963353455e-05,
+      "loss": 0.385,
+      "step": 1822
+    },
+    {
+      "epoch": 1.943496801705757,
+      "grad_norm": 0.45525617704254706,
+      "learning_rate": 6.135056013182969e-05,
+      "loss": 0.3949,
+      "step": 1823
+    },
+    {
+      "epoch": 1.9445628997867805,
+      "grad_norm": 0.41754979578711426,
+      "learning_rate": 6.132537880302412e-05,
+      "loss": 0.3901,
+      "step": 1824
+    },
+    {
+      "epoch": 1.9456289978678039,
+      "grad_norm": 0.36895879752123967,
+      "learning_rate": 6.130018566106702e-05,
+      "loss": 0.3845,
+      "step": 1825
+    },
+    {
+      "epoch": 1.9466950959488272,
+      "grad_norm": 0.2966243721883261,
+      "learning_rate": 6.127498071991406e-05,
+      "loss": 0.3894,
+      "step": 1826
+    },
+    {
+      "epoch": 1.9477611940298507,
+      "grad_norm": 0.29871248582481646,
+      "learning_rate": 6.12497639935275e-05,
+      "loss": 0.3859,
+      "step": 1827
+    },
+    {
+      "epoch": 1.9488272921108742,
+      "grad_norm": 0.3298826130551037,
+      "learning_rate": 6.122453549587603e-05,
+      "loss": 0.3857,
+      "step": 1828
+    },
+    {
+      "epoch": 1.9498933901918978,
+      "grad_norm": 0.3676840915758954,
+      "learning_rate": 6.119929524093499e-05,
+      "loss": 0.3894,
+      "step": 1829
+    },
+    {
+      "epoch": 1.950959488272921,
+      "grad_norm": 0.3611638397068088,
+      "learning_rate": 6.117404324268615e-05,
+      "loss": 0.3917,
+      "step": 1830
+    },
+    {
+      "epoch": 1.9520255863539444,
+      "grad_norm": 0.36356753037609535,
+      "learning_rate": 6.11487795151178e-05,
+      "loss": 0.3885,
+      "step": 1831
+    },
+    {
+      "epoch": 1.953091684434968,
+      "grad_norm": 0.4256434759309199,
+      "learning_rate": 6.112350407222472e-05,
+      "loss": 0.3871,
+      "step": 1832
+    },
+    {
+      "epoch": 1.9541577825159915,
+      "grad_norm": 0.5131488919037342,
+      "learning_rate": 6.109821692800822e-05,
+      "loss": 0.3925,
+      "step": 1833
+    },
+    {
+      "epoch": 1.955223880597015,
+      "grad_norm": 0.5200430235069933,
+      "learning_rate": 6.107291809647603e-05,
+      "loss": 0.3898,
+      "step": 1834
+    },
+    {
+      "epoch": 1.9562899786780383,
+      "grad_norm": 0.44124171195877854,
+      "learning_rate": 6.104760759164242e-05,
+      "loss": 0.3887,
+      "step": 1835
+    },
+    {
+      "epoch": 1.9573560767590619,
+      "grad_norm": 0.3699718554474287,
+      "learning_rate": 6.102228542752809e-05,
+      "loss": 0.3901,
+      "step": 1836
+    },
+    {
+      "epoch": 1.9584221748400852,
+      "grad_norm": 0.3748188674691452,
+      "learning_rate": 6.0996951618160164e-05,
+      "loss": 0.3948,
+      "step": 1837
+    },
+    {
+      "epoch": 1.9594882729211087,
+      "grad_norm": 0.3898699056134908,
+      "learning_rate": 6.097160617757231e-05,
+      "loss": 0.3906,
+      "step": 1838
+    },
+    {
+      "epoch": 1.9605543710021323,
+      "grad_norm": 0.3988044604464649,
+      "learning_rate": 6.094624911980455e-05,
+      "loss": 0.3859,
+      "step": 1839
+    },
+    {
+      "epoch": 1.9616204690831558,
+      "grad_norm": 0.3913844382375237,
+      "learning_rate": 6.0920880458903396e-05,
+      "loss": 0.3861,
+      "step": 1840
+    },
+    {
+      "epoch": 1.962686567164179,
+      "grad_norm": 0.32899811131426077,
+      "learning_rate": 6.089550020892175e-05,
+      "loss": 0.3866,
+      "step": 1841
+    },
+    {
+      "epoch": 1.9637526652452024,
+      "grad_norm": 0.31976863842060194,
+      "learning_rate": 6.0870108383918964e-05,
+      "loss": 0.3827,
+      "step": 1842
+    },
+    {
+      "epoch": 1.964818763326226,
+      "grad_norm": 0.4262488192151024,
+      "learning_rate": 6.084470499796077e-05,
+      "loss": 0.389,
+      "step": 1843
+    },
+    {
+      "epoch": 1.9658848614072495,
+      "grad_norm": 0.4077721083204829,
+      "learning_rate": 6.081929006511935e-05,
+      "loss": 0.3839,
+      "step": 1844
+    },
+    {
+      "epoch": 1.966950959488273,
+      "grad_norm": 0.3347619429742381,
+      "learning_rate": 6.079386359947325e-05,
+      "loss": 0.3831,
+      "step": 1845
+    },
+    {
+      "epoch": 1.9680170575692963,
+      "grad_norm": 0.24957267266383398,
+      "learning_rate": 6.07684256151074e-05,
+      "loss": 0.3821,
+      "step": 1846
+    },
+    {
+      "epoch": 1.9690831556503199,
+      "grad_norm": 0.2976405468436129,
+      "learning_rate": 6.074297612611312e-05,
+      "loss": 0.3833,
+      "step": 1847
+    },
+    {
+      "epoch": 1.9701492537313432,
+      "grad_norm": 0.4249489763296379,
+      "learning_rate": 6.071751514658811e-05,
+      "loss": 0.39,
+      "step": 1848
+    },
+    {
+      "epoch": 1.9712153518123667,
+      "grad_norm": 0.45599309991133324,
+      "learning_rate": 6.069204269063644e-05,
+      "loss": 0.3922,
+      "step": 1849
+    },
+    {
+      "epoch": 1.9722814498933903,
+      "grad_norm": 0.38232128582738506,
+      "learning_rate": 6.066655877236851e-05,
+      "loss": 0.3889,
+      "step": 1850
+    },
+    {
+      "epoch": 1.9733475479744138,
+      "grad_norm": 0.34695673920538606,
+      "learning_rate": 6.06410634059011e-05,
+      "loss": 0.3894,
+      "step": 1851
+    },
+    {
+      "epoch": 1.9744136460554371,
+      "grad_norm": 0.4535787896409531,
+      "learning_rate": 6.06155566053573e-05,
+      "loss": 0.3886,
+      "step": 1852
+    },
+    {
+      "epoch": 1.9754797441364604,
+      "grad_norm": 0.5028521097525349,
+      "learning_rate": 6.059003838486657e-05,
+      "loss": 0.3914,
+      "step": 1853
+    },
+    {
+      "epoch": 1.976545842217484,
+      "grad_norm": 0.43918619216494054,
+      "learning_rate": 6.056450875856467e-05,
+      "loss": 0.3849,
+      "step": 1854
+    },
+    {
+      "epoch": 1.9776119402985075,
+      "grad_norm": 0.45035177429943046,
+      "learning_rate": 6.053896774059368e-05,
+      "loss": 0.391,
+      "step": 1855
+    },
+    {
+      "epoch": 1.978678038379531,
+      "grad_norm": 0.5612538235109835,
+      "learning_rate": 6.051341534510201e-05,
+      "loss": 0.3929,
+      "step": 1856
+    },
+    {
+      "epoch": 1.9797441364605544,
+      "grad_norm": 0.6454231135981549,
+      "learning_rate": 6.048785158624436e-05,
+      "loss": 0.3871,
+      "step": 1857
+    },
+    {
+      "epoch": 1.9808102345415777,
+      "grad_norm": 0.7284809133654999,
+      "learning_rate": 6.0462276478181696e-05,
+      "loss": 0.391,
+      "step": 1858
+    },
+    {
+      "epoch": 1.9818763326226012,
+      "grad_norm": 0.7042942526408297,
+      "learning_rate": 6.043669003508134e-05,
+      "loss": 0.3827,
+      "step": 1859
+    },
+    {
+      "epoch": 1.9829424307036247,
+      "grad_norm": 0.6203970586107856,
+      "learning_rate": 6.0411092271116815e-05,
+      "loss": 0.393,
+      "step": 1860
+    },
+    {
+      "epoch": 1.9840085287846483,
+      "grad_norm": 0.5381848565811816,
+      "learning_rate": 6.038548320046797e-05,
+      "loss": 0.3884,
+      "step": 1861
+    },
+    {
+      "epoch": 1.9850746268656716,
+      "grad_norm": 0.3673908310108203,
+      "learning_rate": 6.035986283732091e-05,
+      "loss": 0.3858,
+      "step": 1862
+    },
+    {
+      "epoch": 1.9861407249466951,
+      "grad_norm": 0.2992259271172484,
+      "learning_rate": 6.0334231195867954e-05,
+      "loss": 0.3876,
+      "step": 1863
+    },
+    {
+      "epoch": 1.9872068230277184,
+      "grad_norm": 0.5048589299680658,
+      "learning_rate": 6.030858829030773e-05,
+      "loss": 0.3915,
+      "step": 1864
+    },
+    {
+      "epoch": 1.988272921108742,
+      "grad_norm": 0.6398688231174923,
+      "learning_rate": 6.0282934134845055e-05,
+      "loss": 0.3871,
+      "step": 1865
+    },
+    {
+      "epoch": 1.9893390191897655,
+      "grad_norm": 0.6041356427710289,
+      "learning_rate": 6.025726874369101e-05,
+      "loss": 0.3906,
+      "step": 1866
+    },
+    {
+      "epoch": 1.990405117270789,
+      "grad_norm": 0.436971182640899,
+      "learning_rate": 6.023159213106288e-05,
+      "loss": 0.3859,
+      "step": 1867
+    },
+    {
+      "epoch": 1.9914712153518124,
+      "grad_norm": 0.3014364362613937,
+      "learning_rate": 6.020590431118417e-05,
+      "loss": 0.3871,
+      "step": 1868
+    },
+    {
+      "epoch": 1.9925373134328357,
+      "grad_norm": 0.34001157186356745,
+      "learning_rate": 6.018020529828461e-05,
+      "loss": 0.3852,
+      "step": 1869
+    },
+    {
+      "epoch": 1.9936034115138592,
+      "grad_norm": 0.4404981881926425,
+      "learning_rate": 6.0154495106600094e-05,
+      "loss": 0.3844,
+      "step": 1870
+    },
+    {
+      "epoch": 1.9946695095948828,
+      "grad_norm": 0.48771270351289164,
+      "learning_rate": 6.012877375037278e-05,
+      "loss": 0.3866,
+      "step": 1871
+    },
+    {
+      "epoch": 1.9957356076759063,
+      "grad_norm": 0.4990494207144115,
+      "learning_rate": 6.01030412438509e-05,
+      "loss": 0.3858,
+      "step": 1872
+    },
+    {
+      "epoch": 1.9968017057569296,
+      "grad_norm": 0.468499732746224,
+      "learning_rate": 6.007729760128898e-05,
+      "loss": 0.3843,
+      "step": 1873
+    },
+    {
+      "epoch": 1.997867803837953,
+      "grad_norm": 0.41065101398677445,
+      "learning_rate": 6.0051542836947625e-05,
+      "loss": 0.3886,
+      "step": 1874
+    },
+    {
+      "epoch": 1.9989339019189765,
+      "grad_norm": 0.444948728644674,
+      "learning_rate": 6.002577696509365e-05,
+      "loss": 0.3855,
+      "step": 1875
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.5080421482170958,
+      "learning_rate": 6.000000000000001e-05,
+      "loss": 0.3761,
+      "step": 1876
+    },
+    {
+      "epoch": 2.0010660980810235,
+      "grad_norm": 0.5368826782068931,
+      "learning_rate": 5.9974211955945795e-05,
+      "loss": 0.3645,
+      "step": 1877
+    },
+    {
+      "epoch": 2.002132196162047,
+      "grad_norm": 0.5340141540515194,
+      "learning_rate": 5.9948412847216255e-05,
+      "loss": 0.3729,
+      "step": 1878
+    },
+    {
+      "epoch": 2.00319829424307,
+      "grad_norm": 0.48248902079325257,
+      "learning_rate": 5.992260268810273e-05,
+      "loss": 0.3696,
+      "step": 1879
+    },
+    {
+      "epoch": 2.0042643923240937,
+      "grad_norm": 0.4867113475226342,
+      "learning_rate": 5.989678149290274e-05,
+      "loss": 0.3736,
+      "step": 1880
+    },
+    {
+      "epoch": 2.0053304904051172,
+      "grad_norm": 0.5127638734663647,
+      "learning_rate": 5.987094927591987e-05,
+      "loss": 0.3655,
+      "step": 1881
+    },
+    {
+      "epoch": 2.0063965884861408,
+      "grad_norm": 0.5333958878222103,
+      "learning_rate": 5.9845106051463815e-05,
+      "loss": 0.3731,
+      "step": 1882
+    },
+    {
+      "epoch": 2.0074626865671643,
+      "grad_norm": 0.496391847282843,
+      "learning_rate": 5.9819251833850395e-05,
+      "loss": 0.3634,
+      "step": 1883
+    },
+    {
+      "epoch": 2.008528784648188,
+      "grad_norm": 0.3918376522955584,
+      "learning_rate": 5.979338663740149e-05,
+      "loss": 0.3748,
+      "step": 1884
+    },
+    {
+      "epoch": 2.009594882729211,
+      "grad_norm": 0.3947052793986647,
+      "learning_rate": 5.9767510476445097e-05,
+      "loss": 0.3721,
+      "step": 1885
+    },
+    {
+      "epoch": 2.0106609808102345,
+      "grad_norm": 0.41802301612396614,
+      "learning_rate": 5.974162336531522e-05,
+      "loss": 0.368,
+      "step": 1886
+    },
+    {
+      "epoch": 2.011727078891258,
+      "grad_norm": 0.4284040217771835,
+      "learning_rate": 5.9715725318352024e-05,
+      "loss": 0.366,
+      "step": 1887
+    },
+    {
+      "epoch": 2.0127931769722816,
+      "grad_norm": 0.4723022945666149,
+      "learning_rate": 5.968981634990164e-05,
+      "loss": 0.3688,
+      "step": 1888
+    },
+    {
+      "epoch": 2.013859275053305,
+      "grad_norm": 0.44534446364431024,
+      "learning_rate": 5.9663896474316325e-05,
+      "loss": 0.3685,
+      "step": 1889
+    },
+    {
+      "epoch": 2.014925373134328,
+      "grad_norm": 0.4512122720314313,
+      "learning_rate": 5.9637965705954316e-05,
+      "loss": 0.3675,
+      "step": 1890
+    },
+    {
+      "epoch": 2.0159914712153517,
+      "grad_norm": 0.41496594298021805,
+      "learning_rate": 5.961202405917993e-05,
+      "loss": 0.3708,
+      "step": 1891
+    },
+    {
+      "epoch": 2.0170575692963753,
+      "grad_norm": 0.4351254782206402,
+      "learning_rate": 5.9586071548363475e-05,
+      "loss": 0.3649,
+      "step": 1892
+    },
+    {
+      "epoch": 2.018123667377399,
+      "grad_norm": 0.42476919414349634,
+      "learning_rate": 5.95601081878813e-05,
+      "loss": 0.3663,
+      "step": 1893
+    },
+    {
+      "epoch": 2.0191897654584223,
+      "grad_norm": 0.3868685718287423,
+      "learning_rate": 5.9534133992115766e-05,
+      "loss": 0.3696,
+      "step": 1894
+    },
+    {
+      "epoch": 2.0202558635394454,
+      "grad_norm": 0.3249684583348523,
+      "learning_rate": 5.9508148975455216e-05,
+      "loss": 0.368,
+      "step": 1895
+    },
+    {
+      "epoch": 2.021321961620469,
+      "grad_norm": 0.35899462535259774,
+      "learning_rate": 5.9482153152294e-05,
+      "loss": 0.3673,
+      "step": 1896
+    },
+    {
+      "epoch": 2.0223880597014925,
+      "grad_norm": 0.3843529775756281,
+      "learning_rate": 5.945614653703245e-05,
+      "loss": 0.3671,
+      "step": 1897
+    },
+    {
+      "epoch": 2.023454157782516,
+      "grad_norm": 0.4180113835206048,
+      "learning_rate": 5.9430129144076894e-05,
+      "loss": 0.3731,
+      "step": 1898
+    },
+    {
+      "epoch": 2.0245202558635396,
+      "grad_norm": 0.37475722594758115,
+      "learning_rate": 5.9404100987839594e-05,
+      "loss": 0.3649,
+      "step": 1899
+    },
+    {
+      "epoch": 2.025586353944563,
+      "grad_norm": 0.28629598775419896,
+      "learning_rate": 5.937806208273881e-05,
+      "loss": 0.3679,
+      "step": 1900
+    },
+    {
+      "epoch": 2.026652452025586,
+      "grad_norm": 0.300199898611435,
+      "learning_rate": 5.9352012443198744e-05,
+      "loss": 0.3751,
+      "step": 1901
+    },
+    {
+      "epoch": 2.0277185501066097,
+      "grad_norm": 0.33065361440169816,
+      "learning_rate": 5.932595208364954e-05,
+      "loss": 0.3674,
+      "step": 1902
+    },
+    {
+      "epoch": 2.0287846481876333,
+      "grad_norm": 0.45055044275702016,
+      "learning_rate": 5.9299881018527286e-05,
+      "loss": 0.3669,
+      "step": 1903
+    },
+    {
+      "epoch": 2.029850746268657,
+      "grad_norm": 0.5359016977011789,
+      "learning_rate": 5.927379926227398e-05,
+      "loss": 0.3736,
+      "step": 1904
+    },
+    {
+      "epoch": 2.0309168443496803,
+      "grad_norm": 0.4728072949818313,
+      "learning_rate": 5.924770682933758e-05,
+      "loss": 0.3673,
+      "step": 1905
+    },
+    {
+      "epoch": 2.0319829424307034,
+      "grad_norm": 0.35779722355951155,
+      "learning_rate": 5.9221603734171916e-05,
+      "loss": 0.3712,
+      "step": 1906
+    },
+    {
+      "epoch": 2.033049040511727,
+      "grad_norm": 0.25778926704533023,
+      "learning_rate": 5.919548999123677e-05,
+      "loss": 0.3605,
+      "step": 1907
+    },
+    {
+      "epoch": 2.0341151385927505,
+      "grad_norm": 0.30958936271010357,
+      "learning_rate": 5.9169365614997786e-05,
+      "loss": 0.3633,
+      "step": 1908
+    },
+    {
+      "epoch": 2.035181236673774,
+      "grad_norm": 0.3966071294002005,
+      "learning_rate": 5.914323061992651e-05,
+      "loss": 0.3712,
+      "step": 1909
+    },
+    {
+      "epoch": 2.0362473347547976,
+      "grad_norm": 0.418075833743344,
+      "learning_rate": 5.9117085020500375e-05,
+      "loss": 0.367,
+      "step": 1910
+    },
+    {
+      "epoch": 2.0373134328358207,
+      "grad_norm": 0.4289422566733278,
+      "learning_rate": 5.909092883120269e-05,
+      "loss": 0.3677,
+      "step": 1911
+    },
+    {
+      "epoch": 2.038379530916844,
+      "grad_norm": 0.38407217216684475,
+      "learning_rate": 5.9064762066522614e-05,
+      "loss": 0.3657,
+      "step": 1912
+    },
+    {
+      "epoch": 2.0394456289978677,
+      "grad_norm": 0.3152326073948108,
+      "learning_rate": 5.9038584740955166e-05,
+      "loss": 0.3775,
+      "step": 1913
+    },
+    {
+      "epoch": 2.0405117270788913,
+      "grad_norm": 0.2902867416768938,
+      "learning_rate": 5.9012396869001255e-05,
+      "loss": 0.3629,
+      "step": 1914
+    },
+    {
+      "epoch": 2.041577825159915,
+      "grad_norm": 0.33254014516271413,
+      "learning_rate": 5.8986198465167566e-05,
+      "loss": 0.3652,
+      "step": 1915
+    },
+    {
+      "epoch": 2.0426439232409384,
+      "grad_norm": 0.3355861259716711,
+      "learning_rate": 5.895998954396669e-05,
+      "loss": 0.3739,
+      "step": 1916
+    },
+    {
+      "epoch": 2.0437100213219614,
+      "grad_norm": 0.33027480184371727,
+      "learning_rate": 5.893377011991696e-05,
+      "loss": 0.3642,
+      "step": 1917
+    },
+    {
+      "epoch": 2.044776119402985,
+      "grad_norm": 0.36437673592398223,
+      "learning_rate": 5.8907540207542616e-05,
+      "loss": 0.3673,
+      "step": 1918
+    },
+    {
+      "epoch": 2.0458422174840085,
+      "grad_norm": 0.37919989065780685,
+      "learning_rate": 5.888129982137364e-05,
+      "loss": 0.3697,
+      "step": 1919
+    },
+    {
+      "epoch": 2.046908315565032,
+      "grad_norm": 0.39804386255773294,
+      "learning_rate": 5.885504897594587e-05,
+      "loss": 0.3706,
+      "step": 1920
+    },
+    {
+      "epoch": 2.0479744136460556,
+      "grad_norm": 0.4048265506757637,
+      "learning_rate": 5.882878768580089e-05,
+      "loss": 0.368,
+      "step": 1921
+    },
+    {
+      "epoch": 2.0490405117270787,
+      "grad_norm": 0.39214048772357035,
+      "learning_rate": 5.880251596548608e-05,
+      "loss": 0.3729,
+      "step": 1922
+    },
+    {
+      "epoch": 2.050106609808102,
+      "grad_norm": 0.3768034830806379,
+      "learning_rate": 5.877623382955463e-05,
+      "loss": 0.366,
+      "step": 1923
+    },
+    {
+      "epoch": 2.0511727078891258,
+      "grad_norm": 0.36912589068028795,
+      "learning_rate": 5.874994129256546e-05,
+      "loss": 0.3704,
+      "step": 1924
+    },
+    {
+      "epoch": 2.0522388059701493,
+      "grad_norm": 0.34472606046121007,
+      "learning_rate": 5.872363836908328e-05,
+      "loss": 0.3605,
+      "step": 1925
+    },
+    {
+      "epoch": 2.053304904051173,
+      "grad_norm": 0.3447390942221336,
+      "learning_rate": 5.869732507367854e-05,
+      "loss": 0.3632,
+      "step": 1926
+    },
+    {
+      "epoch": 2.0543710021321964,
+      "grad_norm": 0.39703909632410184,
+      "learning_rate": 5.867100142092745e-05,
+      "loss": 0.3685,
+      "step": 1927
+    },
+    {
+      "epoch": 2.0554371002132195,
+      "grad_norm": 0.44058895200097586,
+      "learning_rate": 5.864466742541192e-05,
+      "loss": 0.3698,
+      "step": 1928
+    },
+    {
+      "epoch": 2.056503198294243,
+      "grad_norm": 0.4721233594093033,
+      "learning_rate": 5.861832310171963e-05,
+      "loss": 0.3764,
+      "step": 1929
+    },
+    {
+      "epoch": 2.0575692963752665,
+      "grad_norm": 0.4144618882599998,
+      "learning_rate": 5.8591968464443964e-05,
+      "loss": 0.3667,
+      "step": 1930
+    },
+    {
+      "epoch": 2.05863539445629,
+      "grad_norm": 0.302798084281156,
+      "learning_rate": 5.856560352818403e-05,
+      "loss": 0.3753,
+      "step": 1931
+    },
+    {
+      "epoch": 2.0597014925373136,
+      "grad_norm": 0.35481671123309483,
+      "learning_rate": 5.853922830754462e-05,
+      "loss": 0.3725,
+      "step": 1932
+    },
+    {
+      "epoch": 2.0607675906183367,
+      "grad_norm": 0.3547278123247793,
+      "learning_rate": 5.851284281713623e-05,
+      "loss": 0.3718,
+      "step": 1933
+    },
+    {
+      "epoch": 2.0618336886993602,
+      "grad_norm": 0.3404365623104542,
+      "learning_rate": 5.848644707157508e-05,
+      "loss": 0.3768,
+      "step": 1934
+    },
+    {
+      "epoch": 2.0628997867803838,
+      "grad_norm": 0.3550698273564781,
+      "learning_rate": 5.8460041085483004e-05,
+      "loss": 0.3727,
+      "step": 1935
+    },
+    {
+      "epoch": 2.0639658848614073,
+      "grad_norm": 0.35208070245068634,
+      "learning_rate": 5.8433624873487577e-05,
+      "loss": 0.37,
+      "step": 1936
+    },
+    {
+      "epoch": 2.065031982942431,
+      "grad_norm": 0.34247134066073665,
+      "learning_rate": 5.840719845022198e-05,
+      "loss": 0.3687,
+      "step": 1937
+    },
+    {
+      "epoch": 2.066098081023454,
+      "grad_norm": 0.35093264721237094,
+      "learning_rate": 5.8380761830325095e-05,
+      "loss": 0.3736,
+      "step": 1938
+    },
+    {
+      "epoch": 2.0671641791044775,
+      "grad_norm": 0.3638250203992304,
+      "learning_rate": 5.8354315028441434e-05,
+      "loss": 0.368,
+      "step": 1939
+    },
+    {
+      "epoch": 2.068230277185501,
+      "grad_norm": 0.31391019565989364,
+      "learning_rate": 5.832785805922115e-05,
+      "loss": 0.3755,
+      "step": 1940
+    },
+    {
+      "epoch": 2.0692963752665245,
+      "grad_norm": 0.32954228352966447,
+      "learning_rate": 5.830139093732003e-05,
+      "loss": 0.3695,
+      "step": 1941
+    },
+    {
+      "epoch": 2.070362473347548,
+      "grad_norm": 0.39242675952607037,
+      "learning_rate": 5.827491367739948e-05,
+      "loss": 0.3732,
+      "step": 1942
+    },
+    {
+      "epoch": 2.0714285714285716,
+      "grad_norm": 0.45718271088467405,
+      "learning_rate": 5.824842629412653e-05,
+      "loss": 0.3711,
+      "step": 1943
+    },
+    {
+      "epoch": 2.0724946695095947,
+      "grad_norm": 0.449089038740122,
+      "learning_rate": 5.822192880217381e-05,
+      "loss": 0.3682,
+      "step": 1944
+    },
+    {
+      "epoch": 2.0735607675906182,
+      "grad_norm": 0.37218601001021756,
+      "learning_rate": 5.819542121621955e-05,
+      "loss": 0.3672,
+      "step": 1945
+    },
+    {
+      "epoch": 2.074626865671642,
+      "grad_norm": 0.3365200917538621,
+      "learning_rate": 5.8168903550947586e-05,
+      "loss": 0.3667,
+      "step": 1946
+    },
+    {
+      "epoch": 2.0756929637526653,
+      "grad_norm": 0.3776897290391663,
+      "learning_rate": 5.814237582104732e-05,
+      "loss": 0.369,
+      "step": 1947
+    },
+    {
+      "epoch": 2.076759061833689,
+      "grad_norm": 0.4241260819869886,
+      "learning_rate": 5.811583804121373e-05,
+      "loss": 0.3721,
+      "step": 1948
+    },
+    {
+      "epoch": 2.077825159914712,
+      "grad_norm": 0.43159827522631167,
+      "learning_rate": 5.808929022614738e-05,
+      "loss": 0.3707,
+      "step": 1949
+    },
+    {
+      "epoch": 2.0788912579957355,
+      "grad_norm": 0.46068845826576565,
+      "learning_rate": 5.806273239055437e-05,
+      "loss": 0.3665,
+      "step": 1950
+    },
+    {
+      "epoch": 2.079957356076759,
+      "grad_norm": 0.4492734276793744,
+      "learning_rate": 5.803616454914636e-05,
+      "loss": 0.3687,
+      "step": 1951
+    },
+    {
+      "epoch": 2.0810234541577826,
+      "grad_norm": 0.4521877821462862,
+      "learning_rate": 5.800958671664057e-05,
+      "loss": 0.3638,
+      "step": 1952
+    },
+    {
+      "epoch": 2.082089552238806,
+      "grad_norm": 0.41545909536895476,
+      "learning_rate": 5.798299890775971e-05,
+      "loss": 0.3738,
+      "step": 1953
+    },
+    {
+      "epoch": 2.0831556503198296,
+      "grad_norm": 0.2972587597086915,
+      "learning_rate": 5.795640113723207e-05,
+      "loss": 0.367,
+      "step": 1954
+    },
+    {
+      "epoch": 2.0842217484008527,
+      "grad_norm": 0.27641022547500294,
+      "learning_rate": 5.7929793419791416e-05,
+      "loss": 0.3711,
+      "step": 1955
+    },
+    {
+      "epoch": 2.0852878464818763,
+      "grad_norm": 0.39396716547715316,
+      "learning_rate": 5.790317577017705e-05,
+      "loss": 0.3698,
+      "step": 1956
+    },
+    {
+      "epoch": 2.0863539445629,
+      "grad_norm": 0.38769077302687954,
+      "learning_rate": 5.787654820313376e-05,
+      "loss": 0.371,
+      "step": 1957
+    },
+    {
+      "epoch": 2.0874200426439233,
+      "grad_norm": 0.31630048451704645,
+      "learning_rate": 5.784991073341184e-05,
+      "loss": 0.3697,
+      "step": 1958
+    },
+    {
+      "epoch": 2.088486140724947,
+      "grad_norm": 0.23632112461394705,
+      "learning_rate": 5.782326337576705e-05,
+      "loss": 0.3674,
+      "step": 1959
+    },
+    {
+      "epoch": 2.08955223880597,
+      "grad_norm": 0.2503975390702189,
+      "learning_rate": 5.779660614496067e-05,
+      "loss": 0.367,
+      "step": 1960
+    },
+    {
+      "epoch": 2.0906183368869935,
+      "grad_norm": 0.31776105349081096,
+      "learning_rate": 5.776993905575939e-05,
+      "loss": 0.3656,
+      "step": 1961
+    },
+    {
+      "epoch": 2.091684434968017,
+      "grad_norm": 0.29292989917695805,
+      "learning_rate": 5.77432621229354e-05,
+      "loss": 0.3712,
+      "step": 1962
+    },
+    {
+      "epoch": 2.0927505330490406,
+      "grad_norm": 0.2736016829783113,
+      "learning_rate": 5.771657536126634e-05,
+      "loss": 0.3682,
+      "step": 1963
+    },
+    {
+      "epoch": 2.093816631130064,
+      "grad_norm": 0.301999599571911,
+      "learning_rate": 5.768987878553531e-05,
+      "loss": 0.3683,
+      "step": 1964
+    },
+    {
+      "epoch": 2.094882729211087,
+      "grad_norm": 0.3234173819094893,
+      "learning_rate": 5.766317241053077e-05,
+      "loss": 0.369,
+      "step": 1965
+    },
+    {
+      "epoch": 2.0959488272921107,
+      "grad_norm": 0.324842749445397,
+      "learning_rate": 5.763645625104673e-05,
+      "loss": 0.3634,
+      "step": 1966
+    },
+    {
+      "epoch": 2.0970149253731343,
+      "grad_norm": 0.29363848696991995,
+      "learning_rate": 5.7609730321882495e-05,
+      "loss": 0.364,
+      "step": 1967
+    },
+    {
+      "epoch": 2.098081023454158,
+      "grad_norm": 0.2645265350384704,
+      "learning_rate": 5.758299463784287e-05,
+      "loss": 0.365,
+      "step": 1968
+    },
+    {
+      "epoch": 2.0991471215351813,
+      "grad_norm": 0.2474828784122483,
+      "learning_rate": 5.755624921373805e-05,
+      "loss": 0.3743,
+      "step": 1969
+    },
+    {
+      "epoch": 2.100213219616205,
+      "grad_norm": 0.2310435326688506,
+      "learning_rate": 5.752949406438357e-05,
+      "loss": 0.372,
+      "step": 1970
+    },
+    {
+      "epoch": 2.101279317697228,
+      "grad_norm": 0.22201440839590442,
+      "learning_rate": 5.7502729204600416e-05,
+      "loss": 0.3721,
+      "step": 1971
+    },
+    {
+      "epoch": 2.1023454157782515,
+      "grad_norm": 0.2541491754650197,
+      "learning_rate": 5.747595464921493e-05,
+      "loss": 0.3715,
+      "step": 1972
+    },
+    {
+      "epoch": 2.103411513859275,
+      "grad_norm": 0.31165144411877344,
+      "learning_rate": 5.744917041305882e-05,
+      "loss": 0.3654,
+      "step": 1973
+    },
+    {
+      "epoch": 2.1044776119402986,
+      "grad_norm": 0.28922154074403444,
+      "learning_rate": 5.7422376510969165e-05,
+      "loss": 0.3728,
+      "step": 1974
+    },
+    {
+      "epoch": 2.105543710021322,
+      "grad_norm": 0.27560948128394847,
+      "learning_rate": 5.739557295778838e-05,
+      "loss": 0.3695,
+      "step": 1975
+    },
+    {
+      "epoch": 2.106609808102345,
+      "grad_norm": 0.2625778705797159,
+      "learning_rate": 5.736875976836426e-05,
+      "loss": 0.3701,
+      "step": 1976
+    },
+    {
+      "epoch": 2.1076759061833688,
+      "grad_norm": 0.3217515539116153,
+      "learning_rate": 5.73419369575499e-05,
+      "loss": 0.3724,
+      "step": 1977
+    },
+    {
+      "epoch": 2.1087420042643923,
+      "grad_norm": 0.3509916580574595,
+      "learning_rate": 5.731510454020377e-05,
+      "loss": 0.3658,
+      "step": 1978
+    },
+    {
+      "epoch": 2.109808102345416,
+      "grad_norm": 0.3227845858529989,
+      "learning_rate": 5.728826253118961e-05,
+      "loss": 0.3709,
+      "step": 1979
+    },
+    {
+      "epoch": 2.1108742004264394,
+      "grad_norm": 0.3134700723493593,
+      "learning_rate": 5.7261410945376496e-05,
+      "loss": 0.3759,
+      "step": 1980
+    },
+    {
+      "epoch": 2.111940298507463,
+      "grad_norm": 0.31039720803106174,
+      "learning_rate": 5.723454979763882e-05,
+      "loss": 0.3724,
+      "step": 1981
+    },
+    {
+      "epoch": 2.113006396588486,
+      "grad_norm": 0.32706272995692337,
+      "learning_rate": 5.720767910285626e-05,
+      "loss": 0.3722,
+      "step": 1982
+    },
+    {
+      "epoch": 2.1140724946695095,
+      "grad_norm": 0.30080459490071365,
+      "learning_rate": 5.718079887591381e-05,
+      "loss": 0.3733,
+      "step": 1983
+    },
+    {
+      "epoch": 2.115138592750533,
+      "grad_norm": 0.3043060216756326,
+      "learning_rate": 5.715390913170167e-05,
+      "loss": 0.375,
+      "step": 1984
+    },
+    {
+      "epoch": 2.1162046908315566,
+      "grad_norm": 0.28712340226310096,
+      "learning_rate": 5.7127009885115394e-05,
+      "loss": 0.3733,
+      "step": 1985
+    },
+    {
+      "epoch": 2.11727078891258,
+      "grad_norm": 0.31229581889122415,
+      "learning_rate": 5.710010115105576e-05,
+      "loss": 0.3717,
+      "step": 1986
+    },
+    {
+      "epoch": 2.1183368869936032,
+      "grad_norm": 0.3112982453246555,
+      "learning_rate": 5.707318294442881e-05,
+      "loss": 0.375,
+      "step": 1987
+    },
+    {
+      "epoch": 2.1194029850746268,
+      "grad_norm": 0.2935803415699457,
+      "learning_rate": 5.704625528014582e-05,
+      "loss": 0.3688,
+      "step": 1988
+    },
+    {
+      "epoch": 2.1204690831556503,
+      "grad_norm": 0.34557760313152336,
+      "learning_rate": 5.701931817312334e-05,
+      "loss": 0.3654,
+      "step": 1989
+    },
+    {
+      "epoch": 2.121535181236674,
+      "grad_norm": 0.32408207030896125,
+      "learning_rate": 5.6992371638283094e-05,
+      "loss": 0.3681,
+      "step": 1990
+    },
+    {
+      "epoch": 2.1226012793176974,
+      "grad_norm": 0.3209402684629185,
+      "learning_rate": 5.6965415690552083e-05,
+      "loss": 0.3691,
+      "step": 1991
+    },
+    {
+      "epoch": 2.1236673773987205,
+      "grad_norm": 0.37804296374542823,
+      "learning_rate": 5.693845034486251e-05,
+      "loss": 0.3717,
+      "step": 1992
+    },
+    {
+      "epoch": 2.124733475479744,
+      "grad_norm": 0.402239674935612,
+      "learning_rate": 5.691147561615175e-05,
+      "loss": 0.3676,
+      "step": 1993
+    },
+    {
+      "epoch": 2.1257995735607675,
+      "grad_norm": 0.40844484750895815,
+      "learning_rate": 5.688449151936243e-05,
+      "loss": 0.3724,
+      "step": 1994
+    },
+    {
+      "epoch": 2.126865671641791,
+      "grad_norm": 0.42766169546473953,
+      "learning_rate": 5.6857498069442306e-05,
+      "loss": 0.3709,
+      "step": 1995
+    },
+    {
+      "epoch": 2.1279317697228146,
+      "grad_norm": 0.37265857219610005,
+      "learning_rate": 5.683049528134437e-05,
+      "loss": 0.3737,
+      "step": 1996
+    },
+    {
+      "epoch": 2.128997867803838,
+      "grad_norm": 0.3122405362957184,
+      "learning_rate": 5.680348317002676e-05,
+      "loss": 0.3677,
+      "step": 1997
+    },
+    {
+      "epoch": 2.1300639658848612,
+      "grad_norm": 0.38579576284804146,
+      "learning_rate": 5.677646175045276e-05,
+      "loss": 0.3716,
+      "step": 1998
+    },
+    {
+      "epoch": 2.131130063965885,
+      "grad_norm": 0.49480827662022464,
+      "learning_rate": 5.674943103759086e-05,
+      "loss": 0.3778,
+      "step": 1999
+    },
+    {
+      "epoch": 2.1321961620469083,
+      "grad_norm": 0.5608752285357496,
+      "learning_rate": 5.672239104641466e-05,
+      "loss": 0.3734,
+      "step": 2000
+    },
+    {
+      "epoch": 2.133262260127932,
+      "grad_norm": 0.4928793043710862,
+      "learning_rate": 5.669534179190289e-05,
+      "loss": 0.3682,
+      "step": 2001
+    },
+    {
+      "epoch": 2.1343283582089554,
+      "grad_norm": 0.38504633757152384,
+      "learning_rate": 5.666828328903947e-05,
+      "loss": 0.3745,
+      "step": 2002
+    },
+    {
+      "epoch": 2.1353944562899785,
+      "grad_norm": 0.2618645941857424,
+      "learning_rate": 5.664121555281339e-05,
+      "loss": 0.3747,
+      "step": 2003
+    },
+    {
+      "epoch": 2.136460554371002,
+      "grad_norm": 0.2755495972773531,
+      "learning_rate": 5.661413859821874e-05,
+      "loss": 0.3791,
+      "step": 2004
+    },
+    {
+      "epoch": 2.1375266524520256,
+      "grad_norm": 0.34215532694497924,
+      "learning_rate": 5.658705244025479e-05,
+      "loss": 0.3773,
+      "step": 2005
+    },
+    {
+      "epoch": 2.138592750533049,
+      "grad_norm": 0.30880880152023477,
+      "learning_rate": 5.6559957093925826e-05,
+      "loss": 0.3748,
+      "step": 2006
+    },
+    {
+      "epoch": 2.1396588486140726,
+      "grad_norm": 0.3224170651717293,
+      "learning_rate": 5.653285257424129e-05,
+      "loss": 0.3693,
+      "step": 2007
+    },
+    {
+      "epoch": 2.140724946695096,
+      "grad_norm": 0.35605595038557464,
+      "learning_rate": 5.650573889621566e-05,
+      "loss": 0.3713,
+      "step": 2008
+    },
+    {
+      "epoch": 2.1417910447761193,
+      "grad_norm": 0.3377860560809023,
+      "learning_rate": 5.6478616074868506e-05,
+      "loss": 0.376,
+      "step": 2009
+    },
+    {
+      "epoch": 2.142857142857143,
+      "grad_norm": 0.2591813760532827,
+      "learning_rate": 5.645148412522447e-05,
+      "loss": 0.3745,
+      "step": 2010
+    },
+    {
+      "epoch": 2.1439232409381663,
+      "grad_norm": 0.28268845158045547,
+      "learning_rate": 5.642434306231323e-05,
+      "loss": 0.3708,
+      "step": 2011
+    },
+    {
+      "epoch": 2.14498933901919,
+      "grad_norm": 0.35060571169428506,
+      "learning_rate": 5.639719290116954e-05,
+      "loss": 0.3654,
+      "step": 2012
+    },
+    {
+      "epoch": 2.1460554371002134,
+      "grad_norm": 0.3694429698768061,
+      "learning_rate": 5.637003365683317e-05,
+      "loss": 0.3671,
+      "step": 2013
+    },
+    {
+      "epoch": 2.1471215351812365,
+      "grad_norm": 0.34971403364074133,
+      "learning_rate": 5.6342865344348935e-05,
+      "loss": 0.3677,
+      "step": 2014
+    },
+    {
+      "epoch": 2.14818763326226,
+      "grad_norm": 0.27907178596426185,
+      "learning_rate": 5.631568797876665e-05,
+      "loss": 0.3741,
+      "step": 2015
+    },
+    {
+      "epoch": 2.1492537313432836,
+      "grad_norm": 0.27207794051401735,
+      "learning_rate": 5.628850157514118e-05,
+      "loss": 0.3658,
+      "step": 2016
+    },
+    {
+      "epoch": 2.150319829424307,
+      "grad_norm": 0.28041793573401386,
+      "learning_rate": 5.6261306148532377e-05,
+      "loss": 0.3673,
+      "step": 2017
+    },
+    {
+      "epoch": 2.1513859275053306,
+      "grad_norm": 0.24028826632892578,
+      "learning_rate": 5.62341017140051e-05,
+      "loss": 0.3708,
+      "step": 2018
+    },
+    {
+      "epoch": 2.1524520255863537,
+      "grad_norm": 0.3066348756647207,
+      "learning_rate": 5.6206888286629186e-05,
+      "loss": 0.3722,
+      "step": 2019
+    },
+    {
+      "epoch": 2.1535181236673773,
+      "grad_norm": 0.2776104929301282,
+      "learning_rate": 5.6179665881479444e-05,
+      "loss": 0.371,
+      "step": 2020
+    },
+    {
+      "epoch": 2.154584221748401,
+      "grad_norm": 0.2932002920422834,
+      "learning_rate": 5.61524345136357e-05,
+      "loss": 0.3765,
+      "step": 2021
+    },
+    {
+      "epoch": 2.1556503198294243,
+      "grad_norm": 0.42204445414488995,
+      "learning_rate": 5.6125194198182683e-05,
+      "loss": 0.3747,
+      "step": 2022
+    },
+    {
+      "epoch": 2.156716417910448,
+      "grad_norm": 0.4892242798710215,
+      "learning_rate": 5.609794495021016e-05,
+      "loss": 0.3751,
+      "step": 2023
+    },
+    {
+      "epoch": 2.1577825159914714,
+      "grad_norm": 0.43720511006835944,
+      "learning_rate": 5.607068678481274e-05,
+      "loss": 0.368,
+      "step": 2024
+    },
+    {
+      "epoch": 2.1588486140724945,
+      "grad_norm": 0.33287359123355437,
+      "learning_rate": 5.6043419717090075e-05,
+      "loss": 0.3694,
+      "step": 2025
+    },
+    {
+      "epoch": 2.159914712153518,
+      "grad_norm": 0.28896027090363063,
+      "learning_rate": 5.6016143762146685e-05,
+      "loss": 0.3688,
+      "step": 2026
+    },
+    {
+      "epoch": 2.1609808102345416,
+      "grad_norm": 0.2887140479083496,
+      "learning_rate": 5.598885893509203e-05,
+      "loss": 0.3737,
+      "step": 2027
+    },
+    {
+      "epoch": 2.162046908315565,
+      "grad_norm": 0.2695168201652278,
+      "learning_rate": 5.59615652510405e-05,
+      "loss": 0.3705,
+      "step": 2028
+    },
+    {
+      "epoch": 2.1631130063965887,
+      "grad_norm": 0.2407806636654491,
+      "learning_rate": 5.593426272511136e-05,
+      "loss": 0.3698,
+      "step": 2029
+    },
+    {
+      "epoch": 2.1641791044776117,
+      "grad_norm": 0.32911747478553777,
+      "learning_rate": 5.590695137242881e-05,
+      "loss": 0.3752,
+      "step": 2030
+    },
+    {
+      "epoch": 2.1652452025586353,
+      "grad_norm": 0.3868451813947213,
+      "learning_rate": 5.587963120812189e-05,
+      "loss": 0.3794,
+      "step": 2031
+    },
+    {
+      "epoch": 2.166311300639659,
+      "grad_norm": 0.34786974694646033,
+      "learning_rate": 5.585230224732458e-05,
+      "loss": 0.3652,
+      "step": 2032
+    },
+    {
+      "epoch": 2.1673773987206824,
+      "grad_norm": 0.35478109819952885,
+      "learning_rate": 5.582496450517569e-05,
+      "loss": 0.3795,
+      "step": 2033
+    },
+    {
+      "epoch": 2.168443496801706,
+      "grad_norm": 0.3508750515755295,
+      "learning_rate": 5.5797617996818915e-05,
+      "loss": 0.3646,
+      "step": 2034
+    },
+    {
+      "epoch": 2.1695095948827294,
+      "grad_norm": 0.3447886192528013,
+      "learning_rate": 5.57702627374028e-05,
+      "loss": 0.3676,
+      "step": 2035
+    },
+    {
+      "epoch": 2.1705756929637525,
+      "grad_norm": 0.41611174037016146,
+      "learning_rate": 5.5742898742080726e-05,
+      "loss": 0.3703,
+      "step": 2036
+    },
+    {
+      "epoch": 2.171641791044776,
+      "grad_norm": 0.4442311781729241,
+      "learning_rate": 5.5715526026010935e-05,
+      "loss": 0.3659,
+      "step": 2037
+    },
+    {
+      "epoch": 2.1727078891257996,
+      "grad_norm": 0.3670273857196127,
+      "learning_rate": 5.568814460435649e-05,
+      "loss": 0.3768,
+      "step": 2038
+    },
+    {
+      "epoch": 2.173773987206823,
+      "grad_norm": 0.2593513887640986,
+      "learning_rate": 5.5660754492285264e-05,
+      "loss": 0.3669,
+      "step": 2039
+    },
+    {
+      "epoch": 2.1748400852878467,
+      "grad_norm": 0.28282579166446825,
+      "learning_rate": 5.563335570496996e-05,
+      "loss": 0.3739,
+      "step": 2040
+    },
+    {
+      "epoch": 2.1759061833688698,
+      "grad_norm": 0.36977014027652605,
+      "learning_rate": 5.560594825758809e-05,
+      "loss": 0.3615,
+      "step": 2041
+    },
+    {
+      "epoch": 2.1769722814498933,
+      "grad_norm": 0.37906406180450586,
+      "learning_rate": 5.557853216532194e-05,
+      "loss": 0.374,
+      "step": 2042
+    },
+    {
+      "epoch": 2.178038379530917,
+      "grad_norm": 0.3003073120798734,
+      "learning_rate": 5.555110744335863e-05,
+      "loss": 0.3688,
+      "step": 2043
+    },
+    {
+      "epoch": 2.1791044776119404,
+      "grad_norm": 0.30581675422030474,
+      "learning_rate": 5.552367410688999e-05,
+      "loss": 0.3709,
+      "step": 2044
+    },
+    {
+      "epoch": 2.180170575692964,
+      "grad_norm": 0.4243247655082495,
+      "learning_rate": 5.5496232171112703e-05,
+      "loss": 0.3725,
+      "step": 2045
+    },
+    {
+      "epoch": 2.181236673773987,
+      "grad_norm": 0.5418944865918137,
+      "learning_rate": 5.546878165122815e-05,
+      "loss": 0.3747,
+      "step": 2046
+    },
+    {
+      "epoch": 2.1823027718550105,
+      "grad_norm": 0.5717964571682613,
+      "learning_rate": 5.544132256244249e-05,
+      "loss": 0.3701,
+      "step": 2047
+    },
+    {
+      "epoch": 2.183368869936034,
+      "grad_norm": 0.5656819581120287,
+      "learning_rate": 5.5413854919966654e-05,
+      "loss": 0.3721,
+      "step": 2048
+    },
+    {
+      "epoch": 2.1844349680170576,
+      "grad_norm": 0.48959229163286955,
+      "learning_rate": 5.538637873901626e-05,
+      "loss": 0.3697,
+      "step": 2049
+    },
+    {
+      "epoch": 2.185501066098081,
+      "grad_norm": 0.39474515330398696,
+      "learning_rate": 5.5358894034811705e-05,
+      "loss": 0.3723,
+      "step": 2050
+    },
+    {
+      "epoch": 2.1865671641791047,
+      "grad_norm": 0.29811160693240174,
+      "learning_rate": 5.533140082257808e-05,
+      "loss": 0.3765,
+      "step": 2051
+    },
+    {
+      "epoch": 2.1876332622601278,
+      "grad_norm": 0.2398629529634624,
+      "learning_rate": 5.530389911754519e-05,
+      "loss": 0.3702,
+      "step": 2052
+    },
+    {
+      "epoch": 2.1886993603411513,
+      "grad_norm": 0.2641014816643812,
+      "learning_rate": 5.527638893494755e-05,
+      "loss": 0.375,
+      "step": 2053
+    },
+    {
+      "epoch": 2.189765458422175,
+      "grad_norm": 0.26988134759785193,
+      "learning_rate": 5.5248870290024396e-05,
+      "loss": 0.3702,
+      "step": 2054
+    },
+    {
+      "epoch": 2.1908315565031984,
+      "grad_norm": 0.26917332673147915,
+      "learning_rate": 5.5221343198019596e-05,
+      "loss": 0.3695,
+      "step": 2055
+    },
+    {
+      "epoch": 2.191897654584222,
+      "grad_norm": 0.32231195092404136,
+      "learning_rate": 5.519380767418176e-05,
+      "loss": 0.3685,
+      "step": 2056
+    },
+    {
+      "epoch": 2.192963752665245,
+      "grad_norm": 0.3242363910277114,
+      "learning_rate": 5.5166263733764096e-05,
+      "loss": 0.3695,
+      "step": 2057
+    },
+    {
+      "epoch": 2.1940298507462686,
+      "grad_norm": 0.30073683294689246,
+      "learning_rate": 5.5138711392024545e-05,
+      "loss": 0.3704,
+      "step": 2058
+    },
+    {
+      "epoch": 2.195095948827292,
+      "grad_norm": 0.26491871391522587,
+      "learning_rate": 5.5111150664225665e-05,
+      "loss": 0.3715,
+      "step": 2059
+    },
+    {
+      "epoch": 2.1961620469083156,
+      "grad_norm": 0.294788631674408,
+      "learning_rate": 5.508358156563466e-05,
+      "loss": 0.3695,
+      "step": 2060
+    },
+    {
+      "epoch": 2.197228144989339,
+      "grad_norm": 0.3148068642489263,
+      "learning_rate": 5.505600411152341e-05,
+      "loss": 0.3723,
+      "step": 2061
+    },
+    {
+      "epoch": 2.1982942430703627,
+      "grad_norm": 0.2787016615117562,
+      "learning_rate": 5.502841831716833e-05,
+      "loss": 0.3739,
+      "step": 2062
+    },
+    {
+      "epoch": 2.199360341151386,
+      "grad_norm": 0.2643411187566,
+      "learning_rate": 5.5000824197850575e-05,
+      "loss": 0.3739,
+      "step": 2063
+    },
+    {
+      "epoch": 2.2004264392324093,
+      "grad_norm": 0.2991815067944646,
+      "learning_rate": 5.497322176885582e-05,
+      "loss": 0.3734,
+      "step": 2064
+    },
+    {
+      "epoch": 2.201492537313433,
+      "grad_norm": 0.31374929510374666,
+      "learning_rate": 5.494561104547439e-05,
+      "loss": 0.3668,
+      "step": 2065
+    },
+    {
+      "epoch": 2.2025586353944564,
+      "grad_norm": 0.2983982343106373,
+      "learning_rate": 5.491799204300119e-05,
+      "loss": 0.3756,
+      "step": 2066
+    },
+    {
+      "epoch": 2.20362473347548,
+      "grad_norm": 0.3055575842818981,
+      "learning_rate": 5.489036477673571e-05,
+      "loss": 0.3739,
+      "step": 2067
+    },
+    {
+      "epoch": 2.204690831556503,
+      "grad_norm": 0.36274620979363453,
+      "learning_rate": 5.486272926198202e-05,
+      "loss": 0.3701,
+      "step": 2068
+    },
+    {
+      "epoch": 2.2057569296375266,
+      "grad_norm": 0.38535842211347343,
+      "learning_rate": 5.483508551404875e-05,
+      "loss": 0.3683,
+      "step": 2069
+    },
+    {
+      "epoch": 2.20682302771855,
+      "grad_norm": 0.36592619429217743,
+      "learning_rate": 5.4807433548249106e-05,
+      "loss": 0.3719,
+      "step": 2070
+    },
+    {
+      "epoch": 2.2078891257995736,
+      "grad_norm": 0.320538300126649,
+      "learning_rate": 5.4779773379900856e-05,
+      "loss": 0.3681,
+      "step": 2071
+    },
+    {
+      "epoch": 2.208955223880597,
+      "grad_norm": 0.3284949841600627,
+      "learning_rate": 5.4752105024326265e-05,
+      "loss": 0.3722,
+      "step": 2072
+    },
+    {
+      "epoch": 2.2100213219616203,
+      "grad_norm": 0.36352567522233187,
+      "learning_rate": 5.4724428496852184e-05,
+      "loss": 0.3742,
+      "step": 2073
+    },
+    {
+      "epoch": 2.211087420042644,
+      "grad_norm": 0.3708212783274699,
+      "learning_rate": 5.469674381280997e-05,
+      "loss": 0.3723,
+      "step": 2074
+    },
+    {
+      "epoch": 2.2121535181236673,
+      "grad_norm": 0.34144099793040533,
+      "learning_rate": 5.4669050987535504e-05,
+      "loss": 0.3716,
+      "step": 2075
+    },
+    {
+      "epoch": 2.213219616204691,
+      "grad_norm": 0.31816989140951235,
+      "learning_rate": 5.464135003636914e-05,
+      "loss": 0.3651,
+      "step": 2076
+    },
+    {
+      "epoch": 2.2142857142857144,
+      "grad_norm": 0.29852456569861313,
+      "learning_rate": 5.461364097465581e-05,
+      "loss": 0.3711,
+      "step": 2077
+    },
+    {
+      "epoch": 2.2153518123667375,
+      "grad_norm": 0.2455931272133471,
+      "learning_rate": 5.4585923817744864e-05,
+      "loss": 0.3718,
+      "step": 2078
+    },
+    {
+      "epoch": 2.216417910447761,
+      "grad_norm": 0.26413290094993985,
+      "learning_rate": 5.455819858099018e-05,
+      "loss": 0.365,
+      "step": 2079
+    },
+    {
+      "epoch": 2.2174840085287846,
+      "grad_norm": 0.24066490453483533,
+      "learning_rate": 5.4530465279750087e-05,
+      "loss": 0.3705,
+      "step": 2080
+    },
+    {
+      "epoch": 2.218550106609808,
+      "grad_norm": 0.2618435559346739,
+      "learning_rate": 5.450272392938742e-05,
+      "loss": 0.3735,
+      "step": 2081
+    },
+    {
+      "epoch": 2.2196162046908317,
+      "grad_norm": 0.24293872764002158,
+      "learning_rate": 5.4474974545269394e-05,
+      "loss": 0.3717,
+      "step": 2082
+    },
+    {
+      "epoch": 2.220682302771855,
+      "grad_norm": 0.2368196730572011,
+      "learning_rate": 5.444721714276778e-05,
+      "loss": 0.3639,
+      "step": 2083
+    },
+    {
+      "epoch": 2.2217484008528783,
+      "grad_norm": 0.27260123271017317,
+      "learning_rate": 5.44194517372587e-05,
+      "loss": 0.3685,
+      "step": 2084
+    },
+    {
+      "epoch": 2.222814498933902,
+      "grad_norm": 0.270272968775527,
+      "learning_rate": 5.439167834412277e-05,
+      "loss": 0.3672,
+      "step": 2085
+    },
+    {
+      "epoch": 2.2238805970149254,
+      "grad_norm": 0.29536457376389963,
+      "learning_rate": 5.436389697874499e-05,
+      "loss": 0.3656,
+      "step": 2086
+    },
+    {
+      "epoch": 2.224946695095949,
+      "grad_norm": 0.38141850154306833,
+      "learning_rate": 5.4336107656514796e-05,
+      "loss": 0.3718,
+      "step": 2087
+    },
+    {
+      "epoch": 2.2260127931769724,
+      "grad_norm": 0.38555373110953656,
+      "learning_rate": 5.430831039282603e-05,
+      "loss": 0.3653,
+      "step": 2088
+    },
+    {
+      "epoch": 2.227078891257996,
+      "grad_norm": 0.33806578284967886,
+      "learning_rate": 5.428050520307693e-05,
+      "loss": 0.3651,
+      "step": 2089
+    },
+    {
+      "epoch": 2.228144989339019,
+      "grad_norm": 0.3782268091971875,
+      "learning_rate": 5.425269210267013e-05,
+      "loss": 0.3711,
+      "step": 2090
+    },
+    {
+      "epoch": 2.2292110874200426,
+      "grad_norm": 0.4292891226917417,
+      "learning_rate": 5.422487110701263e-05,
+      "loss": 0.3702,
+      "step": 2091
+    },
+    {
+      "epoch": 2.230277185501066,
+      "grad_norm": 0.41142332636108486,
+      "learning_rate": 5.419704223151584e-05,
+      "loss": 0.3661,
+      "step": 2092
+    },
+    {
+      "epoch": 2.2313432835820897,
+      "grad_norm": 0.3326023012950207,
+      "learning_rate": 5.416920549159549e-05,
+      "loss": 0.3721,
+      "step": 2093
+    },
+    {
+      "epoch": 2.232409381663113,
+      "grad_norm": 0.26521116383654986,
+      "learning_rate": 5.4141360902671696e-05,
+      "loss": 0.3723,
+      "step": 2094
+    },
+    {
+      "epoch": 2.2334754797441363,
+      "grad_norm": 0.2612433990558302,
+      "learning_rate": 5.411350848016891e-05,
+      "loss": 0.3601,
+      "step": 2095
+    },
+    {
+      "epoch": 2.23454157782516,
+      "grad_norm": 0.3434339511954728,
+      "learning_rate": 5.4085648239515914e-05,
+      "loss": 0.3692,
+      "step": 2096
+    },
+    {
+      "epoch": 2.2356076759061834,
+      "grad_norm": 0.3918166163519152,
+      "learning_rate": 5.4057780196145856e-05,
+      "loss": 0.3601,
+      "step": 2097
+    },
+    {
+      "epoch": 2.236673773987207,
+      "grad_norm": 0.3818529019674707,
+      "learning_rate": 5.402990436549617e-05,
+      "loss": 0.3703,
+      "step": 2098
+    },
+    {
+      "epoch": 2.2377398720682304,
+      "grad_norm": 0.36098975401837946,
+      "learning_rate": 5.4002020763008624e-05,
+      "loss": 0.3711,
+      "step": 2099
+    },
+    {
+      "epoch": 2.2388059701492535,
+      "grad_norm": 0.31093262243992925,
+      "learning_rate": 5.397412940412927e-05,
+      "loss": 0.3698,
+      "step": 2100
+    },
+    {
+      "epoch": 2.239872068230277,
+      "grad_norm": 0.3234247785754041,
+      "learning_rate": 5.39462303043085e-05,
+      "loss": 0.3699,
+      "step": 2101
+    },
+    {
+      "epoch": 2.2409381663113006,
+      "grad_norm": 0.3388924115506885,
+      "learning_rate": 5.391832347900095e-05,
+      "loss": 0.3714,
+      "step": 2102
+    },
+    {
+      "epoch": 2.242004264392324,
+      "grad_norm": 0.2867861763859992,
+      "learning_rate": 5.389040894366554e-05,
+      "loss": 0.3681,
+      "step": 2103
+    },
+    {
+      "epoch": 2.2430703624733477,
+      "grad_norm": 0.27923094302877083,
+      "learning_rate": 5.386248671376549e-05,
+      "loss": 0.3711,
+      "step": 2104
+    },
+    {
+      "epoch": 2.2441364605543708,
+      "grad_norm": 0.31103472441375013,
+      "learning_rate": 5.383455680476824e-05,
+      "loss": 0.3647,
+      "step": 2105
+    },
+    {
+      "epoch": 2.2452025586353943,
+      "grad_norm": 0.2851322771459375,
+      "learning_rate": 5.380661923214553e-05,
+      "loss": 0.3692,
+      "step": 2106
+    },
+    {
+      "epoch": 2.246268656716418,
+      "grad_norm": 0.32537762914509355,
+      "learning_rate": 5.377867401137332e-05,
+      "loss": 0.372,
+      "step": 2107
+    },
+    {
+      "epoch": 2.2473347547974414,
+      "grad_norm": 0.36286224273560747,
+      "learning_rate": 5.375072115793181e-05,
+      "loss": 0.3736,
+      "step": 2108
+    },
+    {
+      "epoch": 2.248400852878465,
+      "grad_norm": 0.39983878699233394,
+      "learning_rate": 5.3722760687305414e-05,
+      "loss": 0.3726,
+      "step": 2109
+    },
+    {
+      "epoch": 2.2494669509594885,
+      "grad_norm": 0.3468587900831904,
+      "learning_rate": 5.3694792614982794e-05,
+      "loss": 0.3701,
+      "step": 2110
+    },
+    {
+      "epoch": 2.2505330490405115,
+      "grad_norm": 0.29431394759332513,
+      "learning_rate": 5.366681695645681e-05,
+      "loss": 0.3741,
+      "step": 2111
+    },
+    {
+      "epoch": 2.251599147121535,
+      "grad_norm": 0.292448439837048,
+      "learning_rate": 5.363883372722452e-05,
+      "loss": 0.3705,
+      "step": 2112
+    },
+    {
+      "epoch": 2.2526652452025586,
+      "grad_norm": 0.3477064399327982,
+      "learning_rate": 5.3610842942787156e-05,
+      "loss": 0.3698,
+      "step": 2113
+    },
+    {
+      "epoch": 2.253731343283582,
+      "grad_norm": 0.30311348425697426,
+      "learning_rate": 5.3582844618650196e-05,
+      "loss": 0.3734,
+      "step": 2114
+    },
+    {
+      "epoch": 2.2547974413646057,
+      "grad_norm": 0.29982583065786783,
+      "learning_rate": 5.355483877032324e-05,
+      "loss": 0.376,
+      "step": 2115
+    },
+    {
+      "epoch": 2.2558635394456292,
+      "grad_norm": 0.2708619876359533,
+      "learning_rate": 5.352682541332006e-05,
+      "loss": 0.3743,
+      "step": 2116
+    },
+    {
+      "epoch": 2.2569296375266523,
+      "grad_norm": 0.28890106497206663,
+      "learning_rate": 5.349880456315862e-05,
+      "loss": 0.3715,
+      "step": 2117
+    },
+    {
+      "epoch": 2.257995735607676,
+      "grad_norm": 0.2663781257854233,
+      "learning_rate": 5.347077623536099e-05,
+      "loss": 0.3679,
+      "step": 2118
+    },
+    {
+      "epoch": 2.2590618336886994,
+      "grad_norm": 0.2674833564078522,
+      "learning_rate": 5.344274044545344e-05,
+      "loss": 0.3669,
+      "step": 2119
+    },
+    {
+      "epoch": 2.260127931769723,
+      "grad_norm": 0.28497773715596586,
+      "learning_rate": 5.3414697208966315e-05,
+      "loss": 0.3669,
+      "step": 2120
+    },
+    {
+      "epoch": 2.2611940298507465,
+      "grad_norm": 0.26215123696658,
+      "learning_rate": 5.3386646541434126e-05,
+      "loss": 0.3706,
+      "step": 2121
+    },
+    {
+      "epoch": 2.2622601279317696,
+      "grad_norm": 0.32005074927132454,
+      "learning_rate": 5.335858845839546e-05,
+      "loss": 0.3702,
+      "step": 2122
+    },
+    {
+      "epoch": 2.263326226012793,
+      "grad_norm": 0.32411250381064693,
+      "learning_rate": 5.333052297539308e-05,
+      "loss": 0.3699,
+      "step": 2123
+    },
+    {
+      "epoch": 2.2643923240938166,
+      "grad_norm": 0.29000255006330194,
+      "learning_rate": 5.3302450107973766e-05,
+      "loss": 0.368,
+      "step": 2124
+    },
+    {
+      "epoch": 2.26545842217484,
+      "grad_norm": 0.27431618415808595,
+      "learning_rate": 5.327436987168844e-05,
+      "loss": 0.371,
+      "step": 2125
+    },
+    {
+      "epoch": 2.2665245202558637,
+      "grad_norm": 0.378024359549344,
+      "learning_rate": 5.324628228209209e-05,
+      "loss": 0.3721,
+      "step": 2126
+    },
+    {
+      "epoch": 2.267590618336887,
+      "grad_norm": 0.43512140087622997,
+      "learning_rate": 5.321818735474379e-05,
+      "loss": 0.3715,
+      "step": 2127
+    },
+    {
+      "epoch": 2.2686567164179103,
+      "grad_norm": 0.3750669758497868,
+      "learning_rate": 5.3190085105206666e-05,
+      "loss": 0.3764,
+      "step": 2128
+    },
+    {
+      "epoch": 2.269722814498934,
+      "grad_norm": 0.28073557975371866,
+      "learning_rate": 5.31619755490479e-05,
+      "loss": 0.3748,
+      "step": 2129
+    },
+    {
+      "epoch": 2.2707889125799574,
+      "grad_norm": 0.2949560743757881,
+      "learning_rate": 5.3133858701838735e-05,
+      "loss": 0.3709,
+      "step": 2130
+    },
+    {
+      "epoch": 2.271855010660981,
+      "grad_norm": 0.3204542728152737,
+      "learning_rate": 5.310573457915443e-05,
+      "loss": 0.3703,
+      "step": 2131
+    },
+    {
+      "epoch": 2.272921108742004,
+      "grad_norm": 0.2868376825764523,
+      "learning_rate": 5.307760319657429e-05,
+      "loss": 0.3708,
+      "step": 2132
+    },
+    {
+      "epoch": 2.2739872068230276,
+      "grad_norm": 0.251294097973084,
+      "learning_rate": 5.3049464569681654e-05,
+      "loss": 0.3689,
+      "step": 2133
+    },
+    {
+      "epoch": 2.275053304904051,
+      "grad_norm": 0.27463227847743765,
+      "learning_rate": 5.3021318714063834e-05,
+      "loss": 0.3761,
+      "step": 2134
+    },
+    {
+      "epoch": 2.2761194029850746,
+      "grad_norm": 0.33592667040636226,
+      "learning_rate": 5.299316564531219e-05,
+      "loss": 0.3731,
+      "step": 2135
+    },
+    {
+      "epoch": 2.277185501066098,
+      "grad_norm": 0.38903214147148135,
+      "learning_rate": 5.296500537902205e-05,
+      "loss": 0.37,
+      "step": 2136
+    },
+    {
+      "epoch": 2.2782515991471217,
+      "grad_norm": 0.34915279703933755,
+      "learning_rate": 5.293683793079274e-05,
+      "loss": 0.3649,
+      "step": 2137
+    },
+    {
+      "epoch": 2.279317697228145,
+      "grad_norm": 0.3506016673900483,
+      "learning_rate": 5.2908663316227577e-05,
+      "loss": 0.3734,
+      "step": 2138
+    },
+    {
+      "epoch": 2.2803837953091683,
+      "grad_norm": 0.3878071683023028,
+      "learning_rate": 5.28804815509338e-05,
+      "loss": 0.3707,
+      "step": 2139
+    },
+    {
+      "epoch": 2.281449893390192,
+      "grad_norm": 0.35568233275956773,
+      "learning_rate": 5.285229265052268e-05,
+      "loss": 0.3758,
+      "step": 2140
+    },
+    {
+      "epoch": 2.2825159914712154,
+      "grad_norm": 0.3418784225462311,
+      "learning_rate": 5.2824096630609385e-05,
+      "loss": 0.37,
+      "step": 2141
+    },
+    {
+      "epoch": 2.283582089552239,
+      "grad_norm": 0.3607422406160099,
+      "learning_rate": 5.2795893506813024e-05,
+      "loss": 0.3728,
+      "step": 2142
+    },
+    {
+      "epoch": 2.2846481876332625,
+      "grad_norm": 0.3721645564569364,
+      "learning_rate": 5.276768329475671e-05,
+      "loss": 0.3717,
+      "step": 2143
+    },
+    {
+      "epoch": 2.2857142857142856,
+      "grad_norm": 0.3220957092565807,
+      "learning_rate": 5.2739466010067385e-05,
+      "loss": 0.3739,
+      "step": 2144
+    },
+    {
+      "epoch": 2.286780383795309,
+      "grad_norm": 0.2920281939799502,
+      "learning_rate": 5.271124166837599e-05,
+      "loss": 0.366,
+      "step": 2145
+    },
+    {
+      "epoch": 2.2878464818763327,
+      "grad_norm": 0.3080101489292651,
+      "learning_rate": 5.2683010285317333e-05,
+      "loss": 0.3675,
+      "step": 2146
+    },
+    {
+      "epoch": 2.288912579957356,
+      "grad_norm": 0.3042289029809852,
+      "learning_rate": 5.265477187653012e-05,
+      "loss": 0.369,
+      "step": 2147
+    },
+    {
+      "epoch": 2.2899786780383797,
+      "grad_norm": 0.3475082339985931,
+      "learning_rate": 5.262652645765699e-05,
+      "loss": 0.3702,
+      "step": 2148
+    },
+    {
+      "epoch": 2.291044776119403,
+      "grad_norm": 0.3115045075355101,
+      "learning_rate": 5.2598274044344414e-05,
+      "loss": 0.3687,
+      "step": 2149
+    },
+    {
+      "epoch": 2.2921108742004264,
+      "grad_norm": 0.32610593675887717,
+      "learning_rate": 5.257001465224278e-05,
+      "loss": 0.3692,
+      "step": 2150
+    },
+    {
+      "epoch": 2.29317697228145,
+      "grad_norm": 0.37456808053238067,
+      "learning_rate": 5.2541748297006306e-05,
+      "loss": 0.3721,
+      "step": 2151
+    },
+    {
+      "epoch": 2.2942430703624734,
+      "grad_norm": 0.3350461025624831,
+      "learning_rate": 5.251347499429309e-05,
+      "loss": 0.3732,
+      "step": 2152
+    },
+    {
+      "epoch": 2.295309168443497,
+      "grad_norm": 0.2568059568097965,
+      "learning_rate": 5.2485194759765074e-05,
+      "loss": 0.3691,
+      "step": 2153
+    },
+    {
+      "epoch": 2.29637526652452,
+      "grad_norm": 0.27831384466881337,
+      "learning_rate": 5.245690760908803e-05,
+      "loss": 0.3686,
+      "step": 2154
+    },
+    {
+      "epoch": 2.2974413646055436,
+      "grad_norm": 0.26602965147812224,
+      "learning_rate": 5.242861355793157e-05,
+      "loss": 0.3645,
+      "step": 2155
+    },
+    {
+      "epoch": 2.298507462686567,
+      "grad_norm": 0.2587039755436819,
+      "learning_rate": 5.240031262196914e-05,
+      "loss": 0.3757,
+      "step": 2156
+    },
+    {
+      "epoch": 2.2995735607675907,
+      "grad_norm": 0.27151417310136094,
+      "learning_rate": 5.237200481687798e-05,
+      "loss": 0.368,
+      "step": 2157
+    },
+    {
+      "epoch": 2.300639658848614,
+      "grad_norm": 0.26058762590380463,
+      "learning_rate": 5.234369015833914e-05,
+      "loss": 0.3697,
+      "step": 2158
+    },
+    {
+      "epoch": 2.3017057569296373,
+      "grad_norm": 0.33421242107141547,
+      "learning_rate": 5.2315368662037485e-05,
+      "loss": 0.3675,
+      "step": 2159
+    },
+    {
+      "epoch": 2.302771855010661,
+      "grad_norm": 0.3659615591367041,
+      "learning_rate": 5.228704034366162e-05,
+      "loss": 0.3703,
+      "step": 2160
+    },
+    {
+      "epoch": 2.3038379530916844,
+      "grad_norm": 0.27978795201644363,
+      "learning_rate": 5.2258705218904005e-05,
+      "loss": 0.371,
+      "step": 2161
+    },
+    {
+      "epoch": 2.304904051172708,
+      "grad_norm": 0.21852529768913895,
+      "learning_rate": 5.2230363303460794e-05,
+      "loss": 0.3683,
+      "step": 2162
+    },
+    {
+      "epoch": 2.3059701492537314,
+      "grad_norm": 0.3118783856447842,
+      "learning_rate": 5.220201461303193e-05,
+      "loss": 0.3714,
+      "step": 2163
+    },
+    {
+      "epoch": 2.307036247334755,
+      "grad_norm": 0.3392350656900747,
+      "learning_rate": 5.2173659163321145e-05,
+      "loss": 0.3682,
+      "step": 2164
+    },
+    {
+      "epoch": 2.308102345415778,
+      "grad_norm": 0.32734701181884485,
+      "learning_rate": 5.2145296970035846e-05,
+      "loss": 0.3744,
+      "step": 2165
+    },
+    {
+      "epoch": 2.3091684434968016,
+      "grad_norm": 0.303313761017607,
+      "learning_rate": 5.211692804888726e-05,
+      "loss": 0.3744,
+      "step": 2166
+    },
+    {
+      "epoch": 2.310234541577825,
+      "grad_norm": 0.2905681835769858,
+      "learning_rate": 5.2088552415590254e-05,
+      "loss": 0.3713,
+      "step": 2167
+    },
+    {
+      "epoch": 2.3113006396588487,
+      "grad_norm": 0.3534515955282182,
+      "learning_rate": 5.2060170085863484e-05,
+      "loss": 0.3703,
+      "step": 2168
+    },
+    {
+      "epoch": 2.3123667377398722,
+      "grad_norm": 0.4264619477891961,
+      "learning_rate": 5.203178107542925e-05,
+      "loss": 0.3715,
+      "step": 2169
+    },
+    {
+      "epoch": 2.3134328358208958,
+      "grad_norm": 0.47866741810652247,
+      "learning_rate": 5.200338540001364e-05,
+      "loss": 0.3706,
+      "step": 2170
+    },
+    {
+      "epoch": 2.314498933901919,
+      "grad_norm": 0.4733916257306672,
+      "learning_rate": 5.1974983075346335e-05,
+      "loss": 0.3639,
+      "step": 2171
+    },
+    {
+      "epoch": 2.3155650319829424,
+      "grad_norm": 0.42832051016045336,
+      "learning_rate": 5.194657411716076e-05,
+      "loss": 0.3698,
+      "step": 2172
+    },
+    {
+      "epoch": 2.316631130063966,
+      "grad_norm": 0.43789085333356936,
+      "learning_rate": 5.1918158541194014e-05,
+      "loss": 0.3735,
+      "step": 2173
+    },
+    {
+      "epoch": 2.3176972281449895,
+      "grad_norm": 0.4573284771159797,
+      "learning_rate": 5.188973636318684e-05,
+      "loss": 0.3713,
+      "step": 2174
+    },
+    {
+      "epoch": 2.318763326226013,
+      "grad_norm": 0.3916370694042795,
+      "learning_rate": 5.1861307598883644e-05,
+      "loss": 0.3723,
+      "step": 2175
+    },
+    {
+      "epoch": 2.319829424307036,
+      "grad_norm": 0.2585801987726249,
+      "learning_rate": 5.1832872264032495e-05,
+      "loss": 0.37,
+      "step": 2176
+    },
+    {
+      "epoch": 2.3208955223880596,
+      "grad_norm": 0.2602560441928257,
+      "learning_rate": 5.180443037438508e-05,
+      "loss": 0.3711,
+      "step": 2177
+    },
+    {
+      "epoch": 2.321961620469083,
+      "grad_norm": 0.36519338651936206,
+      "learning_rate": 5.1775981945696736e-05,
+      "loss": 0.3701,
+      "step": 2178
+    },
+    {
+      "epoch": 2.3230277185501067,
+      "grad_norm": 0.34363559674089617,
+      "learning_rate": 5.1747526993726406e-05,
+      "loss": 0.3701,
+      "step": 2179
+    },
+    {
+      "epoch": 2.3240938166311302,
+      "grad_norm": 0.34656810990367415,
+      "learning_rate": 5.1719065534236665e-05,
+      "loss": 0.3784,
+      "step": 2180
+    },
+    {
+      "epoch": 2.3251599147121533,
+      "grad_norm": 0.3632003951397555,
+      "learning_rate": 5.169059758299367e-05,
+      "loss": 0.3667,
+      "step": 2181
+    },
+    {
+      "epoch": 2.326226012793177,
+      "grad_norm": 0.33113190305897466,
+      "learning_rate": 5.1662123155767195e-05,
+      "loss": 0.3672,
+      "step": 2182
+    },
+    {
+      "epoch": 2.3272921108742004,
+      "grad_norm": 0.3490341083520184,
+      "learning_rate": 5.163364226833058e-05,
+      "loss": 0.3715,
+      "step": 2183
+    },
+    {
+      "epoch": 2.328358208955224,
+      "grad_norm": 0.30252710599573834,
+      "learning_rate": 5.1605154936460774e-05,
+      "loss": 0.3699,
+      "step": 2184
+    },
+    {
+      "epoch": 2.3294243070362475,
+      "grad_norm": 0.23971322338603304,
+      "learning_rate": 5.1576661175938274e-05,
+      "loss": 0.3684,
+      "step": 2185
+    },
+    {
+      "epoch": 2.3304904051172706,
+      "grad_norm": 0.2935195353785945,
+      "learning_rate": 5.154816100254714e-05,
+      "loss": 0.3728,
+      "step": 2186
+    },
+    {
+      "epoch": 2.331556503198294,
+      "grad_norm": 0.32527117670349187,
+      "learning_rate": 5.151965443207498e-05,
+      "loss": 0.3672,
+      "step": 2187
+    },
+    {
+      "epoch": 2.3326226012793176,
+      "grad_norm": 0.29601539551210887,
+      "learning_rate": 5.149114148031296e-05,
+      "loss": 0.3722,
+      "step": 2188
+    },
+    {
+      "epoch": 2.333688699360341,
+      "grad_norm": 0.29110059563391355,
+      "learning_rate": 5.1462622163055764e-05,
+      "loss": 0.3711,
+      "step": 2189
+    },
+    {
+      "epoch": 2.3347547974413647,
+      "grad_norm": 0.2777132993013124,
+      "learning_rate": 5.143409649610163e-05,
+      "loss": 0.3733,
+      "step": 2190
+    },
+    {
+      "epoch": 2.3358208955223883,
+      "grad_norm": 0.2838065564769729,
+      "learning_rate": 5.14055644952523e-05,
+      "loss": 0.3677,
+      "step": 2191
+    },
+    {
+      "epoch": 2.3368869936034113,
+      "grad_norm": 0.26546607714553994,
+      "learning_rate": 5.137702617631299e-05,
+      "loss": 0.3705,
+      "step": 2192
+    },
+    {
+      "epoch": 2.337953091684435,
+      "grad_norm": 0.2691865136150305,
+      "learning_rate": 5.134848155509245e-05,
+      "loss": 0.3693,
+      "step": 2193
+    },
+    {
+      "epoch": 2.3390191897654584,
+      "grad_norm": 0.3248422617546964,
+      "learning_rate": 5.131993064740293e-05,
+      "loss": 0.373,
+      "step": 2194
+    },
+    {
+      "epoch": 2.340085287846482,
+      "grad_norm": 0.44219209612841087,
+      "learning_rate": 5.1291373469060156e-05,
+      "loss": 0.3746,
+      "step": 2195
+    },
+    {
+      "epoch": 2.3411513859275055,
+      "grad_norm": 0.4487435007125376,
+      "learning_rate": 5.12628100358833e-05,
+      "loss": 0.3693,
+      "step": 2196
+    },
+    {
+      "epoch": 2.342217484008529,
+      "grad_norm": 0.39660870409128357,
+      "learning_rate": 5.123424036369504e-05,
+      "loss": 0.3748,
+      "step": 2197
+    },
+    {
+      "epoch": 2.343283582089552,
+      "grad_norm": 0.35851143039701705,
+      "learning_rate": 5.120566446832146e-05,
+      "loss": 0.3691,
+      "step": 2198
+    },
+    {
+      "epoch": 2.3443496801705757,
+      "grad_norm": 0.334759825605026,
+      "learning_rate": 5.117708236559216e-05,
+      "loss": 0.3673,
+      "step": 2199
+    },
+    {
+      "epoch": 2.345415778251599,
+      "grad_norm": 0.2941272239375261,
+      "learning_rate": 5.114849407134012e-05,
+      "loss": 0.3643,
+      "step": 2200
+    },
+    {
+      "epoch": 2.3464818763326227,
+      "grad_norm": 0.2789608400358062,
+      "learning_rate": 5.111989960140175e-05,
+      "loss": 0.3671,
+      "step": 2201
+    },
+    {
+      "epoch": 2.3475479744136463,
+      "grad_norm": 0.3040283299295704,
+      "learning_rate": 5.109129897161694e-05,
+      "loss": 0.3772,
+      "step": 2202
+    },
+    {
+      "epoch": 2.3486140724946694,
+      "grad_norm": 0.319463979431584,
+      "learning_rate": 5.106269219782891e-05,
+      "loss": 0.3749,
+      "step": 2203
+    },
+    {
+      "epoch": 2.349680170575693,
+      "grad_norm": 0.2870715537863343,
+      "learning_rate": 5.1034079295884366e-05,
+      "loss": 0.3702,
+      "step": 2204
+    },
+    {
+      "epoch": 2.3507462686567164,
+      "grad_norm": 0.23446913387159526,
+      "learning_rate": 5.100546028163334e-05,
+      "loss": 0.3753,
+      "step": 2205
+    },
+    {
+      "epoch": 2.35181236673774,
+      "grad_norm": 0.25437833515945757,
+      "learning_rate": 5.0976835170929296e-05,
+      "loss": 0.3666,
+      "step": 2206
+    },
+    {
+      "epoch": 2.3528784648187635,
+      "grad_norm": 0.30198542630658826,
+      "learning_rate": 5.0948203979629046e-05,
+      "loss": 0.3673,
+      "step": 2207
+    },
+    {
+      "epoch": 2.3539445628997866,
+      "grad_norm": 0.3141893057528808,
+      "learning_rate": 5.091956672359279e-05,
+      "loss": 0.3627,
+      "step": 2208
+    },
+    {
+      "epoch": 2.35501066098081,
+      "grad_norm": 0.2805349073413673,
+      "learning_rate": 5.089092341868407e-05,
+      "loss": 0.3676,
+      "step": 2209
+    },
+    {
+      "epoch": 2.3560767590618337,
+      "grad_norm": 0.2464031232660798,
+      "learning_rate": 5.08622740807698e-05,
+      "loss": 0.3719,
+      "step": 2210
+    },
+    {
+      "epoch": 2.357142857142857,
+      "grad_norm": 0.23438323370618938,
+      "learning_rate": 5.0833618725720214e-05,
+      "loss": 0.3615,
+      "step": 2211
+    },
+    {
+      "epoch": 2.3582089552238807,
+      "grad_norm": 0.2997661424210569,
+      "learning_rate": 5.080495736940889e-05,
+      "loss": 0.3705,
+      "step": 2212
+    },
+    {
+      "epoch": 2.359275053304904,
+      "grad_norm": 0.28620092721857604,
+      "learning_rate": 5.077629002771274e-05,
+      "loss": 0.3713,
+      "step": 2213
+    },
+    {
+      "epoch": 2.3603411513859274,
+      "grad_norm": 0.2468965983835025,
+      "learning_rate": 5.074761671651198e-05,
+      "loss": 0.3696,
+      "step": 2214
+    },
+    {
+      "epoch": 2.361407249466951,
+      "grad_norm": 0.25371334633765574,
+      "learning_rate": 5.071893745169012e-05,
+      "loss": 0.3656,
+      "step": 2215
+    },
+    {
+      "epoch": 2.3624733475479744,
+      "grad_norm": 0.24096589331073096,
+      "learning_rate": 5.0690252249133986e-05,
+      "loss": 0.3709,
+      "step": 2216
+    },
+    {
+      "epoch": 2.363539445628998,
+      "grad_norm": 0.2152100387619128,
+      "learning_rate": 5.066156112473371e-05,
+      "loss": 0.3727,
+      "step": 2217
+    },
+    {
+      "epoch": 2.364605543710021,
+      "grad_norm": 0.19650715541656136,
+      "learning_rate": 5.063286409438265e-05,
+      "loss": 0.3678,
+      "step": 2218
+    },
+    {
+      "epoch": 2.3656716417910446,
+      "grad_norm": 0.21642059764416482,
+      "learning_rate": 5.0604161173977504e-05,
+      "loss": 0.3701,
+      "step": 2219
+    },
+    {
+      "epoch": 2.366737739872068,
+      "grad_norm": 0.27848392769400904,
+      "learning_rate": 5.057545237941818e-05,
+      "loss": 0.3697,
+      "step": 2220
+    },
+    {
+      "epoch": 2.3678038379530917,
+      "grad_norm": 0.3063826642109227,
+      "learning_rate": 5.054673772660785e-05,
+      "loss": 0.3669,
+      "step": 2221
+    },
+    {
+      "epoch": 2.368869936034115,
+      "grad_norm": 0.32561764160202256,
+      "learning_rate": 5.0518017231452965e-05,
+      "loss": 0.369,
+      "step": 2222
+    },
+    {
+      "epoch": 2.3699360341151388,
+      "grad_norm": 0.2857238132037038,
+      "learning_rate": 5.048929090986315e-05,
+      "loss": 0.3658,
+      "step": 2223
+    },
+    {
+      "epoch": 2.3710021321961623,
+      "grad_norm": 0.24657414607956593,
+      "learning_rate": 5.046055877775134e-05,
+      "loss": 0.3689,
+      "step": 2224
+    },
+    {
+      "epoch": 2.3720682302771854,
+      "grad_norm": 0.25182509640136513,
+      "learning_rate": 5.04318208510336e-05,
+      "loss": 0.3737,
+      "step": 2225
+    },
+    {
+      "epoch": 2.373134328358209,
+      "grad_norm": 0.2663592462420767,
+      "learning_rate": 5.040307714562928e-05,
+      "loss": 0.3654,
+      "step": 2226
+    },
+    {
+      "epoch": 2.3742004264392325,
+      "grad_norm": 0.2879071667851132,
+      "learning_rate": 5.0374327677460865e-05,
+      "loss": 0.362,
+      "step": 2227
+    },
+    {
+      "epoch": 2.375266524520256,
+      "grad_norm": 0.31268904527265867,
+      "learning_rate": 5.034557246245411e-05,
+      "loss": 0.3716,
+      "step": 2228
+    },
+    {
+      "epoch": 2.3763326226012795,
+      "grad_norm": 0.33284965204702893,
+      "learning_rate": 5.031681151653788e-05,
+      "loss": 0.3736,
+      "step": 2229
+    },
+    {
+      "epoch": 2.3773987206823026,
+      "grad_norm": 0.33195392118584366,
+      "learning_rate": 5.028804485564424e-05,
+      "loss": 0.3677,
+      "step": 2230
+    },
+    {
+      "epoch": 2.378464818763326,
+      "grad_norm": 0.3258278442178994,
+      "learning_rate": 5.025927249570844e-05,
+      "loss": 0.3693,
+      "step": 2231
+    },
+    {
+      "epoch": 2.3795309168443497,
+      "grad_norm": 0.27343346069158153,
+      "learning_rate": 5.0230494452668864e-05,
+      "loss": 0.3733,
+      "step": 2232
+    },
+    {
+      "epoch": 2.3805970149253732,
+      "grad_norm": 0.25449187697759107,
+      "learning_rate": 5.020171074246707e-05,
+      "loss": 0.3668,
+      "step": 2233
+    },
+    {
+      "epoch": 2.3816631130063968,
+      "grad_norm": 0.30392088734456846,
+      "learning_rate": 5.01729213810477e-05,
+      "loss": 0.3728,
+      "step": 2234
+    },
+    {
+      "epoch": 2.38272921108742,
+      "grad_norm": 0.3421952140379261,
+      "learning_rate": 5.014412638435861e-05,
+      "loss": 0.3684,
+      "step": 2235
+    },
+    {
+      "epoch": 2.3837953091684434,
+      "grad_norm": 0.2857424583185023,
+      "learning_rate": 5.011532576835069e-05,
+      "loss": 0.3629,
+      "step": 2236
+    },
+    {
+      "epoch": 2.384861407249467,
+      "grad_norm": 0.2704981763316653,
+      "learning_rate": 5.008651954897802e-05,
+      "loss": 0.3686,
+      "step": 2237
+    },
+    {
+      "epoch": 2.3859275053304905,
+      "grad_norm": 0.31308011375671224,
+      "learning_rate": 5.005770774219771e-05,
+      "loss": 0.3668,
+      "step": 2238
+    },
+    {
+      "epoch": 2.386993603411514,
+      "grad_norm": 0.3213528795416127,
+      "learning_rate": 5.002889036397005e-05,
+      "loss": 0.3732,
+      "step": 2239
+    },
+    {
+      "epoch": 2.388059701492537,
+      "grad_norm": 0.34520492416317516,
+      "learning_rate": 5.000006743025834e-05,
+      "loss": 0.3745,
+      "step": 2240
+    },
+    {
+      "epoch": 2.3891257995735606,
+      "grad_norm": 0.30507013156130014,
+      "learning_rate": 4.997123895702898e-05,
+      "loss": 0.3705,
+      "step": 2241
+    },
+    {
+      "epoch": 2.390191897654584,
+      "grad_norm": 0.2626181582745989,
+      "learning_rate": 4.994240496025147e-05,
+      "loss": 0.3727,
+      "step": 2242
+    },
+    {
+      "epoch": 2.3912579957356077,
+      "grad_norm": 0.28512789584756165,
+      "learning_rate": 4.9913565455898327e-05,
+      "loss": 0.3756,
+      "step": 2243
+    },
+    {
+      "epoch": 2.3923240938166312,
+      "grad_norm": 0.34534631086868595,
+      "learning_rate": 4.988472045994515e-05,
+      "loss": 0.3713,
+      "step": 2244
+    },
+    {
+      "epoch": 2.3933901918976543,
+      "grad_norm": 0.3429792538710119,
+      "learning_rate": 4.9855869988370566e-05,
+      "loss": 0.3703,
+      "step": 2245
+    },
+    {
+      "epoch": 2.394456289978678,
+      "grad_norm": 0.25240157643598227,
+      "learning_rate": 4.982701405715622e-05,
+      "loss": 0.3677,
+      "step": 2246
+    },
+    {
+      "epoch": 2.3955223880597014,
+      "grad_norm": 0.2549810372821131,
+      "learning_rate": 4.9798152682286824e-05,
+      "loss": 0.3711,
+      "step": 2247
+    },
+    {
+      "epoch": 2.396588486140725,
+      "grad_norm": 0.351895077255566,
+      "learning_rate": 4.976928587975006e-05,
+      "loss": 0.3686,
+      "step": 2248
+    },
+    {
+      "epoch": 2.3976545842217485,
+      "grad_norm": 0.30952942264717365,
+      "learning_rate": 4.974041366553665e-05,
+      "loss": 0.372,
+      "step": 2249
+    },
+    {
+      "epoch": 2.398720682302772,
+      "grad_norm": 0.2747608348019261,
+      "learning_rate": 4.9711536055640285e-05,
+      "loss": 0.3728,
+      "step": 2250
+    },
+    {
+      "epoch": 2.399786780383795,
+      "grad_norm": 0.29969340676529904,
+      "learning_rate": 4.9682653066057676e-05,
+      "loss": 0.3696,
+      "step": 2251
+    },
+    {
+      "epoch": 2.4008528784648187,
+      "grad_norm": 0.32127749277577844,
+      "learning_rate": 4.965376471278848e-05,
+      "loss": 0.3733,
+      "step": 2252
+    },
+    {
+      "epoch": 2.401918976545842,
+      "grad_norm": 0.3048300891627042,
+      "learning_rate": 4.962487101183536e-05,
+      "loss": 0.369,
+      "step": 2253
+    },
+    {
+      "epoch": 2.4029850746268657,
+      "grad_norm": 0.3219753451808507,
+      "learning_rate": 4.959597197920392e-05,
+      "loss": 0.3681,
+      "step": 2254
+    },
+    {
+      "epoch": 2.4040511727078893,
+      "grad_norm": 0.2965082879251037,
+      "learning_rate": 4.956706763090272e-05,
+      "loss": 0.3724,
+      "step": 2255
+    },
+    {
+      "epoch": 2.405117270788913,
+      "grad_norm": 0.275785450919491,
+      "learning_rate": 4.953815798294327e-05,
+      "loss": 0.3675,
+      "step": 2256
+    },
+    {
+      "epoch": 2.406183368869936,
+      "grad_norm": 0.8759881578646651,
+      "learning_rate": 4.950924305134001e-05,
+      "loss": 0.3698,
+      "step": 2257
+    },
+    {
+      "epoch": 2.4072494669509594,
+      "grad_norm": 0.2354192544695265,
+      "learning_rate": 4.948032285211031e-05,
+      "loss": 0.3693,
+      "step": 2258
+    },
+    {
+      "epoch": 2.408315565031983,
+      "grad_norm": 0.28320526847754757,
+      "learning_rate": 4.945139740127444e-05,
+      "loss": 0.3685,
+      "step": 2259
+    },
+    {
+      "epoch": 2.4093816631130065,
+      "grad_norm": 0.27994748944288567,
+      "learning_rate": 4.9422466714855635e-05,
+      "loss": 0.3716,
+      "step": 2260
+    },
+    {
+      "epoch": 2.41044776119403,
+      "grad_norm": 0.2577423187481169,
+      "learning_rate": 4.939353080887996e-05,
+      "loss": 0.3736,
+      "step": 2261
+    },
+    {
+      "epoch": 2.411513859275053,
+      "grad_norm": 0.20189014886728646,
+      "learning_rate": 4.936458969937642e-05,
+      "loss": 0.3634,
+      "step": 2262
+    },
+    {
+      "epoch": 2.4125799573560767,
+      "grad_norm": 0.27831002089150636,
+      "learning_rate": 4.933564340237687e-05,
+      "loss": 0.3698,
+      "step": 2263
+    },
+    {
+      "epoch": 2.4136460554371,
+      "grad_norm": 0.33503367593001443,
+      "learning_rate": 4.930669193391607e-05,
+      "loss": 0.3759,
+      "step": 2264
+    },
+    {
+      "epoch": 2.4147121535181237,
+      "grad_norm": 0.2557075329669632,
+      "learning_rate": 4.927773531003161e-05,
+      "loss": 0.3719,
+      "step": 2265
+    },
+    {
+      "epoch": 2.4157782515991473,
+      "grad_norm": 0.1959575254371725,
+      "learning_rate": 4.9248773546763984e-05,
+      "loss": 0.3666,
+      "step": 2266
+    },
+    {
+      "epoch": 2.4168443496801704,
+      "grad_norm": 0.2566851026971127,
+      "learning_rate": 4.921980666015647e-05,
+      "loss": 0.3722,
+      "step": 2267
+    },
+    {
+      "epoch": 2.417910447761194,
+      "grad_norm": 0.3373730577896526,
+      "learning_rate": 4.919083466625524e-05,
+      "loss": 0.3704,
+      "step": 2268
+    },
+    {
+      "epoch": 2.4189765458422174,
+      "grad_norm": 0.29147620272337893,
+      "learning_rate": 4.916185758110928e-05,
+      "loss": 0.3764,
+      "step": 2269
+    },
+    {
+      "epoch": 2.420042643923241,
+      "grad_norm": 0.23015006905725324,
+      "learning_rate": 4.913287542077035e-05,
+      "loss": 0.3765,
+      "step": 2270
+    },
+    {
+      "epoch": 2.4211087420042645,
+      "grad_norm": 0.2666917022092777,
+      "learning_rate": 4.91038882012931e-05,
+      "loss": 0.3685,
+      "step": 2271
+    },
+    {
+      "epoch": 2.4221748400852876,
+      "grad_norm": 0.2807196321927748,
+      "learning_rate": 4.907489593873493e-05,
+      "loss": 0.3673,
+      "step": 2272
+    },
+    {
+      "epoch": 2.423240938166311,
+      "grad_norm": 0.23362271125449918,
+      "learning_rate": 4.904589864915605e-05,
+      "loss": 0.367,
+      "step": 2273
+    },
+    {
+      "epoch": 2.4243070362473347,
+      "grad_norm": 0.2813715337244537,
+      "learning_rate": 4.901689634861943e-05,
+      "loss": 0.3712,
+      "step": 2274
+    },
+    {
+      "epoch": 2.425373134328358,
+      "grad_norm": 0.29016723056538085,
+      "learning_rate": 4.898788905319087e-05,
+      "loss": 0.373,
+      "step": 2275
+    },
+    {
+      "epoch": 2.4264392324093818,
+      "grad_norm": 0.302081673689402,
+      "learning_rate": 4.895887677893889e-05,
+      "loss": 0.3654,
+      "step": 2276
+    },
+    {
+      "epoch": 2.4275053304904053,
+      "grad_norm": 0.28261113280230443,
+      "learning_rate": 4.892985954193478e-05,
+      "loss": 0.3748,
+      "step": 2277
+    },
+    {
+      "epoch": 2.4285714285714284,
+      "grad_norm": 0.32074482716991354,
+      "learning_rate": 4.890083735825258e-05,
+      "loss": 0.3649,
+      "step": 2278
+    },
+    {
+      "epoch": 2.429637526652452,
+      "grad_norm": 0.3198080528899319,
+      "learning_rate": 4.887181024396907e-05,
+      "loss": 0.3593,
+      "step": 2279
+    },
+    {
+      "epoch": 2.4307036247334755,
+      "grad_norm": 0.25493105066167904,
+      "learning_rate": 4.884277821516377e-05,
+      "loss": 0.3661,
+      "step": 2280
+    },
+    {
+      "epoch": 2.431769722814499,
+      "grad_norm": 0.21652193711164933,
+      "learning_rate": 4.881374128791892e-05,
+      "loss": 0.3712,
+      "step": 2281
+    },
+    {
+      "epoch": 2.4328358208955225,
+      "grad_norm": 0.25095432049789496,
+      "learning_rate": 4.878469947831945e-05,
+      "loss": 0.3704,
+      "step": 2282
+    },
+    {
+      "epoch": 2.433901918976546,
+      "grad_norm": 0.26679049451296216,
+      "learning_rate": 4.875565280245303e-05,
+      "loss": 0.3741,
+      "step": 2283
+    },
+    {
+      "epoch": 2.434968017057569,
+      "grad_norm": 0.2999335227622482,
+      "learning_rate": 4.872660127640999e-05,
+      "loss": 0.3713,
+      "step": 2284
+    },
+    {
+      "epoch": 2.4360341151385927,
+      "grad_norm": 0.2695301789461439,
+      "learning_rate": 4.8697544916283386e-05,
+      "loss": 0.3659,
+      "step": 2285
+    },
+    {
+      "epoch": 2.4371002132196162,
+      "grad_norm": 0.21471471033527165,
+      "learning_rate": 4.866848373816893e-05,
+      "loss": 0.3682,
+      "step": 2286
+    },
+    {
+      "epoch": 2.4381663113006398,
+      "grad_norm": 0.23947760320903563,
+      "learning_rate": 4.863941775816498e-05,
+      "loss": 0.372,
+      "step": 2287
+    },
+    {
+      "epoch": 2.4392324093816633,
+      "grad_norm": 0.267331462460001,
+      "learning_rate": 4.8610346992372603e-05,
+      "loss": 0.3654,
+      "step": 2288
+    },
+    {
+      "epoch": 2.4402985074626864,
+      "grad_norm": 0.20555064622823938,
+      "learning_rate": 4.85812714568955e-05,
+      "loss": 0.3693,
+      "step": 2289
+    },
+    {
+      "epoch": 2.44136460554371,
+      "grad_norm": 0.21294997398754437,
+      "learning_rate": 4.855219116783997e-05,
+      "loss": 0.3728,
+      "step": 2290
+    },
+    {
+      "epoch": 2.4424307036247335,
+      "grad_norm": 0.21976358289050846,
+      "learning_rate": 4.8523106141315005e-05,
+      "loss": 0.3792,
+      "step": 2291
+    },
+    {
+      "epoch": 2.443496801705757,
+      "grad_norm": 0.23884227282875203,
+      "learning_rate": 4.8494016393432205e-05,
+      "loss": 0.3607,
+      "step": 2292
+    },
+    {
+      "epoch": 2.4445628997867805,
+      "grad_norm": 0.2624658049626465,
+      "learning_rate": 4.846492194030577e-05,
+      "loss": 0.3707,
+      "step": 2293
+    },
+    {
+      "epoch": 2.4456289978678036,
+      "grad_norm": 0.28918877258810133,
+      "learning_rate": 4.843582279805251e-05,
+      "loss": 0.3737,
+      "step": 2294
+    },
+    {
+      "epoch": 2.446695095948827,
+      "grad_norm": 0.31404526252223114,
+      "learning_rate": 4.840671898279185e-05,
+      "loss": 0.3659,
+      "step": 2295
+    },
+    {
+      "epoch": 2.4477611940298507,
+      "grad_norm": 0.29838133141624607,
+      "learning_rate": 4.837761051064579e-05,
+      "loss": 0.3687,
+      "step": 2296
+    },
+    {
+      "epoch": 2.4488272921108742,
+      "grad_norm": 0.22887592628169673,
+      "learning_rate": 4.834849739773889e-05,
+      "loss": 0.3722,
+      "step": 2297
+    },
+    {
+      "epoch": 2.449893390191898,
+      "grad_norm": 0.23115344121100373,
+      "learning_rate": 4.8319379660198316e-05,
+      "loss": 0.3737,
+      "step": 2298
+    },
+    {
+      "epoch": 2.450959488272921,
+      "grad_norm": 0.23213200641703416,
+      "learning_rate": 4.829025731415378e-05,
+      "loss": 0.364,
+      "step": 2299
+    },
+    {
+      "epoch": 2.4520255863539444,
+      "grad_norm": 0.2116924844746429,
+      "learning_rate": 4.826113037573756e-05,
+      "loss": 0.3674,
+      "step": 2300
+    },
+    {
+      "epoch": 2.453091684434968,
+      "grad_norm": 0.2559342759140697,
+      "learning_rate": 4.823199886108445e-05,
+      "loss": 0.3752,
+      "step": 2301
+    },
+    {
+      "epoch": 2.4541577825159915,
+      "grad_norm": 0.293845262944136,
+      "learning_rate": 4.82028627863318e-05,
+      "loss": 0.3709,
+      "step": 2302
+    },
+    {
+      "epoch": 2.455223880597015,
+      "grad_norm": 0.29454888718180283,
+      "learning_rate": 4.817372216761948e-05,
+      "loss": 0.3722,
+      "step": 2303
+    },
+    {
+      "epoch": 2.4562899786780386,
+      "grad_norm": 0.24511640910718516,
+      "learning_rate": 4.8144577021089884e-05,
+      "loss": 0.3747,
+      "step": 2304
+    },
+    {
+      "epoch": 2.4573560767590616,
+      "grad_norm": 0.2763149476728892,
+      "learning_rate": 4.81154273628879e-05,
+      "loss": 0.363,
+      "step": 2305
+    },
+    {
+      "epoch": 2.458422174840085,
+      "grad_norm": 0.3243976443000684,
+      "learning_rate": 4.8086273209160936e-05,
+      "loss": 0.3693,
+      "step": 2306
+    },
+    {
+      "epoch": 2.4594882729211087,
+      "grad_norm": 0.28035443109857583,
+      "learning_rate": 4.8057114576058863e-05,
+      "loss": 0.3709,
+      "step": 2307
+    },
+    {
+      "epoch": 2.4605543710021323,
+      "grad_norm": 0.20090541292771277,
+      "learning_rate": 4.802795147973406e-05,
+      "loss": 0.3666,
+      "step": 2308
+    },
+    {
+      "epoch": 2.461620469083156,
+      "grad_norm": 0.2580300489910844,
+      "learning_rate": 4.799878393634136e-05,
+      "loss": 0.3668,
+      "step": 2309
+    },
+    {
+      "epoch": 2.4626865671641793,
+      "grad_norm": 0.25079828700699003,
+      "learning_rate": 4.796961196203806e-05,
+      "loss": 0.3763,
+      "step": 2310
+    },
+    {
+      "epoch": 2.4637526652452024,
+      "grad_norm": 0.24137932132689202,
+      "learning_rate": 4.7940435572983936e-05,
+      "loss": 0.3705,
+      "step": 2311
+    },
+    {
+      "epoch": 2.464818763326226,
+      "grad_norm": 0.24074250093617247,
+      "learning_rate": 4.791125478534118e-05,
+      "loss": 0.3701,
+      "step": 2312
+    },
+    {
+      "epoch": 2.4658848614072495,
+      "grad_norm": 0.21896631390072002,
+      "learning_rate": 4.7882069615274435e-05,
+      "loss": 0.3683,
+      "step": 2313
+    },
+    {
+      "epoch": 2.466950959488273,
+      "grad_norm": 0.22445307173012358,
+      "learning_rate": 4.7852880078950764e-05,
+      "loss": 0.3721,
+      "step": 2314
+    },
+    {
+      "epoch": 2.4680170575692966,
+      "grad_norm": 0.27306888281171515,
+      "learning_rate": 4.782368619253965e-05,
+      "loss": 0.3687,
+      "step": 2315
+    },
+    {
+      "epoch": 2.4690831556503197,
+      "grad_norm": 0.36336555042461727,
+      "learning_rate": 4.7794487972213e-05,
+      "loss": 0.3639,
+      "step": 2316
+    },
+    {
+      "epoch": 2.470149253731343,
+      "grad_norm": 0.3601488690657708,
+      "learning_rate": 4.776528543414511e-05,
+      "loss": 0.3738,
+      "step": 2317
+    },
+    {
+      "epoch": 2.4712153518123667,
+      "grad_norm": 0.35339799743518896,
+      "learning_rate": 4.773607859451265e-05,
+      "loss": 0.3713,
+      "step": 2318
+    },
+    {
+      "epoch": 2.4722814498933903,
+      "grad_norm": 0.3693531153358677,
+      "learning_rate": 4.770686746949472e-05,
+      "loss": 0.3681,
+      "step": 2319
+    },
+    {
+      "epoch": 2.473347547974414,
+      "grad_norm": 0.39311860926438036,
+      "learning_rate": 4.767765207527275e-05,
+      "loss": 0.3696,
+      "step": 2320
+    },
+    {
+      "epoch": 2.474413646055437,
+      "grad_norm": 0.37612079543054977,
+      "learning_rate": 4.764843242803053e-05,
+      "loss": 0.3661,
+      "step": 2321
+    },
+    {
+      "epoch": 2.4754797441364604,
+      "grad_norm": 0.3521543006142998,
+      "learning_rate": 4.761920854395426e-05,
+      "loss": 0.3716,
+      "step": 2322
+    },
+    {
+      "epoch": 2.476545842217484,
+      "grad_norm": 0.26305773466563936,
+      "learning_rate": 4.7589980439232433e-05,
+      "loss": 0.3707,
+      "step": 2323
+    },
+    {
+      "epoch": 2.4776119402985075,
+      "grad_norm": 0.2559553568605443,
+      "learning_rate": 4.756074813005591e-05,
+      "loss": 0.3704,
+      "step": 2324
+    },
+    {
+      "epoch": 2.478678038379531,
+      "grad_norm": 0.2890296801712053,
+      "learning_rate": 4.753151163261787e-05,
+      "loss": 0.3669,
+      "step": 2325
+    },
+    {
+      "epoch": 2.479744136460554,
+      "grad_norm": 0.268964216758793,
+      "learning_rate": 4.75022709631138e-05,
+      "loss": 0.3704,
+      "step": 2326
+    },
+    {
+      "epoch": 2.4808102345415777,
+      "grad_norm": 0.29305303643593816,
+      "learning_rate": 4.747302613774153e-05,
+      "loss": 0.3667,
+      "step": 2327
+    },
+    {
+      "epoch": 2.481876332622601,
+      "grad_norm": 0.3343530498108033,
+      "learning_rate": 4.7443777172701146e-05,
+      "loss": 0.3703,
+      "step": 2328
+    },
+    {
+      "epoch": 2.4829424307036247,
+      "grad_norm": 0.3429097239396157,
+      "learning_rate": 4.74145240841951e-05,
+      "loss": 0.374,
+      "step": 2329
+    },
+    {
+      "epoch": 2.4840085287846483,
+      "grad_norm": 0.27000380770235705,
+      "learning_rate": 4.738526688842803e-05,
+      "loss": 0.371,
+      "step": 2330
+    },
+    {
+      "epoch": 2.485074626865672,
+      "grad_norm": 0.24578044590662604,
+      "learning_rate": 4.735600560160695e-05,
+      "loss": 0.3713,
+      "step": 2331
+    },
+    {
+      "epoch": 2.486140724946695,
+      "grad_norm": 0.2595315485397087,
+      "learning_rate": 4.7326740239941054e-05,
+      "loss": 0.3748,
+      "step": 2332
+    },
+    {
+      "epoch": 2.4872068230277184,
+      "grad_norm": 0.2207117203267181,
+      "learning_rate": 4.729747081964185e-05,
+      "loss": 0.3635,
+      "step": 2333
+    },
+    {
+      "epoch": 2.488272921108742,
+      "grad_norm": 0.22525118617182993,
+      "learning_rate": 4.7268197356923076e-05,
+      "loss": 0.3693,
+      "step": 2334
+    },
+    {
+      "epoch": 2.4893390191897655,
+      "grad_norm": 0.26088710793376507,
+      "learning_rate": 4.7238919868000704e-05,
+      "loss": 0.3609,
+      "step": 2335
+    },
+    {
+      "epoch": 2.490405117270789,
+      "grad_norm": 0.2885428062747875,
+      "learning_rate": 4.720963836909295e-05,
+      "loss": 0.3762,
+      "step": 2336
+    },
+    {
+      "epoch": 2.4914712153518126,
+      "grad_norm": 0.25259361954597753,
+      "learning_rate": 4.718035287642022e-05,
+      "loss": 0.3714,
+      "step": 2337
+    },
+    {
+      "epoch": 2.4925373134328357,
+      "grad_norm": 0.24137669671809825,
+      "learning_rate": 4.715106340620518e-05,
+      "loss": 0.3677,
+      "step": 2338
+    },
+    {
+      "epoch": 2.4936034115138592,
+      "grad_norm": 0.2686347142176362,
+      "learning_rate": 4.712176997467266e-05,
+      "loss": 0.3706,
+      "step": 2339
+    },
+    {
+      "epoch": 2.4946695095948828,
+      "grad_norm": 0.2812137990571536,
+      "learning_rate": 4.709247259804971e-05,
+      "loss": 0.371,
+      "step": 2340
+    },
+    {
+      "epoch": 2.4957356076759063,
+      "grad_norm": 0.24346310160667034,
+      "learning_rate": 4.706317129256554e-05,
+      "loss": 0.3716,
+      "step": 2341
+    },
+    {
+      "epoch": 2.49680170575693,
+      "grad_norm": 0.23214815216215345,
+      "learning_rate": 4.703386607445157e-05,
+      "loss": 0.3712,
+      "step": 2342
+    },
+    {
+      "epoch": 2.497867803837953,
+      "grad_norm": 0.2739545324648638,
+      "learning_rate": 4.7004556959941335e-05,
+      "loss": 0.3651,
+      "step": 2343
+    },
+    {
+      "epoch": 2.4989339019189765,
+      "grad_norm": 0.25915247784626605,
+      "learning_rate": 4.69752439652706e-05,
+      "loss": 0.3663,
+      "step": 2344
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.25618378366217026,
+      "learning_rate": 4.694592710667723e-05,
+      "loss": 0.374,
+      "step": 2345
+    },
+    {
+      "epoch": 2.5010660980810235,
+      "grad_norm": 0.30631526116242996,
+      "learning_rate": 4.69166064004012e-05,
+      "loss": 0.3716,
+      "step": 2346
+    },
+    {
+      "epoch": 2.502132196162047,
+      "grad_norm": 0.3366426624946492,
+      "learning_rate": 4.688728186268472e-05,
+      "loss": 0.3714,
+      "step": 2347
+    },
+    {
+      "epoch": 2.50319829424307,
+      "grad_norm": 0.25931098502063027,
+      "learning_rate": 4.685795350977202e-05,
+      "loss": 0.3715,
+      "step": 2348
+    },
+    {
+      "epoch": 2.5042643923240937,
+      "grad_norm": 0.2123807260445706,
+      "learning_rate": 4.6828621357909494e-05,
+      "loss": 0.3649,
+      "step": 2349
+    },
+    {
+      "epoch": 2.5053304904051172,
+      "grad_norm": 0.19519486858966945,
+      "learning_rate": 4.679928542334564e-05,
+      "loss": 0.3756,
+      "step": 2350
+    },
+    {
+      "epoch": 2.5063965884861408,
+      "grad_norm": 0.24863053693547293,
+      "learning_rate": 4.676994572233101e-05,
+      "loss": 0.3676,
+      "step": 2351
+    },
+    {
+      "epoch": 2.5074626865671643,
+      "grad_norm": 0.25591846224110576,
+      "learning_rate": 4.674060227111831e-05,
+      "loss": 0.3746,
+      "step": 2352
+    },
+    {
+      "epoch": 2.5085287846481874,
+      "grad_norm": 0.2281497146449445,
+      "learning_rate": 4.6711255085962275e-05,
+      "loss": 0.3636,
+      "step": 2353
+    },
+    {
+      "epoch": 2.509594882729211,
+      "grad_norm": 0.2994568082565574,
+      "learning_rate": 4.66819041831197e-05,
+      "loss": 0.373,
+      "step": 2354
+    },
+    {
+      "epoch": 2.5106609808102345,
+      "grad_norm": 0.29113727854820887,
+      "learning_rate": 4.66525495788495e-05,
+      "loss": 0.3723,
+      "step": 2355
+    },
+    {
+      "epoch": 2.511727078891258,
+      "grad_norm": 0.2845367853179552,
+      "learning_rate": 4.662319128941256e-05,
+      "loss": 0.3659,
+      "step": 2356
+    },
+    {
+      "epoch": 2.5127931769722816,
+      "grad_norm": 0.24591581925857822,
+      "learning_rate": 4.6593829331071854e-05,
+      "loss": 0.3777,
+      "step": 2357
+    },
+    {
+      "epoch": 2.5138592750533046,
+      "grad_norm": 0.2322117630084555,
+      "learning_rate": 4.6564463720092405e-05,
+      "loss": 0.3679,
+      "step": 2358
+    },
+    {
+      "epoch": 2.5149253731343286,
+      "grad_norm": 0.22223391656686764,
+      "learning_rate": 4.653509447274121e-05,
+      "loss": 0.3685,
+      "step": 2359
+    },
+    {
+      "epoch": 2.5159914712153517,
+      "grad_norm": 0.2403114798612083,
+      "learning_rate": 4.650572160528733e-05,
+      "loss": 0.3661,
+      "step": 2360
+    },
+    {
+      "epoch": 2.5170575692963753,
+      "grad_norm": 0.2414630524740656,
+      "learning_rate": 4.647634513400178e-05,
+      "loss": 0.3682,
+      "step": 2361
+    },
+    {
+      "epoch": 2.518123667377399,
+      "grad_norm": 0.2006519271395349,
+      "learning_rate": 4.644696507515762e-05,
+      "loss": 0.3654,
+      "step": 2362
+    },
+    {
+      "epoch": 2.5191897654584223,
+      "grad_norm": 0.24007590073272467,
+      "learning_rate": 4.641758144502985e-05,
+      "loss": 0.3659,
+      "step": 2363
+    },
+    {
+      "epoch": 2.520255863539446,
+      "grad_norm": 0.26804872309010386,
+      "learning_rate": 4.638819425989551e-05,
+      "loss": 0.3709,
+      "step": 2364
+    },
+    {
+      "epoch": 2.521321961620469,
+      "grad_norm": 0.24000354250919373,
+      "learning_rate": 4.635880353603356e-05,
+      "loss": 0.3712,
+      "step": 2365
+    },
+    {
+      "epoch": 2.5223880597014925,
+      "grad_norm": 0.2652528639205553,
+      "learning_rate": 4.632940928972491e-05,
+      "loss": 0.3688,
+      "step": 2366
+    },
+    {
+      "epoch": 2.523454157782516,
+      "grad_norm": 0.2580420852617307,
+      "learning_rate": 4.630001153725247e-05,
+      "loss": 0.3703,
+      "step": 2367
+    },
+    {
+      "epoch": 2.5245202558635396,
+      "grad_norm": 0.22089649535392641,
+      "learning_rate": 4.627061029490105e-05,
+      "loss": 0.3711,
+      "step": 2368
+    },
+    {
+      "epoch": 2.525586353944563,
+      "grad_norm": 0.2545584727108516,
+      "learning_rate": 4.6241205578957435e-05,
+      "loss": 0.3737,
+      "step": 2369
+    },
+    {
+      "epoch": 2.526652452025586,
+      "grad_norm": 0.2967974395396264,
+      "learning_rate": 4.6211797405710285e-05,
+      "loss": 0.3712,
+      "step": 2370
+    },
+    {
+      "epoch": 2.5277185501066097,
+      "grad_norm": 0.2791964105337959,
+      "learning_rate": 4.618238579145022e-05,
+      "loss": 0.3732,
+      "step": 2371
+    },
+    {
+      "epoch": 2.5287846481876333,
+      "grad_norm": 0.26473846775672055,
+      "learning_rate": 4.6152970752469716e-05,
+      "loss": 0.3672,
+      "step": 2372
+    },
+    {
+      "epoch": 2.529850746268657,
+      "grad_norm": 0.2049402609403739,
+      "learning_rate": 4.612355230506321e-05,
+      "loss": 0.3691,
+      "step": 2373
+    },
+    {
+      "epoch": 2.5309168443496803,
+      "grad_norm": 0.1856152488020746,
+      "learning_rate": 4.609413046552697e-05,
+      "loss": 0.3741,
+      "step": 2374
+    },
+    {
+      "epoch": 2.5319829424307034,
+      "grad_norm": 0.2258681946492881,
+      "learning_rate": 4.606470525015917e-05,
+      "loss": 0.369,
+      "step": 2375
+    },
+    {
+      "epoch": 2.533049040511727,
+      "grad_norm": 0.25975437520850114,
+      "learning_rate": 4.603527667525987e-05,
+      "loss": 0.3712,
+      "step": 2376
+    },
+    {
+      "epoch": 2.5341151385927505,
+      "grad_norm": 0.2306018312278267,
+      "learning_rate": 4.6005844757130937e-05,
+      "loss": 0.3644,
+      "step": 2377
+    },
+    {
+      "epoch": 2.535181236673774,
+      "grad_norm": 0.24072038440875435,
+      "learning_rate": 4.597640951207615e-05,
+      "loss": 0.3737,
+      "step": 2378
+    },
+    {
+      "epoch": 2.5362473347547976,
+      "grad_norm": 0.2686312117022459,
+      "learning_rate": 4.5946970956401086e-05,
+      "loss": 0.3646,
+      "step": 2379
+    },
+    {
+      "epoch": 2.5373134328358207,
+      "grad_norm": 0.2940976442002358,
+      "learning_rate": 4.59175291064132e-05,
+      "loss": 0.3668,
+      "step": 2380
+    },
+    {
+      "epoch": 2.538379530916844,
+      "grad_norm": 0.2926304524554624,
+      "learning_rate": 4.588808397842172e-05,
+      "loss": 0.3662,
+      "step": 2381
+    },
+    {
+      "epoch": 2.5394456289978677,
+      "grad_norm": 0.24031871433591176,
+      "learning_rate": 4.585863558873774e-05,
+      "loss": 0.365,
+      "step": 2382
+    },
+    {
+      "epoch": 2.5405117270788913,
+      "grad_norm": 0.23386304551259074,
+      "learning_rate": 4.582918395367412e-05,
+      "loss": 0.3708,
+      "step": 2383
+    },
+    {
+      "epoch": 2.541577825159915,
+      "grad_norm": 0.23138470902625616,
+      "learning_rate": 4.5799729089545546e-05,
+      "loss": 0.3666,
+      "step": 2384
+    },
+    {
+      "epoch": 2.542643923240938,
+      "grad_norm": 0.20493257086499,
+      "learning_rate": 4.577027101266847e-05,
+      "loss": 0.3641,
+      "step": 2385
+    },
+    {
+      "epoch": 2.543710021321962,
+      "grad_norm": 0.2742256917065172,
+      "learning_rate": 4.574080973936115e-05,
+      "loss": 0.3656,
+      "step": 2386
+    },
+    {
+      "epoch": 2.544776119402985,
+      "grad_norm": 0.2573784786511887,
+      "learning_rate": 4.5711345285943585e-05,
+      "loss": 0.3722,
+      "step": 2387
+    },
+    {
+      "epoch": 2.5458422174840085,
+      "grad_norm": 0.220635079626304,
+      "learning_rate": 4.568187766873757e-05,
+      "loss": 0.3707,
+      "step": 2388
+    },
+    {
+      "epoch": 2.546908315565032,
+      "grad_norm": 0.24947774232184444,
+      "learning_rate": 4.565240690406661e-05,
+      "loss": 0.3685,
+      "step": 2389
+    },
+    {
+      "epoch": 2.5479744136460556,
+      "grad_norm": 0.22455688901803345,
+      "learning_rate": 4.5622933008256e-05,
+      "loss": 0.3663,
+      "step": 2390
+    },
+    {
+      "epoch": 2.549040511727079,
+      "grad_norm": 0.24210657316421766,
+      "learning_rate": 4.559345599763273e-05,
+      "loss": 0.3715,
+      "step": 2391
+    },
+    {
+      "epoch": 2.550106609808102,
+      "grad_norm": 0.23303940791069144,
+      "learning_rate": 4.556397588852553e-05,
+      "loss": 0.3748,
+      "step": 2392
+    },
+    {
+      "epoch": 2.5511727078891258,
+      "grad_norm": 0.20346922824676458,
+      "learning_rate": 4.553449269726487e-05,
+      "loss": 0.378,
+      "step": 2393
+    },
+    {
+      "epoch": 2.5522388059701493,
+      "grad_norm": 0.25118366195456854,
+      "learning_rate": 4.550500644018289e-05,
+      "loss": 0.3722,
+      "step": 2394
+    },
+    {
+      "epoch": 2.553304904051173,
+      "grad_norm": 0.23799851735663155,
+      "learning_rate": 4.547551713361344e-05,
+      "loss": 0.3654,
+      "step": 2395
+    },
+    {
+      "epoch": 2.5543710021321964,
+      "grad_norm": 0.24008761159127065,
+      "learning_rate": 4.544602479389207e-05,
+      "loss": 0.3715,
+      "step": 2396
+    },
+    {
+      "epoch": 2.5554371002132195,
+      "grad_norm": 0.3224169072622209,
+      "learning_rate": 4.5416529437355996e-05,
+      "loss": 0.3772,
+      "step": 2397
+    },
+    {
+      "epoch": 2.556503198294243,
+      "grad_norm": 0.28620149997931676,
+      "learning_rate": 4.538703108034414e-05,
+      "loss": 0.3678,
+      "step": 2398
+    },
+    {
+      "epoch": 2.5575692963752665,
+      "grad_norm": 0.2192934282403724,
+      "learning_rate": 4.535752973919701e-05,
+      "loss": 0.3703,
+      "step": 2399
+    },
+    {
+      "epoch": 2.55863539445629,
+      "grad_norm": 0.1709676106553397,
+      "learning_rate": 4.532802543025686e-05,
+      "loss": 0.3699,
+      "step": 2400
+    },
+    {
+      "epoch": 2.5597014925373136,
+      "grad_norm": 0.24392777534293011,
+      "learning_rate": 4.529851816986752e-05,
+      "loss": 0.3687,
+      "step": 2401
+    },
+    {
+      "epoch": 2.5607675906183367,
+      "grad_norm": 0.24891562920270666,
+      "learning_rate": 4.5269007974374494e-05,
+      "loss": 0.3697,
+      "step": 2402
+    },
+    {
+      "epoch": 2.5618336886993602,
+      "grad_norm": 0.2384548720836165,
+      "learning_rate": 4.5239494860124895e-05,
+      "loss": 0.3685,
+      "step": 2403
+    },
+    {
+      "epoch": 2.5628997867803838,
+      "grad_norm": 0.27098692831127463,
+      "learning_rate": 4.5209978843467436e-05,
+      "loss": 0.3678,
+      "step": 2404
+    },
+    {
+      "epoch": 2.5639658848614073,
+      "grad_norm": 0.20778244660818992,
+      "learning_rate": 4.5180459940752474e-05,
+      "loss": 0.3667,
+      "step": 2405
+    },
+    {
+      "epoch": 2.565031982942431,
+      "grad_norm": 0.22120855767567077,
+      "learning_rate": 4.515093816833193e-05,
+      "loss": 0.3682,
+      "step": 2406
+    },
+    {
+      "epoch": 2.566098081023454,
+      "grad_norm": 0.2634670562341265,
+      "learning_rate": 4.512141354255935e-05,
+      "loss": 0.3689,
+      "step": 2407
+    },
+    {
+      "epoch": 2.5671641791044775,
+      "grad_norm": 0.2583204612978333,
+      "learning_rate": 4.509188607978983e-05,
+      "loss": 0.365,
+      "step": 2408
+    },
+    {
+      "epoch": 2.568230277185501,
+      "grad_norm": 0.279446147717918,
+      "learning_rate": 4.5062355796380066e-05,
+      "loss": 0.3685,
+      "step": 2409
+    },
+    {
+      "epoch": 2.5692963752665245,
+      "grad_norm": 0.2489838679141536,
+      "learning_rate": 4.503282270868829e-05,
+      "loss": 0.3699,
+      "step": 2410
+    },
+    {
+      "epoch": 2.570362473347548,
+      "grad_norm": 0.2818230446222446,
+      "learning_rate": 4.500328683307428e-05,
+      "loss": 0.3735,
+      "step": 2411
+    },
+    {
+      "epoch": 2.571428571428571,
+      "grad_norm": 0.32345371657550076,
+      "learning_rate": 4.4973748185899416e-05,
+      "loss": 0.3729,
+      "step": 2412
+    },
+    {
+      "epoch": 2.572494669509595,
+      "grad_norm": 0.27902407224490827,
+      "learning_rate": 4.4944206783526536e-05,
+      "loss": 0.3723,
+      "step": 2413
+    },
+    {
+      "epoch": 2.5735607675906182,
+      "grad_norm": 0.2424292615142393,
+      "learning_rate": 4.4914662642320054e-05,
+      "loss": 0.3687,
+      "step": 2414
+    },
+    {
+      "epoch": 2.574626865671642,
+      "grad_norm": 0.251352116384431,
+      "learning_rate": 4.4885115778645895e-05,
+      "loss": 0.3727,
+      "step": 2415
+    },
+    {
+      "epoch": 2.5756929637526653,
+      "grad_norm": 0.24325740030245527,
+      "learning_rate": 4.485556620887148e-05,
+      "loss": 0.3674,
+      "step": 2416
+    },
+    {
+      "epoch": 2.576759061833689,
+      "grad_norm": 0.30803045624118025,
+      "learning_rate": 4.482601394936573e-05,
+      "loss": 0.3707,
+      "step": 2417
+    },
+    {
+      "epoch": 2.5778251599147124,
+      "grad_norm": 0.3292401151984711,
+      "learning_rate": 4.479645901649908e-05,
+      "loss": 0.3732,
+      "step": 2418
+    },
+    {
+      "epoch": 2.5788912579957355,
+      "grad_norm": 0.29582277126522766,
+      "learning_rate": 4.476690142664342e-05,
+      "loss": 0.3744,
+      "step": 2419
+    },
+    {
+      "epoch": 2.579957356076759,
+      "grad_norm": 0.26686256357070676,
+      "learning_rate": 4.47373411961721e-05,
+      "loss": 0.3687,
+      "step": 2420
+    },
+    {
+      "epoch": 2.5810234541577826,
+      "grad_norm": 0.24943332540729626,
+      "learning_rate": 4.470777834145997e-05,
+      "loss": 0.3658,
+      "step": 2421
+    },
+    {
+      "epoch": 2.582089552238806,
+      "grad_norm": 0.2121258978338484,
+      "learning_rate": 4.467821287888331e-05,
+      "loss": 0.3668,
+      "step": 2422
+    },
+    {
+      "epoch": 2.5831556503198296,
+      "grad_norm": 0.2156156797629481,
+      "learning_rate": 4.464864482481984e-05,
+      "loss": 0.3642,
+      "step": 2423
+    },
+    {
+      "epoch": 2.5842217484008527,
+      "grad_norm": 0.2006814924668892,
+      "learning_rate": 4.461907419564874e-05,
+      "loss": 0.3697,
+      "step": 2424
+    },
+    {
+      "epoch": 2.5852878464818763,
+      "grad_norm": 0.210853772833808,
+      "learning_rate": 4.4589501007750595e-05,
+      "loss": 0.3749,
+      "step": 2425
+    },
+    {
+      "epoch": 2.5863539445629,
+      "grad_norm": 0.25717175869949604,
+      "learning_rate": 4.4559925277507416e-05,
+      "loss": 0.3698,
+      "step": 2426
+    },
+    {
+      "epoch": 2.5874200426439233,
+      "grad_norm": 0.25073349786641386,
+      "learning_rate": 4.4530347021302626e-05,
+      "loss": 0.3713,
+      "step": 2427
+    },
+    {
+      "epoch": 2.588486140724947,
+      "grad_norm": 0.2643644113163801,
+      "learning_rate": 4.450076625552102e-05,
+      "loss": 0.3747,
+      "step": 2428
+    },
+    {
+      "epoch": 2.58955223880597,
+      "grad_norm": 0.24060221681359292,
+      "learning_rate": 4.447118299654883e-05,
+      "loss": 0.3599,
+      "step": 2429
+    },
+    {
+      "epoch": 2.5906183368869935,
+      "grad_norm": 0.2720300708647726,
+      "learning_rate": 4.444159726077363e-05,
+      "loss": 0.3718,
+      "step": 2430
+    },
+    {
+      "epoch": 2.591684434968017,
+      "grad_norm": 0.3229511582480334,
+      "learning_rate": 4.4412009064584384e-05,
+      "loss": 0.3666,
+      "step": 2431
+    },
+    {
+      "epoch": 2.5927505330490406,
+      "grad_norm": 0.3888728224230281,
+      "learning_rate": 4.438241842437142e-05,
+      "loss": 0.3654,
+      "step": 2432
+    },
+    {
+      "epoch": 2.593816631130064,
+      "grad_norm": 0.38553138291681316,
+      "learning_rate": 4.435282535652641e-05,
+      "loss": 0.3724,
+      "step": 2433
+    },
+    {
+      "epoch": 2.594882729211087,
+      "grad_norm": 0.31969838656564453,
+      "learning_rate": 4.4323229877442374e-05,
+      "loss": 0.3694,
+      "step": 2434
+    },
+    {
+      "epoch": 2.5959488272921107,
+      "grad_norm": 0.40489827746268775,
+      "learning_rate": 4.429363200351366e-05,
+      "loss": 0.3701,
+      "step": 2435
+    },
+    {
+      "epoch": 2.5970149253731343,
+      "grad_norm": 0.4247590203835174,
+      "learning_rate": 4.426403175113598e-05,
+      "loss": 0.3751,
+      "step": 2436
+    },
+    {
+      "epoch": 2.598081023454158,
+      "grad_norm": 0.3951246174491148,
+      "learning_rate": 4.423442913670631e-05,
+      "loss": 0.3693,
+      "step": 2437
+    },
+    {
+      "epoch": 2.5991471215351813,
+      "grad_norm": 0.3111985733932929,
+      "learning_rate": 4.420482417662297e-05,
+      "loss": 0.3704,
+      "step": 2438
+    },
+    {
+      "epoch": 2.6002132196162044,
+      "grad_norm": 0.18873972891323956,
+      "learning_rate": 4.417521688728556e-05,
+      "loss": 0.3661,
+      "step": 2439
+    },
+    {
+      "epoch": 2.6012793176972284,
+      "grad_norm": 0.2495555063976354,
+      "learning_rate": 4.4145607285095e-05,
+      "loss": 0.3755,
+      "step": 2440
+    },
+    {
+      "epoch": 2.6023454157782515,
+      "grad_norm": 0.36376650288976037,
+      "learning_rate": 4.411599538645345e-05,
+      "loss": 0.3681,
+      "step": 2441
+    },
+    {
+      "epoch": 2.603411513859275,
+      "grad_norm": 0.4693799411527801,
+      "learning_rate": 4.408638120776436e-05,
+      "loss": 0.3705,
+      "step": 2442
+    },
+    {
+      "epoch": 2.6044776119402986,
+      "grad_norm": 0.46887250402317515,
+      "learning_rate": 4.405676476543247e-05,
+      "loss": 0.3671,
+      "step": 2443
+    },
+    {
+      "epoch": 2.605543710021322,
+      "grad_norm": 0.36949597097867015,
+      "learning_rate": 4.402714607586373e-05,
+      "loss": 0.3738,
+      "step": 2444
+    },
+    {
+      "epoch": 2.6066098081023457,
+      "grad_norm": 0.3009704252138755,
+      "learning_rate": 4.399752515546538e-05,
+      "loss": 0.3681,
+      "step": 2445
+    },
+    {
+      "epoch": 2.6076759061833688,
+      "grad_norm": 0.237310962341784,
+      "learning_rate": 4.396790202064583e-05,
+      "loss": 0.3686,
+      "step": 2446
+    },
+    {
+      "epoch": 2.6087420042643923,
+      "grad_norm": 0.23276180946430972,
+      "learning_rate": 4.393827668781478e-05,
+      "loss": 0.3683,
+      "step": 2447
+    },
+    {
+      "epoch": 2.609808102345416,
+      "grad_norm": 0.32501790451361395,
+      "learning_rate": 4.390864917338313e-05,
+      "loss": 0.3708,
+      "step": 2448
+    },
+    {
+      "epoch": 2.6108742004264394,
+      "grad_norm": 0.3613827214092068,
+      "learning_rate": 4.387901949376297e-05,
+      "loss": 0.3651,
+      "step": 2449
+    },
+    {
+      "epoch": 2.611940298507463,
+      "grad_norm": 0.3553110668616482,
+      "learning_rate": 4.3849387665367614e-05,
+      "loss": 0.3728,
+      "step": 2450
+    },
+    {
+      "epoch": 2.613006396588486,
+      "grad_norm": 0.292471368689879,
+      "learning_rate": 4.381975370461155e-05,
+      "loss": 0.3745,
+      "step": 2451
+    },
+    {
+      "epoch": 2.6140724946695095,
+      "grad_norm": 0.2016201008306948,
+      "learning_rate": 4.379011762791045e-05,
+      "loss": 0.3671,
+      "step": 2452
+    },
+    {
+      "epoch": 2.615138592750533,
+      "grad_norm": 0.2616397181578619,
+      "learning_rate": 4.3760479451681164e-05,
+      "loss": 0.3705,
+      "step": 2453
+    },
+    {
+      "epoch": 2.6162046908315566,
+      "grad_norm": 0.3997537994958033,
+      "learning_rate": 4.3730839192341705e-05,
+      "loss": 0.3699,
+      "step": 2454
+    },
+    {
+      "epoch": 2.61727078891258,
+      "grad_norm": 0.4296512845219555,
+      "learning_rate": 4.370119686631123e-05,
+      "loss": 0.3654,
+      "step": 2455
+    },
+    {
+      "epoch": 2.6183368869936032,
+      "grad_norm": 0.327116358825709,
+      "learning_rate": 4.3671552490010036e-05,
+      "loss": 0.372,
+      "step": 2456
+    },
+    {
+      "epoch": 2.6194029850746268,
+      "grad_norm": 0.27235595479261326,
+      "learning_rate": 4.3641906079859584e-05,
+      "loss": 0.3698,
+      "step": 2457
+    },
+    {
+      "epoch": 2.6204690831556503,
+      "grad_norm": 0.35732218737027005,
+      "learning_rate": 4.3612257652282446e-05,
+      "loss": 0.3667,
+      "step": 2458
+    },
+    {
+      "epoch": 2.621535181236674,
+      "grad_norm": 0.37424865085620856,
+      "learning_rate": 4.358260722370229e-05,
+      "loss": 0.364,
+      "step": 2459
+    },
+    {
+      "epoch": 2.6226012793176974,
+      "grad_norm": 0.42605572507707096,
+      "learning_rate": 4.355295481054394e-05,
+      "loss": 0.3735,
+      "step": 2460
+    },
+    {
+      "epoch": 2.6236673773987205,
+      "grad_norm": 0.45443453638837905,
+      "learning_rate": 4.352330042923328e-05,
+      "loss": 0.372,
+      "step": 2461
+    },
+    {
+      "epoch": 2.624733475479744,
+      "grad_norm": 0.41383949227823347,
+      "learning_rate": 4.3493644096197274e-05,
+      "loss": 0.3717,
+      "step": 2462
+    },
+    {
+      "epoch": 2.6257995735607675,
+      "grad_norm": 0.2860714841801369,
+      "learning_rate": 4.3463985827864024e-05,
+      "loss": 0.3646,
+      "step": 2463
+    },
+    {
+      "epoch": 2.626865671641791,
+      "grad_norm": 0.23137129814316412,
+      "learning_rate": 4.343432564066264e-05,
+      "loss": 0.3701,
+      "step": 2464
+    },
+    {
+      "epoch": 2.6279317697228146,
+      "grad_norm": 0.2512842592763747,
+      "learning_rate": 4.340466355102336e-05,
+      "loss": 0.3672,
+      "step": 2465
+    },
+    {
+      "epoch": 2.6289978678038377,
+      "grad_norm": 0.28901573296850475,
+      "learning_rate": 4.3374999575377393e-05,
+      "loss": 0.3651,
+      "step": 2466
+    },
+    {
+      "epoch": 2.6300639658848612,
+      "grad_norm": 0.2993536460359868,
+      "learning_rate": 4.334533373015709e-05,
+      "loss": 0.3688,
+      "step": 2467
+    },
+    {
+      "epoch": 2.631130063965885,
+      "grad_norm": 0.2614300970595663,
+      "learning_rate": 4.3315666031795736e-05,
+      "loss": 0.3678,
+      "step": 2468
+    },
+    {
+      "epoch": 2.6321961620469083,
+      "grad_norm": 0.25189646001058735,
+      "learning_rate": 4.328599649672774e-05,
+      "loss": 0.3712,
+      "step": 2469
+    },
+    {
+      "epoch": 2.633262260127932,
+      "grad_norm": 0.24766325430481256,
+      "learning_rate": 4.3256325141388464e-05,
+      "loss": 0.3641,
+      "step": 2470
+    },
+    {
+      "epoch": 2.6343283582089554,
+      "grad_norm": 0.3035417706772448,
+      "learning_rate": 4.3226651982214286e-05,
+      "loss": 0.3684,
+      "step": 2471
+    },
+    {
+      "epoch": 2.635394456289979,
+      "grad_norm": 0.2636223605711239,
+      "learning_rate": 4.319697703564261e-05,
+      "loss": 0.3688,
+      "step": 2472
+    },
+    {
+      "epoch": 2.636460554371002,
+      "grad_norm": 0.20666359945758703,
+      "learning_rate": 4.3167300318111805e-05,
+      "loss": 0.3636,
+      "step": 2473
+    },
+    {
+      "epoch": 2.6375266524520256,
+      "grad_norm": 0.24577933602515428,
+      "learning_rate": 4.313762184606124e-05,
+      "loss": 0.3681,
+      "step": 2474
+    },
+    {
+      "epoch": 2.638592750533049,
+      "grad_norm": 0.23261579434937743,
+      "learning_rate": 4.310794163593123e-05,
+      "loss": 0.3665,
+      "step": 2475
+    },
+    {
+      "epoch": 2.6396588486140726,
+      "grad_norm": 0.21528938082866045,
+      "learning_rate": 4.307825970416308e-05,
+      "loss": 0.3712,
+      "step": 2476
+    },
+    {
+      "epoch": 2.640724946695096,
+      "grad_norm": 0.2560143422062812,
+      "learning_rate": 4.3048576067199034e-05,
+      "loss": 0.3668,
+      "step": 2477
+    },
+    {
+      "epoch": 2.6417910447761193,
+      "grad_norm": 0.24544814002376064,
+      "learning_rate": 4.3018890741482296e-05,
+      "loss": 0.3705,
+      "step": 2478
+    },
+    {
+      "epoch": 2.642857142857143,
+      "grad_norm": 0.2439487587446761,
+      "learning_rate": 4.298920374345698e-05,
+      "loss": 0.3656,
+      "step": 2479
+    },
+    {
+      "epoch": 2.6439232409381663,
+      "grad_norm": 0.25308388826969747,
+      "learning_rate": 4.295951508956814e-05,
+      "loss": 0.3638,
+      "step": 2480
+    },
+    {
+      "epoch": 2.64498933901919,
+      "grad_norm": 0.30616158821118233,
+      "learning_rate": 4.292982479626175e-05,
+      "loss": 0.3628,
+      "step": 2481
+    },
+    {
+      "epoch": 2.6460554371002134,
+      "grad_norm": 0.2635196635922222,
+      "learning_rate": 4.290013287998469e-05,
+      "loss": 0.369,
+      "step": 2482
+    },
+    {
+      "epoch": 2.6471215351812365,
+      "grad_norm": 0.2737120868957713,
+      "learning_rate": 4.287043935718474e-05,
+      "loss": 0.3729,
+      "step": 2483
+    },
+    {
+      "epoch": 2.64818763326226,
+      "grad_norm": 0.31387047069745105,
+      "learning_rate": 4.2840744244310565e-05,
+      "loss": 0.3749,
+      "step": 2484
+    },
+    {
+      "epoch": 2.6492537313432836,
+      "grad_norm": 0.29584704047191807,
+      "learning_rate": 4.281104755781172e-05,
+      "loss": 0.371,
+      "step": 2485
+    },
+    {
+      "epoch": 2.650319829424307,
+      "grad_norm": 0.31359007568593433,
+      "learning_rate": 4.278134931413862e-05,
+      "loss": 0.3698,
+      "step": 2486
+    },
+    {
+      "epoch": 2.6513859275053306,
+      "grad_norm": 0.2606911686791426,
+      "learning_rate": 4.275164952974256e-05,
+      "loss": 0.3712,
+      "step": 2487
+    },
+    {
+      "epoch": 2.6524520255863537,
+      "grad_norm": 0.2110093354675097,
+      "learning_rate": 4.272194822107566e-05,
+      "loss": 0.3699,
+      "step": 2488
+    },
+    {
+      "epoch": 2.6535181236673773,
+      "grad_norm": 0.1961521426965646,
+      "learning_rate": 4.2692245404590906e-05,
+      "loss": 0.366,
+      "step": 2489
+    },
+    {
+      "epoch": 2.654584221748401,
+      "grad_norm": 0.2515365045262743,
+      "learning_rate": 4.266254109674213e-05,
+      "loss": 0.3683,
+      "step": 2490
+    },
+    {
+      "epoch": 2.6556503198294243,
+      "grad_norm": 0.3249680113246906,
+      "learning_rate": 4.263283531398395e-05,
+      "loss": 0.3637,
+      "step": 2491
+    },
+    {
+      "epoch": 2.656716417910448,
+      "grad_norm": 0.3118378628242833,
+      "learning_rate": 4.260312807277185e-05,
+      "loss": 0.3719,
+      "step": 2492
+    },
+    {
+      "epoch": 2.657782515991471,
+      "grad_norm": 0.22886489658313464,
+      "learning_rate": 4.2573419389562074e-05,
+      "loss": 0.3715,
+      "step": 2493
+    },
+    {
+      "epoch": 2.6588486140724945,
+      "grad_norm": 0.1662131591027839,
+      "learning_rate": 4.254370928081171e-05,
+      "loss": 0.3772,
+      "step": 2494
+    },
+    {
+      "epoch": 2.659914712153518,
+      "grad_norm": 0.1969499093984148,
+      "learning_rate": 4.25139977629786e-05,
+      "loss": 0.3699,
+      "step": 2495
+    },
+    {
+      "epoch": 2.6609808102345416,
+      "grad_norm": 0.25930037314010335,
+      "learning_rate": 4.248428485252139e-05,
+      "loss": 0.3655,
+      "step": 2496
+    },
+    {
+      "epoch": 2.662046908315565,
+      "grad_norm": 0.26415096213448547,
+      "learning_rate": 4.2454570565899476e-05,
+      "loss": 0.3734,
+      "step": 2497
+    },
+    {
+      "epoch": 2.663113006396588,
+      "grad_norm": 0.27116001496004927,
+      "learning_rate": 4.242485491957305e-05,
+      "loss": 0.3657,
+      "step": 2498
+    },
+    {
+      "epoch": 2.664179104477612,
+      "grad_norm": 0.25863792879198805,
+      "learning_rate": 4.239513793000301e-05,
+      "loss": 0.3753,
+      "step": 2499
+    },
+    {
+      "epoch": 2.6652452025586353,
+      "grad_norm": 0.23133617297550452,
+      "learning_rate": 4.2365419613651035e-05,
+      "loss": 0.3737,
+      "step": 2500
+    },
+    {
+      "epoch": 2.666311300639659,
+      "grad_norm": 0.20980481301021206,
+      "learning_rate": 4.233569998697954e-05,
+      "loss": 0.3701,
+      "step": 2501
+    },
+    {
+      "epoch": 2.6673773987206824,
+      "grad_norm": 0.24297415494208804,
+      "learning_rate": 4.2305979066451626e-05,
+      "loss": 0.3677,
+      "step": 2502
+    },
+    {
+      "epoch": 2.668443496801706,
+      "grad_norm": 0.2493426212110305,
+      "learning_rate": 4.2276256868531165e-05,
+      "loss": 0.3652,
+      "step": 2503
+    },
+    {
+      "epoch": 2.6695095948827294,
+      "grad_norm": 0.2303457530378945,
+      "learning_rate": 4.22465334096827e-05,
+      "loss": 0.3771,
+      "step": 2504
+    },
+    {
+      "epoch": 2.6705756929637525,
+      "grad_norm": 0.223769372564616,
+      "learning_rate": 4.221680870637148e-05,
+      "loss": 0.3673,
+      "step": 2505
+    },
+    {
+      "epoch": 2.671641791044776,
+      "grad_norm": 0.2590177839797346,
+      "learning_rate": 4.2187082775063436e-05,
+      "loss": 0.3685,
+      "step": 2506
+    },
+    {
+      "epoch": 2.6727078891257996,
+      "grad_norm": 0.24760592724404568,
+      "learning_rate": 4.2157355632225216e-05,
+      "loss": 0.368,
+      "step": 2507
+    },
+    {
+      "epoch": 2.673773987206823,
+      "grad_norm": 0.22621973085952393,
+      "learning_rate": 4.2127627294324095e-05,
+      "loss": 0.3676,
+      "step": 2508
+    },
+    {
+      "epoch": 2.6748400852878467,
+      "grad_norm": 0.2335594327305558,
+      "learning_rate": 4.2097897777828025e-05,
+      "loss": 0.3743,
+      "step": 2509
+    },
+    {
+      "epoch": 2.6759061833688698,
+      "grad_norm": 0.2281132024099097,
+      "learning_rate": 4.2068167099205625e-05,
+      "loss": 0.3692,
+      "step": 2510
+    },
+    {
+      "epoch": 2.6769722814498933,
+      "grad_norm": 0.22341724033801894,
+      "learning_rate": 4.203843527492613e-05,
+      "loss": 0.3702,
+      "step": 2511
+    },
+    {
+      "epoch": 2.678038379530917,
+      "grad_norm": 0.2424014361387363,
+      "learning_rate": 4.200870232145943e-05,
+      "loss": 0.3706,
+      "step": 2512
+    },
+    {
+      "epoch": 2.6791044776119404,
+      "grad_norm": 0.1953899351329716,
+      "learning_rate": 4.1978968255276043e-05,
+      "loss": 0.3652,
+      "step": 2513
+    },
+    {
+      "epoch": 2.680170575692964,
+      "grad_norm": 0.22170371983312567,
+      "learning_rate": 4.1949233092847095e-05,
+      "loss": 0.3684,
+      "step": 2514
+    },
+    {
+      "epoch": 2.681236673773987,
+      "grad_norm": 0.23765221374061374,
+      "learning_rate": 4.1919496850644316e-05,
+      "loss": 0.3637,
+      "step": 2515
+    },
+    {
+      "epoch": 2.6823027718550105,
+      "grad_norm": 0.23687777883894273,
+      "learning_rate": 4.1889759545140045e-05,
+      "loss": 0.3655,
+      "step": 2516
+    },
+    {
+      "epoch": 2.683368869936034,
+      "grad_norm": 0.22338536825632072,
+      "learning_rate": 4.186002119280718e-05,
+      "loss": 0.3707,
+      "step": 2517
+    },
+    {
+      "epoch": 2.6844349680170576,
+      "grad_norm": 0.22005314669076753,
+      "learning_rate": 4.183028181011927e-05,
+      "loss": 0.374,
+      "step": 2518
+    },
+    {
+      "epoch": 2.685501066098081,
+      "grad_norm": 0.2303045665765263,
+      "learning_rate": 4.180054141355035e-05,
+      "loss": 0.3747,
+      "step": 2519
+    },
+    {
+      "epoch": 2.6865671641791042,
+      "grad_norm": 0.18534817689847655,
+      "learning_rate": 4.177080001957506e-05,
+      "loss": 0.3667,
+      "step": 2520
+    },
+    {
+      "epoch": 2.6876332622601278,
+      "grad_norm": 0.2629553843000235,
+      "learning_rate": 4.174105764466859e-05,
+      "loss": 0.3705,
+      "step": 2521
+    },
+    {
+      "epoch": 2.6886993603411513,
+      "grad_norm": 0.29148782597094064,
+      "learning_rate": 4.1711314305306676e-05,
+      "loss": 0.3704,
+      "step": 2522
+    },
+    {
+      "epoch": 2.689765458422175,
+      "grad_norm": 0.2811624835341636,
+      "learning_rate": 4.168157001796557e-05,
+      "loss": 0.3723,
+      "step": 2523
+    },
+    {
+      "epoch": 2.6908315565031984,
+      "grad_norm": 0.24880403947380741,
+      "learning_rate": 4.165182479912208e-05,
+      "loss": 0.367,
+      "step": 2524
+    },
+    {
+      "epoch": 2.6918976545842215,
+      "grad_norm": 0.2008980746882621,
+      "learning_rate": 4.1622078665253486e-05,
+      "loss": 0.3661,
+      "step": 2525
+    },
+    {
+      "epoch": 2.6929637526652455,
+      "grad_norm": 0.23350481252873526,
+      "learning_rate": 4.159233163283762e-05,
+      "loss": 0.3762,
+      "step": 2526
+    },
+    {
+      "epoch": 2.6940298507462686,
+      "grad_norm": 0.20261861943939447,
+      "learning_rate": 4.156258371835279e-05,
+      "loss": 0.3701,
+      "step": 2527
+    },
+    {
+      "epoch": 2.695095948827292,
+      "grad_norm": 0.25996452870016273,
+      "learning_rate": 4.153283493827777e-05,
+      "loss": 0.3776,
+      "step": 2528
+    },
+    {
+      "epoch": 2.6961620469083156,
+      "grad_norm": 0.19687929153495226,
+      "learning_rate": 4.150308530909187e-05,
+      "loss": 0.3636,
+      "step": 2529
+    },
+    {
+      "epoch": 2.697228144989339,
+      "grad_norm": 0.19556973866759192,
+      "learning_rate": 4.147333484727484e-05,
+      "loss": 0.3586,
+      "step": 2530
+    },
+    {
+      "epoch": 2.6982942430703627,
+      "grad_norm": 0.21721440727421978,
+      "learning_rate": 4.144358356930686e-05,
+      "loss": 0.3685,
+      "step": 2531
+    },
+    {
+      "epoch": 2.699360341151386,
+      "grad_norm": 0.2289561477297396,
+      "learning_rate": 4.141383149166861e-05,
+      "loss": 0.3693,
+      "step": 2532
+    },
+    {
+      "epoch": 2.7004264392324093,
+      "grad_norm": 0.21622654938339614,
+      "learning_rate": 4.138407863084119e-05,
+      "loss": 0.3676,
+      "step": 2533
+    },
+    {
+      "epoch": 2.701492537313433,
+      "grad_norm": 0.20523535800109563,
+      "learning_rate": 4.1354325003306146e-05,
+      "loss": 0.369,
+      "step": 2534
+    },
+    {
+      "epoch": 2.7025586353944564,
+      "grad_norm": 0.24463058005414473,
+      "learning_rate": 4.132457062554543e-05,
+      "loss": 0.3677,
+      "step": 2535
+    },
+    {
+      "epoch": 2.70362473347548,
+      "grad_norm": 0.1851731295878307,
+      "learning_rate": 4.129481551404143e-05,
+      "loss": 0.374,
+      "step": 2536
+    },
+    {
+      "epoch": 2.704690831556503,
+      "grad_norm": 0.19922292111977952,
+      "learning_rate": 4.1265059685276936e-05,
+      "loss": 0.3654,
+      "step": 2537
+    },
+    {
+      "epoch": 2.7057569296375266,
+      "grad_norm": 0.2752693144696535,
+      "learning_rate": 4.123530315573512e-05,
+      "loss": 0.3694,
+      "step": 2538
+    },
+    {
+      "epoch": 2.70682302771855,
+      "grad_norm": 0.23479171085838138,
+      "learning_rate": 4.120554594189955e-05,
+      "loss": 0.3734,
+      "step": 2539
+    },
+    {
+      "epoch": 2.7078891257995736,
+      "grad_norm": 0.1999301041207985,
+      "learning_rate": 4.117578806025419e-05,
+      "loss": 0.3674,
+      "step": 2540
+    },
+    {
+      "epoch": 2.708955223880597,
+      "grad_norm": 0.21844202367376767,
+      "learning_rate": 4.114602952728335e-05,
+      "loss": 0.3695,
+      "step": 2541
+    },
+    {
+      "epoch": 2.7100213219616203,
+      "grad_norm": 0.22666372255765088,
+      "learning_rate": 4.111627035947171e-05,
+      "loss": 0.3675,
+      "step": 2542
+    },
+    {
+      "epoch": 2.711087420042644,
+      "grad_norm": 0.20627199369884083,
+      "learning_rate": 4.108651057330432e-05,
+      "loss": 0.375,
+      "step": 2543
+    },
+    {
+      "epoch": 2.7121535181236673,
+      "grad_norm": 0.17843299855865702,
+      "learning_rate": 4.1056750185266515e-05,
+      "loss": 0.3685,
+      "step": 2544
+    },
+    {
+      "epoch": 2.713219616204691,
+      "grad_norm": 0.2435265879503595,
+      "learning_rate": 4.102698921184405e-05,
+      "loss": 0.37,
+      "step": 2545
+    },
+    {
+      "epoch": 2.7142857142857144,
+      "grad_norm": 0.2682269696770955,
+      "learning_rate": 4.0997227669522924e-05,
+      "loss": 0.3607,
+      "step": 2546
+    },
+    {
+      "epoch": 2.7153518123667375,
+      "grad_norm": 0.23667515131488986,
+      "learning_rate": 4.096746557478949e-05,
+      "loss": 0.372,
+      "step": 2547
+    },
+    {
+      "epoch": 2.716417910447761,
+      "grad_norm": 0.20880475425387418,
+      "learning_rate": 4.0937702944130426e-05,
+      "loss": 0.368,
+      "step": 2548
+    },
+    {
+      "epoch": 2.7174840085287846,
+      "grad_norm": 0.18392143745348416,
+      "learning_rate": 4.0907939794032654e-05,
+      "loss": 0.3733,
+      "step": 2549
+    },
+    {
+      "epoch": 2.718550106609808,
+      "grad_norm": 0.2165652019434693,
+      "learning_rate": 4.087817614098343e-05,
+      "loss": 0.3751,
+      "step": 2550
+    },
+    {
+      "epoch": 2.7196162046908317,
+      "grad_norm": 0.24037086751445425,
+      "learning_rate": 4.084841200147025e-05,
+      "loss": 0.3739,
+      "step": 2551
+    },
+    {
+      "epoch": 2.7206823027718547,
+      "grad_norm": 0.2501714427485925,
+      "learning_rate": 4.0818647391980926e-05,
+      "loss": 0.3737,
+      "step": 2552
+    },
+    {
+      "epoch": 2.7217484008528787,
+      "grad_norm": 0.2344141067398134,
+      "learning_rate": 4.078888232900349e-05,
+      "loss": 0.3698,
+      "step": 2553
+    },
+    {
+      "epoch": 2.722814498933902,
+      "grad_norm": 0.2349463019216503,
+      "learning_rate": 4.075911682902625e-05,
+      "loss": 0.3736,
+      "step": 2554
+    },
+    {
+      "epoch": 2.7238805970149254,
+      "grad_norm": 0.20029237944534903,
+      "learning_rate": 4.0729350908537724e-05,
+      "loss": 0.3757,
+      "step": 2555
+    },
+    {
+      "epoch": 2.724946695095949,
+      "grad_norm": 0.27878488904776033,
+      "learning_rate": 4.069958458402671e-05,
+      "loss": 0.3683,
+      "step": 2556
+    },
+    {
+      "epoch": 2.7260127931769724,
+      "grad_norm": 0.29774631195776696,
+      "learning_rate": 4.06698178719822e-05,
+      "loss": 0.3685,
+      "step": 2557
+    },
+    {
+      "epoch": 2.727078891257996,
+      "grad_norm": 0.2104346445232383,
+      "learning_rate": 4.0640050788893386e-05,
+      "loss": 0.3652,
+      "step": 2558
+    },
+    {
+      "epoch": 2.728144989339019,
+      "grad_norm": 0.2317958367423688,
+      "learning_rate": 4.0610283351249716e-05,
+      "loss": 0.3692,
+      "step": 2559
+    },
+    {
+      "epoch": 2.7292110874200426,
+      "grad_norm": 0.24510244686355384,
+      "learning_rate": 4.058051557554078e-05,
+      "loss": 0.3665,
+      "step": 2560
+    },
+    {
+      "epoch": 2.730277185501066,
+      "grad_norm": 0.19542448243466262,
+      "learning_rate": 4.0550747478256384e-05,
+      "loss": 0.3648,
+      "step": 2561
+    },
+    {
+      "epoch": 2.7313432835820897,
+      "grad_norm": 0.2389772106956452,
+      "learning_rate": 4.052097907588652e-05,
+      "loss": 0.3703,
+      "step": 2562
+    },
+    {
+      "epoch": 2.732409381663113,
+      "grad_norm": 0.3107936261643165,
+      "learning_rate": 4.049121038492131e-05,
+      "loss": 0.3788,
+      "step": 2563
+    },
+    {
+      "epoch": 2.7334754797441363,
+      "grad_norm": 0.2919215370644085,
+      "learning_rate": 4.0461441421851075e-05,
+      "loss": 0.3606,
+      "step": 2564
+    },
+    {
+      "epoch": 2.73454157782516,
+      "grad_norm": 0.24187272238294624,
+      "learning_rate": 4.043167220316628e-05,
+      "loss": 0.3758,
+      "step": 2565
+    },
+    {
+      "epoch": 2.7356076759061834,
+      "grad_norm": 0.24598827183680877,
+      "learning_rate": 4.040190274535752e-05,
+      "loss": 0.3715,
+      "step": 2566
+    },
+    {
+      "epoch": 2.736673773987207,
+      "grad_norm": 0.21275237895976437,
+      "learning_rate": 4.037213306491552e-05,
+      "loss": 0.3739,
+      "step": 2567
+    },
+    {
+      "epoch": 2.7377398720682304,
+      "grad_norm": 0.25491281948796235,
+      "learning_rate": 4.0342363178331146e-05,
+      "loss": 0.3671,
+      "step": 2568
+    },
+    {
+      "epoch": 2.7388059701492535,
+      "grad_norm": 0.2833091097168945,
+      "learning_rate": 4.031259310209536e-05,
+      "loss": 0.371,
+      "step": 2569
+    },
+    {
+      "epoch": 2.739872068230277,
+      "grad_norm": 0.18872311871080574,
+      "learning_rate": 4.028282285269925e-05,
+      "loss": 0.3759,
+      "step": 2570
+    },
+    {
+      "epoch": 2.7409381663113006,
+      "grad_norm": 0.19447861987862541,
+      "learning_rate": 4.0253052446633966e-05,
+      "loss": 0.3713,
+      "step": 2571
+    },
+    {
+      "epoch": 2.742004264392324,
+      "grad_norm": 0.20648269121909482,
+      "learning_rate": 4.022328190039079e-05,
+      "loss": 0.3647,
+      "step": 2572
+    },
+    {
+      "epoch": 2.7430703624733477,
+      "grad_norm": 0.1961509547861183,
+      "learning_rate": 4.019351123046104e-05,
+      "loss": 0.3594,
+      "step": 2573
+    },
+    {
+      "epoch": 2.7441364605543708,
+      "grad_norm": 0.21126880262710795,
+      "learning_rate": 4.0163740453336125e-05,
+      "loss": 0.3675,
+      "step": 2574
+    },
+    {
+      "epoch": 2.7452025586353943,
+      "grad_norm": 0.20602815602021118,
+      "learning_rate": 4.0133969585507514e-05,
+      "loss": 0.3674,
+      "step": 2575
+    },
+    {
+      "epoch": 2.746268656716418,
+      "grad_norm": 0.21819379544686981,
+      "learning_rate": 4.010419864346671e-05,
+      "loss": 0.3759,
+      "step": 2576
+    },
+    {
+      "epoch": 2.7473347547974414,
+      "grad_norm": 0.2076650890871053,
+      "learning_rate": 4.0074427643705274e-05,
+      "loss": 0.3676,
+      "step": 2577
+    },
+    {
+      "epoch": 2.748400852878465,
+      "grad_norm": 0.2017797586755664,
+      "learning_rate": 4.004465660271479e-05,
+      "loss": 0.3685,
+      "step": 2578
+    },
+    {
+      "epoch": 2.749466950959488,
+      "grad_norm": 0.2744626448821499,
+      "learning_rate": 4.001488553698687e-05,
+      "loss": 0.3691,
+      "step": 2579
+    },
+    {
+      "epoch": 2.750533049040512,
+      "grad_norm": 0.22575769767487924,
+      "learning_rate": 3.998511446301315e-05,
+      "loss": 0.3713,
+      "step": 2580
+    },
+    {
+      "epoch": 2.751599147121535,
+      "grad_norm": 0.18090826243386235,
+      "learning_rate": 3.995534339728522e-05,
+      "loss": 0.3643,
+      "step": 2581
+    },
+    {
+      "epoch": 2.7526652452025586,
+      "grad_norm": 0.1810203385011346,
+      "learning_rate": 3.992557235629473e-05,
+      "loss": 0.3679,
+      "step": 2582
+    },
+    {
+      "epoch": 2.753731343283582,
+      "grad_norm": 0.2004645021897271,
+      "learning_rate": 3.989580135653329e-05,
+      "loss": 0.3623,
+      "step": 2583
+    },
+    {
+      "epoch": 2.7547974413646057,
+      "grad_norm": 0.20665259172252912,
+      "learning_rate": 3.98660304144925e-05,
+      "loss": 0.3709,
+      "step": 2584
+    },
+    {
+      "epoch": 2.7558635394456292,
+      "grad_norm": 0.21652327508884206,
+      "learning_rate": 3.983625954666389e-05,
+      "loss": 0.3638,
+      "step": 2585
+    },
+    {
+      "epoch": 2.7569296375266523,
+      "grad_norm": 0.1905584143867772,
+      "learning_rate": 3.9806488769538966e-05,
+      "loss": 0.3698,
+      "step": 2586
+    },
+    {
+      "epoch": 2.757995735607676,
+      "grad_norm": 0.2122311206240018,
+      "learning_rate": 3.977671809960923e-05,
+      "loss": 0.3671,
+      "step": 2587
+    },
+    {
+      "epoch": 2.7590618336886994,
+      "grad_norm": 0.29171836988191924,
+      "learning_rate": 3.974694755336604e-05,
+      "loss": 0.3693,
+      "step": 2588
+    },
+    {
+      "epoch": 2.760127931769723,
+      "grad_norm": 0.27181002702035517,
+      "learning_rate": 3.971717714730076e-05,
+      "loss": 0.3665,
+      "step": 2589
+    },
+    {
+      "epoch": 2.7611940298507465,
+      "grad_norm": 0.21014364337275343,
+      "learning_rate": 3.968740689790464e-05,
+      "loss": 0.3729,
+      "step": 2590
+    },
+    {
+      "epoch": 2.7622601279317696,
+      "grad_norm": 0.2526245805801581,
+      "learning_rate": 3.9657636821668874e-05,
+      "loss": 0.3713,
+      "step": 2591
+    },
+    {
+      "epoch": 2.763326226012793,
+      "grad_norm": 0.25768688989060495,
+      "learning_rate": 3.9627866935084496e-05,
+      "loss": 0.3689,
+      "step": 2592
+    },
+    {
+      "epoch": 2.7643923240938166,
+      "grad_norm": 0.2526926794684304,
+      "learning_rate": 3.959809725464249e-05,
+      "loss": 0.369,
+      "step": 2593
+    },
+    {
+      "epoch": 2.76545842217484,
+      "grad_norm": 0.28824876900515534,
+      "learning_rate": 3.956832779683374e-05,
+      "loss": 0.3716,
+      "step": 2594
+    },
+    {
+      "epoch": 2.7665245202558637,
+      "grad_norm": 0.26584418015578326,
+      "learning_rate": 3.953855857814894e-05,
+      "loss": 0.3703,
+      "step": 2595
+    },
+    {
+      "epoch": 2.767590618336887,
+      "grad_norm": 0.23043100400918048,
+      "learning_rate": 3.950878961507871e-05,
+      "loss": 0.3652,
+      "step": 2596
+    },
+    {
+      "epoch": 2.7686567164179103,
+      "grad_norm": 0.2502510778112517,
+      "learning_rate": 3.94790209241135e-05,
+      "loss": 0.3672,
+      "step": 2597
+    },
+    {
+      "epoch": 2.769722814498934,
+      "grad_norm": 0.2639439073598543,
+      "learning_rate": 3.944925252174363e-05,
+      "loss": 0.3788,
+      "step": 2598
+    },
+    {
+      "epoch": 2.7707889125799574,
+      "grad_norm": 0.2841116000042142,
+      "learning_rate": 3.9419484424459235e-05,
+      "loss": 0.3727,
+      "step": 2599
+    },
+    {
+      "epoch": 2.771855010660981,
+      "grad_norm": 0.2735556859075221,
+      "learning_rate": 3.938971664875029e-05,
+      "loss": 0.3717,
+      "step": 2600
+    },
+    {
+      "epoch": 2.772921108742004,
+      "grad_norm": 0.28637597882164606,
+      "learning_rate": 3.935994921110661e-05,
+      "loss": 0.3688,
+      "step": 2601
+    },
+    {
+      "epoch": 2.7739872068230276,
+      "grad_norm": 0.265199180250517,
+      "learning_rate": 3.933018212801782e-05,
+      "loss": 0.3756,
+      "step": 2602
+    },
+    {
+      "epoch": 2.775053304904051,
+      "grad_norm": 0.24641376266091428,
+      "learning_rate": 3.9300415415973295e-05,
+      "loss": 0.3694,
+      "step": 2603
+    },
+    {
+      "epoch": 2.7761194029850746,
+      "grad_norm": 0.2029121854284716,
+      "learning_rate": 3.9270649091462276e-05,
+      "loss": 0.3619,
+      "step": 2604
+    },
+    {
+      "epoch": 2.777185501066098,
+      "grad_norm": 0.31971921589949576,
+      "learning_rate": 3.9240883170973776e-05,
+      "loss": 0.3664,
+      "step": 2605
+    },
+    {
+      "epoch": 2.7782515991471213,
+      "grad_norm": 0.34026995782754504,
+      "learning_rate": 3.9211117670996524e-05,
+      "loss": 0.3709,
+      "step": 2606
+    },
+    {
+      "epoch": 2.7793176972281453,
+      "grad_norm": 0.31436146070519905,
+      "learning_rate": 3.918135260801908e-05,
+      "loss": 0.3697,
+      "step": 2607
+    },
+    {
+      "epoch": 2.7803837953091683,
+      "grad_norm": 0.30334671667064705,
+      "learning_rate": 3.9151587998529754e-05,
+      "loss": 0.3694,
+      "step": 2608
+    },
+    {
+      "epoch": 2.781449893390192,
+      "grad_norm": 0.26731009437526165,
+      "learning_rate": 3.912182385901659e-05,
+      "loss": 0.368,
+      "step": 2609
+    },
+    {
+      "epoch": 2.7825159914712154,
+      "grad_norm": 0.27147115708275976,
+      "learning_rate": 3.909206020596736e-05,
+      "loss": 0.3626,
+      "step": 2610
+    },
+    {
+      "epoch": 2.783582089552239,
+      "grad_norm": 0.32889706955527953,
+      "learning_rate": 3.906229705586959e-05,
+      "loss": 0.3696,
+      "step": 2611
+    },
+    {
+      "epoch": 2.7846481876332625,
+      "grad_norm": 0.3499202779561899,
+      "learning_rate": 3.903253442521051e-05,
+      "loss": 0.368,
+      "step": 2612
+    },
+    {
+      "epoch": 2.7857142857142856,
+      "grad_norm": 0.2870452604692784,
+      "learning_rate": 3.9002772330477096e-05,
+      "loss": 0.3678,
+      "step": 2613
+    },
+    {
+      "epoch": 2.786780383795309,
+      "grad_norm": 0.2959372810749118,
+      "learning_rate": 3.897301078815597e-05,
+      "loss": 0.3666,
+      "step": 2614
+    },
+    {
+      "epoch": 2.7878464818763327,
+      "grad_norm": 0.251783830117275,
+      "learning_rate": 3.894324981473349e-05,
+      "loss": 0.369,
+      "step": 2615
+    },
+    {
+      "epoch": 2.788912579957356,
+      "grad_norm": 0.20368390077877208,
+      "learning_rate": 3.891348942669571e-05,
+      "loss": 0.3659,
+      "step": 2616
+    },
+    {
+      "epoch": 2.7899786780383797,
+      "grad_norm": 0.3014830515949927,
+      "learning_rate": 3.88837296405283e-05,
+      "loss": 0.3645,
+      "step": 2617
+    },
+    {
+      "epoch": 2.791044776119403,
+      "grad_norm": 0.26718545932498405,
+      "learning_rate": 3.8853970472716656e-05,
+      "loss": 0.365,
+      "step": 2618
+    },
+    {
+      "epoch": 2.7921108742004264,
+      "grad_norm": 0.24899810548394205,
+      "learning_rate": 3.882421193974581e-05,
+      "loss": 0.369,
+      "step": 2619
+    },
+    {
+      "epoch": 2.79317697228145,
+      "grad_norm": 0.2285678956121162,
+      "learning_rate": 3.879445405810047e-05,
+      "loss": 0.3654,
+      "step": 2620
+    },
+    {
+      "epoch": 2.7942430703624734,
+      "grad_norm": 0.25841393123219936,
+      "learning_rate": 3.876469684426489e-05,
+      "loss": 0.3705,
+      "step": 2621
+    },
+    {
+      "epoch": 2.795309168443497,
+      "grad_norm": 0.21957675621749626,
+      "learning_rate": 3.873494031472307e-05,
+      "loss": 0.3706,
+      "step": 2622
+    },
+    {
+      "epoch": 2.79637526652452,
+      "grad_norm": 0.20277378161854126,
+      "learning_rate": 3.870518448595858e-05,
+      "loss": 0.3708,
+      "step": 2623
+    },
+    {
+      "epoch": 2.7974413646055436,
+      "grad_norm": 0.19507078530612013,
+      "learning_rate": 3.867542937445458e-05,
+      "loss": 0.3695,
+      "step": 2624
+    },
+    {
+      "epoch": 2.798507462686567,
+      "grad_norm": 0.22778058725192096,
+      "learning_rate": 3.864567499669387e-05,
+      "loss": 0.3704,
+      "step": 2625
+    },
+    {
+      "epoch": 2.7995735607675907,
+      "grad_norm": 0.2139823499648018,
+      "learning_rate": 3.861592136915881e-05,
+      "loss": 0.368,
+      "step": 2626
+    },
+    {
+      "epoch": 2.800639658848614,
+      "grad_norm": 0.2755772492578195,
+      "learning_rate": 3.858616850833141e-05,
+      "loss": 0.3744,
+      "step": 2627
+    },
+    {
+      "epoch": 2.8017057569296373,
+      "grad_norm": 0.22789410651650585,
+      "learning_rate": 3.855641643069316e-05,
+      "loss": 0.3691,
+      "step": 2628
+    },
+    {
+      "epoch": 2.802771855010661,
+      "grad_norm": 0.19420102545723644,
+      "learning_rate": 3.852666515272517e-05,
+      "loss": 0.3701,
+      "step": 2629
+    },
+    {
+      "epoch": 2.8038379530916844,
+      "grad_norm": 0.22416479381704973,
+      "learning_rate": 3.849691469090814e-05,
+      "loss": 0.3657,
+      "step": 2630
+    },
+    {
+      "epoch": 2.804904051172708,
+      "grad_norm": 0.2226881113118721,
+      "learning_rate": 3.846716506172224e-05,
+      "loss": 0.3712,
+      "step": 2631
+    },
+    {
+      "epoch": 2.8059701492537314,
+      "grad_norm": 0.1957187061751904,
+      "learning_rate": 3.8437416281647226e-05,
+      "loss": 0.3686,
+      "step": 2632
+    },
+    {
+      "epoch": 2.8070362473347545,
+      "grad_norm": 0.196407100670983,
+      "learning_rate": 3.8407668367162397e-05,
+      "loss": 0.3638,
+      "step": 2633
+    },
+    {
+      "epoch": 2.8081023454157785,
+      "grad_norm": 0.1860094576496804,
+      "learning_rate": 3.837792133474653e-05,
+      "loss": 0.364,
+      "step": 2634
+    },
+    {
+      "epoch": 2.8091684434968016,
+      "grad_norm": 0.1954370294920811,
+      "learning_rate": 3.8348175200877934e-05,
+      "loss": 0.3699,
+      "step": 2635
+    },
+    {
+      "epoch": 2.810234541577825,
+      "grad_norm": 0.2202007081382702,
+      "learning_rate": 3.831842998203444e-05,
+      "loss": 0.367,
+      "step": 2636
+    },
+    {
+      "epoch": 2.8113006396588487,
+      "grad_norm": 0.18691426984371892,
+      "learning_rate": 3.828868569469333e-05,
+      "loss": 0.3675,
+      "step": 2637
+    },
+    {
+      "epoch": 2.8123667377398722,
+      "grad_norm": 0.16936594094488316,
+      "learning_rate": 3.825894235533143e-05,
+      "loss": 0.3606,
+      "step": 2638
+    },
+    {
+      "epoch": 2.8134328358208958,
+      "grad_norm": 0.2188119066765755,
+      "learning_rate": 3.8229199980424957e-05,
+      "loss": 0.371,
+      "step": 2639
+    },
+    {
+      "epoch": 2.814498933901919,
+      "grad_norm": 0.19720285663307738,
+      "learning_rate": 3.819945858644966e-05,
+      "loss": 0.3647,
+      "step": 2640
+    },
+    {
+      "epoch": 2.8155650319829424,
+      "grad_norm": 0.2081234890857164,
+      "learning_rate": 3.8169718189880735e-05,
+      "loss": 0.3663,
+      "step": 2641
+    },
+    {
+      "epoch": 2.816631130063966,
+      "grad_norm": 0.1684341668760121,
+      "learning_rate": 3.8139978807192824e-05,
+      "loss": 0.367,
+      "step": 2642
+    },
+    {
+      "epoch": 2.8176972281449895,
+      "grad_norm": 0.1883701323467958,
+      "learning_rate": 3.8110240454859975e-05,
+      "loss": 0.3665,
+      "step": 2643
+    },
+    {
+      "epoch": 2.818763326226013,
+      "grad_norm": 0.18509131366924267,
+      "learning_rate": 3.808050314935569e-05,
+      "loss": 0.3658,
+      "step": 2644
+    },
+    {
+      "epoch": 2.819829424307036,
+      "grad_norm": 0.18175683445064622,
+      "learning_rate": 3.8050766907152925e-05,
+      "loss": 0.3685,
+      "step": 2645
+    },
+    {
+      "epoch": 2.8208955223880596,
+      "grad_norm": 0.1749386564896647,
+      "learning_rate": 3.802103174472397e-05,
+      "loss": 0.364,
+      "step": 2646
+    },
+    {
+      "epoch": 2.821961620469083,
+      "grad_norm": 0.1877228031849092,
+      "learning_rate": 3.799129767854058e-05,
+      "loss": 0.3659,
+      "step": 2647
+    },
+    {
+      "epoch": 2.8230277185501067,
+      "grad_norm": 0.23830061584479004,
+      "learning_rate": 3.796156472507388e-05,
+      "loss": 0.3681,
+      "step": 2648
+    },
+    {
+      "epoch": 2.8240938166311302,
+      "grad_norm": 0.20524827433645737,
+      "learning_rate": 3.79318329007944e-05,
+      "loss": 0.3688,
+      "step": 2649
+    },
+    {
+      "epoch": 2.8251599147121533,
+      "grad_norm": 0.20637998521159778,
+      "learning_rate": 3.790210222217199e-05,
+      "loss": 0.3704,
+      "step": 2650
+    },
+    {
+      "epoch": 2.826226012793177,
+      "grad_norm": 0.18452122451122424,
+      "learning_rate": 3.787237270567591e-05,
+      "loss": 0.3687,
+      "step": 2651
+    },
+    {
+      "epoch": 2.8272921108742004,
+      "grad_norm": 0.2102244859465425,
+      "learning_rate": 3.7842644367774804e-05,
+      "loss": 0.3719,
+      "step": 2652
+    },
+    {
+      "epoch": 2.828358208955224,
+      "grad_norm": 0.17836976761573398,
+      "learning_rate": 3.781291722493657e-05,
+      "loss": 0.3701,
+      "step": 2653
+    },
+    {
+      "epoch": 2.8294243070362475,
+      "grad_norm": 0.20742348576542263,
+      "learning_rate": 3.7783191293628535e-05,
+      "loss": 0.3703,
+      "step": 2654
+    },
+    {
+      "epoch": 2.8304904051172706,
+      "grad_norm": 0.19693977146930702,
+      "learning_rate": 3.775346659031731e-05,
+      "loss": 0.3684,
+      "step": 2655
+    },
+    {
+      "epoch": 2.831556503198294,
+      "grad_norm": 0.2143540021763547,
+      "learning_rate": 3.7723743131468855e-05,
+      "loss": 0.3677,
+      "step": 2656
+    },
+    {
+      "epoch": 2.8326226012793176,
+      "grad_norm": 0.2456330609899814,
+      "learning_rate": 3.769402093354838e-05,
+      "loss": 0.3673,
+      "step": 2657
+    },
+    {
+      "epoch": 2.833688699360341,
+      "grad_norm": 0.2504165053266609,
+      "learning_rate": 3.766430001302047e-05,
+      "loss": 0.3684,
+      "step": 2658
+    },
+    {
+      "epoch": 2.8347547974413647,
+      "grad_norm": 0.20402042897097772,
+      "learning_rate": 3.7634580386348965e-05,
+      "loss": 0.3686,
+      "step": 2659
+    },
+    {
+      "epoch": 2.835820895522388,
+      "grad_norm": 0.2261020174040382,
+      "learning_rate": 3.7604862069997006e-05,
+      "loss": 0.3646,
+      "step": 2660
+    },
+    {
+      "epoch": 2.836886993603412,
+      "grad_norm": 0.18453288873034465,
+      "learning_rate": 3.7575145080426966e-05,
+      "loss": 0.3645,
+      "step": 2661
+    },
+    {
+      "epoch": 2.837953091684435,
+      "grad_norm": 0.22623379991968529,
+      "learning_rate": 3.7545429434100524e-05,
+      "loss": 0.3717,
+      "step": 2662
+    },
+    {
+      "epoch": 2.8390191897654584,
+      "grad_norm": 0.2781308273184132,
+      "learning_rate": 3.751571514747863e-05,
+      "loss": 0.3727,
+      "step": 2663
+    },
+    {
+      "epoch": 2.840085287846482,
+      "grad_norm": 0.2095156006824654,
+      "learning_rate": 3.748600223702141e-05,
+      "loss": 0.3709,
+      "step": 2664
+    },
+    {
+      "epoch": 2.8411513859275055,
+      "grad_norm": 0.19821375447623568,
+      "learning_rate": 3.7456290719188294e-05,
+      "loss": 0.3716,
+      "step": 2665
+    },
+    {
+      "epoch": 2.842217484008529,
+      "grad_norm": 0.19589404239503924,
+      "learning_rate": 3.742658061043793e-05,
+      "loss": 0.3686,
+      "step": 2666
+    },
+    {
+      "epoch": 2.843283582089552,
+      "grad_norm": 0.2663508610327882,
+      "learning_rate": 3.7396871927228165e-05,
+      "loss": 0.3699,
+      "step": 2667
+    },
+    {
+      "epoch": 2.8443496801705757,
+      "grad_norm": 0.22076482818199494,
+      "learning_rate": 3.7367164686016055e-05,
+      "loss": 0.3738,
+      "step": 2668
+    },
+    {
+      "epoch": 2.845415778251599,
+      "grad_norm": 0.15927170037082947,
+      "learning_rate": 3.733745890325788e-05,
+      "loss": 0.3658,
+      "step": 2669
+    },
+    {
+      "epoch": 2.8464818763326227,
+      "grad_norm": 0.20652944107978283,
+      "learning_rate": 3.7307754595409094e-05,
+      "loss": 0.3637,
+      "step": 2670
+    },
+    {
+      "epoch": 2.8475479744136463,
+      "grad_norm": 0.24357837304067567,
+      "learning_rate": 3.727805177892435e-05,
+      "loss": 0.3693,
+      "step": 2671
+    },
+    {
+      "epoch": 2.8486140724946694,
+      "grad_norm": 0.245995883074642,
+      "learning_rate": 3.7248350470257456e-05,
+      "loss": 0.3672,
+      "step": 2672
+    },
+    {
+      "epoch": 2.849680170575693,
+      "grad_norm": 0.19478885206841665,
+      "learning_rate": 3.721865068586138e-05,
+      "loss": 0.37,
+      "step": 2673
+    },
+    {
+      "epoch": 2.8507462686567164,
+      "grad_norm": 0.2530400593412467,
+      "learning_rate": 3.71889524421883e-05,
+      "loss": 0.3649,
+      "step": 2674
+    },
+    {
+      "epoch": 2.85181236673774,
+      "grad_norm": 0.2694796185289994,
+      "learning_rate": 3.715925575568945e-05,
+      "loss": 0.3674,
+      "step": 2675
+    },
+    {
+      "epoch": 2.8528784648187635,
+      "grad_norm": 0.2246054293784152,
+      "learning_rate": 3.712956064281527e-05,
+      "loss": 0.3628,
+      "step": 2676
+    },
+    {
+      "epoch": 2.8539445628997866,
+      "grad_norm": 0.204940141551566,
+      "learning_rate": 3.7099867120015316e-05,
+      "loss": 0.3711,
+      "step": 2677
+    },
+    {
+      "epoch": 2.85501066098081,
+      "grad_norm": 0.2634467161792027,
+      "learning_rate": 3.707017520373827e-05,
+      "loss": 0.3711,
+      "step": 2678
+    },
+    {
+      "epoch": 2.8560767590618337,
+      "grad_norm": 0.2790273311361743,
+      "learning_rate": 3.7040484910431874e-05,
+      "loss": 0.3692,
+      "step": 2679
+    },
+    {
+      "epoch": 2.857142857142857,
+      "grad_norm": 0.1836831295255749,
+      "learning_rate": 3.7010796256543034e-05,
+      "loss": 0.3698,
+      "step": 2680
+    },
+    {
+      "epoch": 2.8582089552238807,
+      "grad_norm": 0.2364870469211339,
+      "learning_rate": 3.6981109258517724e-05,
+      "loss": 0.3703,
+      "step": 2681
+    },
+    {
+      "epoch": 2.859275053304904,
+      "grad_norm": 0.27041513703480374,
+      "learning_rate": 3.695142393280098e-05,
+      "loss": 0.3665,
+      "step": 2682
+    },
+    {
+      "epoch": 2.8603411513859274,
+      "grad_norm": 0.20251757763705422,
+      "learning_rate": 3.692174029583693e-05,
+      "loss": 0.3744,
+      "step": 2683
+    },
+    {
+      "epoch": 2.861407249466951,
+      "grad_norm": 0.23004902486067985,
+      "learning_rate": 3.689205836406878e-05,
+      "loss": 0.3666,
+      "step": 2684
+    },
+    {
+      "epoch": 2.8624733475479744,
+      "grad_norm": 0.210050309914908,
+      "learning_rate": 3.686237815393878e-05,
+      "loss": 0.3726,
+      "step": 2685
+    },
+    {
+      "epoch": 2.863539445628998,
+      "grad_norm": 0.2447284603799976,
+      "learning_rate": 3.683269968188821e-05,
+      "loss": 0.3665,
+      "step": 2686
+    },
+    {
+      "epoch": 2.864605543710021,
+      "grad_norm": 0.1878574669176508,
+      "learning_rate": 3.68030229643574e-05,
+      "loss": 0.3739,
+      "step": 2687
+    },
+    {
+      "epoch": 2.8656716417910446,
+      "grad_norm": 0.20854764527126085,
+      "learning_rate": 3.6773348017785714e-05,
+      "loss": 0.3671,
+      "step": 2688
+    },
+    {
+      "epoch": 2.866737739872068,
+      "grad_norm": 0.20789517420801748,
+      "learning_rate": 3.6743674858611556e-05,
+      "loss": 0.3767,
+      "step": 2689
+    },
+    {
+      "epoch": 2.8678038379530917,
+      "grad_norm": 0.17804786727873217,
+      "learning_rate": 3.6714003503272265e-05,
+      "loss": 0.3627,
+      "step": 2690
+    },
+    {
+      "epoch": 2.868869936034115,
+      "grad_norm": 0.1808758602651246,
+      "learning_rate": 3.668433396820426e-05,
+      "loss": 0.3702,
+      "step": 2691
+    },
+    {
+      "epoch": 2.8699360341151388,
+      "grad_norm": 0.23463363469678927,
+      "learning_rate": 3.665466626984294e-05,
+      "loss": 0.3703,
+      "step": 2692
+    },
+    {
+      "epoch": 2.8710021321961623,
+      "grad_norm": 0.246472560729465,
+      "learning_rate": 3.662500042462262e-05,
+      "loss": 0.361,
+      "step": 2693
+    },
+    {
+      "epoch": 2.8720682302771854,
+      "grad_norm": 0.20564287257852504,
+      "learning_rate": 3.6595336448976655e-05,
+      "loss": 0.3701,
+      "step": 2694
+    },
+    {
+      "epoch": 2.873134328358209,
+      "grad_norm": 0.2106812492583941,
+      "learning_rate": 3.656567435933736e-05,
+      "loss": 0.3628,
+      "step": 2695
+    },
+    {
+      "epoch": 2.8742004264392325,
+      "grad_norm": 0.20453409365105865,
+      "learning_rate": 3.6536014172135996e-05,
+      "loss": 0.3649,
+      "step": 2696
+    },
+    {
+      "epoch": 2.875266524520256,
+      "grad_norm": 0.20646291714378256,
+      "learning_rate": 3.650635590380274e-05,
+      "loss": 0.3666,
+      "step": 2697
+    },
+    {
+      "epoch": 2.8763326226012795,
+      "grad_norm": 0.1836576189111925,
+      "learning_rate": 3.647669957076674e-05,
+      "loss": 0.3686,
+      "step": 2698
+    },
+    {
+      "epoch": 2.8773987206823026,
+      "grad_norm": 0.21929175594887504,
+      "learning_rate": 3.644704518945607e-05,
+      "loss": 0.3747,
+      "step": 2699
+    },
+    {
+      "epoch": 2.878464818763326,
+      "grad_norm": 0.18711665960997673,
+      "learning_rate": 3.641739277629772e-05,
+      "loss": 0.3641,
+      "step": 2700
+    },
+    {
+      "epoch": 2.8795309168443497,
+      "grad_norm": 0.17522013442035012,
+      "learning_rate": 3.638774234771757e-05,
+      "loss": 0.3799,
+      "step": 2701
+    },
+    {
+      "epoch": 2.8805970149253732,
+      "grad_norm": 0.17988902594802456,
+      "learning_rate": 3.635809392014042e-05,
+      "loss": 0.3703,
+      "step": 2702
+    },
+    {
+      "epoch": 2.8816631130063968,
+      "grad_norm": 0.17377049259647168,
+      "learning_rate": 3.632844750998998e-05,
+      "loss": 0.3755,
+      "step": 2703
+    },
+    {
+      "epoch": 2.88272921108742,
+      "grad_norm": 0.19092248556619265,
+      "learning_rate": 3.629880313368879e-05,
+      "loss": 0.3632,
+      "step": 2704
+    },
+    {
+      "epoch": 2.8837953091684434,
+      "grad_norm": 0.18451802560202984,
+      "learning_rate": 3.6269160807658315e-05,
+      "loss": 0.3685,
+      "step": 2705
+    },
+    {
+      "epoch": 2.884861407249467,
+      "grad_norm": 0.16498456390879312,
+      "learning_rate": 3.6239520548318836e-05,
+      "loss": 0.3645,
+      "step": 2706
+    },
+    {
+      "epoch": 2.8859275053304905,
+      "grad_norm": 0.19518509163645606,
+      "learning_rate": 3.620988237208956e-05,
+      "loss": 0.3779,
+      "step": 2707
+    },
+    {
+      "epoch": 2.886993603411514,
+      "grad_norm": 0.2093799236223249,
+      "learning_rate": 3.6180246295388465e-05,
+      "loss": 0.3683,
+      "step": 2708
+    },
+    {
+      "epoch": 2.888059701492537,
+      "grad_norm": 0.18513792614540311,
+      "learning_rate": 3.615061233463239e-05,
+      "loss": 0.3673,
+      "step": 2709
+    },
+    {
+      "epoch": 2.8891257995735606,
+      "grad_norm": 0.18716040338636258,
+      "learning_rate": 3.612098050623705e-05,
+      "loss": 0.3648,
+      "step": 2710
+    },
+    {
+      "epoch": 2.890191897654584,
+      "grad_norm": 0.1570679754787949,
+      "learning_rate": 3.6091350826616886e-05,
+      "loss": 0.3602,
+      "step": 2711
+    },
+    {
+      "epoch": 2.8912579957356077,
+      "grad_norm": 0.1646466111643218,
+      "learning_rate": 3.606172331218523e-05,
+      "loss": 0.3715,
+      "step": 2712
+    },
+    {
+      "epoch": 2.8923240938166312,
+      "grad_norm": 0.19255818967368557,
+      "learning_rate": 3.603209797935418e-05,
+      "loss": 0.3643,
+      "step": 2713
+    },
+    {
+      "epoch": 2.8933901918976543,
+      "grad_norm": 0.23514975868653648,
+      "learning_rate": 3.600247484453465e-05,
+      "loss": 0.3767,
+      "step": 2714
+    },
+    {
+      "epoch": 2.894456289978678,
+      "grad_norm": 0.1715556714292468,
+      "learning_rate": 3.597285392413628e-05,
+      "loss": 0.3653,
+      "step": 2715
+    },
+    {
+      "epoch": 2.8955223880597014,
+      "grad_norm": 0.19278886011586072,
+      "learning_rate": 3.5943235234567534e-05,
+      "loss": 0.3688,
+      "step": 2716
+    },
+    {
+      "epoch": 2.896588486140725,
+      "grad_norm": 0.20799089269786109,
+      "learning_rate": 3.591361879223564e-05,
+      "loss": 0.3684,
+      "step": 2717
+    },
+    {
+      "epoch": 2.8976545842217485,
+      "grad_norm": 0.18186928796074078,
+      "learning_rate": 3.588400461354657e-05,
+      "loss": 0.3641,
+      "step": 2718
+    },
+    {
+      "epoch": 2.8987206823027716,
+      "grad_norm": 0.2508899499795832,
+      "learning_rate": 3.5854392714905015e-05,
+      "loss": 0.3701,
+      "step": 2719
+    },
+    {
+      "epoch": 2.8997867803837956,
+      "grad_norm": 0.24536404303538248,
+      "learning_rate": 3.582478311271445e-05,
+      "loss": 0.3684,
+      "step": 2720
+    },
+    {
+      "epoch": 2.9008528784648187,
+      "grad_norm": 0.21642088139256604,
+      "learning_rate": 3.579517582337705e-05,
+      "loss": 0.3717,
+      "step": 2721
+    },
+    {
+      "epoch": 2.901918976545842,
+      "grad_norm": 0.24159904518479744,
+      "learning_rate": 3.57655708632937e-05,
+      "loss": 0.3698,
+      "step": 2722
+    },
+    {
+      "epoch": 2.9029850746268657,
+      "grad_norm": 0.20752353712283125,
+      "learning_rate": 3.573596824886403e-05,
+      "loss": 0.3687,
+      "step": 2723
+    },
+    {
+      "epoch": 2.9040511727078893,
+      "grad_norm": 0.23067889172808448,
+      "learning_rate": 3.570636799648634e-05,
+      "loss": 0.3699,
+      "step": 2724
+    },
+    {
+      "epoch": 2.905117270788913,
+      "grad_norm": 0.20818039787632228,
+      "learning_rate": 3.5676770122557646e-05,
+      "loss": 0.3684,
+      "step": 2725
+    },
+    {
+      "epoch": 2.906183368869936,
+      "grad_norm": 0.21491825417478544,
+      "learning_rate": 3.5647174643473605e-05,
+      "loss": 0.369,
+      "step": 2726
+    },
+    {
+      "epoch": 2.9072494669509594,
+      "grad_norm": 0.19045430642939612,
+      "learning_rate": 3.5617581575628586e-05,
+      "loss": 0.3727,
+      "step": 2727
+    },
+    {
+      "epoch": 2.908315565031983,
+      "grad_norm": 0.2022812002014471,
+      "learning_rate": 3.5587990935415616e-05,
+      "loss": 0.3632,
+      "step": 2728
+    },
+    {
+      "epoch": 2.9093816631130065,
+      "grad_norm": 0.23520827934411384,
+      "learning_rate": 3.555840273922638e-05,
+      "loss": 0.3698,
+      "step": 2729
+    },
+    {
+      "epoch": 2.91044776119403,
+      "grad_norm": 0.22589615939607646,
+      "learning_rate": 3.552881700345118e-05,
+      "loss": 0.3674,
+      "step": 2730
+    },
+    {
+      "epoch": 2.911513859275053,
+      "grad_norm": 0.17879660376241868,
+      "learning_rate": 3.5499233744478986e-05,
+      "loss": 0.3711,
+      "step": 2731
+    },
+    {
+      "epoch": 2.9125799573560767,
+      "grad_norm": 0.1701981830839262,
+      "learning_rate": 3.54696529786974e-05,
+      "loss": 0.373,
+      "step": 2732
+    },
+    {
+      "epoch": 2.9136460554371,
+      "grad_norm": 0.17388147057917147,
+      "learning_rate": 3.54400747224926e-05,
+      "loss": 0.3694,
+      "step": 2733
+    },
+    {
+      "epoch": 2.9147121535181237,
+      "grad_norm": 0.198912284935368,
+      "learning_rate": 3.541049899224941e-05,
+      "loss": 0.3638,
+      "step": 2734
+    },
+    {
+      "epoch": 2.9157782515991473,
+      "grad_norm": 0.1991247673513127,
+      "learning_rate": 3.538092580435127e-05,
+      "loss": 0.3689,
+      "step": 2735
+    },
+    {
+      "epoch": 2.9168443496801704,
+      "grad_norm": 0.21753207227603405,
+      "learning_rate": 3.5351355175180175e-05,
+      "loss": 0.3677,
+      "step": 2736
+    },
+    {
+      "epoch": 2.917910447761194,
+      "grad_norm": 0.19550372191277038,
+      "learning_rate": 3.53217871211167e-05,
+      "loss": 0.3635,
+      "step": 2737
+    },
+    {
+      "epoch": 2.9189765458422174,
+      "grad_norm": 0.18563885663326216,
+      "learning_rate": 3.529222165854005e-05,
+      "loss": 0.3649,
+      "step": 2738
+    },
+    {
+      "epoch": 2.920042643923241,
+      "grad_norm": 0.22917352839550473,
+      "learning_rate": 3.526265880382791e-05,
+      "loss": 0.3781,
+      "step": 2739
+    },
+    {
+      "epoch": 2.9211087420042645,
+      "grad_norm": 0.1757204517843413,
+      "learning_rate": 3.523309857335659e-05,
+      "loss": 0.3714,
+      "step": 2740
+    },
+    {
+      "epoch": 2.9221748400852876,
+      "grad_norm": 0.19335871653526074,
+      "learning_rate": 3.5203540983500925e-05,
+      "loss": 0.364,
+      "step": 2741
+    },
+    {
+      "epoch": 2.923240938166311,
+      "grad_norm": 0.2516292353731809,
+      "learning_rate": 3.517398605063426e-05,
+      "loss": 0.3645,
+      "step": 2742
+    },
+    {
+      "epoch": 2.9243070362473347,
+      "grad_norm": 0.20904716360457082,
+      "learning_rate": 3.514443379112853e-05,
+      "loss": 0.3732,
+      "step": 2743
+    },
+    {
+      "epoch": 2.925373134328358,
+      "grad_norm": 0.22653010522456504,
+      "learning_rate": 3.511488422135412e-05,
+      "loss": 0.3732,
+      "step": 2744
+    },
+    {
+      "epoch": 2.9264392324093818,
+      "grad_norm": 0.21041114421861964,
+      "learning_rate": 3.508533735767995e-05,
+      "loss": 0.3677,
+      "step": 2745
+    },
+    {
+      "epoch": 2.927505330490405,
+      "grad_norm": 0.19966036046378136,
+      "learning_rate": 3.505579321647347e-05,
+      "loss": 0.3655,
+      "step": 2746
+    },
+    {
+      "epoch": 2.928571428571429,
+      "grad_norm": 0.1739262049297657,
+      "learning_rate": 3.5026251814100604e-05,
+      "loss": 0.3688,
+      "step": 2747
+    },
+    {
+      "epoch": 2.929637526652452,
+      "grad_norm": 0.18367778908948013,
+      "learning_rate": 3.4996713166925724e-05,
+      "loss": 0.3673,
+      "step": 2748
+    },
+    {
+      "epoch": 2.9307036247334755,
+      "grad_norm": 0.2188532499554241,
+      "learning_rate": 3.496717729131172e-05,
+      "loss": 0.3669,
+      "step": 2749
+    },
+    {
+      "epoch": 2.931769722814499,
+      "grad_norm": 0.20187002981612032,
+      "learning_rate": 3.493764420361995e-05,
+      "loss": 0.3705,
+      "step": 2750
+    },
+    {
+      "epoch": 2.9328358208955225,
+      "grad_norm": 0.16322026861126518,
+      "learning_rate": 3.490811392021018e-05,
+      "loss": 0.3611,
+      "step": 2751
+    },
+    {
+      "epoch": 2.933901918976546,
+      "grad_norm": 0.19108916099257472,
+      "learning_rate": 3.4878586457440655e-05,
+      "loss": 0.3699,
+      "step": 2752
+    },
+    {
+      "epoch": 2.934968017057569,
+      "grad_norm": 0.21458163159618987,
+      "learning_rate": 3.484906183166807e-05,
+      "loss": 0.3687,
+      "step": 2753
+    },
+    {
+      "epoch": 2.9360341151385927,
+      "grad_norm": 0.1904908632261195,
+      "learning_rate": 3.481954005924755e-05,
+      "loss": 0.3657,
+      "step": 2754
+    },
+    {
+      "epoch": 2.9371002132196162,
+      "grad_norm": 0.18685652839997016,
+      "learning_rate": 3.4790021156532585e-05,
+      "loss": 0.3688,
+      "step": 2755
+    },
+    {
+      "epoch": 2.9381663113006398,
+      "grad_norm": 0.19812866208833532,
+      "learning_rate": 3.476050513987512e-05,
+      "loss": 0.3665,
+      "step": 2756
+    },
+    {
+      "epoch": 2.9392324093816633,
+      "grad_norm": 0.19248170405628037,
+      "learning_rate": 3.4730992025625506e-05,
+      "loss": 0.3705,
+      "step": 2757
+    },
+    {
+      "epoch": 2.9402985074626864,
+      "grad_norm": 0.17211944249351815,
+      "learning_rate": 3.4701481830132486e-05,
+      "loss": 0.3644,
+      "step": 2758
+    },
+    {
+      "epoch": 2.94136460554371,
+      "grad_norm": 0.1964548683909922,
+      "learning_rate": 3.467197456974315e-05,
+      "loss": 0.3652,
+      "step": 2759
+    },
+    {
+      "epoch": 2.9424307036247335,
+      "grad_norm": 0.20371355156065427,
+      "learning_rate": 3.4642470260802986e-05,
+      "loss": 0.3749,
+      "step": 2760
+    },
+    {
+      "epoch": 2.943496801705757,
+      "grad_norm": 0.1469689938993747,
+      "learning_rate": 3.4612968919655886e-05,
+      "loss": 0.3682,
+      "step": 2761
+    },
+    {
+      "epoch": 2.9445628997867805,
+      "grad_norm": 0.1964620626188304,
+      "learning_rate": 3.458347056264401e-05,
+      "loss": 0.3726,
+      "step": 2762
+    },
+    {
+      "epoch": 2.9456289978678036,
+      "grad_norm": 0.18358666019195752,
+      "learning_rate": 3.4553975206107944e-05,
+      "loss": 0.3668,
+      "step": 2763
+    },
+    {
+      "epoch": 2.946695095948827,
+      "grad_norm": 0.16170424306429584,
+      "learning_rate": 3.452448286638657e-05,
+      "loss": 0.3735,
+      "step": 2764
+    },
+    {
+      "epoch": 2.9477611940298507,
+      "grad_norm": 0.18591540468001805,
+      "learning_rate": 3.4494993559817134e-05,
+      "loss": 0.37,
+      "step": 2765
+    },
+    {
+      "epoch": 2.9488272921108742,
+      "grad_norm": 0.22260097019813457,
+      "learning_rate": 3.446550730273514e-05,
+      "loss": 0.3656,
+      "step": 2766
+    },
+    {
+      "epoch": 2.949893390191898,
+      "grad_norm": 0.2482136891258535,
+      "learning_rate": 3.4436024111474475e-05,
+      "loss": 0.3683,
+      "step": 2767
+    },
+    {
+      "epoch": 2.950959488272921,
+      "grad_norm": 0.19730939109015577,
+      "learning_rate": 3.440654400236729e-05,
+      "loss": 0.378,
+      "step": 2768
+    },
+    {
+      "epoch": 2.9520255863539444,
+      "grad_norm": 0.20731684739497086,
+      "learning_rate": 3.437706699174402e-05,
+      "loss": 0.364,
+      "step": 2769
+    },
+    {
+      "epoch": 2.953091684434968,
+      "grad_norm": 0.22896154197636323,
+      "learning_rate": 3.43475930959334e-05,
+      "loss": 0.3685,
+      "step": 2770
+    },
+    {
+      "epoch": 2.9541577825159915,
+      "grad_norm": 0.20500126289635404,
+      "learning_rate": 3.431812233126245e-05,
+      "loss": 0.3748,
+      "step": 2771
+    },
+    {
+      "epoch": 2.955223880597015,
+      "grad_norm": 0.1633714645059073,
+      "learning_rate": 3.428865471405643e-05,
+      "loss": 0.3591,
+      "step": 2772
+    },
+    {
+      "epoch": 2.956289978678038,
+      "grad_norm": 0.21578628128041608,
+      "learning_rate": 3.425919026063886e-05,
+      "loss": 0.3745,
+      "step": 2773
+    },
+    {
+      "epoch": 2.957356076759062,
+      "grad_norm": 0.16947850465245867,
+      "learning_rate": 3.422972898733154e-05,
+      "loss": 0.3695,
+      "step": 2774
+    },
+    {
+      "epoch": 2.958422174840085,
+      "grad_norm": 0.15700808274896483,
+      "learning_rate": 3.420027091045446e-05,
+      "loss": 0.3627,
+      "step": 2775
+    },
+    {
+      "epoch": 2.9594882729211087,
+      "grad_norm": 0.1897865970091229,
+      "learning_rate": 3.417081604632589e-05,
+      "loss": 0.3624,
+      "step": 2776
+    },
+    {
+      "epoch": 2.9605543710021323,
+      "grad_norm": 0.19673063754579187,
+      "learning_rate": 3.414136441126227e-05,
+      "loss": 0.3692,
+      "step": 2777
+    },
+    {
+      "epoch": 2.961620469083156,
+      "grad_norm": 0.17553015896004215,
+      "learning_rate": 3.4111916021578285e-05,
+      "loss": 0.3678,
+      "step": 2778
+    },
+    {
+      "epoch": 2.9626865671641793,
+      "grad_norm": 0.2183249284173145,
+      "learning_rate": 3.408247089358681e-05,
+      "loss": 0.3747,
+      "step": 2779
+    },
+    {
+      "epoch": 2.9637526652452024,
+      "grad_norm": 0.21927885211369533,
+      "learning_rate": 3.405302904359893e-05,
+      "loss": 0.3673,
+      "step": 2780
+    },
+    {
+      "epoch": 2.964818763326226,
+      "grad_norm": 0.18731814539514857,
+      "learning_rate": 3.402359048792386e-05,
+      "loss": 0.3716,
+      "step": 2781
+    },
+    {
+      "epoch": 2.9658848614072495,
+      "grad_norm": 0.2338787632638593,
+      "learning_rate": 3.399415524286907e-05,
+      "loss": 0.3732,
+      "step": 2782
+    },
+    {
+      "epoch": 2.966950959488273,
+      "grad_norm": 0.21152997804401275,
+      "learning_rate": 3.396472332474015e-05,
+      "loss": 0.3669,
+      "step": 2783
+    },
+    {
+      "epoch": 2.9680170575692966,
+      "grad_norm": 0.17079876375720443,
+      "learning_rate": 3.393529474984083e-05,
+      "loss": 0.3714,
+      "step": 2784
+    },
+    {
+      "epoch": 2.9690831556503197,
+      "grad_norm": 0.20326343887457965,
+      "learning_rate": 3.390586953447304e-05,
+      "loss": 0.3685,
+      "step": 2785
+    },
+    {
+      "epoch": 2.970149253731343,
+      "grad_norm": 0.265596968296806,
+      "learning_rate": 3.38764476949368e-05,
+      "loss": 0.3701,
+      "step": 2786
+    },
+    {
+      "epoch": 2.9712153518123667,
+      "grad_norm": 0.20949193099124558,
+      "learning_rate": 3.38470292475303e-05,
+      "loss": 0.3634,
+      "step": 2787
+    },
+    {
+      "epoch": 2.9722814498933903,
+      "grad_norm": 0.19275153065422299,
+      "learning_rate": 3.3817614208549796e-05,
+      "loss": 0.3652,
+      "step": 2788
+    },
+    {
+      "epoch": 2.973347547974414,
+      "grad_norm": 0.25513016177139153,
+      "learning_rate": 3.378820259428972e-05,
+      "loss": 0.3678,
+      "step": 2789
+    },
+    {
+      "epoch": 2.974413646055437,
+      "grad_norm": 0.20641698835987704,
+      "learning_rate": 3.3758794421042585e-05,
+      "loss": 0.3686,
+      "step": 2790
+    },
+    {
+      "epoch": 2.9754797441364604,
+      "grad_norm": 0.21686027546060258,
+      "learning_rate": 3.3729389705098956e-05,
+      "loss": 0.3721,
+      "step": 2791
+    },
+    {
+      "epoch": 2.976545842217484,
+      "grad_norm": 0.21228392649559347,
+      "learning_rate": 3.3699988462747536e-05,
+      "loss": 0.3619,
+      "step": 2792
+    },
+    {
+      "epoch": 2.9776119402985075,
+      "grad_norm": 0.15906755651360552,
+      "learning_rate": 3.3670590710275095e-05,
+      "loss": 0.3634,
+      "step": 2793
+    },
+    {
+      "epoch": 2.978678038379531,
+      "grad_norm": 0.23673718965759863,
+      "learning_rate": 3.3641196463966466e-05,
+      "loss": 0.3682,
+      "step": 2794
+    },
+    {
+      "epoch": 2.979744136460554,
+      "grad_norm": 0.2179359974386122,
+      "learning_rate": 3.36118057401045e-05,
+      "loss": 0.3661,
+      "step": 2795
+    },
+    {
+      "epoch": 2.9808102345415777,
+      "grad_norm": 0.18870552087602346,
+      "learning_rate": 3.358241855497015e-05,
+      "loss": 0.3692,
+      "step": 2796
+    },
+    {
+      "epoch": 2.981876332622601,
+      "grad_norm": 0.21000830418856412,
+      "learning_rate": 3.35530349248424e-05,
+      "loss": 0.3723,
+      "step": 2797
+    },
+    {
+      "epoch": 2.9829424307036247,
+      "grad_norm": 0.1834568664565684,
+      "learning_rate": 3.352365486599823e-05,
+      "loss": 0.3699,
+      "step": 2798
+    },
+    {
+      "epoch": 2.9840085287846483,
+      "grad_norm": 0.21955054980022862,
+      "learning_rate": 3.349427839471268e-05,
+      "loss": 0.3693,
+      "step": 2799
+    },
+    {
+      "epoch": 2.9850746268656714,
+      "grad_norm": 0.19131385619838284,
+      "learning_rate": 3.346490552725879e-05,
+      "loss": 0.3633,
+      "step": 2800
+    },
+    {
+      "epoch": 2.9861407249466954,
+      "grad_norm": 0.22039397671168487,
+      "learning_rate": 3.3435536279907615e-05,
+      "loss": 0.3668,
+      "step": 2801
+    },
+    {
+      "epoch": 2.9872068230277184,
+      "grad_norm": 0.25817755908994,
+      "learning_rate": 3.340617066892815e-05,
+      "loss": 0.3694,
+      "step": 2802
+    },
+    {
+      "epoch": 2.988272921108742,
+      "grad_norm": 0.17992068437591005,
+      "learning_rate": 3.3376808710587456e-05,
+      "loss": 0.3608,
+      "step": 2803
+    },
+    {
+      "epoch": 2.9893390191897655,
+      "grad_norm": 0.19725409347741943,
+      "learning_rate": 3.334745042115052e-05,
+      "loss": 0.3674,
+      "step": 2804
+    },
+    {
+      "epoch": 2.990405117270789,
+      "grad_norm": 0.2199411430669973,
+      "learning_rate": 3.331809581688031e-05,
+      "loss": 0.3685,
+      "step": 2805
+    },
+    {
+      "epoch": 2.9914712153518126,
+      "grad_norm": 0.186450392495569,
+      "learning_rate": 3.328874491403774e-05,
+      "loss": 0.3652,
+      "step": 2806
+    },
+    {
+      "epoch": 2.9925373134328357,
+      "grad_norm": 0.2010767437452293,
+      "learning_rate": 3.32593977288817e-05,
+      "loss": 0.3795,
+      "step": 2807
+    },
+    {
+      "epoch": 2.9936034115138592,
+      "grad_norm": 0.19580642217271252,
+      "learning_rate": 3.3230054277668994e-05,
+      "loss": 0.3688,
+      "step": 2808
+    },
+    {
+      "epoch": 2.9946695095948828,
+      "grad_norm": 0.17801010293169015,
+      "learning_rate": 3.320071457665437e-05,
+      "loss": 0.3694,
+      "step": 2809
+    },
+    {
+      "epoch": 2.9957356076759063,
+      "grad_norm": 0.183255241372143,
+      "learning_rate": 3.317137864209051e-05,
+      "loss": 0.3711,
+      "step": 2810
+    },
+    {
+      "epoch": 2.99680170575693,
+      "grad_norm": 0.1656711458529668,
+      "learning_rate": 3.3142046490227984e-05,
+      "loss": 0.3668,
+      "step": 2811
+    },
+    {
+      "epoch": 2.997867803837953,
+      "grad_norm": 0.15999468329558042,
+      "learning_rate": 3.311271813731529e-05,
+      "loss": 0.3625,
+      "step": 2812
+    },
+    {
+      "epoch": 2.9989339019189765,
+      "grad_norm": 0.2031732214049462,
+      "learning_rate": 3.3083393599598804e-05,
+      "loss": 0.3667,
+      "step": 2813
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.1833937765340318,
+      "learning_rate": 3.305407289332279e-05,
+      "loss": 0.356,
+      "step": 2814
+    },
+    {
+      "epoch": 3.0010660980810235,
+      "grad_norm": 0.16676086865014514,
+      "learning_rate": 3.3024756034729403e-05,
+      "loss": 0.3446,
+      "step": 2815
+    },
+    {
+      "epoch": 3.002132196162047,
+      "grad_norm": 0.1944616035010321,
+      "learning_rate": 3.299544304005867e-05,
+      "loss": 0.3499,
+      "step": 2816
+    },
+    {
+      "epoch": 3.00319829424307,
+      "grad_norm": 0.21726317154600533,
+      "learning_rate": 3.296613392554845e-05,
+      "loss": 0.3447,
+      "step": 2817
+    },
+    {
+      "epoch": 3.0042643923240937,
+      "grad_norm": 0.18162073687276598,
+      "learning_rate": 3.293682870743446e-05,
+      "loss": 0.3449,
+      "step": 2818
+    },
+    {
+      "epoch": 3.0053304904051172,
+      "grad_norm": 0.19337743208072578,
+      "learning_rate": 3.2907527401950314e-05,
+      "loss": 0.3438,
+      "step": 2819
+    },
+    {
+      "epoch": 3.0063965884861408,
+      "grad_norm": 0.20006928027604456,
+      "learning_rate": 3.287823002532735e-05,
+      "loss": 0.348,
+      "step": 2820
+    },
+    {
+      "epoch": 3.0074626865671643,
+      "grad_norm": 0.21921022479851462,
+      "learning_rate": 3.284893659379483e-05,
+      "loss": 0.355,
+      "step": 2821
+    },
+    {
+      "epoch": 3.008528784648188,
+      "grad_norm": 0.19159498596282584,
+      "learning_rate": 3.2819647123579785e-05,
+      "loss": 0.3471,
+      "step": 2822
+    },
+    {
+      "epoch": 3.009594882729211,
+      "grad_norm": 0.1975964755116819,
+      "learning_rate": 3.2790361630907073e-05,
+      "loss": 0.351,
+      "step": 2823
+    },
+    {
+      "epoch": 3.0106609808102345,
+      "grad_norm": 0.19162779591776102,
+      "learning_rate": 3.276108013199931e-05,
+      "loss": 0.3471,
+      "step": 2824
+    },
+    {
+      "epoch": 3.011727078891258,
+      "grad_norm": 0.20069049454309051,
+      "learning_rate": 3.273180264307693e-05,
+      "loss": 0.3473,
+      "step": 2825
+    },
+    {
+      "epoch": 3.0127931769722816,
+      "grad_norm": 0.18162811798879747,
+      "learning_rate": 3.270252918035817e-05,
+      "loss": 0.3497,
+      "step": 2826
+    },
+    {
+      "epoch": 3.013859275053305,
+      "grad_norm": 0.2190938039478538,
+      "learning_rate": 3.2673259760058966e-05,
+      "loss": 0.342,
+      "step": 2827
+    },
+    {
+      "epoch": 3.014925373134328,
+      "grad_norm": 0.20368209134299048,
+      "learning_rate": 3.264399439839307e-05,
+      "loss": 0.3565,
+      "step": 2828
+    },
+    {
+      "epoch": 3.0159914712153517,
+      "grad_norm": 0.21564518091097337,
+      "learning_rate": 3.261473311157197e-05,
+      "loss": 0.3477,
+      "step": 2829
+    },
+    {
+      "epoch": 3.0170575692963753,
+      "grad_norm": 0.22960493863723194,
+      "learning_rate": 3.258547591580493e-05,
+      "loss": 0.355,
+      "step": 2830
+    },
+    {
+      "epoch": 3.018123667377399,
+      "grad_norm": 0.20630353684777758,
+      "learning_rate": 3.255622282729886e-05,
+      "loss": 0.3491,
+      "step": 2831
+    },
+    {
+      "epoch": 3.0191897654584223,
+      "grad_norm": 0.19432289775660633,
+      "learning_rate": 3.252697386225848e-05,
+      "loss": 0.3463,
+      "step": 2832
+    },
+    {
+      "epoch": 3.0202558635394454,
+      "grad_norm": 0.22226714326699387,
+      "learning_rate": 3.24977290368862e-05,
+      "loss": 0.3474,
+      "step": 2833
+    },
+    {
+      "epoch": 3.021321961620469,
+      "grad_norm": 0.20919728574830324,
+      "learning_rate": 3.2468488367382146e-05,
+      "loss": 0.3543,
+      "step": 2834
+    },
+    {
+      "epoch": 3.0223880597014925,
+      "grad_norm": 0.2097539980157466,
+      "learning_rate": 3.2439251869944096e-05,
+      "loss": 0.3546,
+      "step": 2835
+    },
+    {
+      "epoch": 3.023454157782516,
+      "grad_norm": 0.2378340506907076,
+      "learning_rate": 3.2410019560767566e-05,
+      "loss": 0.3507,
+      "step": 2836
+    },
+    {
+      "epoch": 3.0245202558635396,
+      "grad_norm": 0.19806299173295153,
+      "learning_rate": 3.238079145604576e-05,
+      "loss": 0.3497,
+      "step": 2837
+    },
+    {
+      "epoch": 3.025586353944563,
+      "grad_norm": 0.21295221351192548,
+      "learning_rate": 3.235156757196948e-05,
+      "loss": 0.3427,
+      "step": 2838
+    },
+    {
+      "epoch": 3.026652452025586,
+      "grad_norm": 0.22762110787574957,
+      "learning_rate": 3.2322347924727264e-05,
+      "loss": 0.352,
+      "step": 2839
+    },
+    {
+      "epoch": 3.0277185501066097,
+      "grad_norm": 0.20297127463286518,
+      "learning_rate": 3.229313253050529e-05,
+      "loss": 0.3481,
+      "step": 2840
+    },
+    {
+      "epoch": 3.0287846481876333,
+      "grad_norm": 0.1801189363866266,
+      "learning_rate": 3.2263921405487356e-05,
+      "loss": 0.3491,
+      "step": 2841
+    },
+    {
+      "epoch": 3.029850746268657,
+      "grad_norm": 0.2148709514009852,
+      "learning_rate": 3.2234714565854895e-05,
+      "loss": 0.3562,
+      "step": 2842
+    },
+    {
+      "epoch": 3.0309168443496803,
+      "grad_norm": 0.1905295271571919,
+      "learning_rate": 3.2205512027787005e-05,
+      "loss": 0.3468,
+      "step": 2843
+    },
+    {
+      "epoch": 3.0319829424307034,
+      "grad_norm": 0.2416529548256475,
+      "learning_rate": 3.2176313807460355e-05,
+      "loss": 0.3521,
+      "step": 2844
+    },
+    {
+      "epoch": 3.033049040511727,
+      "grad_norm": 0.19945666420393923,
+      "learning_rate": 3.214711992104925e-05,
+      "loss": 0.3519,
+      "step": 2845
+    },
+    {
+      "epoch": 3.0341151385927505,
+      "grad_norm": 0.1969420298529883,
+      "learning_rate": 3.211793038472558e-05,
+      "loss": 0.3489,
+      "step": 2846
+    },
+    {
+      "epoch": 3.035181236673774,
+      "grad_norm": 0.24393952090774143,
+      "learning_rate": 3.208874521465882e-05,
+      "loss": 0.3555,
+      "step": 2847
+    },
+    {
+      "epoch": 3.0362473347547976,
+      "grad_norm": 0.17068118729565857,
+      "learning_rate": 3.205956442701607e-05,
+      "loss": 0.3492,
+      "step": 2848
+    },
+    {
+      "epoch": 3.0373134328358207,
+      "grad_norm": 0.18265684276116836,
+      "learning_rate": 3.203038803796195e-05,
+      "loss": 0.3536,
+      "step": 2849
+    },
+    {
+      "epoch": 3.038379530916844,
+      "grad_norm": 0.1750878986387194,
+      "learning_rate": 3.200121606365865e-05,
+      "loss": 0.3467,
+      "step": 2850
+    },
+    {
+      "epoch": 3.0394456289978677,
+      "grad_norm": 0.1501654395365917,
+      "learning_rate": 3.197204852026595e-05,
+      "loss": 0.3463,
+      "step": 2851
+    },
+    {
+      "epoch": 3.0405117270788913,
+      "grad_norm": 0.1881586833141222,
+      "learning_rate": 3.194288542394116e-05,
+      "loss": 0.3485,
+      "step": 2852
+    },
+    {
+      "epoch": 3.041577825159915,
+      "grad_norm": 0.17391665540804127,
+      "learning_rate": 3.191372679083908e-05,
+      "loss": 0.349,
+      "step": 2853
+    },
+    {
+      "epoch": 3.0426439232409384,
+      "grad_norm": 0.1855229052361205,
+      "learning_rate": 3.18845726371121e-05,
+      "loss": 0.3522,
+      "step": 2854
+    },
+    {
+      "epoch": 3.0437100213219614,
+      "grad_norm": 0.19940822544482417,
+      "learning_rate": 3.1855422978910136e-05,
+      "loss": 0.3482,
+      "step": 2855
+    },
+    {
+      "epoch": 3.044776119402985,
+      "grad_norm": 0.16112705274220399,
+      "learning_rate": 3.182627783238053e-05,
+      "loss": 0.347,
+      "step": 2856
+    },
+    {
+      "epoch": 3.0458422174840085,
+      "grad_norm": 0.20167791211166475,
+      "learning_rate": 3.179713721366821e-05,
+      "loss": 0.3534,
+      "step": 2857
+    },
+    {
+      "epoch": 3.046908315565032,
+      "grad_norm": 0.18479202502339434,
+      "learning_rate": 3.176800113891556e-05,
+      "loss": 0.351,
+      "step": 2858
+    },
+    {
+      "epoch": 3.0479744136460556,
+      "grad_norm": 0.16272058166990602,
+      "learning_rate": 3.173886962426246e-05,
+      "loss": 0.3494,
+      "step": 2859
+    },
+    {
+      "epoch": 3.0490405117270787,
+      "grad_norm": 0.16165195964515638,
+      "learning_rate": 3.1709742685846224e-05,
+      "loss": 0.3473,
+      "step": 2860
+    },
+    {
+      "epoch": 3.050106609808102,
+      "grad_norm": 0.17386330305907433,
+      "learning_rate": 3.168062033980169e-05,
+      "loss": 0.3497,
+      "step": 2861
+    },
+    {
+      "epoch": 3.0511727078891258,
+      "grad_norm": 0.1458661691624279,
+      "learning_rate": 3.165150260226112e-05,
+      "loss": 0.35,
+      "step": 2862
+    },
+    {
+      "epoch": 3.0522388059701493,
+      "grad_norm": 0.20435764871354212,
+      "learning_rate": 3.162238948935423e-05,
+      "loss": 0.3486,
+      "step": 2863
+    },
+    {
+      "epoch": 3.053304904051173,
+      "grad_norm": 0.19071168515585496,
+      "learning_rate": 3.159328101720816e-05,
+      "loss": 0.3471,
+      "step": 2864
+    },
+    {
+      "epoch": 3.0543710021321964,
+      "grad_norm": 0.18693245756126411,
+      "learning_rate": 3.156417720194749e-05,
+      "loss": 0.3408,
+      "step": 2865
+    },
+    {
+      "epoch": 3.0554371002132195,
+      "grad_norm": 0.19376415825853904,
+      "learning_rate": 3.153507805969425e-05,
+      "loss": 0.3462,
+      "step": 2866
+    },
+    {
+      "epoch": 3.056503198294243,
+      "grad_norm": 0.23970364168448857,
+      "learning_rate": 3.150598360656781e-05,
+      "loss": 0.3585,
+      "step": 2867
+    },
+    {
+      "epoch": 3.0575692963752665,
+      "grad_norm": 0.24686273765529104,
+      "learning_rate": 3.1476893858685e-05,
+      "loss": 0.3522,
+      "step": 2868
+    },
+    {
+      "epoch": 3.05863539445629,
+      "grad_norm": 0.20134496461618384,
+      "learning_rate": 3.1447808832160034e-05,
+      "loss": 0.3534,
+      "step": 2869
+    },
+    {
+      "epoch": 3.0597014925373136,
+      "grad_norm": 0.20811412001280552,
+      "learning_rate": 3.141872854310452e-05,
+      "loss": 0.3529,
+      "step": 2870
+    },
+    {
+      "epoch": 3.0607675906183367,
+      "grad_norm": 0.2252811945149839,
+      "learning_rate": 3.13896530076274e-05,
+      "loss": 0.3507,
+      "step": 2871
+    },
+    {
+      "epoch": 3.0618336886993602,
+      "grad_norm": 0.27899336334092645,
+      "learning_rate": 3.1360582241835025e-05,
+      "loss": 0.3445,
+      "step": 2872
+    },
+    {
+      "epoch": 3.0628997867803838,
+      "grad_norm": 0.18208464259977086,
+      "learning_rate": 3.13315162618311e-05,
+      "loss": 0.3512,
+      "step": 2873
+    },
+    {
+      "epoch": 3.0639658848614073,
+      "grad_norm": 0.19082089290473603,
+      "learning_rate": 3.130245508371663e-05,
+      "loss": 0.3528,
+      "step": 2874
+    },
+    {
+      "epoch": 3.065031982942431,
+      "grad_norm": 0.2131583906254748,
+      "learning_rate": 3.127339872359002e-05,
+      "loss": 0.3521,
+      "step": 2875
+    },
+    {
+      "epoch": 3.066098081023454,
+      "grad_norm": 0.17934067948301102,
+      "learning_rate": 3.1244347197546986e-05,
+      "loss": 0.3484,
+      "step": 2876
+    },
+    {
+      "epoch": 3.0671641791044775,
+      "grad_norm": 0.21329220073762686,
+      "learning_rate": 3.1215300521680564e-05,
+      "loss": 0.3564,
+      "step": 2877
+    },
+    {
+      "epoch": 3.068230277185501,
+      "grad_norm": 0.21979298064989125,
+      "learning_rate": 3.118625871208109e-05,
+      "loss": 0.3441,
+      "step": 2878
+    },
+    {
+      "epoch": 3.0692963752665245,
+      "grad_norm": 0.1899827348463154,
+      "learning_rate": 3.115722178483624e-05,
+      "loss": 0.3548,
+      "step": 2879
+    },
+    {
+      "epoch": 3.070362473347548,
+      "grad_norm": 0.19208909798825324,
+      "learning_rate": 3.1128189756030934e-05,
+      "loss": 0.3472,
+      "step": 2880
+    },
+    {
+      "epoch": 3.0714285714285716,
+      "grad_norm": 0.15715982773841544,
+      "learning_rate": 3.109916264174743e-05,
+      "loss": 0.3447,
+      "step": 2881
+    },
+    {
+      "epoch": 3.0724946695095947,
+      "grad_norm": 0.19051789498649285,
+      "learning_rate": 3.1070140458065235e-05,
+      "loss": 0.3478,
+      "step": 2882
+    },
+    {
+      "epoch": 3.0735607675906182,
+      "grad_norm": 0.17970300278684168,
+      "learning_rate": 3.104112322106112e-05,
+      "loss": 0.3468,
+      "step": 2883
+    },
+    {
+      "epoch": 3.074626865671642,
+      "grad_norm": 0.1733355341791082,
+      "learning_rate": 3.1012110946809134e-05,
+      "loss": 0.355,
+      "step": 2884
+    },
+    {
+      "epoch": 3.0756929637526653,
+      "grad_norm": 0.1997157707879784,
+      "learning_rate": 3.0983103651380574e-05,
+      "loss": 0.3472,
+      "step": 2885
+    },
+    {
+      "epoch": 3.076759061833689,
+      "grad_norm": 0.1683478624827942,
+      "learning_rate": 3.0954101350843966e-05,
+      "loss": 0.3472,
+      "step": 2886
+    },
+    {
+      "epoch": 3.077825159914712,
+      "grad_norm": 0.17517827837363226,
+      "learning_rate": 3.092510406126508e-05,
+      "loss": 0.3466,
+      "step": 2887
+    },
+    {
+      "epoch": 3.0788912579957355,
+      "grad_norm": 0.18556306694129251,
+      "learning_rate": 3.089611179870691e-05,
+      "loss": 0.3498,
+      "step": 2888
+    },
+    {
+      "epoch": 3.079957356076759,
+      "grad_norm": 0.17804006539183986,
+      "learning_rate": 3.086712457922966e-05,
+      "loss": 0.3459,
+      "step": 2889
+    },
+    {
+      "epoch": 3.0810234541577826,
+      "grad_norm": 0.17926971743225267,
+      "learning_rate": 3.083814241889074e-05,
+      "loss": 0.3485,
+      "step": 2890
+    },
+    {
+      "epoch": 3.082089552238806,
+      "grad_norm": 0.18765772737419348,
+      "learning_rate": 3.0809165333744765e-05,
+      "loss": 0.3555,
+      "step": 2891
+    },
+    {
+      "epoch": 3.0831556503198296,
+      "grad_norm": 0.1665179916020946,
+      "learning_rate": 3.0780193339843545e-05,
+      "loss": 0.346,
+      "step": 2892
+    },
+    {
+      "epoch": 3.0842217484008527,
+      "grad_norm": 0.16253422719498084,
+      "learning_rate": 3.075122645323603e-05,
+      "loss": 0.347,
+      "step": 2893
+    },
+    {
+      "epoch": 3.0852878464818763,
+      "grad_norm": 0.1705949117676651,
+      "learning_rate": 3.072226468996839e-05,
+      "loss": 0.3531,
+      "step": 2894
+    },
+    {
+      "epoch": 3.0863539445629,
+      "grad_norm": 0.16669859994578132,
+      "learning_rate": 3.0693308066083954e-05,
+      "loss": 0.3489,
+      "step": 2895
+    },
+    {
+      "epoch": 3.0874200426439233,
+      "grad_norm": 0.15555031037234748,
+      "learning_rate": 3.0664356597623144e-05,
+      "loss": 0.3505,
+      "step": 2896
+    },
+    {
+      "epoch": 3.088486140724947,
+      "grad_norm": 0.18596041745219372,
+      "learning_rate": 3.0635410300623596e-05,
+      "loss": 0.354,
+      "step": 2897
+    },
+    {
+      "epoch": 3.08955223880597,
+      "grad_norm": 0.1905420596562174,
+      "learning_rate": 3.060646919112004e-05,
+      "loss": 0.3481,
+      "step": 2898
+    },
+    {
+      "epoch": 3.0906183368869935,
+      "grad_norm": 0.18015092645672678,
+      "learning_rate": 3.057753328514438e-05,
+      "loss": 0.3492,
+      "step": 2899
+    },
+    {
+      "epoch": 3.091684434968017,
+      "grad_norm": 0.1773289970330081,
+      "learning_rate": 3.0548602598725564e-05,
+      "loss": 0.3523,
+      "step": 2900
+    },
+    {
+      "epoch": 3.0927505330490406,
+      "grad_norm": 0.19120781604022777,
+      "learning_rate": 3.0519677147889705e-05,
+      "loss": 0.3533,
+      "step": 2901
+    },
+    {
+      "epoch": 3.093816631130064,
+      "grad_norm": 0.18705169216756135,
+      "learning_rate": 3.0490756948660017e-05,
+      "loss": 0.3486,
+      "step": 2902
+    },
+    {
+      "epoch": 3.094882729211087,
+      "grad_norm": 0.21344733721957102,
+      "learning_rate": 3.046184201705675e-05,
+      "loss": 0.3546,
+      "step": 2903
+    },
+    {
+      "epoch": 3.0959488272921107,
+      "grad_norm": 0.16522989881515734,
+      "learning_rate": 3.043293236909729e-05,
+      "loss": 0.3486,
+      "step": 2904
+    },
+    {
+      "epoch": 3.0970149253731343,
+      "grad_norm": 0.1819572191490791,
+      "learning_rate": 3.0404028020796087e-05,
+      "loss": 0.3496,
+      "step": 2905
+    },
+    {
+      "epoch": 3.098081023454158,
+      "grad_norm": 0.15871225159309935,
+      "learning_rate": 3.0375128988164655e-05,
+      "loss": 0.3476,
+      "step": 2906
+    },
+    {
+      "epoch": 3.0991471215351813,
+      "grad_norm": 0.17994107490800862,
+      "learning_rate": 3.0346235287211532e-05,
+      "loss": 0.3509,
+      "step": 2907
+    },
+    {
+      "epoch": 3.100213219616205,
+      "grad_norm": 0.16605077988407455,
+      "learning_rate": 3.0317346933942337e-05,
+      "loss": 0.3563,
+      "step": 2908
+    },
+    {
+      "epoch": 3.101279317697228,
+      "grad_norm": 0.16665287698505823,
+      "learning_rate": 3.028846394435973e-05,
+      "loss": 0.3468,
+      "step": 2909
+    },
+    {
+      "epoch": 3.1023454157782515,
+      "grad_norm": 0.19464014834448382,
+      "learning_rate": 3.0259586334463366e-05,
+      "loss": 0.3461,
+      "step": 2910
+    },
+    {
+      "epoch": 3.103411513859275,
+      "grad_norm": 0.1387449385010934,
+      "learning_rate": 3.0230714120249947e-05,
+      "loss": 0.351,
+      "step": 2911
+    },
+    {
+      "epoch": 3.1044776119402986,
+      "grad_norm": 0.17872611874146532,
+      "learning_rate": 3.020184731771319e-05,
+      "loss": 0.3488,
+      "step": 2912
+    },
+    {
+      "epoch": 3.105543710021322,
+      "grad_norm": 0.18416826482656345,
+      "learning_rate": 3.017298594284379e-05,
+      "loss": 0.3487,
+      "step": 2913
+    },
+    {
+      "epoch": 3.106609808102345,
+      "grad_norm": 0.17275054090026867,
+      "learning_rate": 3.0144130011629448e-05,
+      "loss": 0.354,
+      "step": 2914
+    },
+    {
+      "epoch": 3.1076759061833688,
+      "grad_norm": 0.17763788392460803,
+      "learning_rate": 3.011527954005486e-05,
+      "loss": 0.3485,
+      "step": 2915
+    },
+    {
+      "epoch": 3.1087420042643923,
+      "grad_norm": 0.1945341807582054,
+      "learning_rate": 3.0086434544101677e-05,
+      "loss": 0.3489,
+      "step": 2916
+    },
+    {
+      "epoch": 3.109808102345416,
+      "grad_norm": 0.14591971732114,
+      "learning_rate": 3.005759503974854e-05,
+      "loss": 0.3499,
+      "step": 2917
+    },
+    {
+      "epoch": 3.1108742004264394,
+      "grad_norm": 0.16096483078908078,
+      "learning_rate": 3.0028761042971028e-05,
+      "loss": 0.3446,
+      "step": 2918
+    },
+    {
+      "epoch": 3.111940298507463,
+      "grad_norm": 0.1877410826166536,
+      "learning_rate": 2.9999932569741673e-05,
+      "loss": 0.3543,
+      "step": 2919
+    },
+    {
+      "epoch": 3.113006396588486,
+      "grad_norm": 0.1583922934703547,
+      "learning_rate": 2.9971109636029952e-05,
+      "loss": 0.3454,
+      "step": 2920
+    },
+    {
+      "epoch": 3.1140724946695095,
+      "grad_norm": 0.16306779868577526,
+      "learning_rate": 2.994229225780229e-05,
+      "loss": 0.3419,
+      "step": 2921
+    },
+    {
+      "epoch": 3.115138592750533,
+      "grad_norm": 0.1919727737485944,
+      "learning_rate": 2.991348045102199e-05,
+      "loss": 0.3479,
+      "step": 2922
+    },
+    {
+      "epoch": 3.1162046908315566,
+      "grad_norm": 0.14431740367185825,
+      "learning_rate": 2.988467423164931e-05,
+      "loss": 0.3519,
+      "step": 2923
+    },
+    {
+      "epoch": 3.11727078891258,
+      "grad_norm": 0.19602916545574073,
+      "learning_rate": 2.9855873615641414e-05,
+      "loss": 0.3489,
+      "step": 2924
+    },
+    {
+      "epoch": 3.1183368869936032,
+      "grad_norm": 0.15941172314892962,
+      "learning_rate": 2.982707861895231e-05,
+      "loss": 0.3473,
+      "step": 2925
+    },
+    {
+      "epoch": 3.1194029850746268,
+      "grad_norm": 0.15400208783886424,
+      "learning_rate": 2.9798289257532946e-05,
+      "loss": 0.347,
+      "step": 2926
+    },
+    {
+      "epoch": 3.1204690831556503,
+      "grad_norm": 0.16733091547516227,
+      "learning_rate": 2.976950554733114e-05,
+      "loss": 0.3487,
+      "step": 2927
+    },
+    {
+      "epoch": 3.121535181236674,
+      "grad_norm": 0.16131815693424373,
+      "learning_rate": 2.9740727504291577e-05,
+      "loss": 0.3515,
+      "step": 2928
+    },
+    {
+      "epoch": 3.1226012793176974,
+      "grad_norm": 0.18988500484746434,
+      "learning_rate": 2.9711955144355776e-05,
+      "loss": 0.3475,
+      "step": 2929
+    },
+    {
+      "epoch": 3.1236673773987205,
+      "grad_norm": 0.16873039013955726,
+      "learning_rate": 2.9683188483462135e-05,
+      "loss": 0.3506,
+      "step": 2930
+    },
+    {
+      "epoch": 3.124733475479744,
+      "grad_norm": 0.18997760424236068,
+      "learning_rate": 2.9654427537545915e-05,
+      "loss": 0.3527,
+      "step": 2931
+    },
+    {
+      "epoch": 3.1257995735607675,
+      "grad_norm": 0.16015025466863764,
+      "learning_rate": 2.962567232253914e-05,
+      "loss": 0.349,
+      "step": 2932
+    },
+    {
+      "epoch": 3.126865671641791,
+      "grad_norm": 0.14231292340170754,
+      "learning_rate": 2.9596922854370737e-05,
+      "loss": 0.3487,
+      "step": 2933
+    },
+    {
+      "epoch": 3.1279317697228146,
+      "grad_norm": 0.16799598139417324,
+      "learning_rate": 2.9568179148966406e-05,
+      "loss": 0.3504,
+      "step": 2934
+    },
+    {
+      "epoch": 3.128997867803838,
+      "grad_norm": 0.13092435516555,
+      "learning_rate": 2.9539441222248685e-05,
+      "loss": 0.3502,
+      "step": 2935
+    },
+    {
+      "epoch": 3.1300639658848612,
+      "grad_norm": 0.16201697102720602,
+      "learning_rate": 2.9510709090136855e-05,
+      "loss": 0.3507,
+      "step": 2936
+    },
+    {
+      "epoch": 3.131130063965885,
+      "grad_norm": 0.13920375866502643,
+      "learning_rate": 2.9481982768547048e-05,
+      "loss": 0.3484,
+      "step": 2937
+    },
+    {
+      "epoch": 3.1321961620469083,
+      "grad_norm": 0.1608670042161187,
+      "learning_rate": 2.945326227339215e-05,
+      "loss": 0.3505,
+      "step": 2938
+    },
+    {
+      "epoch": 3.133262260127932,
+      "grad_norm": 0.15274868977241857,
+      "learning_rate": 2.942454762058184e-05,
+      "loss": 0.353,
+      "step": 2939
+    },
+    {
+      "epoch": 3.1343283582089554,
+      "grad_norm": 0.1728035181021134,
+      "learning_rate": 2.939583882602251e-05,
+      "loss": 0.3537,
+      "step": 2940
+    },
+    {
+      "epoch": 3.1353944562899785,
+      "grad_norm": 0.20061129935035046,
+      "learning_rate": 2.936713590561735e-05,
+      "loss": 0.3537,
+      "step": 2941
+    },
+    {
+      "epoch": 3.136460554371002,
+      "grad_norm": 0.16121457578954346,
+      "learning_rate": 2.9338438875266315e-05,
+      "loss": 0.3493,
+      "step": 2942
+    },
+    {
+      "epoch": 3.1375266524520256,
+      "grad_norm": 0.1985681459680183,
+      "learning_rate": 2.930974775086602e-05,
+      "loss": 0.3494,
+      "step": 2943
+    },
+    {
+      "epoch": 3.138592750533049,
+      "grad_norm": 0.13586545126970054,
+      "learning_rate": 2.928106254830989e-05,
+      "loss": 0.3485,
+      "step": 2944
+    },
+    {
+      "epoch": 3.1396588486140726,
+      "grad_norm": 0.193935638265073,
+      "learning_rate": 2.9252383283488038e-05,
+      "loss": 0.3506,
+      "step": 2945
+    },
+    {
+      "epoch": 3.140724946695096,
+      "grad_norm": 0.15997507985832404,
+      "learning_rate": 2.9223709972287274e-05,
+      "loss": 0.3501,
+      "step": 2946
+    },
+    {
+      "epoch": 3.1417910447761193,
+      "grad_norm": 0.16394293111831423,
+      "learning_rate": 2.9195042630591115e-05,
+      "loss": 0.3428,
+      "step": 2947
+    },
+    {
+      "epoch": 3.142857142857143,
+      "grad_norm": 0.17865359760016075,
+      "learning_rate": 2.9166381274279803e-05,
+      "loss": 0.3519,
+      "step": 2948
+    },
+    {
+      "epoch": 3.1439232409381663,
+      "grad_norm": 0.17991378591531254,
+      "learning_rate": 2.913772591923021e-05,
+      "loss": 0.3485,
+      "step": 2949
+    },
+    {
+      "epoch": 3.14498933901919,
+      "grad_norm": 0.1693801011865618,
+      "learning_rate": 2.9109076581315937e-05,
+      "loss": 0.3559,
+      "step": 2950
+    },
+    {
+      "epoch": 3.1460554371002134,
+      "grad_norm": 0.14463952221236143,
+      "learning_rate": 2.908043327640723e-05,
+      "loss": 0.3579,
+      "step": 2951
+    },
+    {
+      "epoch": 3.1471215351812365,
+      "grad_norm": 0.1573143821262481,
+      "learning_rate": 2.9051796020370964e-05,
+      "loss": 0.3488,
+      "step": 2952
+    },
+    {
+      "epoch": 3.14818763326226,
+      "grad_norm": 0.16678149027123818,
+      "learning_rate": 2.9023164829070718e-05,
+      "loss": 0.3559,
+      "step": 2953
+    },
+    {
+      "epoch": 3.1492537313432836,
+      "grad_norm": 0.140262490078489,
+      "learning_rate": 2.8994539718366672e-05,
+      "loss": 0.3535,
+      "step": 2954
+    },
+    {
+      "epoch": 3.150319829424307,
+      "grad_norm": 0.2046721039374505,
+      "learning_rate": 2.8965920704115644e-05,
+      "loss": 0.3511,
+      "step": 2955
+    },
+    {
+      "epoch": 3.1513859275053306,
+      "grad_norm": 0.1438158341906599,
+      "learning_rate": 2.8937307802171085e-05,
+      "loss": 0.3443,
+      "step": 2956
+    },
+    {
+      "epoch": 3.1524520255863537,
+      "grad_norm": 0.14253271019539046,
+      "learning_rate": 2.8908701028383084e-05,
+      "loss": 0.3475,
+      "step": 2957
+    },
+    {
+      "epoch": 3.1535181236673773,
+      "grad_norm": 0.14190886636267147,
+      "learning_rate": 2.888010039859826e-05,
+      "loss": 0.3465,
+      "step": 2958
+    },
+    {
+      "epoch": 3.154584221748401,
+      "grad_norm": 0.1635398501114063,
+      "learning_rate": 2.8851505928659896e-05,
+      "loss": 0.3521,
+      "step": 2959
+    },
+    {
+      "epoch": 3.1556503198294243,
+      "grad_norm": 0.1520685576715207,
+      "learning_rate": 2.8822917634407858e-05,
+      "loss": 0.3463,
+      "step": 2960
+    },
+    {
+      "epoch": 3.156716417910448,
+      "grad_norm": 0.1723966595861352,
+      "learning_rate": 2.8794335531678545e-05,
+      "loss": 0.3481,
+      "step": 2961
+    },
+    {
+      "epoch": 3.1577825159914714,
+      "grad_norm": 0.16208068603676945,
+      "learning_rate": 2.8765759636304973e-05,
+      "loss": 0.3575,
+      "step": 2962
+    },
+    {
+      "epoch": 3.1588486140724945,
+      "grad_norm": 0.1453800336738485,
+      "learning_rate": 2.8737189964116705e-05,
+      "loss": 0.3471,
+      "step": 2963
+    },
+    {
+      "epoch": 3.159914712153518,
+      "grad_norm": 0.1858091083588778,
+      "learning_rate": 2.8708626530939865e-05,
+      "loss": 0.3506,
+      "step": 2964
+    },
+    {
+      "epoch": 3.1609808102345416,
+      "grad_norm": 0.18692761383630035,
+      "learning_rate": 2.868006935259708e-05,
+      "loss": 0.3467,
+      "step": 2965
+    },
+    {
+      "epoch": 3.162046908315565,
+      "grad_norm": 0.14846566192818872,
+      "learning_rate": 2.8651518444907556e-05,
+      "loss": 0.3471,
+      "step": 2966
+    },
+    {
+      "epoch": 3.1631130063965887,
+      "grad_norm": 0.16153996741706544,
+      "learning_rate": 2.862297382368702e-05,
+      "loss": 0.3542,
+      "step": 2967
+    },
+    {
+      "epoch": 3.1641791044776117,
+      "grad_norm": 0.16957351241623203,
+      "learning_rate": 2.8594435504747724e-05,
+      "loss": 0.3525,
+      "step": 2968
+    },
+    {
+      "epoch": 3.1652452025586353,
+      "grad_norm": 0.13030665419204615,
+      "learning_rate": 2.856590350389837e-05,
+      "loss": 0.3463,
+      "step": 2969
+    },
+    {
+      "epoch": 3.166311300639659,
+      "grad_norm": 0.13338420990478728,
+      "learning_rate": 2.8537377836944232e-05,
+      "loss": 0.3566,
+      "step": 2970
+    },
+    {
+      "epoch": 3.1673773987206824,
+      "grad_norm": 0.15453541002399912,
+      "learning_rate": 2.850885851968706e-05,
+      "loss": 0.3455,
+      "step": 2971
+    },
+    {
+      "epoch": 3.168443496801706,
+      "grad_norm": 0.1539275910876287,
+      "learning_rate": 2.8480345567925036e-05,
+      "loss": 0.3505,
+      "step": 2972
+    },
+    {
+      "epoch": 3.1695095948827294,
+      "grad_norm": 0.1601347292242495,
+      "learning_rate": 2.8451838997452875e-05,
+      "loss": 0.355,
+      "step": 2973
+    },
+    {
+      "epoch": 3.1705756929637525,
+      "grad_norm": 0.16109585673660984,
+      "learning_rate": 2.8423338824061732e-05,
+      "loss": 0.3494,
+      "step": 2974
+    },
+    {
+      "epoch": 3.171641791044776,
+      "grad_norm": 0.15883551225228731,
+      "learning_rate": 2.839484506353924e-05,
+      "loss": 0.3461,
+      "step": 2975
+    },
+    {
+      "epoch": 3.1727078891257996,
+      "grad_norm": 0.18127651611679274,
+      "learning_rate": 2.836635773166943e-05,
+      "loss": 0.3533,
+      "step": 2976
+    },
+    {
+      "epoch": 3.173773987206823,
+      "grad_norm": 0.14832179685684777,
+      "learning_rate": 2.833787684423282e-05,
+      "loss": 0.3512,
+      "step": 2977
+    },
+    {
+      "epoch": 3.1748400852878467,
+      "grad_norm": 0.1833917506129557,
+      "learning_rate": 2.8309402417006344e-05,
+      "loss": 0.3488,
+      "step": 2978
+    },
+    {
+      "epoch": 3.1759061833688698,
+      "grad_norm": 0.16898776691695933,
+      "learning_rate": 2.8280934465763352e-05,
+      "loss": 0.3451,
+      "step": 2979
+    },
+    {
+      "epoch": 3.1769722814498933,
+      "grad_norm": 0.14677271644886078,
+      "learning_rate": 2.82524730062736e-05,
+      "loss": 0.3509,
+      "step": 2980
+    },
+    {
+      "epoch": 3.178038379530917,
+      "grad_norm": 0.19057648965710483,
+      "learning_rate": 2.8224018054303278e-05,
+      "loss": 0.3551,
+      "step": 2981
+    },
+    {
+      "epoch": 3.1791044776119404,
+      "grad_norm": 0.13538971093481636,
+      "learning_rate": 2.8195569625614933e-05,
+      "loss": 0.3456,
+      "step": 2982
+    },
+    {
+      "epoch": 3.180170575692964,
+      "grad_norm": 0.1390167814695544,
+      "learning_rate": 2.816712773596751e-05,
+      "loss": 0.3531,
+      "step": 2983
+    },
+    {
+      "epoch": 3.181236673773987,
+      "grad_norm": 0.15503459765163632,
+      "learning_rate": 2.8138692401116366e-05,
+      "loss": 0.3496,
+      "step": 2984
+    },
+    {
+      "epoch": 3.1823027718550105,
+      "grad_norm": 0.14462829278663678,
+      "learning_rate": 2.811026363681317e-05,
+      "loss": 0.3553,
+      "step": 2985
+    },
+    {
+      "epoch": 3.183368869936034,
+      "grad_norm": 0.17695463207505402,
+      "learning_rate": 2.8081841458806002e-05,
+      "loss": 0.3533,
+      "step": 2986
+    },
+    {
+      "epoch": 3.1844349680170576,
+      "grad_norm": 0.1591547097918123,
+      "learning_rate": 2.8053425882839252e-05,
+      "loss": 0.3471,
+      "step": 2987
+    },
+    {
+      "epoch": 3.185501066098081,
+      "grad_norm": 0.23045461569297743,
+      "learning_rate": 2.802501692465368e-05,
+      "loss": 0.3558,
+      "step": 2988
+    },
+    {
+      "epoch": 3.1865671641791047,
+      "grad_norm": 0.17997948094924288,
+      "learning_rate": 2.799661459998638e-05,
+      "loss": 0.3451,
+      "step": 2989
+    },
+    {
+      "epoch": 3.1876332622601278,
+      "grad_norm": 0.15073757914110497,
+      "learning_rate": 2.7968218924570757e-05,
+      "loss": 0.3499,
+      "step": 2990
+    },
+    {
+      "epoch": 3.1886993603411513,
+      "grad_norm": 0.149109855986733,
+      "learning_rate": 2.7939829914136533e-05,
+      "loss": 0.3505,
+      "step": 2991
+    },
+    {
+      "epoch": 3.189765458422175,
+      "grad_norm": 0.1828583325585577,
+      "learning_rate": 2.791144758440975e-05,
+      "loss": 0.3534,
+      "step": 2992
+    },
+    {
+      "epoch": 3.1908315565031984,
+      "grad_norm": 0.1835467829513033,
+      "learning_rate": 2.788307195111276e-05,
+      "loss": 0.3571,
+      "step": 2993
+    },
+    {
+      "epoch": 3.191897654584222,
+      "grad_norm": 0.14112165071165333,
+      "learning_rate": 2.7854703029964157e-05,
+      "loss": 0.349,
+      "step": 2994
+    },
+    {
+      "epoch": 3.192963752665245,
+      "grad_norm": 0.14994590904845737,
+      "learning_rate": 2.7826340836678868e-05,
+      "loss": 0.3512,
+      "step": 2995
+    },
+    {
+      "epoch": 3.1940298507462686,
+      "grad_norm": 0.15001218329008587,
+      "learning_rate": 2.779798538696807e-05,
+      "loss": 0.3508,
+      "step": 2996
+    },
+    {
+      "epoch": 3.195095948827292,
+      "grad_norm": 0.14167484348076181,
+      "learning_rate": 2.776963669653923e-05,
+      "loss": 0.3531,
+      "step": 2997
+    },
+    {
+      "epoch": 3.1961620469083156,
+      "grad_norm": 0.16415685787588938,
+      "learning_rate": 2.7741294781096008e-05,
+      "loss": 0.3475,
+      "step": 2998
+    },
+    {
+      "epoch": 3.197228144989339,
+      "grad_norm": 0.18767514637702254,
+      "learning_rate": 2.7712959656338375e-05,
+      "loss": 0.3503,
+      "step": 2999
+    },
+    {
+      "epoch": 3.1982942430703627,
+      "grad_norm": 0.1318270842245805,
+      "learning_rate": 2.7684631337962535e-05,
+      "loss": 0.3483,
+      "step": 3000
+    },
+    {
+      "epoch": 3.199360341151386,
+      "grad_norm": 0.17743086770734862,
+      "learning_rate": 2.7656309841660864e-05,
+      "loss": 0.3497,
+      "step": 3001
+    },
+    {
+      "epoch": 3.2004264392324093,
+      "grad_norm": 0.20226460565691426,
+      "learning_rate": 2.7627995183122025e-05,
+      "loss": 0.3548,
+      "step": 3002
+    },
+    {
+      "epoch": 3.201492537313433,
+      "grad_norm": 0.1404795955564027,
+      "learning_rate": 2.7599687378030862e-05,
+      "loss": 0.3456,
+      "step": 3003
+    },
+    {
+      "epoch": 3.2025586353944564,
+      "grad_norm": 0.1664198795587612,
+      "learning_rate": 2.7571386442068443e-05,
+      "loss": 0.348,
+      "step": 3004
+    },
+    {
+      "epoch": 3.20362473347548,
+      "grad_norm": 0.1871561566919217,
+      "learning_rate": 2.754309239091199e-05,
+      "loss": 0.3487,
+      "step": 3005
+    },
+    {
+      "epoch": 3.204690831556503,
+      "grad_norm": 0.1762158383532131,
+      "learning_rate": 2.7514805240234942e-05,
+      "loss": 0.3489,
+      "step": 3006
+    },
+    {
+      "epoch": 3.2057569296375266,
+      "grad_norm": 0.1409821970880933,
+      "learning_rate": 2.7486525005706915e-05,
+      "loss": 0.3462,
+      "step": 3007
+    },
+    {
+      "epoch": 3.20682302771855,
+      "grad_norm": 0.1919928381276622,
+      "learning_rate": 2.745825170299371e-05,
+      "loss": 0.3431,
+      "step": 3008
+    },
+    {
+      "epoch": 3.2078891257995736,
+      "grad_norm": 0.14051119559763037,
+      "learning_rate": 2.7429985347757232e-05,
+      "loss": 0.3448,
+      "step": 3009
+    },
+    {
+      "epoch": 3.208955223880597,
+      "grad_norm": 0.1581183828916611,
+      "learning_rate": 2.7401725955655582e-05,
+      "loss": 0.3494,
+      "step": 3010
+    },
+    {
+      "epoch": 3.2100213219616203,
+      "grad_norm": 0.16965574848535128,
+      "learning_rate": 2.7373473542343023e-05,
+      "loss": 0.3447,
+      "step": 3011
+    },
+    {
+      "epoch": 3.211087420042644,
+      "grad_norm": 0.1596889434072171,
+      "learning_rate": 2.7345228123469886e-05,
+      "loss": 0.3483,
+      "step": 3012
+    },
+    {
+      "epoch": 3.2121535181236673,
+      "grad_norm": 0.17408804946301867,
+      "learning_rate": 2.731698971468268e-05,
+      "loss": 0.3567,
+      "step": 3013
+    },
+    {
+      "epoch": 3.213219616204691,
+      "grad_norm": 0.16081014646576602,
+      "learning_rate": 2.7288758331624025e-05,
+      "loss": 0.3447,
+      "step": 3014
+    },
+    {
+      "epoch": 3.2142857142857144,
+      "grad_norm": 0.1387088635896051,
+      "learning_rate": 2.7260533989932628e-05,
+      "loss": 0.3512,
+      "step": 3015
+    },
+    {
+      "epoch": 3.2153518123667375,
+      "grad_norm": 0.15123466339681416,
+      "learning_rate": 2.7232316705243305e-05,
+      "loss": 0.3481,
+      "step": 3016
+    },
+    {
+      "epoch": 3.216417910447761,
+      "grad_norm": 0.15290558325073092,
+      "learning_rate": 2.720410649318698e-05,
+      "loss": 0.3488,
+      "step": 3017
+    },
+    {
+      "epoch": 3.2174840085287846,
+      "grad_norm": 0.17149168701297954,
+      "learning_rate": 2.7175903369390638e-05,
+      "loss": 0.3517,
+      "step": 3018
+    },
+    {
+      "epoch": 3.218550106609808,
+      "grad_norm": 0.16114167670939453,
+      "learning_rate": 2.7147707349477327e-05,
+      "loss": 0.352,
+      "step": 3019
+    },
+    {
+      "epoch": 3.2196162046908317,
+      "grad_norm": 0.14097634690323843,
+      "learning_rate": 2.7119518449066205e-05,
+      "loss": 0.349,
+      "step": 3020
+    },
+    {
+      "epoch": 3.220682302771855,
+      "grad_norm": 0.16246129452859187,
+      "learning_rate": 2.7091336683772437e-05,
+      "loss": 0.3496,
+      "step": 3021
+    },
+    {
+      "epoch": 3.2217484008528783,
+      "grad_norm": 0.16336429392891585,
+      "learning_rate": 2.7063162069207262e-05,
+      "loss": 0.3474,
+      "step": 3022
+    },
+    {
+      "epoch": 3.222814498933902,
+      "grad_norm": 0.14988643488196962,
+      "learning_rate": 2.7034994620977965e-05,
+      "loss": 0.3477,
+      "step": 3023
+    },
+    {
+      "epoch": 3.2238805970149254,
+      "grad_norm": 0.16792178240326763,
+      "learning_rate": 2.700683435468782e-05,
+      "loss": 0.3515,
+      "step": 3024
+    },
+    {
+      "epoch": 3.224946695095949,
+      "grad_norm": 0.14865516690552993,
+      "learning_rate": 2.6978681285936176e-05,
+      "loss": 0.3535,
+      "step": 3025
+    },
+    {
+      "epoch": 3.2260127931769724,
+      "grad_norm": 0.15929449524500977,
+      "learning_rate": 2.6950535430318373e-05,
+      "loss": 0.3519,
+      "step": 3026
+    },
+    {
+      "epoch": 3.227078891257996,
+      "grad_norm": 0.1599091744019697,
+      "learning_rate": 2.692239680342572e-05,
+      "loss": 0.3529,
+      "step": 3027
+    },
+    {
+      "epoch": 3.228144989339019,
+      "grad_norm": 0.1484966105614425,
+      "learning_rate": 2.689426542084558e-05,
+      "loss": 0.3579,
+      "step": 3028
+    },
+    {
+      "epoch": 3.2292110874200426,
+      "grad_norm": 0.1385274858632669,
+      "learning_rate": 2.686614129816129e-05,
+      "loss": 0.3497,
+      "step": 3029
+    },
+    {
+      "epoch": 3.230277185501066,
+      "grad_norm": 0.14991864182995415,
+      "learning_rate": 2.683802445095211e-05,
+      "loss": 0.3525,
+      "step": 3030
+    },
+    {
+      "epoch": 3.2313432835820897,
+      "grad_norm": 0.15476961485342514,
+      "learning_rate": 2.6809914894793344e-05,
+      "loss": 0.3424,
+      "step": 3031
+    },
+    {
+      "epoch": 3.232409381663113,
+      "grad_norm": 0.12397588075589204,
+      "learning_rate": 2.6781812645256216e-05,
+      "loss": 0.3426,
+      "step": 3032
+    },
+    {
+      "epoch": 3.2334754797441363,
+      "grad_norm": 0.15055980357741625,
+      "learning_rate": 2.6753717717907925e-05,
+      "loss": 0.3515,
+      "step": 3033
+    },
+    {
+      "epoch": 3.23454157782516,
+      "grad_norm": 0.14872467882760348,
+      "learning_rate": 2.672563012831158e-05,
+      "loss": 0.3456,
+      "step": 3034
+    },
+    {
+      "epoch": 3.2356076759061834,
+      "grad_norm": 0.14261242873484498,
+      "learning_rate": 2.6697549892026247e-05,
+      "loss": 0.3544,
+      "step": 3035
+    },
+    {
+      "epoch": 3.236673773987207,
+      "grad_norm": 0.15857980357826498,
+      "learning_rate": 2.666947702460693e-05,
+      "loss": 0.3546,
+      "step": 3036
+    },
+    {
+      "epoch": 3.2377398720682304,
+      "grad_norm": 0.16110840435766394,
+      "learning_rate": 2.6641411541604544e-05,
+      "loss": 0.3536,
+      "step": 3037
+    },
+    {
+      "epoch": 3.2388059701492535,
+      "grad_norm": 0.16212693241700432,
+      "learning_rate": 2.6613353458565887e-05,
+      "loss": 0.3511,
+      "step": 3038
+    },
+    {
+      "epoch": 3.239872068230277,
+      "grad_norm": 0.14587623221049775,
+      "learning_rate": 2.6585302791033688e-05,
+      "loss": 0.3522,
+      "step": 3039
+    },
+    {
+      "epoch": 3.2409381663113006,
+      "grad_norm": 0.13597680071298357,
+      "learning_rate": 2.6557259554546577e-05,
+      "loss": 0.3447,
+      "step": 3040
+    },
+    {
+      "epoch": 3.242004264392324,
+      "grad_norm": 0.16519790924407182,
+      "learning_rate": 2.6529223764639013e-05,
+      "loss": 0.3531,
+      "step": 3041
+    },
+    {
+      "epoch": 3.2430703624733477,
+      "grad_norm": 0.12865148005181068,
+      "learning_rate": 2.650119543684139e-05,
+      "loss": 0.3453,
+      "step": 3042
+    },
+    {
+      "epoch": 3.2441364605543708,
+      "grad_norm": 0.16931532502182053,
+      "learning_rate": 2.6473174586679947e-05,
+      "loss": 0.3481,
+      "step": 3043
+    },
+    {
+      "epoch": 3.2452025586353943,
+      "grad_norm": 0.20153378061586566,
+      "learning_rate": 2.644516122967678e-05,
+      "loss": 0.3474,
+      "step": 3044
+    },
+    {
+      "epoch": 3.246268656716418,
+      "grad_norm": 0.15391567159601363,
+      "learning_rate": 2.6417155381349814e-05,
+      "loss": 0.3555,
+      "step": 3045
+    },
+    {
+      "epoch": 3.2473347547974414,
+      "grad_norm": 0.18485138313641797,
+      "learning_rate": 2.638915705721284e-05,
+      "loss": 0.348,
+      "step": 3046
+    },
+    {
+      "epoch": 3.248400852878465,
+      "grad_norm": 0.19279732548240924,
+      "learning_rate": 2.6361166272775503e-05,
+      "loss": 0.3473,
+      "step": 3047
+    },
+    {
+      "epoch": 3.2494669509594885,
+      "grad_norm": 0.14914227409372177,
+      "learning_rate": 2.6333183043543207e-05,
+      "loss": 0.3491,
+      "step": 3048
+    },
+    {
+      "epoch": 3.2505330490405115,
+      "grad_norm": 0.198163678392842,
+      "learning_rate": 2.630520738501721e-05,
+      "loss": 0.3537,
+      "step": 3049
+    },
+    {
+      "epoch": 3.251599147121535,
+      "grad_norm": 0.1495697517853574,
+      "learning_rate": 2.6277239312694596e-05,
+      "loss": 0.3521,
+      "step": 3050
+    },
+    {
+      "epoch": 3.2526652452025586,
+      "grad_norm": 0.18049308488761015,
+      "learning_rate": 2.624927884206821e-05,
+      "loss": 0.3559,
+      "step": 3051
+    },
+    {
+      "epoch": 3.253731343283582,
+      "grad_norm": 0.1554880781007249,
+      "learning_rate": 2.6221325988626686e-05,
+      "loss": 0.3474,
+      "step": 3052
+    },
+    {
+      "epoch": 3.2547974413646057,
+      "grad_norm": 0.17014877400454104,
+      "learning_rate": 2.619338076785448e-05,
+      "loss": 0.3471,
+      "step": 3053
+    },
+    {
+      "epoch": 3.2558635394456292,
+      "grad_norm": 0.2175697746157974,
+      "learning_rate": 2.6165443195231763e-05,
+      "loss": 0.3526,
+      "step": 3054
+    },
+    {
+      "epoch": 3.2569296375266523,
+      "grad_norm": 0.1533228757448852,
+      "learning_rate": 2.6137513286234528e-05,
+      "loss": 0.355,
+      "step": 3055
+    },
+    {
+      "epoch": 3.257995735607676,
+      "grad_norm": 0.170158319003947,
+      "learning_rate": 2.6109591056334474e-05,
+      "loss": 0.3497,
+      "step": 3056
+    },
+    {
+      "epoch": 3.2590618336886994,
+      "grad_norm": 0.2110756154467986,
+      "learning_rate": 2.608167652099906e-05,
+      "loss": 0.3439,
+      "step": 3057
+    },
+    {
+      "epoch": 3.260127931769723,
+      "grad_norm": 0.14837278713007337,
+      "learning_rate": 2.6053769695691507e-05,
+      "loss": 0.3423,
+      "step": 3058
+    },
+    {
+      "epoch": 3.2611940298507465,
+      "grad_norm": 0.1809824101682613,
+      "learning_rate": 2.6025870595870733e-05,
+      "loss": 0.3487,
+      "step": 3059
+    },
+    {
+      "epoch": 3.2622601279317696,
+      "grad_norm": 0.21033652261955751,
+      "learning_rate": 2.5997979236991386e-05,
+      "loss": 0.3527,
+      "step": 3060
+    },
+    {
+      "epoch": 3.263326226012793,
+      "grad_norm": 0.21776809754829038,
+      "learning_rate": 2.5970095634503833e-05,
+      "loss": 0.3472,
+      "step": 3061
+    },
+    {
+      "epoch": 3.2643923240938166,
+      "grad_norm": 0.15831570347972593,
+      "learning_rate": 2.5942219803854168e-05,
+      "loss": 0.3437,
+      "step": 3062
+    },
+    {
+      "epoch": 3.26545842217484,
+      "grad_norm": 0.17838265493780678,
+      "learning_rate": 2.59143517604841e-05,
+      "loss": 0.3476,
+      "step": 3063
+    },
+    {
+      "epoch": 3.2665245202558637,
+      "grad_norm": 0.20607722523788694,
+      "learning_rate": 2.588649151983111e-05,
+      "loss": 0.3472,
+      "step": 3064
+    },
+    {
+      "epoch": 3.267590618336887,
+      "grad_norm": 0.13745662304832107,
+      "learning_rate": 2.5858639097328314e-05,
+      "loss": 0.351,
+      "step": 3065
+    },
+    {
+      "epoch": 3.2686567164179103,
+      "grad_norm": 0.1941806894488011,
+      "learning_rate": 2.583079450840453e-05,
+      "loss": 0.3537,
+      "step": 3066
+    },
+    {
+      "epoch": 3.269722814498934,
+      "grad_norm": 0.21824169614328015,
+      "learning_rate": 2.5802957768484173e-05,
+      "loss": 0.3461,
+      "step": 3067
+    },
+    {
+      "epoch": 3.2707889125799574,
+      "grad_norm": 0.16071832337632452,
+      "learning_rate": 2.5775128892987368e-05,
+      "loss": 0.3505,
+      "step": 3068
+    },
+    {
+      "epoch": 3.271855010660981,
+      "grad_norm": 0.163987167314734,
+      "learning_rate": 2.574730789732989e-05,
+      "loss": 0.345,
+      "step": 3069
+    },
+    {
+      "epoch": 3.272921108742004,
+      "grad_norm": 0.19905464540458992,
+      "learning_rate": 2.5719494796923085e-05,
+      "loss": 0.3574,
+      "step": 3070
+    },
+    {
+      "epoch": 3.2739872068230276,
+      "grad_norm": 0.16383647964999232,
+      "learning_rate": 2.569168960717398e-05,
+      "loss": 0.3495,
+      "step": 3071
+    },
+    {
+      "epoch": 3.275053304904051,
+      "grad_norm": 0.17693978324786094,
+      "learning_rate": 2.5663892343485214e-05,
+      "loss": 0.3499,
+      "step": 3072
+    },
+    {
+      "epoch": 3.2761194029850746,
+      "grad_norm": 0.18500426824572122,
+      "learning_rate": 2.5636103021255026e-05,
+      "loss": 0.3549,
+      "step": 3073
+    },
+    {
+      "epoch": 3.277185501066098,
+      "grad_norm": 0.19504056601794661,
+      "learning_rate": 2.5608321655877243e-05,
+      "loss": 0.3525,
+      "step": 3074
+    },
+    {
+      "epoch": 3.2782515991471217,
+      "grad_norm": 0.16037712681498448,
+      "learning_rate": 2.5580548262741304e-05,
+      "loss": 0.3474,
+      "step": 3075
+    },
+    {
+      "epoch": 3.279317697228145,
+      "grad_norm": 0.17838055168396916,
+      "learning_rate": 2.5552782857232238e-05,
+      "loss": 0.3519,
+      "step": 3076
+    },
+    {
+      "epoch": 3.2803837953091683,
+      "grad_norm": 0.18837820237068092,
+      "learning_rate": 2.5525025454730612e-05,
+      "loss": 0.3497,
+      "step": 3077
+    },
+    {
+      "epoch": 3.281449893390192,
+      "grad_norm": 0.1403381013776277,
+      "learning_rate": 2.54972760706126e-05,
+      "loss": 0.3505,
+      "step": 3078
+    },
+    {
+      "epoch": 3.2825159914712154,
+      "grad_norm": 0.234296775090172,
+      "learning_rate": 2.546953472024991e-05,
+      "loss": 0.3502,
+      "step": 3079
+    },
+    {
+      "epoch": 3.283582089552239,
+      "grad_norm": 0.17074806676345755,
+      "learning_rate": 2.5441801419009835e-05,
+      "loss": 0.3449,
+      "step": 3080
+    },
+    {
+      "epoch": 3.2846481876332625,
+      "grad_norm": 0.17367375722672437,
+      "learning_rate": 2.541407618225515e-05,
+      "loss": 0.3503,
+      "step": 3081
+    },
+    {
+      "epoch": 3.2857142857142856,
+      "grad_norm": 0.22550229833239124,
+      "learning_rate": 2.53863590253442e-05,
+      "loss": 0.3477,
+      "step": 3082
+    },
+    {
+      "epoch": 3.286780383795309,
+      "grad_norm": 0.13550221130573722,
+      "learning_rate": 2.5358649963630867e-05,
+      "loss": 0.3449,
+      "step": 3083
+    },
+    {
+      "epoch": 3.2878464818763327,
+      "grad_norm": 0.2501969239886366,
+      "learning_rate": 2.533094901246452e-05,
+      "loss": 0.3528,
+      "step": 3084
+    },
+    {
+      "epoch": 3.288912579957356,
+      "grad_norm": 0.20047206624632516,
+      "learning_rate": 2.5303256187190038e-05,
+      "loss": 0.3466,
+      "step": 3085
+    },
+    {
+      "epoch": 3.2899786780383797,
+      "grad_norm": 0.14385706915661878,
+      "learning_rate": 2.527557150314783e-05,
+      "loss": 0.348,
+      "step": 3086
+    },
+    {
+      "epoch": 3.291044776119403,
+      "grad_norm": 0.23455244794599786,
+      "learning_rate": 2.524789497567375e-05,
+      "loss": 0.3575,
+      "step": 3087
+    },
+    {
+      "epoch": 3.2921108742004264,
+      "grad_norm": 0.19720110623267217,
+      "learning_rate": 2.522022662009916e-05,
+      "loss": 0.355,
+      "step": 3088
+    },
+    {
+      "epoch": 3.29317697228145,
+      "grad_norm": 0.16009116202356855,
+      "learning_rate": 2.5192566451750904e-05,
+      "loss": 0.3483,
+      "step": 3089
+    },
+    {
+      "epoch": 3.2942430703624734,
+      "grad_norm": 0.19300246066961,
+      "learning_rate": 2.516491448595126e-05,
+      "loss": 0.3537,
+      "step": 3090
+    },
+    {
+      "epoch": 3.295309168443497,
+      "grad_norm": 0.17653040799060407,
+      "learning_rate": 2.5137270738018e-05,
+      "loss": 0.3463,
+      "step": 3091
+    },
+    {
+      "epoch": 3.29637526652452,
+      "grad_norm": 0.19537049403171847,
+      "learning_rate": 2.5109635223264305e-05,
+      "loss": 0.3424,
+      "step": 3092
+    },
+    {
+      "epoch": 3.2974413646055436,
+      "grad_norm": 0.18015777011327852,
+      "learning_rate": 2.5082007956998817e-05,
+      "loss": 0.3476,
+      "step": 3093
+    },
+    {
+      "epoch": 3.298507462686567,
+      "grad_norm": 0.20459512311591024,
+      "learning_rate": 2.505438895452562e-05,
+      "loss": 0.351,
+      "step": 3094
+    },
+    {
+      "epoch": 3.2995735607675907,
+      "grad_norm": 0.12366732649317048,
+      "learning_rate": 2.5026778231144194e-05,
+      "loss": 0.354,
+      "step": 3095
+    },
+    {
+      "epoch": 3.300639658848614,
+      "grad_norm": 0.18890110077159672,
+      "learning_rate": 2.4999175802149438e-05,
+      "loss": 0.3523,
+      "step": 3096
+    },
+    {
+      "epoch": 3.3017057569296373,
+      "grad_norm": 0.1626946147400098,
+      "learning_rate": 2.4971581682831668e-05,
+      "loss": 0.3527,
+      "step": 3097
+    },
+    {
+      "epoch": 3.302771855010661,
+      "grad_norm": 0.14801464468188533,
+      "learning_rate": 2.494399588847662e-05,
+      "loss": 0.3533,
+      "step": 3098
+    },
+    {
+      "epoch": 3.3038379530916844,
+      "grad_norm": 0.1593500938211409,
+      "learning_rate": 2.4916418434365346e-05,
+      "loss": 0.3528,
+      "step": 3099
+    },
+    {
+      "epoch": 3.304904051172708,
+      "grad_norm": 0.19644211588346217,
+      "learning_rate": 2.488884933577434e-05,
+      "loss": 0.3454,
+      "step": 3100
+    },
+    {
+      "epoch": 3.3059701492537314,
+      "grad_norm": 0.1353438588547366,
+      "learning_rate": 2.4861288607975458e-05,
+      "loss": 0.3516,
+      "step": 3101
+    },
+    {
+      "epoch": 3.307036247334755,
+      "grad_norm": 0.17358657320991785,
+      "learning_rate": 2.4833736266235917e-05,
+      "loss": 0.3499,
+      "step": 3102
+    },
+    {
+      "epoch": 3.308102345415778,
+      "grad_norm": 0.15831348237018122,
+      "learning_rate": 2.4806192325818258e-05,
+      "loss": 0.3531,
+      "step": 3103
+    },
+    {
+      "epoch": 3.3091684434968016,
+      "grad_norm": 0.13584971050526615,
+      "learning_rate": 2.47786568019804e-05,
+      "loss": 0.3593,
+      "step": 3104
+    },
+    {
+      "epoch": 3.310234541577825,
+      "grad_norm": 0.15418755306420462,
+      "learning_rate": 2.475112970997562e-05,
+      "loss": 0.341,
+      "step": 3105
+    },
+    {
+      "epoch": 3.3113006396588487,
+      "grad_norm": 0.1615129100696353,
+      "learning_rate": 2.472361106505245e-05,
+      "loss": 0.3514,
+      "step": 3106
+    },
+    {
+      "epoch": 3.3123667377398722,
+      "grad_norm": 0.15930398883446692,
+      "learning_rate": 2.4696100882454817e-05,
+      "loss": 0.351,
+      "step": 3107
+    },
+    {
+      "epoch": 3.3134328358208958,
+      "grad_norm": 0.1403230514001666,
+      "learning_rate": 2.466859917742193e-05,
+      "loss": 0.3505,
+      "step": 3108
+    },
+    {
+      "epoch": 3.314498933901919,
+      "grad_norm": 0.17920374641541106,
+      "learning_rate": 2.464110596518831e-05,
+      "loss": 0.3449,
+      "step": 3109
+    },
+    {
+      "epoch": 3.3155650319829424,
+      "grad_norm": 0.1254186924644907,
+      "learning_rate": 2.4613621260983755e-05,
+      "loss": 0.3461,
+      "step": 3110
+    },
+    {
+      "epoch": 3.316631130063966,
+      "grad_norm": 0.13780163296736875,
+      "learning_rate": 2.458614508003336e-05,
+      "loss": 0.3511,
+      "step": 3111
+    },
+    {
+      "epoch": 3.3176972281449895,
+      "grad_norm": 0.143814949009123,
+      "learning_rate": 2.455867743755751e-05,
+      "loss": 0.3505,
+      "step": 3112
+    },
+    {
+      "epoch": 3.318763326226013,
+      "grad_norm": 0.12209116221359133,
+      "learning_rate": 2.4531218348771866e-05,
+      "loss": 0.3443,
+      "step": 3113
+    },
+    {
+      "epoch": 3.319829424307036,
+      "grad_norm": 0.14799939510773816,
+      "learning_rate": 2.450376782888731e-05,
+      "loss": 0.3471,
+      "step": 3114
+    },
+    {
+      "epoch": 3.3208955223880596,
+      "grad_norm": 0.14123259883639513,
+      "learning_rate": 2.4476325893110008e-05,
+      "loss": 0.3502,
+      "step": 3115
+    },
+    {
+      "epoch": 3.321961620469083,
+      "grad_norm": 0.14375390587624387,
+      "learning_rate": 2.4448892556641393e-05,
+      "loss": 0.3463,
+      "step": 3116
+    },
+    {
+      "epoch": 3.3230277185501067,
+      "grad_norm": 0.17421629802465335,
+      "learning_rate": 2.4421467834678067e-05,
+      "loss": 0.3475,
+      "step": 3117
+    },
+    {
+      "epoch": 3.3240938166311302,
+      "grad_norm": 0.14674808581087775,
+      "learning_rate": 2.439405174241192e-05,
+      "loss": 0.3451,
+      "step": 3118
+    },
+    {
+      "epoch": 3.3251599147121533,
+      "grad_norm": 0.21187149354636753,
+      "learning_rate": 2.4366644295030054e-05,
+      "loss": 0.3483,
+      "step": 3119
+    },
+    {
+      "epoch": 3.326226012793177,
+      "grad_norm": 0.2012599089170034,
+      "learning_rate": 2.433924550771476e-05,
+      "loss": 0.3483,
+      "step": 3120
+    },
+    {
+      "epoch": 3.3272921108742004,
+      "grad_norm": 0.16780980496521763,
+      "learning_rate": 2.4311855395643527e-05,
+      "loss": 0.3462,
+      "step": 3121
+    },
+    {
+      "epoch": 3.328358208955224,
+      "grad_norm": 0.16005417142229794,
+      "learning_rate": 2.428447397398908e-05,
+      "loss": 0.3533,
+      "step": 3122
+    },
+    {
+      "epoch": 3.3294243070362475,
+      "grad_norm": 0.15099935350620822,
+      "learning_rate": 2.425710125791929e-05,
+      "loss": 0.35,
+      "step": 3123
+    },
+    {
+      "epoch": 3.3304904051172706,
+      "grad_norm": 0.15768848577018113,
+      "learning_rate": 2.4229737262597216e-05,
+      "loss": 0.3498,
+      "step": 3124
+    },
+    {
+      "epoch": 3.331556503198294,
+      "grad_norm": 0.155851400636398,
+      "learning_rate": 2.4202382003181098e-05,
+      "loss": 0.3557,
+      "step": 3125
+    },
+    {
+      "epoch": 3.3326226012793176,
+      "grad_norm": 0.14267053890933395,
+      "learning_rate": 2.4175035494824316e-05,
+      "loss": 0.3486,
+      "step": 3126
+    },
+    {
+      "epoch": 3.333688699360341,
+      "grad_norm": 0.15843384932692273,
+      "learning_rate": 2.4147697752675433e-05,
+      "loss": 0.3515,
+      "step": 3127
+    },
+    {
+      "epoch": 3.3347547974413647,
+      "grad_norm": 0.17371867125289828,
+      "learning_rate": 2.4120368791878122e-05,
+      "loss": 0.3491,
+      "step": 3128
+    },
+    {
+      "epoch": 3.3358208955223883,
+      "grad_norm": 0.16084952359302623,
+      "learning_rate": 2.4093048627571206e-05,
+      "loss": 0.3529,
+      "step": 3129
+    },
+    {
+      "epoch": 3.3368869936034113,
+      "grad_norm": 0.14408159186617614,
+      "learning_rate": 2.4065737274888646e-05,
+      "loss": 0.3525,
+      "step": 3130
+    },
+    {
+      "epoch": 3.337953091684435,
+      "grad_norm": 0.1656284179555205,
+      "learning_rate": 2.403843474895952e-05,
+      "loss": 0.3486,
+      "step": 3131
+    },
+    {
+      "epoch": 3.3390191897654584,
+      "grad_norm": 0.14146974935511217,
+      "learning_rate": 2.401114106490798e-05,
+      "loss": 0.3482,
+      "step": 3132
+    },
+    {
+      "epoch": 3.340085287846482,
+      "grad_norm": 0.1573564750602216,
+      "learning_rate": 2.3983856237853322e-05,
+      "loss": 0.3543,
+      "step": 3133
+    },
+    {
+      "epoch": 3.3411513859275055,
+      "grad_norm": 0.1757062911156589,
+      "learning_rate": 2.395658028290995e-05,
+      "loss": 0.3538,
+      "step": 3134
+    },
+    {
+      "epoch": 3.342217484008529,
+      "grad_norm": 0.13319696217453428,
+      "learning_rate": 2.3929313215187274e-05,
+      "loss": 0.3488,
+      "step": 3135
+    },
+    {
+      "epoch": 3.343283582089552,
+      "grad_norm": 0.15216598233478518,
+      "learning_rate": 2.390205504978986e-05,
+      "loss": 0.3574,
+      "step": 3136
+    },
+    {
+      "epoch": 3.3443496801705757,
+      "grad_norm": 0.14909080713095296,
+      "learning_rate": 2.3874805801817313e-05,
+      "loss": 0.3488,
+      "step": 3137
+    },
+    {
+      "epoch": 3.345415778251599,
+      "grad_norm": 0.12886400979675366,
+      "learning_rate": 2.384756548636432e-05,
+      "loss": 0.3474,
+      "step": 3138
+    },
+    {
+      "epoch": 3.3464818763326227,
+      "grad_norm": 0.17178937827304538,
+      "learning_rate": 2.3820334118520566e-05,
+      "loss": 0.3487,
+      "step": 3139
+    },
+    {
+      "epoch": 3.3475479744136463,
+      "grad_norm": 0.14792535442811752,
+      "learning_rate": 2.3793111713370824e-05,
+      "loss": 0.3493,
+      "step": 3140
+    },
+    {
+      "epoch": 3.3486140724946694,
+      "grad_norm": 0.13063840167660581,
+      "learning_rate": 2.3765898285994898e-05,
+      "loss": 0.3436,
+      "step": 3141
+    },
+    {
+      "epoch": 3.349680170575693,
+      "grad_norm": 0.1415902890105865,
+      "learning_rate": 2.3738693851467627e-05,
+      "loss": 0.3449,
+      "step": 3142
+    },
+    {
+      "epoch": 3.3507462686567164,
+      "grad_norm": 0.14468858858404718,
+      "learning_rate": 2.371149842485882e-05,
+      "loss": 0.3515,
+      "step": 3143
+    },
+    {
+      "epoch": 3.35181236673774,
+      "grad_norm": 0.12569229542636323,
+      "learning_rate": 2.3684312021233353e-05,
+      "loss": 0.3531,
+      "step": 3144
+    },
+    {
+      "epoch": 3.3528784648187635,
+      "grad_norm": 0.13410217015876344,
+      "learning_rate": 2.3657134655651085e-05,
+      "loss": 0.3493,
+      "step": 3145
+    },
+    {
+      "epoch": 3.3539445628997866,
+      "grad_norm": 0.12050748503992254,
+      "learning_rate": 2.3629966343166836e-05,
+      "loss": 0.3475,
+      "step": 3146
+    },
+    {
+      "epoch": 3.35501066098081,
+      "grad_norm": 0.13625653337833818,
+      "learning_rate": 2.3602807098830462e-05,
+      "loss": 0.3527,
+      "step": 3147
+    },
+    {
+      "epoch": 3.3560767590618337,
+      "grad_norm": 0.1270892484979118,
+      "learning_rate": 2.3575656937686765e-05,
+      "loss": 0.3451,
+      "step": 3148
+    },
+    {
+      "epoch": 3.357142857142857,
+      "grad_norm": 0.1292034055414621,
+      "learning_rate": 2.3548515874775547e-05,
+      "loss": 0.3468,
+      "step": 3149
+    },
+    {
+      "epoch": 3.3582089552238807,
+      "grad_norm": 0.13474465507291278,
+      "learning_rate": 2.3521383925131508e-05,
+      "loss": 0.3475,
+      "step": 3150
+    },
+    {
+      "epoch": 3.359275053304904,
+      "grad_norm": 0.14535330585186346,
+      "learning_rate": 2.349426110378435e-05,
+      "loss": 0.349,
+      "step": 3151
+    },
+    {
+      "epoch": 3.3603411513859274,
+      "grad_norm": 0.14304977788811093,
+      "learning_rate": 2.3467147425758735e-05,
+      "loss": 0.3497,
+      "step": 3152
+    },
+    {
+      "epoch": 3.361407249466951,
+      "grad_norm": 0.13088921289643768,
+      "learning_rate": 2.3440042906074187e-05,
+      "loss": 0.3476,
+      "step": 3153
+    },
+    {
+      "epoch": 3.3624733475479744,
+      "grad_norm": 0.13796584925331404,
+      "learning_rate": 2.3412947559745226e-05,
+      "loss": 0.3463,
+      "step": 3154
+    },
+    {
+      "epoch": 3.363539445628998,
+      "grad_norm": 0.1388788022895029,
+      "learning_rate": 2.338586140178127e-05,
+      "loss": 0.3485,
+      "step": 3155
+    },
+    {
+      "epoch": 3.364605543710021,
+      "grad_norm": 0.12618907252498665,
+      "learning_rate": 2.335878444718663e-05,
+      "loss": 0.3547,
+      "step": 3156
+    },
+    {
+      "epoch": 3.3656716417910446,
+      "grad_norm": 0.14196252167717818,
+      "learning_rate": 2.3331716710960536e-05,
+      "loss": 0.3495,
+      "step": 3157
+    },
+    {
+      "epoch": 3.366737739872068,
+      "grad_norm": 0.13807258992426674,
+      "learning_rate": 2.3304658208097105e-05,
+      "loss": 0.3498,
+      "step": 3158
+    },
+    {
+      "epoch": 3.3678038379530917,
+      "grad_norm": 0.12981419614673304,
+      "learning_rate": 2.3277608953585346e-05,
+      "loss": 0.3441,
+      "step": 3159
+    },
+    {
+      "epoch": 3.368869936034115,
+      "grad_norm": 0.16657533516126402,
+      "learning_rate": 2.3250568962409155e-05,
+      "loss": 0.353,
+      "step": 3160
+    },
+    {
+      "epoch": 3.3699360341151388,
+      "grad_norm": 0.12148830285530396,
+      "learning_rate": 2.322353824954725e-05,
+      "loss": 0.3523,
+      "step": 3161
+    },
+    {
+      "epoch": 3.3710021321961623,
+      "grad_norm": 0.1706380629428814,
+      "learning_rate": 2.319651682997325e-05,
+      "loss": 0.3487,
+      "step": 3162
+    },
+    {
+      "epoch": 3.3720682302771854,
+      "grad_norm": 0.14429216152158644,
+      "learning_rate": 2.316950471865564e-05,
+      "loss": 0.3511,
+      "step": 3163
+    },
+    {
+      "epoch": 3.373134328358209,
+      "grad_norm": 0.13448516843067299,
+      "learning_rate": 2.31425019305577e-05,
+      "loss": 0.3542,
+      "step": 3164
+    },
+    {
+      "epoch": 3.3742004264392325,
+      "grad_norm": 0.16508680435233514,
+      "learning_rate": 2.3115508480637575e-05,
+      "loss": 0.3477,
+      "step": 3165
+    },
+    {
+      "epoch": 3.375266524520256,
+      "grad_norm": 0.1326808111603962,
+      "learning_rate": 2.308852438384824e-05,
+      "loss": 0.349,
+      "step": 3166
+    },
+    {
+      "epoch": 3.3763326226012795,
+      "grad_norm": 0.14047757472242012,
+      "learning_rate": 2.3061549655137498e-05,
+      "loss": 0.3513,
+      "step": 3167
+    },
+    {
+      "epoch": 3.3773987206823026,
+      "grad_norm": 0.1441083727611898,
+      "learning_rate": 2.3034584309447913e-05,
+      "loss": 0.3479,
+      "step": 3168
+    },
+    {
+      "epoch": 3.378464818763326,
+      "grad_norm": 0.13591931890052086,
+      "learning_rate": 2.3007628361716902e-05,
+      "loss": 0.3515,
+      "step": 3169
+    },
+    {
+      "epoch": 3.3795309168443497,
+      "grad_norm": 0.1255576121479874,
+      "learning_rate": 2.298068182687666e-05,
+      "loss": 0.3519,
+      "step": 3170
+    },
+    {
+      "epoch": 3.3805970149253732,
+      "grad_norm": 0.14967290050741613,
+      "learning_rate": 2.295374471985418e-05,
+      "loss": 0.3482,
+      "step": 3171
+    },
+    {
+      "epoch": 3.3816631130063968,
+      "grad_norm": 0.12421430145394535,
+      "learning_rate": 2.2926817055571194e-05,
+      "loss": 0.3526,
+      "step": 3172
+    },
+    {
+      "epoch": 3.38272921108742,
+      "grad_norm": 0.17262897437577504,
+      "learning_rate": 2.289989884894425e-05,
+      "loss": 0.3442,
+      "step": 3173
+    },
+    {
+      "epoch": 3.3837953091684434,
+      "grad_norm": 0.14546213806847885,
+      "learning_rate": 2.287299011488461e-05,
+      "loss": 0.3531,
+      "step": 3174
+    },
+    {
+      "epoch": 3.384861407249467,
+      "grad_norm": 0.1574355823375362,
+      "learning_rate": 2.2846090868298333e-05,
+      "loss": 0.3474,
+      "step": 3175
+    },
+    {
+      "epoch": 3.3859275053304905,
+      "grad_norm": 0.15336705073042523,
+      "learning_rate": 2.2819201124086216e-05,
+      "loss": 0.3503,
+      "step": 3176
+    },
+    {
+      "epoch": 3.386993603411514,
+      "grad_norm": 0.13640452357535515,
+      "learning_rate": 2.279232089714374e-05,
+      "loss": 0.3505,
+      "step": 3177
+    },
+    {
+      "epoch": 3.388059701492537,
+      "grad_norm": 0.15004013469173,
+      "learning_rate": 2.2765450202361186e-05,
+      "loss": 0.3519,
+      "step": 3178
+    },
+    {
+      "epoch": 3.3891257995735606,
+      "grad_norm": 0.14151494212433288,
+      "learning_rate": 2.273858905462353e-05,
+      "loss": 0.3518,
+      "step": 3179
+    },
+    {
+      "epoch": 3.390191897654584,
+      "grad_norm": 0.1345161110814909,
+      "learning_rate": 2.2711737468810418e-05,
+      "loss": 0.3481,
+      "step": 3180
+    },
+    {
+      "epoch": 3.3912579957356077,
+      "grad_norm": 0.12660333007474672,
+      "learning_rate": 2.268489545979625e-05,
+      "loss": 0.3484,
+      "step": 3181
+    },
+    {
+      "epoch": 3.3923240938166312,
+      "grad_norm": 0.13060878781996574,
+      "learning_rate": 2.265806304245012e-05,
+      "loss": 0.349,
+      "step": 3182
+    },
+    {
+      "epoch": 3.3933901918976543,
+      "grad_norm": 0.13348764262187832,
+      "learning_rate": 2.263124023163576e-05,
+      "loss": 0.3471,
+      "step": 3183
+    },
+    {
+      "epoch": 3.394456289978678,
+      "grad_norm": 0.12898484952700903,
+      "learning_rate": 2.2604427042211633e-05,
+      "loss": 0.3521,
+      "step": 3184
+    },
+    {
+      "epoch": 3.3955223880597014,
+      "grad_norm": 0.14350540973892006,
+      "learning_rate": 2.2577623489030865e-05,
+      "loss": 0.3485,
+      "step": 3185
+    },
+    {
+      "epoch": 3.396588486140725,
+      "grad_norm": 0.11886459992083558,
+      "learning_rate": 2.25508295869412e-05,
+      "loss": 0.3518,
+      "step": 3186
+    },
+    {
+      "epoch": 3.3976545842217485,
+      "grad_norm": 0.12711816829481948,
+      "learning_rate": 2.2524045350785088e-05,
+      "loss": 0.3508,
+      "step": 3187
+    },
+    {
+      "epoch": 3.398720682302772,
+      "grad_norm": 0.12078532866593578,
+      "learning_rate": 2.2497270795399598e-05,
+      "loss": 0.3519,
+      "step": 3188
+    },
+    {
+      "epoch": 3.399786780383795,
+      "grad_norm": 0.14360930786740936,
+      "learning_rate": 2.2470505935616457e-05,
+      "loss": 0.3496,
+      "step": 3189
+    },
+    {
+      "epoch": 3.4008528784648187,
+      "grad_norm": 0.12963425574071996,
+      "learning_rate": 2.2443750786261976e-05,
+      "loss": 0.3553,
+      "step": 3190
+    },
+    {
+      "epoch": 3.401918976545842,
+      "grad_norm": 0.13049268581835108,
+      "learning_rate": 2.2417005362157135e-05,
+      "loss": 0.3517,
+      "step": 3191
+    },
+    {
+      "epoch": 3.4029850746268657,
+      "grad_norm": 0.15413582639564263,
+      "learning_rate": 2.2390269678117525e-05,
+      "loss": 0.3517,
+      "step": 3192
+    },
+    {
+      "epoch": 3.4040511727078893,
+      "grad_norm": 0.12663844292929444,
+      "learning_rate": 2.2363543748953296e-05,
+      "loss": 0.3466,
+      "step": 3193
+    },
+    {
+      "epoch": 3.405117270788913,
+      "grad_norm": 0.13821512264352942,
+      "learning_rate": 2.2336827589469232e-05,
+      "loss": 0.3523,
+      "step": 3194
+    },
+    {
+      "epoch": 3.406183368869936,
+      "grad_norm": 0.11822096798559724,
+      "learning_rate": 2.2310121214464706e-05,
+      "loss": 0.3439,
+      "step": 3195
+    },
+    {
+      "epoch": 3.4072494669509594,
+      "grad_norm": 0.12664000688547125,
+      "learning_rate": 2.228342463873367e-05,
+      "loss": 0.3566,
+      "step": 3196
+    },
+    {
+      "epoch": 3.408315565031983,
+      "grad_norm": 0.13892075241413948,
+      "learning_rate": 2.2256737877064607e-05,
+      "loss": 0.3459,
+      "step": 3197
+    },
+    {
+      "epoch": 3.4093816631130065,
+      "grad_norm": 0.12083777967142596,
+      "learning_rate": 2.2230060944240623e-05,
+      "loss": 0.3519,
+      "step": 3198
+    },
+    {
+      "epoch": 3.41044776119403,
+      "grad_norm": 0.14125853339995179,
+      "learning_rate": 2.220339385503934e-05,
+      "loss": 0.351,
+      "step": 3199
+    },
+    {
+      "epoch": 3.411513859275053,
+      "grad_norm": 0.13762034609216564,
+      "learning_rate": 2.2176736624232964e-05,
+      "loss": 0.3427,
+      "step": 3200
+    },
+    {
+      "epoch": 3.4125799573560767,
+      "grad_norm": 0.17537017660770718,
+      "learning_rate": 2.2150089266588173e-05,
+      "loss": 0.356,
+      "step": 3201
+    },
+    {
+      "epoch": 3.4136460554371,
+      "grad_norm": 0.13074766627409623,
+      "learning_rate": 2.2123451796866247e-05,
+      "loss": 0.3531,
+      "step": 3202
+    },
+    {
+      "epoch": 3.4147121535181237,
+      "grad_norm": 0.15980968926250733,
+      "learning_rate": 2.2096824229822973e-05,
+      "loss": 0.348,
+      "step": 3203
+    },
+    {
+      "epoch": 3.4157782515991473,
+      "grad_norm": 0.1180753762858572,
+      "learning_rate": 2.2070206580208598e-05,
+      "loss": 0.3505,
+      "step": 3204
+    },
+    {
+      "epoch": 3.4168443496801704,
+      "grad_norm": 0.14397710716700302,
+      "learning_rate": 2.2043598862767937e-05,
+      "loss": 0.3539,
+      "step": 3205
+    },
+    {
+      "epoch": 3.417910447761194,
+      "grad_norm": 0.15418073714942476,
+      "learning_rate": 2.2017001092240288e-05,
+      "loss": 0.349,
+      "step": 3206
+    },
+    {
+      "epoch": 3.4189765458422174,
+      "grad_norm": 0.15081085999678343,
+      "learning_rate": 2.1990413283359447e-05,
+      "loss": 0.3483,
+      "step": 3207
+    },
+    {
+      "epoch": 3.420042643923241,
+      "grad_norm": 0.11082213481326116,
+      "learning_rate": 2.1963835450853646e-05,
+      "loss": 0.3508,
+      "step": 3208
+    },
+    {
+      "epoch": 3.4211087420042645,
+      "grad_norm": 0.15412690956286193,
+      "learning_rate": 2.1937267609445634e-05,
+      "loss": 0.3537,
+      "step": 3209
+    },
+    {
+      "epoch": 3.4221748400852876,
+      "grad_norm": 0.13940387381283997,
+      "learning_rate": 2.191070977385264e-05,
+      "loss": 0.3509,
+      "step": 3210
+    },
+    {
+      "epoch": 3.423240938166311,
+      "grad_norm": 0.12890974324334237,
+      "learning_rate": 2.1884161958786283e-05,
+      "loss": 0.3516,
+      "step": 3211
+    },
+    {
+      "epoch": 3.4243070362473347,
+      "grad_norm": 0.14545024164766443,
+      "learning_rate": 2.1857624178952693e-05,
+      "loss": 0.3447,
+      "step": 3212
+    },
+    {
+      "epoch": 3.425373134328358,
+      "grad_norm": 0.1330648345375272,
+      "learning_rate": 2.1831096449052424e-05,
+      "loss": 0.3448,
+      "step": 3213
+    },
+    {
+      "epoch": 3.4264392324093818,
+      "grad_norm": 0.12671343057546725,
+      "learning_rate": 2.1804578783780465e-05,
+      "loss": 0.3516,
+      "step": 3214
+    },
+    {
+      "epoch": 3.4275053304904053,
+      "grad_norm": 0.14329897961042323,
+      "learning_rate": 2.177807119782621e-05,
+      "loss": 0.3501,
+      "step": 3215
+    },
+    {
+      "epoch": 3.4285714285714284,
+      "grad_norm": 0.1337530920580984,
+      "learning_rate": 2.175157370587348e-05,
+      "loss": 0.3466,
+      "step": 3216
+    },
+    {
+      "epoch": 3.429637526652452,
+      "grad_norm": 0.16912612472386718,
+      "learning_rate": 2.1725086322600526e-05,
+      "loss": 0.3507,
+      "step": 3217
+    },
+    {
+      "epoch": 3.4307036247334755,
+      "grad_norm": 0.14266480292217754,
+      "learning_rate": 2.1698609062679985e-05,
+      "loss": 0.3545,
+      "step": 3218
+    },
+    {
+      "epoch": 3.431769722814499,
+      "grad_norm": 0.14472729674524853,
+      "learning_rate": 2.167214194077886e-05,
+      "loss": 0.3496,
+      "step": 3219
+    },
+    {
+      "epoch": 3.4328358208955225,
+      "grad_norm": 0.16298798619576543,
+      "learning_rate": 2.1645684971558572e-05,
+      "loss": 0.3474,
+      "step": 3220
+    },
+    {
+      "epoch": 3.433901918976546,
+      "grad_norm": 0.1468219073239334,
+      "learning_rate": 2.1619238169674918e-05,
+      "loss": 0.353,
+      "step": 3221
+    },
+    {
+      "epoch": 3.434968017057569,
+      "grad_norm": 0.19187830974162423,
+      "learning_rate": 2.1592801549778034e-05,
+      "loss": 0.3518,
+      "step": 3222
+    },
+    {
+      "epoch": 3.4360341151385927,
+      "grad_norm": 0.17741138849918361,
+      "learning_rate": 2.1566375126512437e-05,
+      "loss": 0.3464,
+      "step": 3223
+    },
+    {
+      "epoch": 3.4371002132196162,
+      "grad_norm": 0.1603692083176825,
+      "learning_rate": 2.1539958914517e-05,
+      "loss": 0.3476,
+      "step": 3224
+    },
+    {
+      "epoch": 3.4381663113006398,
+      "grad_norm": 0.1663268863315761,
+      "learning_rate": 2.151355292842494e-05,
+      "loss": 0.3496,
+      "step": 3225
+    },
+    {
+      "epoch": 3.4392324093816633,
+      "grad_norm": 0.1767760891428736,
+      "learning_rate": 2.1487157182863773e-05,
+      "loss": 0.3465,
+      "step": 3226
+    },
+    {
+      "epoch": 3.4402985074626864,
+      "grad_norm": 0.14846390418886782,
+      "learning_rate": 2.1460771692455388e-05,
+      "loss": 0.3572,
+      "step": 3227
+    },
+    {
+      "epoch": 3.44136460554371,
+      "grad_norm": 0.1805328018495273,
+      "learning_rate": 2.143439647181597e-05,
+      "loss": 0.3521,
+      "step": 3228
+    },
+    {
+      "epoch": 3.4424307036247335,
+      "grad_norm": 0.15018285087651392,
+      "learning_rate": 2.140803153555604e-05,
+      "loss": 0.356,
+      "step": 3229
+    },
+    {
+      "epoch": 3.443496801705757,
+      "grad_norm": 0.16606433943154644,
+      "learning_rate": 2.1381676898280372e-05,
+      "loss": 0.3473,
+      "step": 3230
+    },
+    {
+      "epoch": 3.4445628997867805,
+      "grad_norm": 0.16838391423331878,
+      "learning_rate": 2.135533257458808e-05,
+      "loss": 0.3482,
+      "step": 3231
+    },
+    {
+      "epoch": 3.4456289978678036,
+      "grad_norm": 0.1327187010033925,
+      "learning_rate": 2.1328998579072566e-05,
+      "loss": 0.3562,
+      "step": 3232
+    },
+    {
+      "epoch": 3.446695095948827,
+      "grad_norm": 0.15065705283311742,
+      "learning_rate": 2.130267492632146e-05,
+      "loss": 0.3471,
+      "step": 3233
+    },
+    {
+      "epoch": 3.4477611940298507,
+      "grad_norm": 0.11852053817686456,
+      "learning_rate": 2.1276361630916718e-05,
+      "loss": 0.3542,
+      "step": 3234
+    },
+    {
+      "epoch": 3.4488272921108742,
+      "grad_norm": 0.13152597541802122,
+      "learning_rate": 2.125005870743453e-05,
+      "loss": 0.3539,
+      "step": 3235
+    },
+    {
+      "epoch": 3.449893390191898,
+      "grad_norm": 0.14103478135014386,
+      "learning_rate": 2.1223766170445383e-05,
+      "loss": 0.3508,
+      "step": 3236
+    },
+    {
+      "epoch": 3.450959488272921,
+      "grad_norm": 0.11147740547754476,
+      "learning_rate": 2.1197484034513927e-05,
+      "loss": 0.3453,
+      "step": 3237
+    },
+    {
+      "epoch": 3.4520255863539444,
+      "grad_norm": 0.12515850742634446,
+      "learning_rate": 2.1171212314199117e-05,
+      "loss": 0.3512,
+      "step": 3238
+    },
+    {
+      "epoch": 3.453091684434968,
+      "grad_norm": 0.12854446210826828,
+      "learning_rate": 2.1144951024054144e-05,
+      "loss": 0.3471,
+      "step": 3239
+    },
+    {
+      "epoch": 3.4541577825159915,
+      "grad_norm": 0.12150324184398577,
+      "learning_rate": 2.111870017862636e-05,
+      "loss": 0.3472,
+      "step": 3240
+    },
+    {
+      "epoch": 3.455223880597015,
+      "grad_norm": 0.1402826179243876,
+      "learning_rate": 2.1092459792457384e-05,
+      "loss": 0.355,
+      "step": 3241
+    },
+    {
+      "epoch": 3.4562899786780386,
+      "grad_norm": 0.12712753288646708,
+      "learning_rate": 2.1066229880083035e-05,
+      "loss": 0.35,
+      "step": 3242
+    },
+    {
+      "epoch": 3.4573560767590616,
+      "grad_norm": 0.13839576676511672,
+      "learning_rate": 2.104001045603333e-05,
+      "loss": 0.3527,
+      "step": 3243
+    },
+    {
+      "epoch": 3.458422174840085,
+      "grad_norm": 0.13644673504349322,
+      "learning_rate": 2.1013801534832434e-05,
+      "loss": 0.3509,
+      "step": 3244
+    },
+    {
+      "epoch": 3.4594882729211087,
+      "grad_norm": 0.14306136604732483,
+      "learning_rate": 2.0987603130998745e-05,
+      "loss": 0.3507,
+      "step": 3245
+    },
+    {
+      "epoch": 3.4605543710021323,
+      "grad_norm": 0.15950680626735908,
+      "learning_rate": 2.096141525904484e-05,
+      "loss": 0.3485,
+      "step": 3246
+    },
+    {
+      "epoch": 3.461620469083156,
+      "grad_norm": 0.11275460669446143,
+      "learning_rate": 2.09352379334774e-05,
+      "loss": 0.3529,
+      "step": 3247
+    },
+    {
+      "epoch": 3.4626865671641793,
+      "grad_norm": 0.1745871935155396,
+      "learning_rate": 2.0909071168797332e-05,
+      "loss": 0.3509,
+      "step": 3248
+    },
+    {
+      "epoch": 3.4637526652452024,
+      "grad_norm": 0.12473410126901693,
+      "learning_rate": 2.0882914979499635e-05,
+      "loss": 0.353,
+      "step": 3249
+    },
+    {
+      "epoch": 3.464818763326226,
+      "grad_norm": 0.15474894939169337,
+      "learning_rate": 2.0856769380073497e-05,
+      "loss": 0.3544,
+      "step": 3250
+    },
+    {
+      "epoch": 3.4658848614072495,
+      "grad_norm": 0.14452008614896242,
+      "learning_rate": 2.0830634385002234e-05,
+      "loss": 0.3528,
+      "step": 3251
+    },
+    {
+      "epoch": 3.466950959488273,
+      "grad_norm": 0.1418673442984818,
+      "learning_rate": 2.080451000876325e-05,
+      "loss": 0.3476,
+      "step": 3252
+    },
+    {
+      "epoch": 3.4680170575692966,
+      "grad_norm": 0.16943424994947526,
+      "learning_rate": 2.0778396265828097e-05,
+      "loss": 0.356,
+      "step": 3253
+    },
+    {
+      "epoch": 3.4690831556503197,
+      "grad_norm": 0.14164905105719902,
+      "learning_rate": 2.0752293170662453e-05,
+      "loss": 0.3492,
+      "step": 3254
+    },
+    {
+      "epoch": 3.470149253731343,
+      "grad_norm": 0.17297216086140343,
+      "learning_rate": 2.0726200737726044e-05,
+      "loss": 0.3456,
+      "step": 3255
+    },
+    {
+      "epoch": 3.4712153518123667,
+      "grad_norm": 0.13358545880627057,
+      "learning_rate": 2.0700118981472737e-05,
+      "loss": 0.35,
+      "step": 3256
+    },
+    {
+      "epoch": 3.4722814498933903,
+      "grad_norm": 0.130949084809917,
+      "learning_rate": 2.0674047916350472e-05,
+      "loss": 0.3485,
+      "step": 3257
+    },
+    {
+      "epoch": 3.473347547974414,
+      "grad_norm": 0.15140183416957245,
+      "learning_rate": 2.0647987556801276e-05,
+      "loss": 0.3488,
+      "step": 3258
+    },
+    {
+      "epoch": 3.474413646055437,
+      "grad_norm": 0.13537402309349822,
+      "learning_rate": 2.0621937917261202e-05,
+      "loss": 0.3526,
+      "step": 3259
+    },
+    {
+      "epoch": 3.4754797441364604,
+      "grad_norm": 0.12927192577901372,
+      "learning_rate": 2.059589901216042e-05,
+      "loss": 0.3578,
+      "step": 3260
+    },
+    {
+      "epoch": 3.476545842217484,
+      "grad_norm": 0.13927688677691658,
+      "learning_rate": 2.0569870855923133e-05,
+      "loss": 0.3514,
+      "step": 3261
+    },
+    {
+      "epoch": 3.4776119402985075,
+      "grad_norm": 0.1446653742467127,
+      "learning_rate": 2.0543853462967568e-05,
+      "loss": 0.3476,
+      "step": 3262
+    },
+    {
+      "epoch": 3.478678038379531,
+      "grad_norm": 0.14655142804794563,
+      "learning_rate": 2.0517846847706018e-05,
+      "loss": 0.3521,
+      "step": 3263
+    },
+    {
+      "epoch": 3.479744136460554,
+      "grad_norm": 0.15775326994652664,
+      "learning_rate": 2.0491851024544798e-05,
+      "loss": 0.3515,
+      "step": 3264
+    },
+    {
+      "epoch": 3.4808102345415777,
+      "grad_norm": 0.14124595672889267,
+      "learning_rate": 2.0465866007884254e-05,
+      "loss": 0.3519,
+      "step": 3265
+    },
+    {
+      "epoch": 3.481876332622601,
+      "grad_norm": 0.14159196279556782,
+      "learning_rate": 2.0439891812118713e-05,
+      "loss": 0.3467,
+      "step": 3266
+    },
+    {
+      "epoch": 3.4829424307036247,
+      "grad_norm": 0.15538385639162236,
+      "learning_rate": 2.0413928451636532e-05,
+      "loss": 0.3478,
+      "step": 3267
+    },
+    {
+      "epoch": 3.4840085287846483,
+      "grad_norm": 0.12201107022654291,
+      "learning_rate": 2.038797594082009e-05,
+      "loss": 0.3468,
+      "step": 3268
+    },
+    {
+      "epoch": 3.485074626865672,
+      "grad_norm": 0.13632684112183946,
+      "learning_rate": 2.0362034294045694e-05,
+      "loss": 0.3461,
+      "step": 3269
+    },
+    {
+      "epoch": 3.486140724946695,
+      "grad_norm": 0.14100032656692765,
+      "learning_rate": 2.0336103525683685e-05,
+      "loss": 0.3519,
+      "step": 3270
+    },
+    {
+      "epoch": 3.4872068230277184,
+      "grad_norm": 0.13444787195465405,
+      "learning_rate": 2.0310183650098357e-05,
+      "loss": 0.3495,
+      "step": 3271
+    },
+    {
+      "epoch": 3.488272921108742,
+      "grad_norm": 0.12769235515457522,
+      "learning_rate": 2.0284274681647993e-05,
+      "loss": 0.3453,
+      "step": 3272
+    },
+    {
+      "epoch": 3.4893390191897655,
+      "grad_norm": 0.12220035736690575,
+      "learning_rate": 2.0258376634684786e-05,
+      "loss": 0.3496,
+      "step": 3273
+    },
+    {
+      "epoch": 3.490405117270789,
+      "grad_norm": 0.1568430272842312,
+      "learning_rate": 2.023248952355492e-05,
+      "loss": 0.354,
+      "step": 3274
+    },
+    {
+      "epoch": 3.4914712153518126,
+      "grad_norm": 0.11426477449378872,
+      "learning_rate": 2.0206613362598507e-05,
+      "loss": 0.3476,
+      "step": 3275
+    },
+    {
+      "epoch": 3.4925373134328357,
+      "grad_norm": 0.15450922714432697,
+      "learning_rate": 2.018074816614962e-05,
+      "loss": 0.351,
+      "step": 3276
+    },
+    {
+      "epoch": 3.4936034115138592,
+      "grad_norm": 0.13239244497688343,
+      "learning_rate": 2.0154893948536195e-05,
+      "loss": 0.3516,
+      "step": 3277
+    },
+    {
+      "epoch": 3.4946695095948828,
+      "grad_norm": 0.1591598013421958,
+      "learning_rate": 2.0129050724080138e-05,
+      "loss": 0.3499,
+      "step": 3278
+    },
+    {
+      "epoch": 3.4957356076759063,
+      "grad_norm": 0.15772096557717954,
+      "learning_rate": 2.0103218507097274e-05,
+      "loss": 0.3497,
+      "step": 3279
+    },
+    {
+      "epoch": 3.49680170575693,
+      "grad_norm": 0.12425309706820227,
+      "learning_rate": 2.0077397311897274e-05,
+      "loss": 0.3474,
+      "step": 3280
+    },
+    {
+      "epoch": 3.497867803837953,
+      "grad_norm": 0.1506011178175042,
+      "learning_rate": 2.005158715278376e-05,
+      "loss": 0.3472,
+      "step": 3281
+    },
+    {
+      "epoch": 3.4989339019189765,
+      "grad_norm": 0.11149966822962049,
+      "learning_rate": 2.0025788044054212e-05,
+      "loss": 0.3504,
+      "step": 3282
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 0.15273982349883355,
+      "learning_rate": 2.0000000000000012e-05,
+      "loss": 0.3485,
+      "step": 3283
+    },
+    {
+      "epoch": 3.5010660980810235,
+      "grad_norm": 0.12108827495133537,
+      "learning_rate": 1.9974223034906362e-05,
+      "loss": 0.3521,
+      "step": 3284
+    },
+    {
+      "epoch": 3.502132196162047,
+      "grad_norm": 0.14237030818834184,
+      "learning_rate": 1.9948457163052385e-05,
+      "loss": 0.3519,
+      "step": 3285
+    },
+    {
+      "epoch": 3.50319829424307,
+      "grad_norm": 0.11471865230712094,
+      "learning_rate": 1.9922702398711026e-05,
+      "loss": 0.3536,
+      "step": 3286
+    },
+    {
+      "epoch": 3.5042643923240937,
+      "grad_norm": 0.13762079947986805,
+      "learning_rate": 1.989695875614911e-05,
+      "loss": 0.3496,
+      "step": 3287
+    },
+    {
+      "epoch": 3.5053304904051172,
+      "grad_norm": 0.12762020300370305,
+      "learning_rate": 1.987122624962724e-05,
+      "loss": 0.3483,
+      "step": 3288
+    },
+    {
+      "epoch": 3.5063965884861408,
+      "grad_norm": 0.13891339336483755,
+      "learning_rate": 1.9845504893399906e-05,
+      "loss": 0.3494,
+      "step": 3289
+    },
+    {
+      "epoch": 3.5074626865671643,
+      "grad_norm": 0.12926973335164987,
+      "learning_rate": 1.9819794701715412e-05,
+      "loss": 0.3489,
+      "step": 3290
+    },
+    {
+      "epoch": 3.5085287846481874,
+      "grad_norm": 0.11926976710182068,
+      "learning_rate": 1.9794095688815846e-05,
+      "loss": 0.3496,
+      "step": 3291
+    },
+    {
+      "epoch": 3.509594882729211,
+      "grad_norm": 0.1169531192619173,
+      "learning_rate": 1.9768407868937136e-05,
+      "loss": 0.3536,
+      "step": 3292
+    },
+    {
+      "epoch": 3.5106609808102345,
+      "grad_norm": 0.11709738090603788,
+      "learning_rate": 1.9742731256308997e-05,
+      "loss": 0.3505,
+      "step": 3293
+    },
+    {
+      "epoch": 3.511727078891258,
+      "grad_norm": 0.11720868076698791,
+      "learning_rate": 1.9717065865154962e-05,
+      "loss": 0.3555,
+      "step": 3294
+    },
+    {
+      "epoch": 3.5127931769722816,
+      "grad_norm": 0.11598807611455429,
+      "learning_rate": 1.969141170969228e-05,
+      "loss": 0.3492,
+      "step": 3295
+    },
+    {
+      "epoch": 3.5138592750533046,
+      "grad_norm": 0.11677301458081207,
+      "learning_rate": 1.9665768804132046e-05,
+      "loss": 0.3513,
+      "step": 3296
+    },
+    {
+      "epoch": 3.5149253731343286,
+      "grad_norm": 0.11769361312971446,
+      "learning_rate": 1.9640137162679108e-05,
+      "loss": 0.358,
+      "step": 3297
+    },
+    {
+      "epoch": 3.5159914712153517,
+      "grad_norm": 0.11945244360571523,
+      "learning_rate": 1.9614516799532035e-05,
+      "loss": 0.3508,
+      "step": 3298
+    },
+    {
+      "epoch": 3.5170575692963753,
+      "grad_norm": 0.10553148957707399,
+      "learning_rate": 1.9588907728883192e-05,
+      "loss": 0.3527,
+      "step": 3299
+    },
+    {
+      "epoch": 3.518123667377399,
+      "grad_norm": 0.12506068947753565,
+      "learning_rate": 1.9563309964918664e-05,
+      "loss": 0.3525,
+      "step": 3300
+    },
+    {
+      "epoch": 3.5191897654584223,
+      "grad_norm": 0.1274298809654222,
+      "learning_rate": 1.9537723521818307e-05,
+      "loss": 0.3586,
+      "step": 3301
+    },
+    {
+      "epoch": 3.520255863539446,
+      "grad_norm": 0.11963014257012117,
+      "learning_rate": 1.9512148413755653e-05,
+      "loss": 0.3554,
+      "step": 3302
+    },
+    {
+      "epoch": 3.521321961620469,
+      "grad_norm": 0.1426495501126215,
+      "learning_rate": 1.9486584654897987e-05,
+      "loss": 0.3544,
+      "step": 3303
+    },
+    {
+      "epoch": 3.5223880597014925,
+      "grad_norm": 0.11378394285714087,
+      "learning_rate": 1.9461032259406317e-05,
+      "loss": 0.357,
+      "step": 3304
+    },
+    {
+      "epoch": 3.523454157782516,
+      "grad_norm": 0.13151158677780098,
+      "learning_rate": 1.9435491241435343e-05,
+      "loss": 0.3504,
+      "step": 3305
+    },
+    {
+      "epoch": 3.5245202558635396,
+      "grad_norm": 0.11047177913070025,
+      "learning_rate": 1.9409961615133435e-05,
+      "loss": 0.3449,
+      "step": 3306
+    },
+    {
+      "epoch": 3.525586353944563,
+      "grad_norm": 0.14518194022909464,
+      "learning_rate": 1.9384443394642697e-05,
+      "loss": 0.3537,
+      "step": 3307
+    },
+    {
+      "epoch": 3.526652452025586,
+      "grad_norm": 0.13288265424403722,
+      "learning_rate": 1.9358936594098915e-05,
+      "loss": 0.3472,
+      "step": 3308
+    },
+    {
+      "epoch": 3.5277185501066097,
+      "grad_norm": 0.13481285136461793,
+      "learning_rate": 1.9333441227631494e-05,
+      "loss": 0.3522,
+      "step": 3309
+    },
+    {
+      "epoch": 3.5287846481876333,
+      "grad_norm": 0.12955597370536356,
+      "learning_rate": 1.9307957309363562e-05,
+      "loss": 0.3591,
+      "step": 3310
+    },
+    {
+      "epoch": 3.529850746268657,
+      "grad_norm": 0.13070491932515202,
+      "learning_rate": 1.928248485341188e-05,
+      "loss": 0.3523,
+      "step": 3311
+    },
+    {
+      "epoch": 3.5309168443496803,
+      "grad_norm": 0.12150544834528254,
+      "learning_rate": 1.9257023873886885e-05,
+      "loss": 0.3548,
+      "step": 3312
+    },
+    {
+      "epoch": 3.5319829424307034,
+      "grad_norm": 0.12389814641248816,
+      "learning_rate": 1.9231574384892608e-05,
+      "loss": 0.3491,
+      "step": 3313
+    },
+    {
+      "epoch": 3.533049040511727,
+      "grad_norm": 0.11812191547579137,
+      "learning_rate": 1.9206136400526753e-05,
+      "loss": 0.3552,
+      "step": 3314
+    },
+    {
+      "epoch": 3.5341151385927505,
+      "grad_norm": 0.11705609586266169,
+      "learning_rate": 1.9180709934880657e-05,
+      "loss": 0.3567,
+      "step": 3315
+    },
+    {
+      "epoch": 3.535181236673774,
+      "grad_norm": 0.12705382222301062,
+      "learning_rate": 1.915529500203923e-05,
+      "loss": 0.3522,
+      "step": 3316
+    },
+    {
+      "epoch": 3.5362473347547976,
+      "grad_norm": 0.1195951170844615,
+      "learning_rate": 1.9129891616081045e-05,
+      "loss": 0.3459,
+      "step": 3317
+    },
+    {
+      "epoch": 3.5373134328358207,
+      "grad_norm": 0.11156152984847112,
+      "learning_rate": 1.910449979107827e-05,
+      "loss": 0.3524,
+      "step": 3318
+    },
+    {
+      "epoch": 3.538379530916844,
+      "grad_norm": 0.12607688725487942,
+      "learning_rate": 1.907911954109662e-05,
+      "loss": 0.3505,
+      "step": 3319
+    },
+    {
+      "epoch": 3.5394456289978677,
+      "grad_norm": 0.12385238816738517,
+      "learning_rate": 1.9053750880195453e-05,
+      "loss": 0.353,
+      "step": 3320
+    },
+    {
+      "epoch": 3.5405117270788913,
+      "grad_norm": 0.11986136600785428,
+      "learning_rate": 1.9028393822427707e-05,
+      "loss": 0.3534,
+      "step": 3321
+    },
+    {
+      "epoch": 3.541577825159915,
+      "grad_norm": 0.1306944687194926,
+      "learning_rate": 1.900304838183984e-05,
+      "loss": 0.3451,
+      "step": 3322
+    },
+    {
+      "epoch": 3.542643923240938,
+      "grad_norm": 0.12521720177645257,
+      "learning_rate": 1.8977714572471942e-05,
+      "loss": 0.3463,
+      "step": 3323
+    },
+    {
+      "epoch": 3.543710021321962,
+      "grad_norm": 0.14676298985824088,
+      "learning_rate": 1.8952392408357596e-05,
+      "loss": 0.348,
+      "step": 3324
+    },
+    {
+      "epoch": 3.544776119402985,
+      "grad_norm": 0.1270324568921544,
+      "learning_rate": 1.892708190352398e-05,
+      "loss": 0.3469,
+      "step": 3325
+    },
+    {
+      "epoch": 3.5458422174840085,
+      "grad_norm": 0.13384984086554236,
+      "learning_rate": 1.890178307199181e-05,
+      "loss": 0.3586,
+      "step": 3326
+    },
+    {
+      "epoch": 3.546908315565032,
+      "grad_norm": 0.1403291759377033,
+      "learning_rate": 1.88764959277753e-05,
+      "loss": 0.3544,
+      "step": 3327
+    },
+    {
+      "epoch": 3.5479744136460556,
+      "grad_norm": 0.1578563606515396,
+      "learning_rate": 1.8851220484882223e-05,
+      "loss": 0.3484,
+      "step": 3328
+    },
+    {
+      "epoch": 3.549040511727079,
+      "grad_norm": 0.1332388726289066,
+      "learning_rate": 1.8825956757313864e-05,
+      "loss": 0.3546,
+      "step": 3329
+    },
+    {
+      "epoch": 3.550106609808102,
+      "grad_norm": 0.14974892308226678,
+      "learning_rate": 1.8800704759065027e-05,
+      "loss": 0.3465,
+      "step": 3330
+    },
+    {
+      "epoch": 3.5511727078891258,
+      "grad_norm": 0.13311408301746736,
+      "learning_rate": 1.877546450412398e-05,
+      "loss": 0.3425,
+      "step": 3331
+    },
+    {
+      "epoch": 3.5522388059701493,
+      "grad_norm": 0.1486101241297523,
+      "learning_rate": 1.8750236006472525e-05,
+      "loss": 0.3448,
+      "step": 3332
+    },
+    {
+      "epoch": 3.553304904051173,
+      "grad_norm": 0.126555084429222,
+      "learning_rate": 1.872501928008594e-05,
+      "loss": 0.3476,
+      "step": 3333
+    },
+    {
+      "epoch": 3.5543710021321964,
+      "grad_norm": 0.1288483891538286,
+      "learning_rate": 1.8699814338933e-05,
+      "loss": 0.3517,
+      "step": 3334
+    },
+    {
+      "epoch": 3.5554371002132195,
+      "grad_norm": 0.13278424016806864,
+      "learning_rate": 1.8674621196975892e-05,
+      "loss": 0.352,
+      "step": 3335
+    },
+    {
+      "epoch": 3.556503198294243,
+      "grad_norm": 0.13632340052187167,
+      "learning_rate": 1.864943986817033e-05,
+      "loss": 0.3486,
+      "step": 3336
+    },
+    {
+      "epoch": 3.5575692963752665,
+      "grad_norm": 0.12666253550466774,
+      "learning_rate": 1.8624270366465476e-05,
+      "loss": 0.351,
+      "step": 3337
+    },
+    {
+      "epoch": 3.55863539445629,
+      "grad_norm": 0.13872865389172592,
+      "learning_rate": 1.8599112705803894e-05,
+      "loss": 0.3546,
+      "step": 3338
+    },
+    {
+      "epoch": 3.5597014925373136,
+      "grad_norm": 0.11819669860101821,
+      "learning_rate": 1.857396690012163e-05,
+      "loss": 0.3496,
+      "step": 3339
+    },
+    {
+      "epoch": 3.5607675906183367,
+      "grad_norm": 0.1324098142656928,
+      "learning_rate": 1.8548832963348167e-05,
+      "loss": 0.3512,
+      "step": 3340
+    },
+    {
+      "epoch": 3.5618336886993602,
+      "grad_norm": 0.1213227275635731,
+      "learning_rate": 1.8523710909406408e-05,
+      "loss": 0.3529,
+      "step": 3341
+    },
+    {
+      "epoch": 3.5628997867803838,
+      "grad_norm": 0.13139162851354952,
+      "learning_rate": 1.8498600752212633e-05,
+      "loss": 0.3463,
+      "step": 3342
+    },
+    {
+      "epoch": 3.5639658848614073,
+      "grad_norm": 0.12022593351799091,
+      "learning_rate": 1.847350250567658e-05,
+      "loss": 0.349,
+      "step": 3343
+    },
+    {
+      "epoch": 3.565031982942431,
+      "grad_norm": 0.14848931387258837,
+      "learning_rate": 1.8448416183701387e-05,
+      "loss": 0.3523,
+      "step": 3344
+    },
+    {
+      "epoch": 3.566098081023454,
+      "grad_norm": 0.13293877643754679,
+      "learning_rate": 1.8423341800183547e-05,
+      "loss": 0.3548,
+      "step": 3345
+    },
+    {
+      "epoch": 3.5671641791044775,
+      "grad_norm": 0.13819112943650344,
+      "learning_rate": 1.8398279369012975e-05,
+      "loss": 0.3586,
+      "step": 3346
+    },
+    {
+      "epoch": 3.568230277185501,
+      "grad_norm": 0.15221572285365978,
+      "learning_rate": 1.8373228904072958e-05,
+      "loss": 0.3487,
+      "step": 3347
+    },
+    {
+      "epoch": 3.5692963752665245,
+      "grad_norm": 0.12335264810660222,
+      "learning_rate": 1.8348190419240168e-05,
+      "loss": 0.3576,
+      "step": 3348
+    },
+    {
+      "epoch": 3.570362473347548,
+      "grad_norm": 0.14402755325400285,
+      "learning_rate": 1.8323163928384597e-05,
+      "loss": 0.3557,
+      "step": 3349
+    },
+    {
+      "epoch": 3.571428571428571,
+      "grad_norm": 0.121086447817541,
+      "learning_rate": 1.829814944536963e-05,
+      "loss": 0.3545,
+      "step": 3350
+    },
+    {
+      "epoch": 3.572494669509595,
+      "grad_norm": 0.1456232992893829,
+      "learning_rate": 1.8273146984051987e-05,
+      "loss": 0.3508,
+      "step": 3351
+    },
+    {
+      "epoch": 3.5735607675906182,
+      "grad_norm": 0.13621927041516765,
+      "learning_rate": 1.8248156558281756e-05,
+      "loss": 0.3479,
+      "step": 3352
+    },
+    {
+      "epoch": 3.574626865671642,
+      "grad_norm": 0.12188450870001802,
+      "learning_rate": 1.8223178181902296e-05,
+      "loss": 0.3514,
+      "step": 3353
+    },
+    {
+      "epoch": 3.5756929637526653,
+      "grad_norm": 0.13329743005861297,
+      "learning_rate": 1.8198211868750352e-05,
+      "loss": 0.3495,
+      "step": 3354
+    },
+    {
+      "epoch": 3.576759061833689,
+      "grad_norm": 0.118410852419534,
+      "learning_rate": 1.8173257632655973e-05,
+      "loss": 0.3482,
+      "step": 3355
+    },
+    {
+      "epoch": 3.5778251599147124,
+      "grad_norm": 0.12392097571524742,
+      "learning_rate": 1.8148315487442482e-05,
+      "loss": 0.3509,
+      "step": 3356
+    },
+    {
+      "epoch": 3.5788912579957355,
+      "grad_norm": 0.149797120649777,
+      "learning_rate": 1.8123385446926546e-05,
+      "loss": 0.351,
+      "step": 3357
+    },
+    {
+      "epoch": 3.579957356076759,
+      "grad_norm": 0.11637490638409546,
+      "learning_rate": 1.8098467524918114e-05,
+      "loss": 0.348,
+      "step": 3358
+    },
+    {
+      "epoch": 3.5810234541577826,
+      "grad_norm": 0.1302506928123163,
+      "learning_rate": 1.807356173522043e-05,
+      "loss": 0.3473,
+      "step": 3359
+    },
+    {
+      "epoch": 3.582089552238806,
+      "grad_norm": 0.11734314593339212,
+      "learning_rate": 1.8048668091629976e-05,
+      "loss": 0.3556,
+      "step": 3360
+    },
+    {
+      "epoch": 3.5831556503198296,
+      "grad_norm": 0.13699195492761537,
+      "learning_rate": 1.802378660793656e-05,
+      "loss": 0.3509,
+      "step": 3361
+    },
+    {
+      "epoch": 3.5842217484008527,
+      "grad_norm": 0.1192014977364567,
+      "learning_rate": 1.7998917297923224e-05,
+      "loss": 0.3471,
+      "step": 3362
+    },
+    {
+      "epoch": 3.5852878464818763,
+      "grad_norm": 0.1296315171549434,
+      "learning_rate": 1.7974060175366288e-05,
+      "loss": 0.3465,
+      "step": 3363
+    },
+    {
+      "epoch": 3.5863539445629,
+      "grad_norm": 0.13858112246619184,
+      "learning_rate": 1.794921525403528e-05,
+      "loss": 0.3497,
+      "step": 3364
+    },
+    {
+      "epoch": 3.5874200426439233,
+      "grad_norm": 0.1119373071168273,
+      "learning_rate": 1.7924382547693006e-05,
+      "loss": 0.3545,
+      "step": 3365
+    },
+    {
+      "epoch": 3.588486140724947,
+      "grad_norm": 0.13479994066024156,
+      "learning_rate": 1.7899562070095517e-05,
+      "loss": 0.3459,
+      "step": 3366
+    },
+    {
+      "epoch": 3.58955223880597,
+      "grad_norm": 0.11961904622503035,
+      "learning_rate": 1.7874753834992023e-05,
+      "loss": 0.3486,
+      "step": 3367
+    },
+    {
+      "epoch": 3.5906183368869935,
+      "grad_norm": 0.1295577113450289,
+      "learning_rate": 1.7849957856125032e-05,
+      "loss": 0.3437,
+      "step": 3368
+    },
+    {
+      "epoch": 3.591684434968017,
+      "grad_norm": 0.13384088967192623,
+      "learning_rate": 1.7825174147230212e-05,
+      "loss": 0.3508,
+      "step": 3369
+    },
+    {
+      "epoch": 3.5927505330490406,
+      "grad_norm": 0.14860888147610143,
+      "learning_rate": 1.780040272203647e-05,
+      "loss": 0.3456,
+      "step": 3370
+    },
+    {
+      "epoch": 3.593816631130064,
+      "grad_norm": 0.12817477422964021,
+      "learning_rate": 1.7775643594265858e-05,
+      "loss": 0.3502,
+      "step": 3371
+    },
+    {
+      "epoch": 3.594882729211087,
+      "grad_norm": 0.15577728782871172,
+      "learning_rate": 1.775089677763366e-05,
+      "loss": 0.3537,
+      "step": 3372
+    },
+    {
+      "epoch": 3.5959488272921107,
+      "grad_norm": 0.12786761103356425,
+      "learning_rate": 1.7726162285848343e-05,
+      "loss": 0.3447,
+      "step": 3373
+    },
+    {
+      "epoch": 3.5970149253731343,
+      "grad_norm": 0.14028709617612561,
+      "learning_rate": 1.77014401326115e-05,
+      "loss": 0.3463,
+      "step": 3374
+    },
+    {
+      "epoch": 3.598081023454158,
+      "grad_norm": 0.12954870341900906,
+      "learning_rate": 1.7676730331617934e-05,
+      "loss": 0.3519,
+      "step": 3375
+    },
+    {
+      "epoch": 3.5991471215351813,
+      "grad_norm": 0.12691468222674968,
+      "learning_rate": 1.765203289655559e-05,
+      "loss": 0.3491,
+      "step": 3376
+    },
+    {
+      "epoch": 3.6002132196162044,
+      "grad_norm": 0.15452666879821078,
+      "learning_rate": 1.7627347841105575e-05,
+      "loss": 0.3469,
+      "step": 3377
+    },
+    {
+      "epoch": 3.6012793176972284,
+      "grad_norm": 0.12502978476826676,
+      "learning_rate": 1.7602675178942102e-05,
+      "loss": 0.3441,
+      "step": 3378
+    },
+    {
+      "epoch": 3.6023454157782515,
+      "grad_norm": 0.13006787492356886,
+      "learning_rate": 1.7578014923732558e-05,
+      "loss": 0.3501,
+      "step": 3379
+    },
+    {
+      "epoch": 3.603411513859275,
+      "grad_norm": 0.10724477831469756,
+      "learning_rate": 1.7553367089137438e-05,
+      "loss": 0.3501,
+      "step": 3380
+    },
+    {
+      "epoch": 3.6044776119402986,
+      "grad_norm": 0.13170260076841805,
+      "learning_rate": 1.7528731688810383e-05,
+      "loss": 0.3505,
+      "step": 3381
+    },
+    {
+      "epoch": 3.605543710021322,
+      "grad_norm": 0.11121721943362509,
+      "learning_rate": 1.7504108736398087e-05,
+      "loss": 0.3523,
+      "step": 3382
+    },
+    {
+      "epoch": 3.6066098081023457,
+      "grad_norm": 0.14589447941044803,
+      "learning_rate": 1.747949824554041e-05,
+      "loss": 0.3546,
+      "step": 3383
+    },
+    {
+      "epoch": 3.6076759061833688,
+      "grad_norm": 0.13319282485716943,
+      "learning_rate": 1.745490022987029e-05,
+      "loss": 0.3496,
+      "step": 3384
+    },
+    {
+      "epoch": 3.6087420042643923,
+      "grad_norm": 0.12337788148881837,
+      "learning_rate": 1.7430314703013727e-05,
+      "loss": 0.3504,
+      "step": 3385
+    },
+    {
+      "epoch": 3.609808102345416,
+      "grad_norm": 0.14541264966288076,
+      "learning_rate": 1.7405741678589838e-05,
+      "loss": 0.3507,
+      "step": 3386
+    },
+    {
+      "epoch": 3.6108742004264394,
+      "grad_norm": 0.11959380106376388,
+      "learning_rate": 1.7381181170210814e-05,
+      "loss": 0.3511,
+      "step": 3387
+    },
+    {
+      "epoch": 3.611940298507463,
+      "grad_norm": 0.1245152479458827,
+      "learning_rate": 1.7356633191481877e-05,
+      "loss": 0.3469,
+      "step": 3388
+    },
+    {
+      "epoch": 3.613006396588486,
+      "grad_norm": 0.10647154581436648,
+      "learning_rate": 1.7332097756001335e-05,
+      "loss": 0.3512,
+      "step": 3389
+    },
+    {
+      "epoch": 3.6140724946695095,
+      "grad_norm": 0.11205355107663482,
+      "learning_rate": 1.730757487736057e-05,
+      "loss": 0.3449,
+      "step": 3390
+    },
+    {
+      "epoch": 3.615138592750533,
+      "grad_norm": 0.1071566104296763,
+      "learning_rate": 1.7283064569143947e-05,
+      "loss": 0.349,
+      "step": 3391
+    },
+    {
+      "epoch": 3.6162046908315566,
+      "grad_norm": 0.10713350170160939,
+      "learning_rate": 1.7258566844928915e-05,
+      "loss": 0.3495,
+      "step": 3392
+    },
+    {
+      "epoch": 3.61727078891258,
+      "grad_norm": 0.11331625037340512,
+      "learning_rate": 1.7234081718285965e-05,
+      "loss": 0.351,
+      "step": 3393
+    },
+    {
+      "epoch": 3.6183368869936032,
+      "grad_norm": 0.115641631530902,
+      "learning_rate": 1.7209609202778542e-05,
+      "loss": 0.3487,
+      "step": 3394
+    },
+    {
+      "epoch": 3.6194029850746268,
+      "grad_norm": 0.12437375198658421,
+      "learning_rate": 1.7185149311963186e-05,
+      "loss": 0.3513,
+      "step": 3395
+    },
+    {
+      "epoch": 3.6204690831556503,
+      "grad_norm": 0.11954602193026391,
+      "learning_rate": 1.716070205938938e-05,
+      "loss": 0.3523,
+      "step": 3396
+    },
+    {
+      "epoch": 3.621535181236674,
+      "grad_norm": 0.1310194394078668,
+      "learning_rate": 1.7136267458599633e-05,
+      "loss": 0.3495,
+      "step": 3397
+    },
+    {
+      "epoch": 3.6226012793176974,
+      "grad_norm": 0.11894966733116948,
+      "learning_rate": 1.711184552312945e-05,
+      "loss": 0.3491,
+      "step": 3398
+    },
+    {
+      "epoch": 3.6236673773987205,
+      "grad_norm": 0.1260933584074009,
+      "learning_rate": 1.7087436266507333e-05,
+      "loss": 0.3516,
+      "step": 3399
+    },
+    {
+      "epoch": 3.624733475479744,
+      "grad_norm": 0.11495853607822465,
+      "learning_rate": 1.706303970225471e-05,
+      "loss": 0.3507,
+      "step": 3400
+    },
+    {
+      "epoch": 3.6257995735607675,
+      "grad_norm": 0.1580722155106881,
+      "learning_rate": 1.703865584388602e-05,
+      "loss": 0.349,
+      "step": 3401
+    },
+    {
+      "epoch": 3.626865671641791,
+      "grad_norm": 0.1166417464300119,
+      "learning_rate": 1.7014284704908673e-05,
+      "loss": 0.3504,
+      "step": 3402
+    },
+    {
+      "epoch": 3.6279317697228146,
+      "grad_norm": 0.13094944084926075,
+      "learning_rate": 1.6989926298822977e-05,
+      "loss": 0.3517,
+      "step": 3403
+    },
+    {
+      "epoch": 3.6289978678038377,
+      "grad_norm": 0.1091018787371315,
+      "learning_rate": 1.6965580639122247e-05,
+      "loss": 0.347,
+      "step": 3404
+    },
+    {
+      "epoch": 3.6300639658848612,
+      "grad_norm": 0.13573951697931824,
+      "learning_rate": 1.694124773929271e-05,
+      "loss": 0.3506,
+      "step": 3405
+    },
+    {
+      "epoch": 3.631130063965885,
+      "grad_norm": 0.11997387496896626,
+      "learning_rate": 1.691692761281354e-05,
+      "loss": 0.3505,
+      "step": 3406
+    },
+    {
+      "epoch": 3.6321961620469083,
+      "grad_norm": 0.1581248499900699,
+      "learning_rate": 1.6892620273156795e-05,
+      "loss": 0.351,
+      "step": 3407
+    },
+    {
+      "epoch": 3.633262260127932,
+      "grad_norm": 0.1253330836795441,
+      "learning_rate": 1.686832573378749e-05,
+      "loss": 0.3527,
+      "step": 3408
+    },
+    {
+      "epoch": 3.6343283582089554,
+      "grad_norm": 0.12308261327495958,
+      "learning_rate": 1.684404400816354e-05,
+      "loss": 0.3485,
+      "step": 3409
+    },
+    {
+      "epoch": 3.635394456289979,
+      "grad_norm": 0.1094772181271851,
+      "learning_rate": 1.6819775109735777e-05,
+      "loss": 0.3526,
+      "step": 3410
+    },
+    {
+      "epoch": 3.636460554371002,
+      "grad_norm": 0.14783525680446527,
+      "learning_rate": 1.6795519051947877e-05,
+      "loss": 0.3493,
+      "step": 3411
+    },
+    {
+      "epoch": 3.6375266524520256,
+      "grad_norm": 0.11897120110784674,
+      "learning_rate": 1.6771275848236447e-05,
+      "loss": 0.3482,
+      "step": 3412
+    },
+    {
+      "epoch": 3.638592750533049,
+      "grad_norm": 0.1346934227512216,
+      "learning_rate": 1.674704551203098e-05,
+      "loss": 0.3498,
+      "step": 3413
+    },
+    {
+      "epoch": 3.6396588486140726,
+      "grad_norm": 0.11012404823074266,
+      "learning_rate": 1.6722828056753794e-05,
+      "loss": 0.35,
+      "step": 3414
+    },
+    {
+      "epoch": 3.640724946695096,
+      "grad_norm": 0.13021214801697548,
+      "learning_rate": 1.6698623495820117e-05,
+      "loss": 0.3483,
+      "step": 3415
+    },
+    {
+      "epoch": 3.6417910447761193,
+      "grad_norm": 0.10956855386418592,
+      "learning_rate": 1.6674431842638014e-05,
+      "loss": 0.3492,
+      "step": 3416
+    },
+    {
+      "epoch": 3.642857142857143,
+      "grad_norm": 0.11548747621938685,
+      "learning_rate": 1.6650253110608415e-05,
+      "loss": 0.345,
+      "step": 3417
+    },
+    {
+      "epoch": 3.6439232409381663,
+      "grad_norm": 0.10744538789719912,
+      "learning_rate": 1.662608731312506e-05,
+      "loss": 0.3513,
+      "step": 3418
+    },
+    {
+      "epoch": 3.64498933901919,
+      "grad_norm": 0.12523119814259084,
+      "learning_rate": 1.6601934463574553e-05,
+      "loss": 0.349,
+      "step": 3419
+    },
+    {
+      "epoch": 3.6460554371002134,
+      "grad_norm": 0.1183786159156495,
+      "learning_rate": 1.657779457533632e-05,
+      "loss": 0.3497,
+      "step": 3420
+    },
+    {
+      "epoch": 3.6471215351812365,
+      "grad_norm": 0.1258499101414295,
+      "learning_rate": 1.6553667661782624e-05,
+      "loss": 0.3471,
+      "step": 3421
+    },
+    {
+      "epoch": 3.64818763326226,
+      "grad_norm": 0.11610530838109377,
+      "learning_rate": 1.652955373627848e-05,
+      "loss": 0.3533,
+      "step": 3422
+    },
+    {
+      "epoch": 3.6492537313432836,
+      "grad_norm": 0.12286306836120377,
+      "learning_rate": 1.6505452812181775e-05,
+      "loss": 0.3483,
+      "step": 3423
+    },
+    {
+      "epoch": 3.650319829424307,
+      "grad_norm": 0.09734440175920124,
+      "learning_rate": 1.648136490284318e-05,
+      "loss": 0.344,
+      "step": 3424
+    },
+    {
+      "epoch": 3.6513859275053306,
+      "grad_norm": 0.13317697498320796,
+      "learning_rate": 1.645729002160611e-05,
+      "loss": 0.3485,
+      "step": 3425
+    },
+    {
+      "epoch": 3.6524520255863537,
+      "grad_norm": 0.11968047549008518,
+      "learning_rate": 1.643322818180682e-05,
+      "loss": 0.3504,
+      "step": 3426
+    },
+    {
+      "epoch": 3.6535181236673773,
+      "grad_norm": 0.12970016078696814,
+      "learning_rate": 1.6409179396774317e-05,
+      "loss": 0.3494,
+      "step": 3427
+    },
+    {
+      "epoch": 3.654584221748401,
+      "grad_norm": 0.10979013596405256,
+      "learning_rate": 1.638514367983039e-05,
+      "loss": 0.3475,
+      "step": 3428
+    },
+    {
+      "epoch": 3.6556503198294243,
+      "grad_norm": 0.11958509569515018,
+      "learning_rate": 1.6361121044289553e-05,
+      "loss": 0.3424,
+      "step": 3429
+    },
+    {
+      "epoch": 3.656716417910448,
+      "grad_norm": 0.1210201578994878,
+      "learning_rate": 1.6337111503459104e-05,
+      "loss": 0.3443,
+      "step": 3430
+    },
+    {
+      "epoch": 3.657782515991471,
+      "grad_norm": 0.11828921119105222,
+      "learning_rate": 1.63131150706391e-05,
+      "loss": 0.3442,
+      "step": 3431
+    },
+    {
+      "epoch": 3.6588486140724945,
+      "grad_norm": 0.12968866909782567,
+      "learning_rate": 1.6289131759122292e-05,
+      "loss": 0.3488,
+      "step": 3432
+    },
+    {
+      "epoch": 3.659914712153518,
+      "grad_norm": 0.11513145282740704,
+      "learning_rate": 1.62651615821942e-05,
+      "loss": 0.3515,
+      "step": 3433
+    },
+    {
+      "epoch": 3.6609808102345416,
+      "grad_norm": 0.10887008435723226,
+      "learning_rate": 1.6241204553133054e-05,
+      "loss": 0.3422,
+      "step": 3434
+    },
+    {
+      "epoch": 3.662046908315565,
+      "grad_norm": 0.10848962774276776,
+      "learning_rate": 1.6217260685209815e-05,
+      "loss": 0.3508,
+      "step": 3435
+    },
+    {
+      "epoch": 3.663113006396588,
+      "grad_norm": 0.10984253318601138,
+      "learning_rate": 1.619332999168812e-05,
+      "loss": 0.3473,
+      "step": 3436
+    },
+    {
+      "epoch": 3.664179104477612,
+      "grad_norm": 0.10354031929218735,
+      "learning_rate": 1.6169412485824343e-05,
+      "loss": 0.3511,
+      "step": 3437
+    },
+    {
+      "epoch": 3.6652452025586353,
+      "grad_norm": 0.11266732298785025,
+      "learning_rate": 1.6145508180867537e-05,
+      "loss": 0.3438,
+      "step": 3438
+    },
+    {
+      "epoch": 3.666311300639659,
+      "grad_norm": 0.11007091054264355,
+      "learning_rate": 1.6121617090059455e-05,
+      "loss": 0.3519,
+      "step": 3439
+    },
+    {
+      "epoch": 3.6673773987206824,
+      "grad_norm": 0.11010507294324125,
+      "learning_rate": 1.6097739226634494e-05,
+      "loss": 0.3503,
+      "step": 3440
+    },
+    {
+      "epoch": 3.668443496801706,
+      "grad_norm": 0.1107608574897804,
+      "learning_rate": 1.6073874603819767e-05,
+      "loss": 0.3444,
+      "step": 3441
+    },
+    {
+      "epoch": 3.6695095948827294,
+      "grad_norm": 0.09996660829470029,
+      "learning_rate": 1.605002323483505e-05,
+      "loss": 0.3508,
+      "step": 3442
+    },
+    {
+      "epoch": 3.6705756929637525,
+      "grad_norm": 0.10541897341054957,
+      "learning_rate": 1.6026185132892722e-05,
+      "loss": 0.3473,
+      "step": 3443
+    },
+    {
+      "epoch": 3.671641791044776,
+      "grad_norm": 0.09364677982992888,
+      "learning_rate": 1.6002360311197874e-05,
+      "loss": 0.345,
+      "step": 3444
+    },
+    {
+      "epoch": 3.6727078891257996,
+      "grad_norm": 0.11633515138576979,
+      "learning_rate": 1.5978548782948228e-05,
+      "loss": 0.3537,
+      "step": 3445
+    },
+    {
+      "epoch": 3.673773987206823,
+      "grad_norm": 0.10455339919916513,
+      "learning_rate": 1.595475056133413e-05,
+      "loss": 0.3441,
+      "step": 3446
+    },
+    {
+      "epoch": 3.6748400852878467,
+      "grad_norm": 0.10817600283132285,
+      "learning_rate": 1.5930965659538547e-05,
+      "loss": 0.3513,
+      "step": 3447
+    },
+    {
+      "epoch": 3.6759061833688698,
+      "grad_norm": 0.11724298481759776,
+      "learning_rate": 1.590719409073708e-05,
+      "loss": 0.3471,
+      "step": 3448
+    },
+    {
+      "epoch": 3.6769722814498933,
+      "grad_norm": 0.11034160049080355,
+      "learning_rate": 1.5883435868097942e-05,
+      "loss": 0.3494,
+      "step": 3449
+    },
+    {
+      "epoch": 3.678038379530917,
+      "grad_norm": 0.10935682929664699,
+      "learning_rate": 1.5859691004781977e-05,
+      "loss": 0.3531,
+      "step": 3450
+    },
+    {
+      "epoch": 3.6791044776119404,
+      "grad_norm": 0.10669839160222076,
+      "learning_rate": 1.5835959513942577e-05,
+      "loss": 0.3464,
+      "step": 3451
+    },
+    {
+      "epoch": 3.680170575692964,
+      "grad_norm": 0.10499647433795345,
+      "learning_rate": 1.5812241408725757e-05,
+      "loss": 0.3484,
+      "step": 3452
+    },
+    {
+      "epoch": 3.681236673773987,
+      "grad_norm": 0.10722258869475547,
+      "learning_rate": 1.5788536702270136e-05,
+      "loss": 0.3498,
+      "step": 3453
+    },
+    {
+      "epoch": 3.6823027718550105,
+      "grad_norm": 0.1120874379953065,
+      "learning_rate": 1.5764845407706863e-05,
+      "loss": 0.3561,
+      "step": 3454
+    },
+    {
+      "epoch": 3.683368869936034,
+      "grad_norm": 0.11848268184587527,
+      "learning_rate": 1.5741167538159697e-05,
+      "loss": 0.3464,
+      "step": 3455
+    },
+    {
+      "epoch": 3.6844349680170576,
+      "grad_norm": 0.12047369839392386,
+      "learning_rate": 1.5717503106744957e-05,
+      "loss": 0.3572,
+      "step": 3456
+    },
+    {
+      "epoch": 3.685501066098081,
+      "grad_norm": 0.12166580342753797,
+      "learning_rate": 1.569385212657149e-05,
+      "loss": 0.3489,
+      "step": 3457
+    },
+    {
+      "epoch": 3.6865671641791042,
+      "grad_norm": 0.12106373222633703,
+      "learning_rate": 1.567021461074071e-05,
+      "loss": 0.3516,
+      "step": 3458
+    },
+    {
+      "epoch": 3.6876332622601278,
+      "grad_norm": 0.11484973241212469,
+      "learning_rate": 1.5646590572346596e-05,
+      "loss": 0.3524,
+      "step": 3459
+    },
+    {
+      "epoch": 3.6886993603411513,
+      "grad_norm": 0.10937051777873967,
+      "learning_rate": 1.5622980024475608e-05,
+      "loss": 0.3439,
+      "step": 3460
+    },
+    {
+      "epoch": 3.689765458422175,
+      "grad_norm": 0.12137658902823265,
+      "learning_rate": 1.5599382980206773e-05,
+      "loss": 0.3516,
+      "step": 3461
+    },
+    {
+      "epoch": 3.6908315565031984,
+      "grad_norm": 0.12974179145102452,
+      "learning_rate": 1.5575799452611647e-05,
+      "loss": 0.3545,
+      "step": 3462
+    },
+    {
+      "epoch": 3.6918976545842215,
+      "grad_norm": 0.13163480419478382,
+      "learning_rate": 1.5552229454754245e-05,
+      "loss": 0.3462,
+      "step": 3463
+    },
+    {
+      "epoch": 3.6929637526652455,
+      "grad_norm": 0.09995629474690315,
+      "learning_rate": 1.5528672999691137e-05,
+      "loss": 0.3463,
+      "step": 3464
+    },
+    {
+      "epoch": 3.6940298507462686,
+      "grad_norm": 0.12284728073753405,
+      "learning_rate": 1.550513010047139e-05,
+      "loss": 0.3478,
+      "step": 3465
+    },
+    {
+      "epoch": 3.695095948827292,
+      "grad_norm": 0.11912635744633349,
+      "learning_rate": 1.5481600770136512e-05,
+      "loss": 0.3469,
+      "step": 3466
+    },
+    {
+      "epoch": 3.6961620469083156,
+      "grad_norm": 0.1046940554664763,
+      "learning_rate": 1.5458085021720557e-05,
+      "loss": 0.349,
+      "step": 3467
+    },
+    {
+      "epoch": 3.697228144989339,
+      "grad_norm": 0.10784839095846516,
+      "learning_rate": 1.543458286825003e-05,
+      "loss": 0.3533,
+      "step": 3468
+    },
+    {
+      "epoch": 3.6982942430703627,
+      "grad_norm": 0.1195343323249887,
+      "learning_rate": 1.5411094322743875e-05,
+      "loss": 0.3508,
+      "step": 3469
+    },
+    {
+      "epoch": 3.699360341151386,
+      "grad_norm": 0.10962783595704657,
+      "learning_rate": 1.5387619398213543e-05,
+      "loss": 0.3471,
+      "step": 3470
+    },
+    {
+      "epoch": 3.7004264392324093,
+      "grad_norm": 0.1150816905348978,
+      "learning_rate": 1.5364158107662935e-05,
+      "loss": 0.3518,
+      "step": 3471
+    },
+    {
+      "epoch": 3.701492537313433,
+      "grad_norm": 0.13829883854757222,
+      "learning_rate": 1.534071046408836e-05,
+      "loss": 0.3521,
+      "step": 3472
+    },
+    {
+      "epoch": 3.7025586353944564,
+      "grad_norm": 0.12445105825556152,
+      "learning_rate": 1.531727648047861e-05,
+      "loss": 0.3554,
+      "step": 3473
+    },
+    {
+      "epoch": 3.70362473347548,
+      "grad_norm": 0.1157531505696079,
+      "learning_rate": 1.5293856169814885e-05,
+      "loss": 0.3489,
+      "step": 3474
+    },
+    {
+      "epoch": 3.704690831556503,
+      "grad_norm": 0.1330840415067106,
+      "learning_rate": 1.527044954507084e-05,
+      "loss": 0.35,
+      "step": 3475
+    },
+    {
+      "epoch": 3.7057569296375266,
+      "grad_norm": 0.10784494484991011,
+      "learning_rate": 1.5247056619212507e-05,
+      "loss": 0.3488,
+      "step": 3476
+    },
+    {
+      "epoch": 3.70682302771855,
+      "grad_norm": 0.12127485364434097,
+      "learning_rate": 1.5223677405198354e-05,
+      "loss": 0.3499,
+      "step": 3477
+    },
+    {
+      "epoch": 3.7078891257995736,
+      "grad_norm": 0.11013802869572574,
+      "learning_rate": 1.5200311915979255e-05,
+      "loss": 0.3421,
+      "step": 3478
+    },
+    {
+      "epoch": 3.708955223880597,
+      "grad_norm": 0.12178309778215979,
+      "learning_rate": 1.517696016449849e-05,
+      "loss": 0.3527,
+      "step": 3479
+    },
+    {
+      "epoch": 3.7100213219616203,
+      "grad_norm": 0.11504029087431382,
+      "learning_rate": 1.515362216369169e-05,
+      "loss": 0.3483,
+      "step": 3480
+    },
+    {
+      "epoch": 3.711087420042644,
+      "grad_norm": 0.12440538915523314,
+      "learning_rate": 1.5130297926486908e-05,
+      "loss": 0.3481,
+      "step": 3481
+    },
+    {
+      "epoch": 3.7121535181236673,
+      "grad_norm": 0.1157756581499601,
+      "learning_rate": 1.5106987465804572e-05,
+      "loss": 0.3472,
+      "step": 3482
+    },
+    {
+      "epoch": 3.713219616204691,
+      "grad_norm": 0.13884491519651773,
+      "learning_rate": 1.5083690794557435e-05,
+      "loss": 0.3491,
+      "step": 3483
+    },
+    {
+      "epoch": 3.7142857142857144,
+      "grad_norm": 0.11160024061874728,
+      "learning_rate": 1.5060407925650662e-05,
+      "loss": 0.3442,
+      "step": 3484
+    },
+    {
+      "epoch": 3.7153518123667375,
+      "grad_norm": 0.1388304685706655,
+      "learning_rate": 1.503713887198175e-05,
+      "loss": 0.3524,
+      "step": 3485
+    },
+    {
+      "epoch": 3.716417910447761,
+      "grad_norm": 0.10291882827022071,
+      "learning_rate": 1.5013883646440555e-05,
+      "loss": 0.3469,
+      "step": 3486
+    },
+    {
+      "epoch": 3.7174840085287846,
+      "grad_norm": 0.11581595497958068,
+      "learning_rate": 1.499064226190924e-05,
+      "loss": 0.3478,
+      "step": 3487
+    },
+    {
+      "epoch": 3.718550106609808,
+      "grad_norm": 0.12039433257965831,
+      "learning_rate": 1.4967414731262339e-05,
+      "loss": 0.3476,
+      "step": 3488
+    },
+    {
+      "epoch": 3.7196162046908317,
+      "grad_norm": 0.11994551705543056,
+      "learning_rate": 1.494420106736671e-05,
+      "loss": 0.349,
+      "step": 3489
+    },
+    {
+      "epoch": 3.7206823027718547,
+      "grad_norm": 0.10943678193661488,
+      "learning_rate": 1.4921001283081488e-05,
+      "loss": 0.3443,
+      "step": 3490
+    },
+    {
+      "epoch": 3.7217484008528787,
+      "grad_norm": 0.12365465123838881,
+      "learning_rate": 1.489781539125816e-05,
+      "loss": 0.3428,
+      "step": 3491
+    },
+    {
+      "epoch": 3.722814498933902,
+      "grad_norm": 0.12996918119430528,
+      "learning_rate": 1.4874643404740505e-05,
+      "loss": 0.3492,
+      "step": 3492
+    },
+    {
+      "epoch": 3.7238805970149254,
+      "grad_norm": 0.12320823229427783,
+      "learning_rate": 1.4851485336364616e-05,
+      "loss": 0.3555,
+      "step": 3493
+    },
+    {
+      "epoch": 3.724946695095949,
+      "grad_norm": 0.11983024758002496,
+      "learning_rate": 1.4828341198958827e-05,
+      "loss": 0.3529,
+      "step": 3494
+    },
+    {
+      "epoch": 3.7260127931769724,
+      "grad_norm": 0.11090723612747737,
+      "learning_rate": 1.4805211005343804e-05,
+      "loss": 0.3497,
+      "step": 3495
+    },
+    {
+      "epoch": 3.727078891257996,
+      "grad_norm": 0.11889325003301451,
+      "learning_rate": 1.4782094768332477e-05,
+      "loss": 0.3475,
+      "step": 3496
+    },
+    {
+      "epoch": 3.728144989339019,
+      "grad_norm": 0.11039401592347903,
+      "learning_rate": 1.4758992500730047e-05,
+      "loss": 0.3506,
+      "step": 3497
+    },
+    {
+      "epoch": 3.7292110874200426,
+      "grad_norm": 0.11723348316498812,
+      "learning_rate": 1.4735904215333942e-05,
+      "loss": 0.3491,
+      "step": 3498
+    },
+    {
+      "epoch": 3.730277185501066,
+      "grad_norm": 0.1096667948027839,
+      "learning_rate": 1.4712829924933888e-05,
+      "loss": 0.35,
+      "step": 3499
+    },
+    {
+      "epoch": 3.7313432835820897,
+      "grad_norm": 0.12827667700436324,
+      "learning_rate": 1.4689769642311862e-05,
+      "loss": 0.3488,
+      "step": 3500
+    },
+    {
+      "epoch": 3.732409381663113,
+      "grad_norm": 0.11597291937955526,
+      "learning_rate": 1.466672338024202e-05,
+      "loss": 0.3529,
+      "step": 3501
+    },
+    {
+      "epoch": 3.7334754797441363,
+      "grad_norm": 0.12098185152644861,
+      "learning_rate": 1.4643691151490825e-05,
+      "loss": 0.3543,
+      "step": 3502
+    },
+    {
+      "epoch": 3.73454157782516,
+      "grad_norm": 0.1264853588754496,
+      "learning_rate": 1.462067296881692e-05,
+      "loss": 0.3494,
+      "step": 3503
+    },
+    {
+      "epoch": 3.7356076759061834,
+      "grad_norm": 0.11554640924569769,
+      "learning_rate": 1.4597668844971203e-05,
+      "loss": 0.3477,
+      "step": 3504
+    },
+    {
+      "epoch": 3.736673773987207,
+      "grad_norm": 0.1263291190867985,
+      "learning_rate": 1.4574678792696735e-05,
+      "loss": 0.3464,
+      "step": 3505
+    },
+    {
+      "epoch": 3.7377398720682304,
+      "grad_norm": 0.1140310154067713,
+      "learning_rate": 1.4551702824728824e-05,
+      "loss": 0.3534,
+      "step": 3506
+    },
+    {
+      "epoch": 3.7388059701492535,
+      "grad_norm": 0.1374330707269001,
+      "learning_rate": 1.452874095379496e-05,
+      "loss": 0.3525,
+      "step": 3507
+    },
+    {
+      "epoch": 3.739872068230277,
+      "grad_norm": 0.10373341206088472,
+      "learning_rate": 1.4505793192614838e-05,
+      "loss": 0.3487,
+      "step": 3508
+    },
+    {
+      "epoch": 3.7409381663113006,
+      "grad_norm": 0.12296501711153075,
+      "learning_rate": 1.4482859553900302e-05,
+      "loss": 0.3523,
+      "step": 3509
+    },
+    {
+      "epoch": 3.742004264392324,
+      "grad_norm": 0.09959644973450633,
+      "learning_rate": 1.4459940050355412e-05,
+      "loss": 0.3459,
+      "step": 3510
+    },
+    {
+      "epoch": 3.7430703624733477,
+      "grad_norm": 0.12381271538841246,
+      "learning_rate": 1.4437034694676388e-05,
+      "loss": 0.3442,
+      "step": 3511
+    },
+    {
+      "epoch": 3.7441364605543708,
+      "grad_norm": 0.11420172023496247,
+      "learning_rate": 1.4414143499551583e-05,
+      "loss": 0.3435,
+      "step": 3512
+    },
+    {
+      "epoch": 3.7452025586353943,
+      "grad_norm": 0.12761045293800952,
+      "learning_rate": 1.4391266477661545e-05,
+      "loss": 0.3502,
+      "step": 3513
+    },
+    {
+      "epoch": 3.746268656716418,
+      "grad_norm": 0.12938266287837766,
+      "learning_rate": 1.4368403641678951e-05,
+      "loss": 0.353,
+      "step": 3514
+    },
+    {
+      "epoch": 3.7473347547974414,
+      "grad_norm": 0.12147529927692077,
+      "learning_rate": 1.434555500426864e-05,
+      "loss": 0.3529,
+      "step": 3515
+    },
+    {
+      "epoch": 3.748400852878465,
+      "grad_norm": 0.12991621986150606,
+      "learning_rate": 1.4322720578087546e-05,
+      "loss": 0.3508,
+      "step": 3516
+    },
+    {
+      "epoch": 3.749466950959488,
+      "grad_norm": 0.11210702326795871,
+      "learning_rate": 1.4299900375784761e-05,
+      "loss": 0.3491,
+      "step": 3517
+    },
+    {
+      "epoch": 3.750533049040512,
+      "grad_norm": 0.12221491366601378,
+      "learning_rate": 1.4277094410001508e-05,
+      "loss": 0.3486,
+      "step": 3518
+    },
+    {
+      "epoch": 3.751599147121535,
+      "grad_norm": 0.11758736893333697,
+      "learning_rate": 1.4254302693371083e-05,
+      "loss": 0.3525,
+      "step": 3519
+    },
+    {
+      "epoch": 3.7526652452025586,
+      "grad_norm": 0.11720998722860178,
+      "learning_rate": 1.4231525238518917e-05,
+      "loss": 0.3526,
+      "step": 3520
+    },
+    {
+      "epoch": 3.753731343283582,
+      "grad_norm": 0.11702149986171113,
+      "learning_rate": 1.4208762058062546e-05,
+      "loss": 0.3567,
+      "step": 3521
+    },
+    {
+      "epoch": 3.7547974413646057,
+      "grad_norm": 0.11781947922119294,
+      "learning_rate": 1.4186013164611593e-05,
+      "loss": 0.3476,
+      "step": 3522
+    },
+    {
+      "epoch": 3.7558635394456292,
+      "grad_norm": 0.11028285471866768,
+      "learning_rate": 1.4163278570767744e-05,
+      "loss": 0.3472,
+      "step": 3523
+    },
+    {
+      "epoch": 3.7569296375266523,
+      "grad_norm": 0.11480800401974497,
+      "learning_rate": 1.4140558289124795e-05,
+      "loss": 0.3498,
+      "step": 3524
+    },
+    {
+      "epoch": 3.757995735607676,
+      "grad_norm": 0.10511476764554756,
+      "learning_rate": 1.411785233226861e-05,
+      "loss": 0.346,
+      "step": 3525
+    },
+    {
+      "epoch": 3.7590618336886994,
+      "grad_norm": 0.11246301899459499,
+      "learning_rate": 1.4095160712777087e-05,
+      "loss": 0.3463,
+      "step": 3526
+    },
+    {
+      "epoch": 3.760127931769723,
+      "grad_norm": 0.11075751087710273,
+      "learning_rate": 1.4072483443220213e-05,
+      "loss": 0.351,
+      "step": 3527
+    },
+    {
+      "epoch": 3.7611940298507465,
+      "grad_norm": 0.11457686956190802,
+      "learning_rate": 1.4049820536160033e-05,
+      "loss": 0.3507,
+      "step": 3528
+    },
+    {
+      "epoch": 3.7622601279317696,
+      "grad_norm": 0.13257820543611734,
+      "learning_rate": 1.4027172004150594e-05,
+      "loss": 0.3483,
+      "step": 3529
+    },
+    {
+      "epoch": 3.763326226012793,
+      "grad_norm": 0.09934268832616149,
+      "learning_rate": 1.400453785973801e-05,
+      "loss": 0.3471,
+      "step": 3530
+    },
+    {
+      "epoch": 3.7643923240938166,
+      "grad_norm": 0.1217605278748467,
+      "learning_rate": 1.3981918115460448e-05,
+      "loss": 0.3452,
+      "step": 3531
+    },
+    {
+      "epoch": 3.76545842217484,
+      "grad_norm": 0.1025857492577431,
+      "learning_rate": 1.3959312783848033e-05,
+      "loss": 0.3496,
+      "step": 3532
+    },
+    {
+      "epoch": 3.7665245202558637,
+      "grad_norm": 0.12375183362556455,
+      "learning_rate": 1.3936721877422965e-05,
+      "loss": 0.3532,
+      "step": 3533
+    },
+    {
+      "epoch": 3.767590618336887,
+      "grad_norm": 0.10188334610518712,
+      "learning_rate": 1.3914145408699437e-05,
+      "loss": 0.3436,
+      "step": 3534
+    },
+    {
+      "epoch": 3.7686567164179103,
+      "grad_norm": 0.12657161432596883,
+      "learning_rate": 1.3891583390183621e-05,
+      "loss": 0.3517,
+      "step": 3535
+    },
+    {
+      "epoch": 3.769722814498934,
+      "grad_norm": 0.11261285266481309,
+      "learning_rate": 1.3869035834373712e-05,
+      "loss": 0.3533,
+      "step": 3536
+    },
+    {
+      "epoch": 3.7707889125799574,
+      "grad_norm": 0.11006622035677963,
+      "learning_rate": 1.3846502753759899e-05,
+      "loss": 0.3475,
+      "step": 3537
+    },
+    {
+      "epoch": 3.771855010660981,
+      "grad_norm": 0.10803610024542715,
+      "learning_rate": 1.3823984160824306e-05,
+      "loss": 0.3521,
+      "step": 3538
+    },
+    {
+      "epoch": 3.772921108742004,
+      "grad_norm": 0.10308686673663463,
+      "learning_rate": 1.3801480068041083e-05,
+      "loss": 0.3486,
+      "step": 3539
+    },
+    {
+      "epoch": 3.7739872068230276,
+      "grad_norm": 0.11418577872966433,
+      "learning_rate": 1.3778990487876338e-05,
+      "loss": 0.3527,
+      "step": 3540
+    },
+    {
+      "epoch": 3.775053304904051,
+      "grad_norm": 0.115083278681719,
+      "learning_rate": 1.3756515432788105e-05,
+      "loss": 0.3547,
+      "step": 3541
+    },
+    {
+      "epoch": 3.7761194029850746,
+      "grad_norm": 0.105339094797037,
+      "learning_rate": 1.3734054915226405e-05,
+      "loss": 0.3509,
+      "step": 3542
+    },
+    {
+      "epoch": 3.777185501066098,
+      "grad_norm": 0.11586176556566104,
+      "learning_rate": 1.3711608947633202e-05,
+      "loss": 0.3508,
+      "step": 3543
+    },
+    {
+      "epoch": 3.7782515991471213,
+      "grad_norm": 0.1066837228197777,
+      "learning_rate": 1.3689177542442406e-05,
+      "loss": 0.3477,
+      "step": 3544
+    },
+    {
+      "epoch": 3.7793176972281453,
+      "grad_norm": 0.11175162407532467,
+      "learning_rate": 1.3666760712079828e-05,
+      "loss": 0.3486,
+      "step": 3545
+    },
+    {
+      "epoch": 3.7803837953091683,
+      "grad_norm": 0.10538041603411623,
+      "learning_rate": 1.3644358468963233e-05,
+      "loss": 0.3477,
+      "step": 3546
+    },
+    {
+      "epoch": 3.781449893390192,
+      "grad_norm": 0.12393679042543049,
+      "learning_rate": 1.3621970825502317e-05,
+      "loss": 0.3535,
+      "step": 3547
+    },
+    {
+      "epoch": 3.7825159914712154,
+      "grad_norm": 0.10491158418048802,
+      "learning_rate": 1.3599597794098648e-05,
+      "loss": 0.3562,
+      "step": 3548
+    },
+    {
+      "epoch": 3.783582089552239,
+      "grad_norm": 0.12508143946693426,
+      "learning_rate": 1.3577239387145729e-05,
+      "loss": 0.3513,
+      "step": 3549
+    },
+    {
+      "epoch": 3.7846481876332625,
+      "grad_norm": 0.11761432608648041,
+      "learning_rate": 1.3554895617028958e-05,
+      "loss": 0.3502,
+      "step": 3550
+    },
+    {
+      "epoch": 3.7857142857142856,
+      "grad_norm": 0.12468188783060895,
+      "learning_rate": 1.3532566496125634e-05,
+      "loss": 0.349,
+      "step": 3551
+    },
+    {
+      "epoch": 3.786780383795309,
+      "grad_norm": 0.11776026667363994,
+      "learning_rate": 1.3510252036804907e-05,
+      "loss": 0.3496,
+      "step": 3552
+    },
+    {
+      "epoch": 3.7878464818763327,
+      "grad_norm": 0.10743422269996004,
+      "learning_rate": 1.348795225142784e-05,
+      "loss": 0.3512,
+      "step": 3553
+    },
+    {
+      "epoch": 3.788912579957356,
+      "grad_norm": 0.10720339132099257,
+      "learning_rate": 1.3465667152347353e-05,
+      "loss": 0.346,
+      "step": 3554
+    },
+    {
+      "epoch": 3.7899786780383797,
+      "grad_norm": 0.10732063941636186,
+      "learning_rate": 1.3443396751908243e-05,
+      "loss": 0.3459,
+      "step": 3555
+    },
+    {
+      "epoch": 3.791044776119403,
+      "grad_norm": 0.1113815053455615,
+      "learning_rate": 1.3421141062447136e-05,
+      "loss": 0.3491,
+      "step": 3556
+    },
+    {
+      "epoch": 3.7921108742004264,
+      "grad_norm": 0.11139797758194125,
+      "learning_rate": 1.3398900096292535e-05,
+      "loss": 0.3513,
+      "step": 3557
+    },
+    {
+      "epoch": 3.79317697228145,
+      "grad_norm": 0.11752612657415666,
+      "learning_rate": 1.3376673865764796e-05,
+      "loss": 0.354,
+      "step": 3558
+    },
+    {
+      "epoch": 3.7942430703624734,
+      "grad_norm": 0.11709263878140849,
+      "learning_rate": 1.3354462383176064e-05,
+      "loss": 0.3547,
+      "step": 3559
+    },
+    {
+      "epoch": 3.795309168443497,
+      "grad_norm": 0.12982606895742937,
+      "learning_rate": 1.3332265660830364e-05,
+      "loss": 0.3467,
+      "step": 3560
+    },
+    {
+      "epoch": 3.79637526652452,
+      "grad_norm": 0.11565736542146775,
+      "learning_rate": 1.3310083711023527e-05,
+      "loss": 0.3516,
+      "step": 3561
+    },
+    {
+      "epoch": 3.7974413646055436,
+      "grad_norm": 0.09828691895649047,
+      "learning_rate": 1.3287916546043209e-05,
+      "loss": 0.3487,
+      "step": 3562
+    },
+    {
+      "epoch": 3.798507462686567,
+      "grad_norm": 0.1155849088844352,
+      "learning_rate": 1.3265764178168841e-05,
+      "loss": 0.3512,
+      "step": 3563
+    },
+    {
+      "epoch": 3.7995735607675907,
+      "grad_norm": 0.10134447160803774,
+      "learning_rate": 1.3243626619671704e-05,
+      "loss": 0.3541,
+      "step": 3564
+    },
+    {
+      "epoch": 3.800639658848614,
+      "grad_norm": 0.10380565104455901,
+      "learning_rate": 1.3221503882814846e-05,
+      "loss": 0.3466,
+      "step": 3565
+    },
+    {
+      "epoch": 3.8017057569296373,
+      "grad_norm": 0.10299695475323739,
+      "learning_rate": 1.3199395979853132e-05,
+      "loss": 0.3503,
+      "step": 3566
+    },
+    {
+      "epoch": 3.802771855010661,
+      "grad_norm": 0.131615477948487,
+      "learning_rate": 1.3177302923033164e-05,
+      "loss": 0.3515,
+      "step": 3567
+    },
+    {
+      "epoch": 3.8038379530916844,
+      "grad_norm": 0.09761956730459642,
+      "learning_rate": 1.3155224724593364e-05,
+      "loss": 0.3516,
+      "step": 3568
+    },
+    {
+      "epoch": 3.804904051172708,
+      "grad_norm": 0.10306669274819454,
+      "learning_rate": 1.3133161396763909e-05,
+      "loss": 0.3456,
+      "step": 3569
+    },
+    {
+      "epoch": 3.8059701492537314,
+      "grad_norm": 0.10145722545444867,
+      "learning_rate": 1.311111295176672e-05,
+      "loss": 0.3439,
+      "step": 3570
+    },
+    {
+      "epoch": 3.8070362473347545,
+      "grad_norm": 0.10770148837143485,
+      "learning_rate": 1.3089079401815497e-05,
+      "loss": 0.3496,
+      "step": 3571
+    },
+    {
+      "epoch": 3.8081023454157785,
+      "grad_norm": 0.11320765417229854,
+      "learning_rate": 1.3067060759115684e-05,
+      "loss": 0.3542,
+      "step": 3572
+    },
+    {
+      "epoch": 3.8091684434968016,
+      "grad_norm": 0.1163579650557563,
+      "learning_rate": 1.3045057035864477e-05,
+      "loss": 0.3554,
+      "step": 3573
+    },
+    {
+      "epoch": 3.810234541577825,
+      "grad_norm": 0.09905809774744204,
+      "learning_rate": 1.302306824425077e-05,
+      "loss": 0.3486,
+      "step": 3574
+    },
+    {
+      "epoch": 3.8113006396588487,
+      "grad_norm": 0.12608559744995004,
+      "learning_rate": 1.3001094396455223e-05,
+      "loss": 0.3514,
+      "step": 3575
+    },
+    {
+      "epoch": 3.8123667377398722,
+      "grad_norm": 0.09595835889524401,
+      "learning_rate": 1.297913550465022e-05,
+      "loss": 0.3462,
+      "step": 3576
+    },
+    {
+      "epoch": 3.8134328358208958,
+      "grad_norm": 0.13266355171728472,
+      "learning_rate": 1.2957191580999821e-05,
+      "loss": 0.3542,
+      "step": 3577
+    },
+    {
+      "epoch": 3.814498933901919,
+      "grad_norm": 0.09688667994875337,
+      "learning_rate": 1.2935262637659824e-05,
+      "loss": 0.3512,
+      "step": 3578
+    },
+    {
+      "epoch": 3.8155650319829424,
+      "grad_norm": 0.13077081224692166,
+      "learning_rate": 1.2913348686777734e-05,
+      "loss": 0.3463,
+      "step": 3579
+    },
+    {
+      "epoch": 3.816631130063966,
+      "grad_norm": 0.1000818706820891,
+      "learning_rate": 1.2891449740492749e-05,
+      "loss": 0.3499,
+      "step": 3580
+    },
+    {
+      "epoch": 3.8176972281449895,
+      "grad_norm": 0.12244332857777668,
+      "learning_rate": 1.2869565810935724e-05,
+      "loss": 0.3483,
+      "step": 3581
+    },
+    {
+      "epoch": 3.818763326226013,
+      "grad_norm": 0.10413603824918938,
+      "learning_rate": 1.2847696910229228e-05,
+      "loss": 0.3513,
+      "step": 3582
+    },
+    {
+      "epoch": 3.819829424307036,
+      "grad_norm": 0.10738770551253633,
+      "learning_rate": 1.2825843050487495e-05,
+      "loss": 0.3474,
+      "step": 3583
+    },
+    {
+      "epoch": 3.8208955223880596,
+      "grad_norm": 0.1181068563058338,
+      "learning_rate": 1.2804004243816444e-05,
+      "loss": 0.3541,
+      "step": 3584
+    },
+    {
+      "epoch": 3.821961620469083,
+      "grad_norm": 0.10846504895224321,
+      "learning_rate": 1.2782180502313609e-05,
+      "loss": 0.3499,
+      "step": 3585
+    },
+    {
+      "epoch": 3.8230277185501067,
+      "grad_norm": 0.10970803247483393,
+      "learning_rate": 1.2760371838068228e-05,
+      "loss": 0.3463,
+      "step": 3586
+    },
+    {
+      "epoch": 3.8240938166311302,
+      "grad_norm": 0.11300737915406556,
+      "learning_rate": 1.2738578263161175e-05,
+      "loss": 0.3497,
+      "step": 3587
+    },
+    {
+      "epoch": 3.8251599147121533,
+      "grad_norm": 0.10850041595269576,
+      "learning_rate": 1.2716799789664931e-05,
+      "loss": 0.3503,
+      "step": 3588
+    },
+    {
+      "epoch": 3.826226012793177,
+      "grad_norm": 0.1242287822226434,
+      "learning_rate": 1.2695036429643657e-05,
+      "loss": 0.3464,
+      "step": 3589
+    },
+    {
+      "epoch": 3.8272921108742004,
+      "grad_norm": 0.11234434571174157,
+      "learning_rate": 1.2673288195153118e-05,
+      "loss": 0.349,
+      "step": 3590
+    },
+    {
+      "epoch": 3.828358208955224,
+      "grad_norm": 0.11607729586786202,
+      "learning_rate": 1.2651555098240724e-05,
+      "loss": 0.3502,
+      "step": 3591
+    },
+    {
+      "epoch": 3.8294243070362475,
+      "grad_norm": 0.11570224496079395,
+      "learning_rate": 1.2629837150945447e-05,
+      "loss": 0.3499,
+      "step": 3592
+    },
+    {
+      "epoch": 3.8304904051172706,
+      "grad_norm": 0.12136791006711244,
+      "learning_rate": 1.2608134365297922e-05,
+      "loss": 0.354,
+      "step": 3593
+    },
+    {
+      "epoch": 3.831556503198294,
+      "grad_norm": 0.10914056771119587,
+      "learning_rate": 1.2586446753320374e-05,
+      "loss": 0.3538,
+      "step": 3594
+    },
+    {
+      "epoch": 3.8326226012793176,
+      "grad_norm": 0.12699655682552763,
+      "learning_rate": 1.256477432702659e-05,
+      "loss": 0.3469,
+      "step": 3595
+    },
+    {
+      "epoch": 3.833688699360341,
+      "grad_norm": 0.11727167103613097,
+      "learning_rate": 1.2543117098421976e-05,
+      "loss": 0.345,
+      "step": 3596
+    },
+    {
+      "epoch": 3.8347547974413647,
+      "grad_norm": 0.1207939527624419,
+      "learning_rate": 1.2521475079503524e-05,
+      "loss": 0.3477,
+      "step": 3597
+    },
+    {
+      "epoch": 3.835820895522388,
+      "grad_norm": 0.11622965193657621,
+      "learning_rate": 1.2499848282259767e-05,
+      "loss": 0.3489,
+      "step": 3598
+    },
+    {
+      "epoch": 3.836886993603412,
+      "grad_norm": 0.13564634838395118,
+      "learning_rate": 1.2478236718670834e-05,
+      "loss": 0.3498,
+      "step": 3599
+    },
+    {
+      "epoch": 3.837953091684435,
+      "grad_norm": 0.11550765632689088,
+      "learning_rate": 1.2456640400708424e-05,
+      "loss": 0.3515,
+      "step": 3600
+    },
+    {
+      "epoch": 3.8390191897654584,
+      "grad_norm": 0.10993793619669771,
+      "learning_rate": 1.2435059340335753e-05,
+      "loss": 0.3435,
+      "step": 3601
+    },
+    {
+      "epoch": 3.840085287846482,
+      "grad_norm": 0.11694097576572215,
+      "learning_rate": 1.241349354950761e-05,
+      "loss": 0.3487,
+      "step": 3602
+    },
+    {
+      "epoch": 3.8411513859275055,
+      "grad_norm": 0.10341436996831099,
+      "learning_rate": 1.2391943040170343e-05,
+      "loss": 0.348,
+      "step": 3603
+    },
+    {
+      "epoch": 3.842217484008529,
+      "grad_norm": 0.12769414945414873,
+      "learning_rate": 1.2370407824261785e-05,
+      "loss": 0.3489,
+      "step": 3604
+    },
+    {
+      "epoch": 3.843283582089552,
+      "grad_norm": 0.09401881452558605,
+      "learning_rate": 1.2348887913711343e-05,
+      "loss": 0.3522,
+      "step": 3605
+    },
+    {
+      "epoch": 3.8443496801705757,
+      "grad_norm": 0.11795172219684597,
+      "learning_rate": 1.2327383320439937e-05,
+      "loss": 0.3449,
+      "step": 3606
+    },
+    {
+      "epoch": 3.845415778251599,
+      "grad_norm": 0.1051555043312947,
+      "learning_rate": 1.2305894056359967e-05,
+      "loss": 0.3499,
+      "step": 3607
+    },
+    {
+      "epoch": 3.8464818763326227,
+      "grad_norm": 0.12236877442049936,
+      "learning_rate": 1.2284420133375385e-05,
+      "loss": 0.3522,
+      "step": 3608
+    },
+    {
+      "epoch": 3.8475479744136463,
+      "grad_norm": 0.10450987023264936,
+      "learning_rate": 1.2262961563381643e-05,
+      "loss": 0.35,
+      "step": 3609
+    },
+    {
+      "epoch": 3.8486140724946694,
+      "grad_norm": 0.11632851457098174,
+      "learning_rate": 1.224151835826565e-05,
+      "loss": 0.3425,
+      "step": 3610
+    },
+    {
+      "epoch": 3.849680170575693,
+      "grad_norm": 0.1313585390733969,
+      "learning_rate": 1.222009052990583e-05,
+      "loss": 0.3501,
+      "step": 3611
+    },
+    {
+      "epoch": 3.8507462686567164,
+      "grad_norm": 0.09885089090066136,
+      "learning_rate": 1.2198678090172096e-05,
+      "loss": 0.3476,
+      "step": 3612
+    },
+    {
+      "epoch": 3.85181236673774,
+      "grad_norm": 0.12386048454800534,
+      "learning_rate": 1.2177281050925829e-05,
+      "loss": 0.3533,
+      "step": 3613
+    },
+    {
+      "epoch": 3.8528784648187635,
+      "grad_norm": 0.10109383676324435,
+      "learning_rate": 1.2155899424019864e-05,
+      "loss": 0.3479,
+      "step": 3614
+    },
+    {
+      "epoch": 3.8539445628997866,
+      "grad_norm": 0.11463340028361436,
+      "learning_rate": 1.2134533221298517e-05,
+      "loss": 0.3547,
+      "step": 3615
+    },
+    {
+      "epoch": 3.85501066098081,
+      "grad_norm": 0.10969960215345126,
+      "learning_rate": 1.2113182454597565e-05,
+      "loss": 0.3514,
+      "step": 3616
+    },
+    {
+      "epoch": 3.8560767590618337,
+      "grad_norm": 0.11112827116259213,
+      "learning_rate": 1.2091847135744198e-05,
+      "loss": 0.349,
+      "step": 3617
+    },
+    {
+      "epoch": 3.857142857142857,
+      "grad_norm": 0.11843720838928255,
+      "learning_rate": 1.2070527276557092e-05,
+      "loss": 0.3441,
+      "step": 3618
+    },
+    {
+      "epoch": 3.8582089552238807,
+      "grad_norm": 0.10458805732550369,
+      "learning_rate": 1.2049222888846334e-05,
+      "loss": 0.3561,
+      "step": 3619
+    },
+    {
+      "epoch": 3.859275053304904,
+      "grad_norm": 0.11054412675997474,
+      "learning_rate": 1.2027933984413469e-05,
+      "loss": 0.3527,
+      "step": 3620
+    },
+    {
+      "epoch": 3.8603411513859274,
+      "grad_norm": 0.10285700468758723,
+      "learning_rate": 1.2006660575051407e-05,
+      "loss": 0.3471,
+      "step": 3621
+    },
+    {
+      "epoch": 3.861407249466951,
+      "grad_norm": 0.1136506392149834,
+      "learning_rate": 1.1985402672544532e-05,
+      "loss": 0.3453,
+      "step": 3622
+    },
+    {
+      "epoch": 3.8624733475479744,
+      "grad_norm": 0.09831280484139478,
+      "learning_rate": 1.1964160288668629e-05,
+      "loss": 0.3506,
+      "step": 3623
+    },
+    {
+      "epoch": 3.863539445628998,
+      "grad_norm": 0.11187723045529688,
+      "learning_rate": 1.1942933435190845e-05,
+      "loss": 0.3499,
+      "step": 3624
+    },
+    {
+      "epoch": 3.864605543710021,
+      "grad_norm": 0.09266608981403165,
+      "learning_rate": 1.1921722123869773e-05,
+      "loss": 0.351,
+      "step": 3625
+    },
+    {
+      "epoch": 3.8656716417910446,
+      "grad_norm": 0.11511057624858669,
+      "learning_rate": 1.1900526366455369e-05,
+      "loss": 0.3468,
+      "step": 3626
+    },
+    {
+      "epoch": 3.866737739872068,
+      "grad_norm": 0.10956860731954124,
+      "learning_rate": 1.1879346174689e-05,
+      "loss": 0.3492,
+      "step": 3627
+    },
+    {
+      "epoch": 3.8678038379530917,
+      "grad_norm": 0.10708959205937109,
+      "learning_rate": 1.1858181560303366e-05,
+      "loss": 0.3469,
+      "step": 3628
+    },
+    {
+      "epoch": 3.868869936034115,
+      "grad_norm": 0.11831696552244149,
+      "learning_rate": 1.183703253502257e-05,
+      "loss": 0.3462,
+      "step": 3629
+    },
+    {
+      "epoch": 3.8699360341151388,
+      "grad_norm": 0.11785222740074164,
+      "learning_rate": 1.1815899110562081e-05,
+      "loss": 0.3504,
+      "step": 3630
+    },
+    {
+      "epoch": 3.8710021321961623,
+      "grad_norm": 0.11362853709274134,
+      "learning_rate": 1.1794781298628725e-05,
+      "loss": 0.349,
+      "step": 3631
+    },
+    {
+      "epoch": 3.8720682302771854,
+      "grad_norm": 0.11822864612991446,
+      "learning_rate": 1.1773679110920648e-05,
+      "loss": 0.3477,
+      "step": 3632
+    },
+    {
+      "epoch": 3.873134328358209,
+      "grad_norm": 0.10399345332909624,
+      "learning_rate": 1.1752592559127378e-05,
+      "loss": 0.3473,
+      "step": 3633
+    },
+    {
+      "epoch": 3.8742004264392325,
+      "grad_norm": 0.10956934208464444,
+      "learning_rate": 1.1731521654929785e-05,
+      "loss": 0.3497,
+      "step": 3634
+    },
+    {
+      "epoch": 3.875266524520256,
+      "grad_norm": 0.12301783671574228,
+      "learning_rate": 1.1710466410000021e-05,
+      "loss": 0.3505,
+      "step": 3635
+    },
+    {
+      "epoch": 3.8763326226012795,
+      "grad_norm": 0.11590042052538112,
+      "learning_rate": 1.1689426836001618e-05,
+      "loss": 0.3478,
+      "step": 3636
+    },
+    {
+      "epoch": 3.8773987206823026,
+      "grad_norm": 0.10524997397293638,
+      "learning_rate": 1.1668402944589405e-05,
+      "loss": 0.3476,
+      "step": 3637
+    },
+    {
+      "epoch": 3.878464818763326,
+      "grad_norm": 0.13955869940094448,
+      "learning_rate": 1.1647394747409538e-05,
+      "loss": 0.3486,
+      "step": 3638
+    },
+    {
+      "epoch": 3.8795309168443497,
+      "grad_norm": 0.09733253071530201,
+      "learning_rate": 1.1626402256099439e-05,
+      "loss": 0.3501,
+      "step": 3639
+    },
+    {
+      "epoch": 3.8805970149253732,
+      "grad_norm": 0.13743368067281408,
+      "learning_rate": 1.1605425482287869e-05,
+      "loss": 0.3448,
+      "step": 3640
+    },
+    {
+      "epoch": 3.8816631130063968,
+      "grad_norm": 0.13018722986242173,
+      "learning_rate": 1.1584464437594875e-05,
+      "loss": 0.3507,
+      "step": 3641
+    },
+    {
+      "epoch": 3.88272921108742,
+      "grad_norm": 0.11788165392769949,
+      "learning_rate": 1.1563519133631793e-05,
+      "loss": 0.3506,
+      "step": 3642
+    },
+    {
+      "epoch": 3.8837953091684434,
+      "grad_norm": 0.13701749120457338,
+      "learning_rate": 1.1542589582001215e-05,
+      "loss": 0.3456,
+      "step": 3643
+    },
+    {
+      "epoch": 3.884861407249467,
+      "grad_norm": 0.10813589381609363,
+      "learning_rate": 1.1521675794297028e-05,
+      "loss": 0.3483,
+      "step": 3644
+    },
+    {
+      "epoch": 3.8859275053304905,
+      "grad_norm": 0.1224953861353445,
+      "learning_rate": 1.1500777782104406e-05,
+      "loss": 0.3417,
+      "step": 3645
+    },
+    {
+      "epoch": 3.886993603411514,
+      "grad_norm": 0.11152538352587757,
+      "learning_rate": 1.1479895556999731e-05,
+      "loss": 0.3514,
+      "step": 3646
+    },
+    {
+      "epoch": 3.888059701492537,
+      "grad_norm": 0.11316497383817867,
+      "learning_rate": 1.145902913055068e-05,
+      "loss": 0.3523,
+      "step": 3647
+    },
+    {
+      "epoch": 3.8891257995735606,
+      "grad_norm": 0.11019891396827555,
+      "learning_rate": 1.1438178514316181e-05,
+      "loss": 0.3498,
+      "step": 3648
+    },
+    {
+      "epoch": 3.890191897654584,
+      "grad_norm": 0.10842084319722904,
+      "learning_rate": 1.1417343719846387e-05,
+      "loss": 0.354,
+      "step": 3649
+    },
+    {
+      "epoch": 3.8912579957356077,
+      "grad_norm": 0.1056924090606256,
+      "learning_rate": 1.1396524758682678e-05,
+      "loss": 0.3456,
+      "step": 3650
+    },
+    {
+      "epoch": 3.8923240938166312,
+      "grad_norm": 0.10490460770794618,
+      "learning_rate": 1.137572164235769e-05,
+      "loss": 0.3492,
+      "step": 3651
+    },
+    {
+      "epoch": 3.8933901918976543,
+      "grad_norm": 0.09975927894418671,
+      "learning_rate": 1.1354934382395272e-05,
+      "loss": 0.3509,
+      "step": 3652
+    },
+    {
+      "epoch": 3.894456289978678,
+      "grad_norm": 0.10145902183398905,
+      "learning_rate": 1.1334162990310471e-05,
+      "loss": 0.3447,
+      "step": 3653
+    },
+    {
+      "epoch": 3.8955223880597014,
+      "grad_norm": 0.10481224542330943,
+      "learning_rate": 1.1313407477609561e-05,
+      "loss": 0.347,
+      "step": 3654
+    },
+    {
+      "epoch": 3.896588486140725,
+      "grad_norm": 0.10001648400892982,
+      "learning_rate": 1.1292667855790027e-05,
+      "loss": 0.3513,
+      "step": 3655
+    },
+    {
+      "epoch": 3.8976545842217485,
+      "grad_norm": 0.09792997968349304,
+      "learning_rate": 1.1271944136340544e-05,
+      "loss": 0.3423,
+      "step": 3656
+    },
+    {
+      "epoch": 3.8987206823027716,
+      "grad_norm": 0.1087579485635863,
+      "learning_rate": 1.1251236330740962e-05,
+      "loss": 0.3518,
+      "step": 3657
+    },
+    {
+      "epoch": 3.8997867803837956,
+      "grad_norm": 0.10273076928541128,
+      "learning_rate": 1.123054445046233e-05,
+      "loss": 0.3403,
+      "step": 3658
+    },
+    {
+      "epoch": 3.9008528784648187,
+      "grad_norm": 0.11360397048458296,
+      "learning_rate": 1.1209868506966881e-05,
+      "loss": 0.3479,
+      "step": 3659
+    },
+    {
+      "epoch": 3.901918976545842,
+      "grad_norm": 0.09676431212644183,
+      "learning_rate": 1.118920851170803e-05,
+      "loss": 0.3412,
+      "step": 3660
+    },
+    {
+      "epoch": 3.9029850746268657,
+      "grad_norm": 0.09307333554900554,
+      "learning_rate": 1.1168564476130301e-05,
+      "loss": 0.3482,
+      "step": 3661
+    },
+    {
+      "epoch": 3.9040511727078893,
+      "grad_norm": 0.10397754347232523,
+      "learning_rate": 1.1147936411669446e-05,
+      "loss": 0.3481,
+      "step": 3662
+    },
+    {
+      "epoch": 3.905117270788913,
+      "grad_norm": 0.08937777918666276,
+      "learning_rate": 1.1127324329752342e-05,
+      "loss": 0.3483,
+      "step": 3663
+    },
+    {
+      "epoch": 3.906183368869936,
+      "grad_norm": 0.09042966313131054,
+      "learning_rate": 1.110672824179699e-05,
+      "loss": 0.3487,
+      "step": 3664
+    },
+    {
+      "epoch": 3.9072494669509594,
+      "grad_norm": 0.09557462612752123,
+      "learning_rate": 1.1086148159212562e-05,
+      "loss": 0.3536,
+      "step": 3665
+    },
+    {
+      "epoch": 3.908315565031983,
+      "grad_norm": 0.09035818870121247,
+      "learning_rate": 1.1065584093399373e-05,
+      "loss": 0.3519,
+      "step": 3666
+    },
+    {
+      "epoch": 3.9093816631130065,
+      "grad_norm": 0.0942584757184624,
+      "learning_rate": 1.1045036055748817e-05,
+      "loss": 0.3448,
+      "step": 3667
+    },
+    {
+      "epoch": 3.91044776119403,
+      "grad_norm": 0.09364257142264377,
+      "learning_rate": 1.102450405764345e-05,
+      "loss": 0.3504,
+      "step": 3668
+    },
+    {
+      "epoch": 3.911513859275053,
+      "grad_norm": 0.09007339517858659,
+      "learning_rate": 1.100398811045695e-05,
+      "loss": 0.3501,
+      "step": 3669
+    },
+    {
+      "epoch": 3.9125799573560767,
+      "grad_norm": 0.08911386586011126,
+      "learning_rate": 1.0983488225554053e-05,
+      "loss": 0.3493,
+      "step": 3670
+    },
+    {
+      "epoch": 3.9136460554371,
+      "grad_norm": 0.09675034869903551,
+      "learning_rate": 1.0963004414290653e-05,
+      "loss": 0.3572,
+      "step": 3671
+    },
+    {
+      "epoch": 3.9147121535181237,
+      "grad_norm": 0.09188043129655729,
+      "learning_rate": 1.0942536688013713e-05,
+      "loss": 0.3541,
+      "step": 3672
+    },
+    {
+      "epoch": 3.9157782515991473,
+      "grad_norm": 0.09937464367903363,
+      "learning_rate": 1.0922085058061285e-05,
+      "loss": 0.3482,
+      "step": 3673
+    },
+    {
+      "epoch": 3.9168443496801704,
+      "grad_norm": 0.09926679761687571,
+      "learning_rate": 1.0901649535762506e-05,
+      "loss": 0.3567,
+      "step": 3674
+    },
+    {
+      "epoch": 3.917910447761194,
+      "grad_norm": 0.09965290213224968,
+      "learning_rate": 1.0881230132437608e-05,
+      "loss": 0.3462,
+      "step": 3675
+    },
+    {
+      "epoch": 3.9189765458422174,
+      "grad_norm": 0.1003530954835924,
+      "learning_rate": 1.0860826859397862e-05,
+      "loss": 0.3472,
+      "step": 3676
+    },
+    {
+      "epoch": 3.920042643923241,
+      "grad_norm": 0.09596106891446698,
+      "learning_rate": 1.0840439727945626e-05,
+      "loss": 0.3479,
+      "step": 3677
+    },
+    {
+      "epoch": 3.9211087420042645,
+      "grad_norm": 0.09919666795826218,
+      "learning_rate": 1.0820068749374327e-05,
+      "loss": 0.3449,
+      "step": 3678
+    },
+    {
+      "epoch": 3.9221748400852876,
+      "grad_norm": 0.1012132428938292,
+      "learning_rate": 1.0799713934968406e-05,
+      "loss": 0.3512,
+      "step": 3679
+    },
+    {
+      "epoch": 3.923240938166311,
+      "grad_norm": 0.09755542244495895,
+      "learning_rate": 1.0779375296003374e-05,
+      "loss": 0.352,
+      "step": 3680
+    },
+    {
+      "epoch": 3.9243070362473347,
+      "grad_norm": 0.09000645950013296,
+      "learning_rate": 1.0759052843745806e-05,
+      "loss": 0.3468,
+      "step": 3681
+    },
+    {
+      "epoch": 3.925373134328358,
+      "grad_norm": 0.09876937193470242,
+      "learning_rate": 1.073874658945325e-05,
+      "loss": 0.3504,
+      "step": 3682
+    },
+    {
+      "epoch": 3.9264392324093818,
+      "grad_norm": 0.09239651298132856,
+      "learning_rate": 1.0718456544374333e-05,
+      "loss": 0.3491,
+      "step": 3683
+    },
+    {
+      "epoch": 3.927505330490405,
+      "grad_norm": 0.10058467408628505,
+      "learning_rate": 1.0698182719748682e-05,
+      "loss": 0.3511,
+      "step": 3684
+    },
+    {
+      "epoch": 3.928571428571429,
+      "grad_norm": 0.08600735325668393,
+      "learning_rate": 1.0677925126806956e-05,
+      "loss": 0.3421,
+      "step": 3685
+    },
+    {
+      "epoch": 3.929637526652452,
+      "grad_norm": 0.09083257850771187,
+      "learning_rate": 1.0657683776770788e-05,
+      "loss": 0.3493,
+      "step": 3686
+    },
+    {
+      "epoch": 3.9307036247334755,
+      "grad_norm": 0.09314333955394852,
+      "learning_rate": 1.0637458680852841e-05,
+      "loss": 0.3533,
+      "step": 3687
+    },
+    {
+      "epoch": 3.931769722814499,
+      "grad_norm": 0.09360495013685566,
+      "learning_rate": 1.0617249850256766e-05,
+      "loss": 0.349,
+      "step": 3688
+    },
+    {
+      "epoch": 3.9328358208955225,
+      "grad_norm": 0.09627416992471283,
+      "learning_rate": 1.0597057296177225e-05,
+      "loss": 0.3516,
+      "step": 3689
+    },
+    {
+      "epoch": 3.933901918976546,
+      "grad_norm": 0.09541087809563052,
+      "learning_rate": 1.0576881029799808e-05,
+      "loss": 0.3543,
+      "step": 3690
+    },
+    {
+      "epoch": 3.934968017057569,
+      "grad_norm": 0.10348458965898262,
+      "learning_rate": 1.0556721062301141e-05,
+      "loss": 0.3567,
+      "step": 3691
+    },
+    {
+      "epoch": 3.9360341151385927,
+      "grad_norm": 0.09872750858387926,
+      "learning_rate": 1.05365774048488e-05,
+      "loss": 0.3491,
+      "step": 3692
+    },
+    {
+      "epoch": 3.9371002132196162,
+      "grad_norm": 0.10350018663429181,
+      "learning_rate": 1.051645006860131e-05,
+      "loss": 0.3498,
+      "step": 3693
+    },
+    {
+      "epoch": 3.9381663113006398,
+      "grad_norm": 0.10594157752622746,
+      "learning_rate": 1.0496339064708172e-05,
+      "loss": 0.3506,
+      "step": 3694
+    },
+    {
+      "epoch": 3.9392324093816633,
+      "grad_norm": 0.10372816420096216,
+      "learning_rate": 1.0476244404309846e-05,
+      "loss": 0.3514,
+      "step": 3695
+    },
+    {
+      "epoch": 3.9402985074626864,
+      "grad_norm": 0.10601590445244279,
+      "learning_rate": 1.0456166098537737e-05,
+      "loss": 0.3444,
+      "step": 3696
+    },
+    {
+      "epoch": 3.94136460554371,
+      "grad_norm": 0.11008051797371131,
+      "learning_rate": 1.0436104158514158e-05,
+      "loss": 0.3473,
+      "step": 3697
+    },
+    {
+      "epoch": 3.9424307036247335,
+      "grad_norm": 0.09606351168418183,
+      "learning_rate": 1.0416058595352391e-05,
+      "loss": 0.3484,
+      "step": 3698
+    },
+    {
+      "epoch": 3.943496801705757,
+      "grad_norm": 0.1071996803306346,
+      "learning_rate": 1.039602942015664e-05,
+      "loss": 0.3476,
+      "step": 3699
+    },
+    {
+      "epoch": 3.9445628997867805,
+      "grad_norm": 0.09623667142559411,
+      "learning_rate": 1.0376016644022044e-05,
+      "loss": 0.3447,
+      "step": 3700
+    },
+    {
+      "epoch": 3.9456289978678036,
+      "grad_norm": 0.10721079752870356,
+      "learning_rate": 1.035602027803461e-05,
+      "loss": 0.3469,
+      "step": 3701
+    },
+    {
+      "epoch": 3.946695095948827,
+      "grad_norm": 0.1025843055534788,
+      "learning_rate": 1.0336040333271295e-05,
+      "loss": 0.3447,
+      "step": 3702
+    },
+    {
+      "epoch": 3.9477611940298507,
+      "grad_norm": 0.09871032111134458,
+      "learning_rate": 1.0316076820799968e-05,
+      "loss": 0.3529,
+      "step": 3703
+    },
+    {
+      "epoch": 3.9488272921108742,
+      "grad_norm": 0.10311671354974446,
+      "learning_rate": 1.029612975167935e-05,
+      "loss": 0.3474,
+      "step": 3704
+    },
+    {
+      "epoch": 3.949893390191898,
+      "grad_norm": 0.10336281514714324,
+      "learning_rate": 1.0276199136959097e-05,
+      "loss": 0.3512,
+      "step": 3705
+    },
+    {
+      "epoch": 3.950959488272921,
+      "grad_norm": 0.10601850200687606,
+      "learning_rate": 1.025628498767973e-05,
+      "loss": 0.3514,
+      "step": 3706
+    },
+    {
+      "epoch": 3.9520255863539444,
+      "grad_norm": 0.08908113002105139,
+      "learning_rate": 1.0236387314872664e-05,
+      "loss": 0.3435,
+      "step": 3707
+    },
+    {
+      "epoch": 3.953091684434968,
+      "grad_norm": 0.10671007341009846,
+      "learning_rate": 1.0216506129560155e-05,
+      "loss": 0.3551,
+      "step": 3708
+    },
+    {
+      "epoch": 3.9541577825159915,
+      "grad_norm": 0.09784872373144313,
+      "learning_rate": 1.0196641442755354e-05,
+      "loss": 0.3466,
+      "step": 3709
+    },
+    {
+      "epoch": 3.955223880597015,
+      "grad_norm": 0.11088241430701273,
+      "learning_rate": 1.0176793265462282e-05,
+      "loss": 0.3455,
+      "step": 3710
+    },
+    {
+      "epoch": 3.956289978678038,
+      "grad_norm": 0.0910427685127364,
+      "learning_rate": 1.0156961608675768e-05,
+      "loss": 0.3505,
+      "step": 3711
+    },
+    {
+      "epoch": 3.957356076759062,
+      "grad_norm": 0.10461641359651551,
+      "learning_rate": 1.0137146483381538e-05,
+      "loss": 0.354,
+      "step": 3712
+    },
+    {
+      "epoch": 3.958422174840085,
+      "grad_norm": 0.09627545711712145,
+      "learning_rate": 1.0117347900556137e-05,
+      "loss": 0.3513,
+      "step": 3713
+    },
+    {
+      "epoch": 3.9594882729211087,
+      "grad_norm": 0.1073622509994735,
+      "learning_rate": 1.0097565871166961e-05,
+      "loss": 0.3546,
+      "step": 3714
+    },
+    {
+      "epoch": 3.9605543710021323,
+      "grad_norm": 0.09484625771277952,
+      "learning_rate": 1.0077800406172207e-05,
+      "loss": 0.3524,
+      "step": 3715
+    },
+    {
+      "epoch": 3.961620469083156,
+      "grad_norm": 0.08894481152860242,
+      "learning_rate": 1.0058051516520929e-05,
+      "loss": 0.3474,
+      "step": 3716
+    },
+    {
+      "epoch": 3.9626865671641793,
+      "grad_norm": 0.09733312978906855,
+      "learning_rate": 1.0038319213152979e-05,
+      "loss": 0.3495,
+      "step": 3717
+    },
+    {
+      "epoch": 3.9637526652452024,
+      "grad_norm": 0.09864930949926093,
+      "learning_rate": 1.001860350699904e-05,
+      "loss": 0.3496,
+      "step": 3718
+    },
+    {
+      "epoch": 3.964818763326226,
+      "grad_norm": 0.0981894103889243,
+      "learning_rate": 9.99890440898057e-06,
+      "loss": 0.3529,
+      "step": 3719
+    },
+    {
+      "epoch": 3.9658848614072495,
+      "grad_norm": 0.10482896841184723,
+      "learning_rate": 9.97922193000985e-06,
+      "loss": 0.3519,
+      "step": 3720
+    },
+    {
+      "epoch": 3.966950959488273,
+      "grad_norm": 0.10498811386677907,
+      "learning_rate": 9.959556080989973e-06,
+      "loss": 0.3488,
+      "step": 3721
+    },
+    {
+      "epoch": 3.9680170575692966,
+      "grad_norm": 0.11366999497073144,
+      "learning_rate": 9.939906872814764e-06,
+      "loss": 0.3535,
+      "step": 3722
+    },
+    {
+      "epoch": 3.9690831556503197,
+      "grad_norm": 0.09665112411107446,
+      "learning_rate": 9.920274316368879e-06,
+      "loss": 0.3511,
+      "step": 3723
+    },
+    {
+      "epoch": 3.970149253731343,
+      "grad_norm": 0.10146451503365203,
+      "learning_rate": 9.900658422527734e-06,
+      "loss": 0.346,
+      "step": 3724
+    },
+    {
+      "epoch": 3.9712153518123667,
+      "grad_norm": 0.10925095180031567,
+      "learning_rate": 9.881059202157525e-06,
+      "loss": 0.3507,
+      "step": 3725
+    },
+    {
+      "epoch": 3.9722814498933903,
+      "grad_norm": 0.09225433257638278,
+      "learning_rate": 9.86147666611518e-06,
+      "loss": 0.3482,
+      "step": 3726
+    },
+    {
+      "epoch": 3.973347547974414,
+      "grad_norm": 0.0943825856008904,
+      "learning_rate": 9.841910825248412e-06,
+      "loss": 0.3464,
+      "step": 3727
+    },
+    {
+      "epoch": 3.974413646055437,
+      "grad_norm": 0.09290035671178411,
+      "learning_rate": 9.82236169039569e-06,
+      "loss": 0.3449,
+      "step": 3728
+    },
+    {
+      "epoch": 3.9754797441364604,
+      "grad_norm": 0.0921926282044607,
+      "learning_rate": 9.802829272386227e-06,
+      "loss": 0.3554,
+      "step": 3729
+    },
+    {
+      "epoch": 3.976545842217484,
+      "grad_norm": 0.0960555525918967,
+      "learning_rate": 9.783313582039935e-06,
+      "loss": 0.3483,
+      "step": 3730
+    },
+    {
+      "epoch": 3.9776119402985075,
+      "grad_norm": 0.0917415114727513,
+      "learning_rate": 9.763814630167516e-06,
+      "loss": 0.3458,
+      "step": 3731
+    },
+    {
+      "epoch": 3.978678038379531,
+      "grad_norm": 0.08801795546770015,
+      "learning_rate": 9.744332427570384e-06,
+      "loss": 0.3436,
+      "step": 3732
+    },
+    {
+      "epoch": 3.979744136460554,
+      "grad_norm": 0.09754716872379321,
+      "learning_rate": 9.72486698504064e-06,
+      "loss": 0.3406,
+      "step": 3733
+    },
+    {
+      "epoch": 3.9808102345415777,
+      "grad_norm": 0.09042401583271688,
+      "learning_rate": 9.705418313361141e-06,
+      "loss": 0.3486,
+      "step": 3734
+    },
+    {
+      "epoch": 3.981876332622601,
+      "grad_norm": 0.08028758020380314,
+      "learning_rate": 9.685986423305449e-06,
+      "loss": 0.3494,
+      "step": 3735
+    },
+    {
+      "epoch": 3.9829424307036247,
+      "grad_norm": 0.09007532159921652,
+      "learning_rate": 9.666571325637806e-06,
+      "loss": 0.3482,
+      "step": 3736
+    },
+    {
+      "epoch": 3.9840085287846483,
+      "grad_norm": 0.09260516061644863,
+      "learning_rate": 9.647173031113173e-06,
+      "loss": 0.3452,
+      "step": 3737
+    },
+    {
+      "epoch": 3.9850746268656714,
+      "grad_norm": 0.09089810495372147,
+      "learning_rate": 9.627791550477209e-06,
+      "loss": 0.3462,
+      "step": 3738
+    },
+    {
+      "epoch": 3.9861407249466954,
+      "grad_norm": 0.09771736284518276,
+      "learning_rate": 9.608426894466225e-06,
+      "loss": 0.3532,
+      "step": 3739
+    },
+    {
+      "epoch": 3.9872068230277184,
+      "grad_norm": 0.1259794070436244,
+      "learning_rate": 9.589079073807244e-06,
+      "loss": 0.349,
+      "step": 3740
+    },
+    {
+      "epoch": 3.988272921108742,
+      "grad_norm": 0.09140483649945423,
+      "learning_rate": 9.569748099217962e-06,
+      "loss": 0.3433,
+      "step": 3741
+    },
+    {
+      "epoch": 3.9893390191897655,
+      "grad_norm": 0.12248710790523741,
+      "learning_rate": 9.55043398140672e-06,
+      "loss": 0.3522,
+      "step": 3742
+    },
+    {
+      "epoch": 3.990405117270789,
+      "grad_norm": 0.09833501236467464,
+      "learning_rate": 9.53113673107254e-06,
+      "loss": 0.3531,
+      "step": 3743
+    },
+    {
+      "epoch": 3.9914712153518126,
+      "grad_norm": 0.10722069561647879,
+      "learning_rate": 9.511856358905108e-06,
+      "loss": 0.3513,
+      "step": 3744
+    },
+    {
+      "epoch": 3.9925373134328357,
+      "grad_norm": 0.12373429125821357,
+      "learning_rate": 9.492592875584728e-06,
+      "loss": 0.3498,
+      "step": 3745
+    },
+    {
+      "epoch": 3.9936034115138592,
+      "grad_norm": 0.10792510560456024,
+      "learning_rate": 9.473346291782376e-06,
+      "loss": 0.3532,
+      "step": 3746
+    },
+    {
+      "epoch": 3.9946695095948828,
+      "grad_norm": 0.10969296551778704,
+      "learning_rate": 9.454116618159675e-06,
+      "loss": 0.3481,
+      "step": 3747
+    },
+    {
+      "epoch": 3.9957356076759063,
+      "grad_norm": 0.10095085119982077,
+      "learning_rate": 9.434903865368837e-06,
+      "loss": 0.3484,
+      "step": 3748
+    },
+    {
+      "epoch": 3.99680170575693,
+      "grad_norm": 0.1008779164840716,
+      "learning_rate": 9.415708044052744e-06,
+      "loss": 0.3515,
+      "step": 3749
+    },
+    {
+      "epoch": 3.997867803837953,
+      "grad_norm": 0.09334484031468043,
+      "learning_rate": 9.396529164844893e-06,
+      "loss": 0.3536,
+      "step": 3750
+    },
+    {
+      "epoch": 3.9989339019189765,
+      "grad_norm": 0.08642194713513292,
+      "learning_rate": 9.377367238369368e-06,
+      "loss": 0.3451,
+      "step": 3751
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.1134408652616496,
+      "learning_rate": 9.358222275240884e-06,
+      "loss": 0.336,
+      "step": 3752
+    },
+    {
+      "epoch": 4.001066098081023,
+      "grad_norm": 0.104623946486961,
+      "learning_rate": 9.33909428606476e-06,
+      "loss": 0.3372,
+      "step": 3753
+    },
+    {
+      "epoch": 4.002132196162047,
+      "grad_norm": 0.10983574887630808,
+      "learning_rate": 9.31998328143692e-06,
+      "loss": 0.3316,
+      "step": 3754
+    },
+    {
+      "epoch": 4.00319829424307,
+      "grad_norm": 0.10886704749727195,
+      "learning_rate": 9.30088927194384e-06,
+      "loss": 0.3346,
+      "step": 3755
+    },
+    {
+      "epoch": 4.004264392324094,
+      "grad_norm": 0.11057981475076427,
+      "learning_rate": 9.281812268162626e-06,
+      "loss": 0.3377,
+      "step": 3756
+    },
+    {
+      "epoch": 4.005330490405117,
+      "grad_norm": 0.12400384882704346,
+      "learning_rate": 9.262752280660944e-06,
+      "loss": 0.3342,
+      "step": 3757
+    },
+    {
+      "epoch": 4.00639658848614,
+      "grad_norm": 0.11298478774918766,
+      "learning_rate": 9.243709319997047e-06,
+      "loss": 0.338,
+      "step": 3758
+    },
+    {
+      "epoch": 4.007462686567164,
+      "grad_norm": 0.13596408021241022,
+      "learning_rate": 9.224683396719728e-06,
+      "loss": 0.335,
+      "step": 3759
+    },
+    {
+      "epoch": 4.008528784648187,
+      "grad_norm": 0.11672755460945555,
+      "learning_rate": 9.205674521368362e-06,
+      "loss": 0.3316,
+      "step": 3760
+    },
+    {
+      "epoch": 4.009594882729211,
+      "grad_norm": 0.11331525635027023,
+      "learning_rate": 9.186682704472898e-06,
+      "loss": 0.3294,
+      "step": 3761
+    },
+    {
+      "epoch": 4.0106609808102345,
+      "grad_norm": 0.1195438195692172,
+      "learning_rate": 9.167707956553787e-06,
+      "loss": 0.3343,
+      "step": 3762
+    },
+    {
+      "epoch": 4.011727078891258,
+      "grad_norm": 0.1091957644805815,
+      "learning_rate": 9.148750288122063e-06,
+      "loss": 0.3348,
+      "step": 3763
+    },
+    {
+      "epoch": 4.0127931769722816,
+      "grad_norm": 0.11204673273341918,
+      "learning_rate": 9.129809709679297e-06,
+      "loss": 0.3321,
+      "step": 3764
+    },
+    {
+      "epoch": 4.013859275053305,
+      "grad_norm": 0.10282221122253991,
+      "learning_rate": 9.110886231717595e-06,
+      "loss": 0.3339,
+      "step": 3765
+    },
+    {
+      "epoch": 4.014925373134329,
+      "grad_norm": 0.10864476435323264,
+      "learning_rate": 9.09197986471955e-06,
+      "loss": 0.3355,
+      "step": 3766
+    },
+    {
+      "epoch": 4.015991471215352,
+      "grad_norm": 0.10772893714735123,
+      "learning_rate": 9.073090619158322e-06,
+      "loss": 0.3306,
+      "step": 3767
+    },
+    {
+      "epoch": 4.017057569296376,
+      "grad_norm": 0.11249780372035993,
+      "learning_rate": 9.054218505497587e-06,
+      "loss": 0.3343,
+      "step": 3768
+    },
+    {
+      "epoch": 4.018123667377399,
+      "grad_norm": 0.1126010464588701,
+      "learning_rate": 9.035363534191486e-06,
+      "loss": 0.3312,
+      "step": 3769
+    },
+    {
+      "epoch": 4.019189765458422,
+      "grad_norm": 0.12907028042017912,
+      "learning_rate": 9.016525715684711e-06,
+      "loss": 0.3327,
+      "step": 3770
+    },
+    {
+      "epoch": 4.020255863539446,
+      "grad_norm": 0.11807831450158258,
+      "learning_rate": 8.99770506041243e-06,
+      "loss": 0.3386,
+      "step": 3771
+    },
+    {
+      "epoch": 4.021321961620469,
+      "grad_norm": 0.10963014735265454,
+      "learning_rate": 8.978901578800316e-06,
+      "loss": 0.3309,
+      "step": 3772
+    },
+    {
+      "epoch": 4.022388059701493,
+      "grad_norm": 0.11867805956803665,
+      "learning_rate": 8.960115281264507e-06,
+      "loss": 0.3423,
+      "step": 3773
+    },
+    {
+      "epoch": 4.023454157782516,
+      "grad_norm": 0.11598119651875034,
+      "learning_rate": 8.941346178211639e-06,
+      "loss": 0.3281,
+      "step": 3774
+    },
+    {
+      "epoch": 4.024520255863539,
+      "grad_norm": 0.10424229151793267,
+      "learning_rate": 8.922594280038823e-06,
+      "loss": 0.323,
+      "step": 3775
+    },
+    {
+      "epoch": 4.025586353944563,
+      "grad_norm": 0.10762211731941046,
+      "learning_rate": 8.903859597133646e-06,
+      "loss": 0.3344,
+      "step": 3776
+    },
+    {
+      "epoch": 4.026652452025586,
+      "grad_norm": 0.11394090140716011,
+      "learning_rate": 8.88514213987413e-06,
+      "loss": 0.3331,
+      "step": 3777
+    },
+    {
+      "epoch": 4.02771855010661,
+      "grad_norm": 0.09343973744188745,
+      "learning_rate": 8.866441918628777e-06,
+      "loss": 0.333,
+      "step": 3778
+    },
+    {
+      "epoch": 4.028784648187633,
+      "grad_norm": 0.11583112915467723,
+      "learning_rate": 8.847758943756556e-06,
+      "loss": 0.3385,
+      "step": 3779
+    },
+    {
+      "epoch": 4.029850746268656,
+      "grad_norm": 0.10675906671526333,
+      "learning_rate": 8.829093225606842e-06,
+      "loss": 0.3338,
+      "step": 3780
+    },
+    {
+      "epoch": 4.03091684434968,
+      "grad_norm": 0.10418803649968214,
+      "learning_rate": 8.810444774519475e-06,
+      "loss": 0.3383,
+      "step": 3781
+    },
+    {
+      "epoch": 4.031982942430703,
+      "grad_norm": 0.12106960416467054,
+      "learning_rate": 8.791813600824742e-06,
+      "loss": 0.3378,
+      "step": 3782
+    },
+    {
+      "epoch": 4.033049040511727,
+      "grad_norm": 0.11769429634565491,
+      "learning_rate": 8.773199714843339e-06,
+      "loss": 0.3326,
+      "step": 3783
+    },
+    {
+      "epoch": 4.0341151385927505,
+      "grad_norm": 0.11355409773845418,
+      "learning_rate": 8.754603126886385e-06,
+      "loss": 0.3351,
+      "step": 3784
+    },
+    {
+      "epoch": 4.035181236673774,
+      "grad_norm": 0.09880842808847773,
+      "learning_rate": 8.736023847255426e-06,
+      "loss": 0.3356,
+      "step": 3785
+    },
+    {
+      "epoch": 4.036247334754798,
+      "grad_norm": 0.11158321630050068,
+      "learning_rate": 8.71746188624242e-06,
+      "loss": 0.3363,
+      "step": 3786
+    },
+    {
+      "epoch": 4.037313432835821,
+      "grad_norm": 0.1161769963643131,
+      "learning_rate": 8.698917254129732e-06,
+      "loss": 0.3328,
+      "step": 3787
+    },
+    {
+      "epoch": 4.038379530916845,
+      "grad_norm": 0.11351757271831203,
+      "learning_rate": 8.680389961190116e-06,
+      "loss": 0.3311,
+      "step": 3788
+    },
+    {
+      "epoch": 4.039445628997868,
+      "grad_norm": 0.10521775128019155,
+      "learning_rate": 8.66188001768673e-06,
+      "loss": 0.3346,
+      "step": 3789
+    },
+    {
+      "epoch": 4.040511727078891,
+      "grad_norm": 0.10112600207752875,
+      "learning_rate": 8.64338743387314e-06,
+      "loss": 0.3365,
+      "step": 3790
+    },
+    {
+      "epoch": 4.041577825159915,
+      "grad_norm": 0.10617915170837067,
+      "learning_rate": 8.624912219993248e-06,
+      "loss": 0.3332,
+      "step": 3791
+    },
+    {
+      "epoch": 4.042643923240938,
+      "grad_norm": 0.10169862043274906,
+      "learning_rate": 8.606454386281368e-06,
+      "loss": 0.3276,
+      "step": 3792
+    },
+    {
+      "epoch": 4.043710021321962,
+      "grad_norm": 0.10543707108073325,
+      "learning_rate": 8.588013942962195e-06,
+      "loss": 0.337,
+      "step": 3793
+    },
+    {
+      "epoch": 4.044776119402985,
+      "grad_norm": 0.10597483996500817,
+      "learning_rate": 8.569590900250775e-06,
+      "loss": 0.3389,
+      "step": 3794
+    },
+    {
+      "epoch": 4.045842217484009,
+      "grad_norm": 0.10326713662353049,
+      "learning_rate": 8.551185268352502e-06,
+      "loss": 0.3337,
+      "step": 3795
+    },
+    {
+      "epoch": 4.046908315565032,
+      "grad_norm": 0.0966216388278116,
+      "learning_rate": 8.532797057463145e-06,
+      "loss": 0.3398,
+      "step": 3796
+    },
+    {
+      "epoch": 4.047974413646055,
+      "grad_norm": 0.09422934050818582,
+      "learning_rate": 8.51442627776883e-06,
+      "loss": 0.3373,
+      "step": 3797
+    },
+    {
+      "epoch": 4.049040511727079,
+      "grad_norm": 0.1036405776308619,
+      "learning_rate": 8.496072939445997e-06,
+      "loss": 0.3336,
+      "step": 3798
+    },
+    {
+      "epoch": 4.050106609808102,
+      "grad_norm": 0.10297234333425688,
+      "learning_rate": 8.477737052661444e-06,
+      "loss": 0.3357,
+      "step": 3799
+    },
+    {
+      "epoch": 4.051172707889126,
+      "grad_norm": 0.09051843440055769,
+      "learning_rate": 8.459418627572304e-06,
+      "loss": 0.3356,
+      "step": 3800
+    },
+    {
+      "epoch": 4.052238805970149,
+      "grad_norm": 0.09979444399952614,
+      "learning_rate": 8.44111767432604e-06,
+      "loss": 0.3383,
+      "step": 3801
+    },
+    {
+      "epoch": 4.053304904051172,
+      "grad_norm": 0.09217956240397418,
+      "learning_rate": 8.422834203060418e-06,
+      "loss": 0.3324,
+      "step": 3802
+    },
+    {
+      "epoch": 4.054371002132196,
+      "grad_norm": 0.09053551898130377,
+      "learning_rate": 8.404568223903529e-06,
+      "loss": 0.3336,
+      "step": 3803
+    },
+    {
+      "epoch": 4.0554371002132195,
+      "grad_norm": 0.09177723542384852,
+      "learning_rate": 8.386319746973787e-06,
+      "loss": 0.3339,
+      "step": 3804
+    },
+    {
+      "epoch": 4.056503198294243,
+      "grad_norm": 0.09189669268340435,
+      "learning_rate": 8.36808878237989e-06,
+      "loss": 0.3382,
+      "step": 3805
+    },
+    {
+      "epoch": 4.0575692963752665,
+      "grad_norm": 0.0942192210220161,
+      "learning_rate": 8.349875340220847e-06,
+      "loss": 0.327,
+      "step": 3806
+    },
+    {
+      "epoch": 4.05863539445629,
+      "grad_norm": 0.09403630781253433,
+      "learning_rate": 8.331679430585971e-06,
+      "loss": 0.3406,
+      "step": 3807
+    },
+    {
+      "epoch": 4.059701492537314,
+      "grad_norm": 0.10368304373550687,
+      "learning_rate": 8.313501063554827e-06,
+      "loss": 0.3329,
+      "step": 3808
+    },
+    {
+      "epoch": 4.060767590618337,
+      "grad_norm": 0.09734440431718942,
+      "learning_rate": 8.295340249197301e-06,
+      "loss": 0.3339,
+      "step": 3809
+    },
+    {
+      "epoch": 4.061833688699361,
+      "grad_norm": 0.09789072226393697,
+      "learning_rate": 8.277196997573545e-06,
+      "loss": 0.3392,
+      "step": 3810
+    },
+    {
+      "epoch": 4.062899786780384,
+      "grad_norm": 0.11027941919044942,
+      "learning_rate": 8.259071318733962e-06,
+      "loss": 0.3326,
+      "step": 3811
+    },
+    {
+      "epoch": 4.063965884861407,
+      "grad_norm": 0.09341818511371246,
+      "learning_rate": 8.240963222719243e-06,
+      "loss": 0.3304,
+      "step": 3812
+    },
+    {
+      "epoch": 4.065031982942431,
+      "grad_norm": 0.0985386753548714,
+      "learning_rate": 8.222872719560339e-06,
+      "loss": 0.3373,
+      "step": 3813
+    },
+    {
+      "epoch": 4.066098081023454,
+      "grad_norm": 0.09601340900666522,
+      "learning_rate": 8.204799819278438e-06,
+      "loss": 0.3352,
+      "step": 3814
+    },
+    {
+      "epoch": 4.067164179104478,
+      "grad_norm": 0.11111998488141665,
+      "learning_rate": 8.186744531884989e-06,
+      "loss": 0.3399,
+      "step": 3815
+    },
+    {
+      "epoch": 4.068230277185501,
+      "grad_norm": 0.09039086409906084,
+      "learning_rate": 8.168706867381692e-06,
+      "loss": 0.3316,
+      "step": 3816
+    },
+    {
+      "epoch": 4.069296375266524,
+      "grad_norm": 0.09819313076741153,
+      "learning_rate": 8.150686835760467e-06,
+      "loss": 0.3389,
+      "step": 3817
+    },
+    {
+      "epoch": 4.070362473347548,
+      "grad_norm": 0.10154340371524484,
+      "learning_rate": 8.132684447003471e-06,
+      "loss": 0.3348,
+      "step": 3818
+    },
+    {
+      "epoch": 4.071428571428571,
+      "grad_norm": 0.08699449860021097,
+      "learning_rate": 8.114699711083113e-06,
+      "loss": 0.3377,
+      "step": 3819
+    },
+    {
+      "epoch": 4.072494669509595,
+      "grad_norm": 0.10459986672211814,
+      "learning_rate": 8.096732637961974e-06,
+      "loss": 0.3366,
+      "step": 3820
+    },
+    {
+      "epoch": 4.073560767590618,
+      "grad_norm": 0.09394697895679324,
+      "learning_rate": 8.078783237592894e-06,
+      "loss": 0.3309,
+      "step": 3821
+    },
+    {
+      "epoch": 4.074626865671641,
+      "grad_norm": 0.0914538941685346,
+      "learning_rate": 8.060851519918901e-06,
+      "loss": 0.3361,
+      "step": 3822
+    },
+    {
+      "epoch": 4.075692963752665,
+      "grad_norm": 0.08804168096705205,
+      "learning_rate": 8.042937494873255e-06,
+      "loss": 0.3342,
+      "step": 3823
+    },
+    {
+      "epoch": 4.076759061833688,
+      "grad_norm": 0.08507323171593308,
+      "learning_rate": 8.025041172379366e-06,
+      "loss": 0.3338,
+      "step": 3824
+    },
+    {
+      "epoch": 4.077825159914712,
+      "grad_norm": 0.08903348958715734,
+      "learning_rate": 8.007162562350882e-06,
+      "loss": 0.3372,
+      "step": 3825
+    },
+    {
+      "epoch": 4.0788912579957355,
+      "grad_norm": 0.08728752400888612,
+      "learning_rate": 7.989301674691634e-06,
+      "loss": 0.3299,
+      "step": 3826
+    },
+    {
+      "epoch": 4.0799573560767595,
+      "grad_norm": 0.10425110765874443,
+      "learning_rate": 7.971458519295598e-06,
+      "loss": 0.3363,
+      "step": 3827
+    },
+    {
+      "epoch": 4.081023454157783,
+      "grad_norm": 0.08654110423017557,
+      "learning_rate": 7.953633106046971e-06,
+      "loss": 0.3362,
+      "step": 3828
+    },
+    {
+      "epoch": 4.082089552238806,
+      "grad_norm": 0.09648199398820004,
+      "learning_rate": 7.935825444820109e-06,
+      "loss": 0.3336,
+      "step": 3829
+    },
+    {
+      "epoch": 4.08315565031983,
+      "grad_norm": 0.09601930808087997,
+      "learning_rate": 7.918035545479532e-06,
+      "loss": 0.3331,
+      "step": 3830
+    },
+    {
+      "epoch": 4.084221748400853,
+      "grad_norm": 0.08913546055906947,
+      "learning_rate": 7.900263417879905e-06,
+      "loss": 0.3333,
+      "step": 3831
+    },
+    {
+      "epoch": 4.085287846481877,
+      "grad_norm": 0.08746186482392278,
+      "learning_rate": 7.882509071866074e-06,
+      "loss": 0.3351,
+      "step": 3832
+    },
+    {
+      "epoch": 4.0863539445629,
+      "grad_norm": 0.09436654519402421,
+      "learning_rate": 7.864772517273019e-06,
+      "loss": 0.3287,
+      "step": 3833
+    },
+    {
+      "epoch": 4.087420042643923,
+      "grad_norm": 0.09443565575400105,
+      "learning_rate": 7.847053763925884e-06,
+      "loss": 0.3306,
+      "step": 3834
+    },
+    {
+      "epoch": 4.088486140724947,
+      "grad_norm": 0.09012184838426746,
+      "learning_rate": 7.829352821639915e-06,
+      "loss": 0.3345,
+      "step": 3835
+    },
+    {
+      "epoch": 4.08955223880597,
+      "grad_norm": 0.08459214899851109,
+      "learning_rate": 7.811669700220523e-06,
+      "loss": 0.3338,
+      "step": 3836
+    },
+    {
+      "epoch": 4.090618336886994,
+      "grad_norm": 0.0940823716359374,
+      "learning_rate": 7.794004409463256e-06,
+      "loss": 0.3333,
+      "step": 3837
+    },
+    {
+      "epoch": 4.091684434968017,
+      "grad_norm": 0.09658670184543283,
+      "learning_rate": 7.77635695915374e-06,
+      "loss": 0.332,
+      "step": 3838
+    },
+    {
+      "epoch": 4.09275053304904,
+      "grad_norm": 0.09760335361225825,
+      "learning_rate": 7.758727359067752e-06,
+      "loss": 0.334,
+      "step": 3839
+    },
+    {
+      "epoch": 4.093816631130064,
+      "grad_norm": 0.09269720585969814,
+      "learning_rate": 7.741115618971182e-06,
+      "loss": 0.3381,
+      "step": 3840
+    },
+    {
+      "epoch": 4.094882729211087,
+      "grad_norm": 0.08955928469567048,
+      "learning_rate": 7.723521748620023e-06,
+      "loss": 0.334,
+      "step": 3841
+    },
+    {
+      "epoch": 4.095948827292111,
+      "grad_norm": 0.09989251564772907,
+      "learning_rate": 7.705945757760349e-06,
+      "loss": 0.3324,
+      "step": 3842
+    },
+    {
+      "epoch": 4.097014925373134,
+      "grad_norm": 0.08940858807005556,
+      "learning_rate": 7.688387656128355e-06,
+      "loss": 0.3462,
+      "step": 3843
+    },
+    {
+      "epoch": 4.098081023454157,
+      "grad_norm": 0.09359897320243363,
+      "learning_rate": 7.670847453450325e-06,
+      "loss": 0.3359,
+      "step": 3844
+    },
+    {
+      "epoch": 4.099147121535181,
+      "grad_norm": 0.09401701464605473,
+      "learning_rate": 7.653325159442597e-06,
+      "loss": 0.3352,
+      "step": 3845
+    },
+    {
+      "epoch": 4.100213219616204,
+      "grad_norm": 0.08829874201235319,
+      "learning_rate": 7.63582078381163e-06,
+      "loss": 0.3375,
+      "step": 3846
+    },
+    {
+      "epoch": 4.101279317697228,
+      "grad_norm": 0.10865348890518493,
+      "learning_rate": 7.618334336253927e-06,
+      "loss": 0.3344,
+      "step": 3847
+    },
+    {
+      "epoch": 4.1023454157782515,
+      "grad_norm": 0.09122759099076949,
+      "learning_rate": 7.60086582645609e-06,
+      "loss": 0.3311,
+      "step": 3848
+    },
+    {
+      "epoch": 4.103411513859275,
+      "grad_norm": 0.08531789969850555,
+      "learning_rate": 7.5834152640947444e-06,
+      "loss": 0.3326,
+      "step": 3849
+    },
+    {
+      "epoch": 4.104477611940299,
+      "grad_norm": 0.108511169138579,
+      "learning_rate": 7.565982658836599e-06,
+      "loss": 0.3299,
+      "step": 3850
+    },
+    {
+      "epoch": 4.105543710021322,
+      "grad_norm": 0.07907053738511821,
+      "learning_rate": 7.548568020338418e-06,
+      "loss": 0.3335,
+      "step": 3851
+    },
+    {
+      "epoch": 4.106609808102346,
+      "grad_norm": 0.09975341694550234,
+      "learning_rate": 7.531171358247009e-06,
+      "loss": 0.3352,
+      "step": 3852
+    },
+    {
+      "epoch": 4.107675906183369,
+      "grad_norm": 0.08769932023249331,
+      "learning_rate": 7.5137926821992055e-06,
+      "loss": 0.334,
+      "step": 3853
+    },
+    {
+      "epoch": 4.108742004264393,
+      "grad_norm": 0.10031133415920226,
+      "learning_rate": 7.496432001821898e-06,
+      "loss": 0.3311,
+      "step": 3854
+    },
+    {
+      "epoch": 4.109808102345416,
+      "grad_norm": 0.09300616655005942,
+      "learning_rate": 7.4790893267320115e-06,
+      "loss": 0.3347,
+      "step": 3855
+    },
+    {
+      "epoch": 4.110874200426439,
+      "grad_norm": 0.09170488621474711,
+      "learning_rate": 7.461764666536471e-06,
+      "loss": 0.3334,
+      "step": 3856
+    },
+    {
+      "epoch": 4.111940298507463,
+      "grad_norm": 0.10055859853923502,
+      "learning_rate": 7.444458030832238e-06,
+      "loss": 0.3395,
+      "step": 3857
+    },
+    {
+      "epoch": 4.113006396588486,
+      "grad_norm": 0.08680033935993048,
+      "learning_rate": 7.427169429206294e-06,
+      "loss": 0.3411,
+      "step": 3858
+    },
+    {
+      "epoch": 4.11407249466951,
+      "grad_norm": 0.09485145644525105,
+      "learning_rate": 7.409898871235639e-06,
+      "loss": 0.333,
+      "step": 3859
+    },
+    {
+      "epoch": 4.115138592750533,
+      "grad_norm": 0.09123152248422452,
+      "learning_rate": 7.39264636648724e-06,
+      "loss": 0.3331,
+      "step": 3860
+    },
+    {
+      "epoch": 4.116204690831556,
+      "grad_norm": 0.10403836531483734,
+      "learning_rate": 7.375411924518099e-06,
+      "loss": 0.328,
+      "step": 3861
+    },
+    {
+      "epoch": 4.11727078891258,
+      "grad_norm": 0.08367336466007777,
+      "learning_rate": 7.358195554875203e-06,
+      "loss": 0.3337,
+      "step": 3862
+    },
+    {
+      "epoch": 4.118336886993603,
+      "grad_norm": 0.10846732273711228,
+      "learning_rate": 7.340997267095535e-06,
+      "loss": 0.3333,
+      "step": 3863
+    },
+    {
+      "epoch": 4.119402985074627,
+      "grad_norm": 0.12164518439766003,
+      "learning_rate": 7.323817070706036e-06,
+      "loss": 0.3341,
+      "step": 3864
+    },
+    {
+      "epoch": 4.12046908315565,
+      "grad_norm": 0.09560739668306972,
+      "learning_rate": 7.3066549752236435e-06,
+      "loss": 0.3364,
+      "step": 3865
+    },
+    {
+      "epoch": 4.121535181236673,
+      "grad_norm": 0.09724336285167177,
+      "learning_rate": 7.289510990155286e-06,
+      "loss": 0.3264,
+      "step": 3866
+    },
+    {
+      "epoch": 4.122601279317697,
+      "grad_norm": 0.1017353520878921,
+      "learning_rate": 7.2723851249978114e-06,
+      "loss": 0.3305,
+      "step": 3867
+    },
+    {
+      "epoch": 4.1236673773987205,
+      "grad_norm": 0.1069625874611212,
+      "learning_rate": 7.255277389238075e-06,
+      "loss": 0.335,
+      "step": 3868
+    },
+    {
+      "epoch": 4.1247334754797444,
+      "grad_norm": 0.09669306285547304,
+      "learning_rate": 7.238187792352871e-06,
+      "loss": 0.328,
+      "step": 3869
+    },
+    {
+      "epoch": 4.1257995735607675,
+      "grad_norm": 0.10282239370897285,
+      "learning_rate": 7.221116343808963e-06,
+      "loss": 0.3311,
+      "step": 3870
+    },
+    {
+      "epoch": 4.126865671641791,
+      "grad_norm": 0.10105373041143043,
+      "learning_rate": 7.2040630530630175e-06,
+      "loss": 0.334,
+      "step": 3871
+    },
+    {
+      "epoch": 4.127931769722815,
+      "grad_norm": 0.0854802255803901,
+      "learning_rate": 7.187027929561683e-06,
+      "loss": 0.3374,
+      "step": 3872
+    },
+    {
+      "epoch": 4.128997867803838,
+      "grad_norm": 0.10520223879341609,
+      "learning_rate": 7.170010982741549e-06,
+      "loss": 0.3332,
+      "step": 3873
+    },
+    {
+      "epoch": 4.130063965884862,
+      "grad_norm": 0.09932971506774815,
+      "learning_rate": 7.153012222029097e-06,
+      "loss": 0.3313,
+      "step": 3874
+    },
+    {
+      "epoch": 4.131130063965885,
+      "grad_norm": 0.09719494548254369,
+      "learning_rate": 7.136031656840763e-06,
+      "loss": 0.3308,
+      "step": 3875
+    },
+    {
+      "epoch": 4.132196162046908,
+      "grad_norm": 0.08740775740958265,
+      "learning_rate": 7.1190692965829126e-06,
+      "loss": 0.3353,
+      "step": 3876
+    },
+    {
+      "epoch": 4.133262260127932,
+      "grad_norm": 0.08603123144210989,
+      "learning_rate": 7.102125150651784e-06,
+      "loss": 0.3288,
+      "step": 3877
+    },
+    {
+      "epoch": 4.134328358208955,
+      "grad_norm": 0.09256501786020642,
+      "learning_rate": 7.085199228433577e-06,
+      "loss": 0.3333,
+      "step": 3878
+    },
+    {
+      "epoch": 4.135394456289979,
+      "grad_norm": 0.09108945673753906,
+      "learning_rate": 7.068291539304368e-06,
+      "loss": 0.3381,
+      "step": 3879
+    },
+    {
+      "epoch": 4.136460554371002,
+      "grad_norm": 0.0796397200521715,
+      "learning_rate": 7.0514020926301285e-06,
+      "loss": 0.3314,
+      "step": 3880
+    },
+    {
+      "epoch": 4.137526652452026,
+      "grad_norm": 0.09620207070657663,
+      "learning_rate": 7.034530897766738e-06,
+      "loss": 0.3348,
+      "step": 3881
+    },
+    {
+      "epoch": 4.138592750533049,
+      "grad_norm": 0.08902821892098343,
+      "learning_rate": 7.017677964059979e-06,
+      "loss": 0.336,
+      "step": 3882
+    },
+    {
+      "epoch": 4.139658848614072,
+      "grad_norm": 0.08377254099774022,
+      "learning_rate": 7.000843300845473e-06,
+      "loss": 0.3312,
+      "step": 3883
+    },
+    {
+      "epoch": 4.140724946695096,
+      "grad_norm": 0.09387188110362456,
+      "learning_rate": 6.984026917448763e-06,
+      "loss": 0.3355,
+      "step": 3884
+    },
+    {
+      "epoch": 4.141791044776119,
+      "grad_norm": 0.08819576291353799,
+      "learning_rate": 6.967228823185257e-06,
+      "loss": 0.3306,
+      "step": 3885
+    },
+    {
+      "epoch": 4.142857142857143,
+      "grad_norm": 0.09756341988481407,
+      "learning_rate": 6.950449027360213e-06,
+      "loss": 0.3359,
+      "step": 3886
+    },
+    {
+      "epoch": 4.143923240938166,
+      "grad_norm": 0.08464278902185567,
+      "learning_rate": 6.9336875392687695e-06,
+      "loss": 0.3336,
+      "step": 3887
+    },
+    {
+      "epoch": 4.144989339019189,
+      "grad_norm": 0.08869325843091481,
+      "learning_rate": 6.91694436819593e-06,
+      "loss": 0.3325,
+      "step": 3888
+    },
+    {
+      "epoch": 4.146055437100213,
+      "grad_norm": 0.08580192830087008,
+      "learning_rate": 6.9002195234165295e-06,
+      "loss": 0.329,
+      "step": 3889
+    },
+    {
+      "epoch": 4.1471215351812365,
+      "grad_norm": 0.09091187799203043,
+      "learning_rate": 6.8835130141952625e-06,
+      "loss": 0.3319,
+      "step": 3890
+    },
+    {
+      "epoch": 4.1481876332622605,
+      "grad_norm": 0.08949013100276429,
+      "learning_rate": 6.866824849786673e-06,
+      "loss": 0.3348,
+      "step": 3891
+    },
+    {
+      "epoch": 4.149253731343284,
+      "grad_norm": 0.07727474516040857,
+      "learning_rate": 6.850155039435145e-06,
+      "loss": 0.3353,
+      "step": 3892
+    },
+    {
+      "epoch": 4.150319829424307,
+      "grad_norm": 0.09632177794333178,
+      "learning_rate": 6.833503592374864e-06,
+      "loss": 0.3387,
+      "step": 3893
+    },
+    {
+      "epoch": 4.151385927505331,
+      "grad_norm": 0.08865373218038916,
+      "learning_rate": 6.8168705178298835e-06,
+      "loss": 0.3309,
+      "step": 3894
+    },
+    {
+      "epoch": 4.152452025586354,
+      "grad_norm": 0.09466620554710381,
+      "learning_rate": 6.800255825014063e-06,
+      "loss": 0.333,
+      "step": 3895
+    },
+    {
+      "epoch": 4.153518123667378,
+      "grad_norm": 0.08200431933116618,
+      "learning_rate": 6.78365952313107e-06,
+      "loss": 0.3304,
+      "step": 3896
+    },
+    {
+      "epoch": 4.154584221748401,
+      "grad_norm": 0.0857938915091229,
+      "learning_rate": 6.767081621374392e-06,
+      "loss": 0.3335,
+      "step": 3897
+    },
+    {
+      "epoch": 4.155650319829424,
+      "grad_norm": 0.09483772251293643,
+      "learning_rate": 6.750522128927332e-06,
+      "loss": 0.3366,
+      "step": 3898
+    },
+    {
+      "epoch": 4.156716417910448,
+      "grad_norm": 0.0963326853227681,
+      "learning_rate": 6.733981054962995e-06,
+      "loss": 0.3287,
+      "step": 3899
+    },
+    {
+      "epoch": 4.157782515991471,
+      "grad_norm": 0.10620078154702665,
+      "learning_rate": 6.717458408644262e-06,
+      "loss": 0.3337,
+      "step": 3900
+    },
+    {
+      "epoch": 4.158848614072495,
+      "grad_norm": 0.0807624229086923,
+      "learning_rate": 6.700954199123821e-06,
+      "loss": 0.3351,
+      "step": 3901
+    },
+    {
+      "epoch": 4.159914712153518,
+      "grad_norm": 0.08529115263300027,
+      "learning_rate": 6.68446843554416e-06,
+      "loss": 0.3324,
+      "step": 3902
+    },
+    {
+      "epoch": 4.160980810234541,
+      "grad_norm": 0.09055299262053584,
+      "learning_rate": 6.66800112703752e-06,
+      "loss": 0.3337,
+      "step": 3903
+    },
+    {
+      "epoch": 4.162046908315565,
+      "grad_norm": 0.10436288204758602,
+      "learning_rate": 6.6515522827259414e-06,
+      "loss": 0.3353,
+      "step": 3904
+    },
+    {
+      "epoch": 4.163113006396588,
+      "grad_norm": 0.09038774397915303,
+      "learning_rate": 6.63512191172123e-06,
+      "loss": 0.3332,
+      "step": 3905
+    },
+    {
+      "epoch": 4.164179104477612,
+      "grad_norm": 0.10833203568207923,
+      "learning_rate": 6.618710023124961e-06,
+      "loss": 0.3335,
+      "step": 3906
+    },
+    {
+      "epoch": 4.165245202558635,
+      "grad_norm": 0.08665653296892867,
+      "learning_rate": 6.6023166260284555e-06,
+      "loss": 0.3384,
+      "step": 3907
+    },
+    {
+      "epoch": 4.166311300639659,
+      "grad_norm": 0.08387664029237912,
+      "learning_rate": 6.585941729512808e-06,
+      "loss": 0.3385,
+      "step": 3908
+    },
+    {
+      "epoch": 4.167377398720682,
+      "grad_norm": 0.08361794082456976,
+      "learning_rate": 6.569585342648861e-06,
+      "loss": 0.3333,
+      "step": 3909
+    },
+    {
+      "epoch": 4.1684434968017055,
+      "grad_norm": 0.08647483778408228,
+      "learning_rate": 6.55324747449722e-06,
+      "loss": 0.3383,
+      "step": 3910
+    },
+    {
+      "epoch": 4.169509594882729,
+      "grad_norm": 0.08605099292944113,
+      "learning_rate": 6.536928134108183e-06,
+      "loss": 0.3436,
+      "step": 3911
+    },
+    {
+      "epoch": 4.1705756929637525,
+      "grad_norm": 0.1009274470047588,
+      "learning_rate": 6.520627330521838e-06,
+      "loss": 0.337,
+      "step": 3912
+    },
+    {
+      "epoch": 4.1716417910447765,
+      "grad_norm": 0.08627645906743135,
+      "learning_rate": 6.504345072767986e-06,
+      "loss": 0.3391,
+      "step": 3913
+    },
+    {
+      "epoch": 4.1727078891258,
+      "grad_norm": 0.08829399505823976,
+      "learning_rate": 6.48808136986613e-06,
+      "loss": 0.3321,
+      "step": 3914
+    },
+    {
+      "epoch": 4.173773987206823,
+      "grad_norm": 0.08315647833828722,
+      "learning_rate": 6.471836230825533e-06,
+      "loss": 0.3386,
+      "step": 3915
+    },
+    {
+      "epoch": 4.174840085287847,
+      "grad_norm": 0.08482870520076757,
+      "learning_rate": 6.455609664645153e-06,
+      "loss": 0.3345,
+      "step": 3916
+    },
+    {
+      "epoch": 4.17590618336887,
+      "grad_norm": 0.08464445899920194,
+      "learning_rate": 6.439401680313677e-06,
+      "loss": 0.3321,
+      "step": 3917
+    },
+    {
+      "epoch": 4.176972281449894,
+      "grad_norm": 0.10394390037142433,
+      "learning_rate": 6.423212286809462e-06,
+      "loss": 0.3308,
+      "step": 3918
+    },
+    {
+      "epoch": 4.178038379530917,
+      "grad_norm": 0.09043061918479862,
+      "learning_rate": 6.407041493100603e-06,
+      "loss": 0.3375,
+      "step": 3919
+    },
+    {
+      "epoch": 4.17910447761194,
+      "grad_norm": 0.08930388093603546,
+      "learning_rate": 6.390889308144879e-06,
+      "loss": 0.3412,
+      "step": 3920
+    },
+    {
+      "epoch": 4.180170575692964,
+      "grad_norm": 0.09541571773419251,
+      "learning_rate": 6.374755740889775e-06,
+      "loss": 0.3381,
+      "step": 3921
+    },
+    {
+      "epoch": 4.181236673773987,
+      "grad_norm": 0.09036041961820482,
+      "learning_rate": 6.3586408002724195e-06,
+      "loss": 0.3319,
+      "step": 3922
+    },
+    {
+      "epoch": 4.182302771855011,
+      "grad_norm": 0.09106038987637816,
+      "learning_rate": 6.342544495219671e-06,
+      "loss": 0.3308,
+      "step": 3923
+    },
+    {
+      "epoch": 4.183368869936034,
+      "grad_norm": 0.08457513382844575,
+      "learning_rate": 6.326466834648055e-06,
+      "loss": 0.3318,
+      "step": 3924
+    },
+    {
+      "epoch": 4.184434968017057,
+      "grad_norm": 0.08404396446642012,
+      "learning_rate": 6.310407827463736e-06,
+      "loss": 0.3295,
+      "step": 3925
+    },
+    {
+      "epoch": 4.185501066098081,
+      "grad_norm": 0.09617345450918036,
+      "learning_rate": 6.29436748256258e-06,
+      "loss": 0.331,
+      "step": 3926
+    },
+    {
+      "epoch": 4.186567164179104,
+      "grad_norm": 0.09146742715237753,
+      "learning_rate": 6.278345808830102e-06,
+      "loss": 0.3346,
+      "step": 3927
+    },
+    {
+      "epoch": 4.187633262260128,
+      "grad_norm": 0.08405415079436931,
+      "learning_rate": 6.262342815141495e-06,
+      "loss": 0.3346,
+      "step": 3928
+    },
+    {
+      "epoch": 4.188699360341151,
+      "grad_norm": 0.09298813400757024,
+      "learning_rate": 6.246358510361559e-06,
+      "loss": 0.3318,
+      "step": 3929
+    },
+    {
+      "epoch": 4.189765458422174,
+      "grad_norm": 0.08261784827703691,
+      "learning_rate": 6.230392903344777e-06,
+      "loss": 0.3314,
+      "step": 3930
+    },
+    {
+      "epoch": 4.190831556503198,
+      "grad_norm": 0.09664506287358952,
+      "learning_rate": 6.214446002935282e-06,
+      "loss": 0.3275,
+      "step": 3931
+    },
+    {
+      "epoch": 4.1918976545842215,
+      "grad_norm": 0.10130324537487397,
+      "learning_rate": 6.198517817966805e-06,
+      "loss": 0.3392,
+      "step": 3932
+    },
+    {
+      "epoch": 4.1929637526652455,
+      "grad_norm": 0.0885774081244104,
+      "learning_rate": 6.182608357262738e-06,
+      "loss": 0.332,
+      "step": 3933
+    },
+    {
+      "epoch": 4.1940298507462686,
+      "grad_norm": 0.09028658637288987,
+      "learning_rate": 6.166717629636103e-06,
+      "loss": 0.3273,
+      "step": 3934
+    },
+    {
+      "epoch": 4.1950959488272925,
+      "grad_norm": 0.09059894957809664,
+      "learning_rate": 6.150845643889542e-06,
+      "loss": 0.3391,
+      "step": 3935
+    },
+    {
+      "epoch": 4.196162046908316,
+      "grad_norm": 0.0947415382692891,
+      "learning_rate": 6.1349924088152905e-06,
+      "loss": 0.3329,
+      "step": 3936
+    },
+    {
+      "epoch": 4.197228144989339,
+      "grad_norm": 0.07775785182191923,
+      "learning_rate": 6.119157933195232e-06,
+      "loss": 0.3337,
+      "step": 3937
+    },
+    {
+      "epoch": 4.198294243070363,
+      "grad_norm": 0.09230865898141519,
+      "learning_rate": 6.1033422258008364e-06,
+      "loss": 0.3356,
+      "step": 3938
+    },
+    {
+      "epoch": 4.199360341151386,
+      "grad_norm": 0.09075042427461232,
+      "learning_rate": 6.087545295393198e-06,
+      "loss": 0.3366,
+      "step": 3939
+    },
+    {
+      "epoch": 4.20042643923241,
+      "grad_norm": 0.08591434886921874,
+      "learning_rate": 6.071767150722974e-06,
+      "loss": 0.3329,
+      "step": 3940
+    },
+    {
+      "epoch": 4.201492537313433,
+      "grad_norm": 0.08354052632216036,
+      "learning_rate": 6.056007800530444e-06,
+      "loss": 0.337,
+      "step": 3941
+    },
+    {
+      "epoch": 4.202558635394456,
+      "grad_norm": 0.09416369800009326,
+      "learning_rate": 6.040267253545482e-06,
+      "loss": 0.3336,
+      "step": 3942
+    },
+    {
+      "epoch": 4.20362473347548,
+      "grad_norm": 0.08465467278990185,
+      "learning_rate": 6.024545518487515e-06,
+      "loss": 0.3357,
+      "step": 3943
+    },
+    {
+      "epoch": 4.204690831556503,
+      "grad_norm": 0.08504324724039129,
+      "learning_rate": 6.0088426040655704e-06,
+      "loss": 0.3328,
+      "step": 3944
+    },
+    {
+      "epoch": 4.205756929637527,
+      "grad_norm": 0.07972041873768856,
+      "learning_rate": 5.993158518978255e-06,
+      "loss": 0.3296,
+      "step": 3945
+    },
+    {
+      "epoch": 4.20682302771855,
+      "grad_norm": 0.08077528320595509,
+      "learning_rate": 5.977493271913739e-06,
+      "loss": 0.331,
+      "step": 3946
+    },
+    {
+      "epoch": 4.207889125799573,
+      "grad_norm": 0.07617274164835192,
+      "learning_rate": 5.961846871549739e-06,
+      "loss": 0.331,
+      "step": 3947
+    },
+    {
+      "epoch": 4.208955223880597,
+      "grad_norm": 0.09002104870840356,
+      "learning_rate": 5.946219326553557e-06,
+      "loss": 0.3404,
+      "step": 3948
+    },
+    {
+      "epoch": 4.21002132196162,
+      "grad_norm": 0.09718872414198558,
+      "learning_rate": 5.930610645582051e-06,
+      "loss": 0.3367,
+      "step": 3949
+    },
+    {
+      "epoch": 4.211087420042644,
+      "grad_norm": 0.08759177753235749,
+      "learning_rate": 5.915020837281602e-06,
+      "loss": 0.3389,
+      "step": 3950
+    },
+    {
+      "epoch": 4.212153518123667,
+      "grad_norm": 0.08550266499801752,
+      "learning_rate": 5.899449910288169e-06,
+      "loss": 0.3314,
+      "step": 3951
+    },
+    {
+      "epoch": 4.21321961620469,
+      "grad_norm": 0.08610889895227287,
+      "learning_rate": 5.883897873227216e-06,
+      "loss": 0.3321,
+      "step": 3952
+    },
+    {
+      "epoch": 4.214285714285714,
+      "grad_norm": 0.07857770022593799,
+      "learning_rate": 5.868364734713776e-06,
+      "loss": 0.3407,
+      "step": 3953
+    },
+    {
+      "epoch": 4.2153518123667375,
+      "grad_norm": 0.10475981108457424,
+      "learning_rate": 5.852850503352407e-06,
+      "loss": 0.3384,
+      "step": 3954
+    },
+    {
+      "epoch": 4.2164179104477615,
+      "grad_norm": 0.07804224672207336,
+      "learning_rate": 5.837355187737172e-06,
+      "loss": 0.3368,
+      "step": 3955
+    },
+    {
+      "epoch": 4.217484008528785,
+      "grad_norm": 0.08058373335851238,
+      "learning_rate": 5.821878796451681e-06,
+      "loss": 0.3322,
+      "step": 3956
+    },
+    {
+      "epoch": 4.218550106609808,
+      "grad_norm": 0.08404369349869568,
+      "learning_rate": 5.806421338069053e-06,
+      "loss": 0.337,
+      "step": 3957
+    },
+    {
+      "epoch": 4.219616204690832,
+      "grad_norm": 0.08080561168777425,
+      "learning_rate": 5.790982821151905e-06,
+      "loss": 0.3339,
+      "step": 3958
+    },
+    {
+      "epoch": 4.220682302771855,
+      "grad_norm": 0.0828839824572233,
+      "learning_rate": 5.7755632542523744e-06,
+      "loss": 0.3308,
+      "step": 3959
+    },
+    {
+      "epoch": 4.221748400852879,
+      "grad_norm": 0.09417702718528696,
+      "learning_rate": 5.7601626459121175e-06,
+      "loss": 0.3337,
+      "step": 3960
+    },
+    {
+      "epoch": 4.222814498933902,
+      "grad_norm": 0.08229090459967245,
+      "learning_rate": 5.744781004662247e-06,
+      "loss": 0.3349,
+      "step": 3961
+    },
+    {
+      "epoch": 4.223880597014926,
+      "grad_norm": 0.08396218859737672,
+      "learning_rate": 5.729418339023407e-06,
+      "loss": 0.3415,
+      "step": 3962
+    },
+    {
+      "epoch": 4.224946695095949,
+      "grad_norm": 0.08423179577148057,
+      "learning_rate": 5.714074657505708e-06,
+      "loss": 0.3349,
+      "step": 3963
+    },
+    {
+      "epoch": 4.226012793176972,
+      "grad_norm": 0.0747196908529032,
+      "learning_rate": 5.6987499686087695e-06,
+      "loss": 0.3361,
+      "step": 3964
+    },
+    {
+      "epoch": 4.227078891257996,
+      "grad_norm": 0.09430188688349524,
+      "learning_rate": 5.683444280821651e-06,
+      "loss": 0.3381,
+      "step": 3965
+    },
+    {
+      "epoch": 4.228144989339019,
+      "grad_norm": 0.07893437331383663,
+      "learning_rate": 5.668157602622914e-06,
+      "loss": 0.333,
+      "step": 3966
+    },
+    {
+      "epoch": 4.229211087420043,
+      "grad_norm": 0.07954371770687908,
+      "learning_rate": 5.6528899424805886e-06,
+      "loss": 0.3354,
+      "step": 3967
+    },
+    {
+      "epoch": 4.230277185501066,
+      "grad_norm": 0.07863279825537235,
+      "learning_rate": 5.637641308852169e-06,
+      "loss": 0.3242,
+      "step": 3968
+    },
+    {
+      "epoch": 4.231343283582089,
+      "grad_norm": 0.08110711920649173,
+      "learning_rate": 5.622411710184592e-06,
+      "loss": 0.3359,
+      "step": 3969
+    },
+    {
+      "epoch": 4.232409381663113,
+      "grad_norm": 0.08076149758379426,
+      "learning_rate": 5.607201154914275e-06,
+      "loss": 0.3315,
+      "step": 3970
+    },
+    {
+      "epoch": 4.233475479744136,
+      "grad_norm": 0.08084146195418195,
+      "learning_rate": 5.592009651467081e-06,
+      "loss": 0.3333,
+      "step": 3971
+    },
+    {
+      "epoch": 4.23454157782516,
+      "grad_norm": 0.07585877249863301,
+      "learning_rate": 5.5768372082582925e-06,
+      "loss": 0.3276,
+      "step": 3972
+    },
+    {
+      "epoch": 4.235607675906183,
+      "grad_norm": 0.07846952562528792,
+      "learning_rate": 5.561683833692666e-06,
+      "loss": 0.335,
+      "step": 3973
+    },
+    {
+      "epoch": 4.2366737739872065,
+      "grad_norm": 0.08010327334926544,
+      "learning_rate": 5.546549536164381e-06,
+      "loss": 0.3341,
+      "step": 3974
+    },
+    {
+      "epoch": 4.23773987206823,
+      "grad_norm": 0.0799875870943478,
+      "learning_rate": 5.531434324057068e-06,
+      "loss": 0.336,
+      "step": 3975
+    },
+    {
+      "epoch": 4.2388059701492535,
+      "grad_norm": 0.08230353675191754,
+      "learning_rate": 5.516338205743745e-06,
+      "loss": 0.3299,
+      "step": 3976
+    },
+    {
+      "epoch": 4.2398720682302775,
+      "grad_norm": 0.08053168122604346,
+      "learning_rate": 5.501261189586889e-06,
+      "loss": 0.3339,
+      "step": 3977
+    },
+    {
+      "epoch": 4.240938166311301,
+      "grad_norm": 0.07665080765192847,
+      "learning_rate": 5.486203283938376e-06,
+      "loss": 0.3365,
+      "step": 3978
+    },
+    {
+      "epoch": 4.242004264392324,
+      "grad_norm": 0.08493277690607401,
+      "learning_rate": 5.471164497139523e-06,
+      "loss": 0.3352,
+      "step": 3979
+    },
+    {
+      "epoch": 4.243070362473348,
+      "grad_norm": 0.08757131176321435,
+      "learning_rate": 5.456144837521012e-06,
+      "loss": 0.336,
+      "step": 3980
+    },
+    {
+      "epoch": 4.244136460554371,
+      "grad_norm": 0.08088307895934461,
+      "learning_rate": 5.441144313402964e-06,
+      "loss": 0.3352,
+      "step": 3981
+    },
+    {
+      "epoch": 4.245202558635395,
+      "grad_norm": 0.10303950476433886,
+      "learning_rate": 5.426162933094898e-06,
+      "loss": 0.3363,
+      "step": 3982
+    },
+    {
+      "epoch": 4.246268656716418,
+      "grad_norm": 0.08551651606703374,
+      "learning_rate": 5.411200704895705e-06,
+      "loss": 0.3419,
+      "step": 3983
+    },
+    {
+      "epoch": 4.247334754797441,
+      "grad_norm": 0.08310317399531018,
+      "learning_rate": 5.396257637093687e-06,
+      "loss": 0.334,
+      "step": 3984
+    },
+    {
+      "epoch": 4.248400852878465,
+      "grad_norm": 0.08656256571909457,
+      "learning_rate": 5.381333737966525e-06,
+      "loss": 0.3345,
+      "step": 3985
+    },
+    {
+      "epoch": 4.249466950959488,
+      "grad_norm": 0.08701071532379277,
+      "learning_rate": 5.3664290157813005e-06,
+      "loss": 0.3326,
+      "step": 3986
+    },
+    {
+      "epoch": 4.250533049040512,
+      "grad_norm": 0.08260458610600334,
+      "learning_rate": 5.3515434787944295e-06,
+      "loss": 0.3357,
+      "step": 3987
+    },
+    {
+      "epoch": 4.251599147121535,
+      "grad_norm": 0.09118839401494544,
+      "learning_rate": 5.336677135251744e-06,
+      "loss": 0.3309,
+      "step": 3988
+    },
+    {
+      "epoch": 4.252665245202559,
+      "grad_norm": 0.0833732616912321,
+      "learning_rate": 5.321829993388421e-06,
+      "loss": 0.3332,
+      "step": 3989
+    },
+    {
+      "epoch": 4.253731343283582,
+      "grad_norm": 0.0770950611344399,
+      "learning_rate": 5.307002061429001e-06,
+      "loss": 0.3295,
+      "step": 3990
+    },
+    {
+      "epoch": 4.254797441364605,
+      "grad_norm": 0.08654137149689174,
+      "learning_rate": 5.292193347587389e-06,
+      "loss": 0.3335,
+      "step": 3991
+    },
+    {
+      "epoch": 4.255863539445629,
+      "grad_norm": 0.08023871846633238,
+      "learning_rate": 5.277403860066841e-06,
+      "loss": 0.3318,
+      "step": 3992
+    },
+    {
+      "epoch": 4.256929637526652,
+      "grad_norm": 0.08681903564804878,
+      "learning_rate": 5.262633607059982e-06,
+      "loss": 0.3386,
+      "step": 3993
+    },
+    {
+      "epoch": 4.257995735607676,
+      "grad_norm": 0.08369350417988806,
+      "learning_rate": 5.247882596748737e-06,
+      "loss": 0.3356,
+      "step": 3994
+    },
+    {
+      "epoch": 4.259061833688699,
+      "grad_norm": 0.0911097900892705,
+      "learning_rate": 5.233150837304415e-06,
+      "loss": 0.3389,
+      "step": 3995
+    },
+    {
+      "epoch": 4.2601279317697225,
+      "grad_norm": 0.08091682403105288,
+      "learning_rate": 5.218438336887643e-06,
+      "loss": 0.3418,
+      "step": 3996
+    },
+    {
+      "epoch": 4.2611940298507465,
+      "grad_norm": 0.07957719556291395,
+      "learning_rate": 5.203745103648392e-06,
+      "loss": 0.3353,
+      "step": 3997
+    },
+    {
+      "epoch": 4.26226012793177,
+      "grad_norm": 0.077328833322233,
+      "learning_rate": 5.189071145725928e-06,
+      "loss": 0.3383,
+      "step": 3998
+    },
+    {
+      "epoch": 4.2633262260127935,
+      "grad_norm": 0.08081742747511964,
+      "learning_rate": 5.174416471248873e-06,
+      "loss": 0.3325,
+      "step": 3999
+    },
+    {
+      "epoch": 4.264392324093817,
+      "grad_norm": 0.07861878616801701,
+      "learning_rate": 5.159781088335161e-06,
+      "loss": 0.3342,
+      "step": 4000
+    },
+    {
+      "epoch": 4.26545842217484,
+      "grad_norm": 0.08400538210654203,
+      "learning_rate": 5.145165005092017e-06,
+      "loss": 0.3327,
+      "step": 4001
+    },
+    {
+      "epoch": 4.266524520255864,
+      "grad_norm": 0.07589313698458251,
+      "learning_rate": 5.130568229616004e-06,
+      "loss": 0.3332,
+      "step": 4002
+    },
+    {
+      "epoch": 4.267590618336887,
+      "grad_norm": 0.08147340980126598,
+      "learning_rate": 5.115990769992971e-06,
+      "loss": 0.3281,
+      "step": 4003
+    },
+    {
+      "epoch": 4.268656716417911,
+      "grad_norm": 0.07835055312388887,
+      "learning_rate": 5.101432634298089e-06,
+      "loss": 0.3339,
+      "step": 4004
+    },
+    {
+      "epoch": 4.269722814498934,
+      "grad_norm": 0.07947896584006585,
+      "learning_rate": 5.086893830595783e-06,
+      "loss": 0.3382,
+      "step": 4005
+    },
+    {
+      "epoch": 4.270788912579957,
+      "grad_norm": 0.07841316295017421,
+      "learning_rate": 5.07237436693981e-06,
+      "loss": 0.333,
+      "step": 4006
+    },
+    {
+      "epoch": 4.271855010660981,
+      "grad_norm": 0.09374263052005652,
+      "learning_rate": 5.057874251373194e-06,
+      "loss": 0.3332,
+      "step": 4007
+    },
+    {
+      "epoch": 4.272921108742004,
+      "grad_norm": 0.07579864101983415,
+      "learning_rate": 5.0433934919282525e-06,
+      "loss": 0.3324,
+      "step": 4008
+    },
+    {
+      "epoch": 4.273987206823028,
+      "grad_norm": 0.0804083559161328,
+      "learning_rate": 5.0289320966265645e-06,
+      "loss": 0.3325,
+      "step": 4009
+    },
+    {
+      "epoch": 4.275053304904051,
+      "grad_norm": 0.08512014086820764,
+      "learning_rate": 5.014490073478993e-06,
+      "loss": 0.342,
+      "step": 4010
+    },
+    {
+      "epoch": 4.276119402985074,
+      "grad_norm": 0.08530022661651851,
+      "learning_rate": 5.00006743048568e-06,
+      "loss": 0.3348,
+      "step": 4011
+    },
+    {
+      "epoch": 4.277185501066098,
+      "grad_norm": 0.07751500270346133,
+      "learning_rate": 4.985664175636e-06,
+      "loss": 0.3377,
+      "step": 4012
+    },
+    {
+      "epoch": 4.278251599147121,
+      "grad_norm": 0.075817452063534,
+      "learning_rate": 4.97128031690862e-06,
+      "loss": 0.3324,
+      "step": 4013
+    },
+    {
+      "epoch": 4.279317697228145,
+      "grad_norm": 0.07989341231264155,
+      "learning_rate": 4.956915862271445e-06,
+      "loss": 0.3371,
+      "step": 4014
+    },
+    {
+      "epoch": 4.280383795309168,
+      "grad_norm": 0.08374182318320397,
+      "learning_rate": 4.942570819681649e-06,
+      "loss": 0.3349,
+      "step": 4015
+    },
+    {
+      "epoch": 4.281449893390192,
+      "grad_norm": 0.08077261991520684,
+      "learning_rate": 4.928245197085626e-06,
+      "loss": 0.3418,
+      "step": 4016
+    },
+    {
+      "epoch": 4.282515991471215,
+      "grad_norm": 0.09142010449739497,
+      "learning_rate": 4.913939002419028e-06,
+      "loss": 0.3403,
+      "step": 4017
+    },
+    {
+      "epoch": 4.2835820895522385,
+      "grad_norm": 0.08261620442873233,
+      "learning_rate": 4.899652243606752e-06,
+      "loss": 0.3342,
+      "step": 4018
+    },
+    {
+      "epoch": 4.2846481876332625,
+      "grad_norm": 0.07949162404713021,
+      "learning_rate": 4.88538492856291e-06,
+      "loss": 0.3342,
+      "step": 4019
+    },
+    {
+      "epoch": 4.285714285714286,
+      "grad_norm": 0.08029027369166614,
+      "learning_rate": 4.871137065190854e-06,
+      "loss": 0.3347,
+      "step": 4020
+    },
+    {
+      "epoch": 4.28678038379531,
+      "grad_norm": 0.0792316332451311,
+      "learning_rate": 4.856908661383175e-06,
+      "loss": 0.339,
+      "step": 4021
+    },
+    {
+      "epoch": 4.287846481876333,
+      "grad_norm": 0.0801437973547572,
+      "learning_rate": 4.842699725021649e-06,
+      "loss": 0.331,
+      "step": 4022
+    },
+    {
+      "epoch": 4.288912579957356,
+      "grad_norm": 0.08488879536646476,
+      "learning_rate": 4.828510263977295e-06,
+      "loss": 0.3306,
+      "step": 4023
+    },
+    {
+      "epoch": 4.28997867803838,
+      "grad_norm": 0.07756702190689371,
+      "learning_rate": 4.814340286110346e-06,
+      "loss": 0.3321,
+      "step": 4024
+    },
+    {
+      "epoch": 4.291044776119403,
+      "grad_norm": 0.09257841600912252,
+      "learning_rate": 4.800189799270221e-06,
+      "loss": 0.3341,
+      "step": 4025
+    },
+    {
+      "epoch": 4.292110874200427,
+      "grad_norm": 0.08241075259625737,
+      "learning_rate": 4.786058811295564e-06,
+      "loss": 0.3379,
+      "step": 4026
+    },
+    {
+      "epoch": 4.29317697228145,
+      "grad_norm": 0.08900515508524041,
+      "learning_rate": 4.771947330014195e-06,
+      "loss": 0.3335,
+      "step": 4027
+    },
+    {
+      "epoch": 4.294243070362473,
+      "grad_norm": 0.08609235279832582,
+      "learning_rate": 4.757855363243149e-06,
+      "loss": 0.3395,
+      "step": 4028
+    },
+    {
+      "epoch": 4.295309168443497,
+      "grad_norm": 0.08284456057342021,
+      "learning_rate": 4.743782918788653e-06,
+      "loss": 0.3291,
+      "step": 4029
+    },
+    {
+      "epoch": 4.29637526652452,
+      "grad_norm": 0.0713526073589266,
+      "learning_rate": 4.729730004446094e-06,
+      "loss": 0.3265,
+      "step": 4030
+    },
+    {
+      "epoch": 4.297441364605544,
+      "grad_norm": 0.08416229502175335,
+      "learning_rate": 4.715696628000057e-06,
+      "loss": 0.3343,
+      "step": 4031
+    },
+    {
+      "epoch": 4.298507462686567,
+      "grad_norm": 0.07860706489090766,
+      "learning_rate": 4.701682797224316e-06,
+      "loss": 0.3342,
+      "step": 4032
+    },
+    {
+      "epoch": 4.29957356076759,
+      "grad_norm": 0.07625660551514413,
+      "learning_rate": 4.687688519881799e-06,
+      "loss": 0.3352,
+      "step": 4033
+    },
+    {
+      "epoch": 4.300639658848614,
+      "grad_norm": 0.08922308300609538,
+      "learning_rate": 4.673713803724602e-06,
+      "loss": 0.3264,
+      "step": 4034
+    },
+    {
+      "epoch": 4.301705756929637,
+      "grad_norm": 0.07964888287437034,
+      "learning_rate": 4.659758656494e-06,
+      "loss": 0.3347,
+      "step": 4035
+    },
+    {
+      "epoch": 4.302771855010661,
+      "grad_norm": 0.08135734469936476,
+      "learning_rate": 4.645823085920409e-06,
+      "loss": 0.3411,
+      "step": 4036
+    },
+    {
+      "epoch": 4.303837953091684,
+      "grad_norm": 0.07699749391475862,
+      "learning_rate": 4.6319070997234315e-06,
+      "loss": 0.3366,
+      "step": 4037
+    },
+    {
+      "epoch": 4.3049040511727075,
+      "grad_norm": 0.079215440831635,
+      "learning_rate": 4.618010705611777e-06,
+      "loss": 0.3404,
+      "step": 4038
+    },
+    {
+      "epoch": 4.3059701492537314,
+      "grad_norm": 0.08102338622902823,
+      "learning_rate": 4.604133911283333e-06,
+      "loss": 0.3379,
+      "step": 4039
+    },
+    {
+      "epoch": 4.3070362473347545,
+      "grad_norm": 0.08630172945404184,
+      "learning_rate": 4.590276724425136e-06,
+      "loss": 0.3319,
+      "step": 4040
+    },
+    {
+      "epoch": 4.3081023454157785,
+      "grad_norm": 0.09511994379660102,
+      "learning_rate": 4.576439152713326e-06,
+      "loss": 0.3315,
+      "step": 4041
+    },
+    {
+      "epoch": 4.309168443496802,
+      "grad_norm": 0.08335455005241084,
+      "learning_rate": 4.562621203813211e-06,
+      "loss": 0.3316,
+      "step": 4042
+    },
+    {
+      "epoch": 4.310234541577826,
+      "grad_norm": 0.08307933355980296,
+      "learning_rate": 4.548822885379212e-06,
+      "loss": 0.336,
+      "step": 4043
+    },
+    {
+      "epoch": 4.311300639658849,
+      "grad_norm": 0.0843309027586609,
+      "learning_rate": 4.535044205054893e-06,
+      "loss": 0.3295,
+      "step": 4044
+    },
+    {
+      "epoch": 4.312366737739872,
+      "grad_norm": 0.08460136640969265,
+      "learning_rate": 4.521285170472904e-06,
+      "loss": 0.336,
+      "step": 4045
+    },
+    {
+      "epoch": 4.313432835820896,
+      "grad_norm": 0.07814369280046486,
+      "learning_rate": 4.507545789255052e-06,
+      "loss": 0.3332,
+      "step": 4046
+    },
+    {
+      "epoch": 4.314498933901919,
+      "grad_norm": 0.0855030516487757,
+      "learning_rate": 4.4938260690122435e-06,
+      "loss": 0.3342,
+      "step": 4047
+    },
+    {
+      "epoch": 4.315565031982943,
+      "grad_norm": 0.08812581908559658,
+      "learning_rate": 4.480126017344471e-06,
+      "loss": 0.3329,
+      "step": 4048
+    },
+    {
+      "epoch": 4.316631130063966,
+      "grad_norm": 0.07749089456051617,
+      "learning_rate": 4.466445641840862e-06,
+      "loss": 0.3333,
+      "step": 4049
+    },
+    {
+      "epoch": 4.317697228144989,
+      "grad_norm": 0.07815807671811865,
+      "learning_rate": 4.45278495007964e-06,
+      "loss": 0.3363,
+      "step": 4050
+    },
+    {
+      "epoch": 4.318763326226013,
+      "grad_norm": 0.07358888097958845,
+      "learning_rate": 4.439143949628118e-06,
+      "loss": 0.3295,
+      "step": 4051
+    },
+    {
+      "epoch": 4.319829424307036,
+      "grad_norm": 0.08431613734886675,
+      "learning_rate": 4.425522648042684e-06,
+      "loss": 0.3275,
+      "step": 4052
+    },
+    {
+      "epoch": 4.32089552238806,
+      "grad_norm": 0.07450901993676554,
+      "learning_rate": 4.411921052868846e-06,
+      "loss": 0.33,
+      "step": 4053
+    },
+    {
+      "epoch": 4.321961620469083,
+      "grad_norm": 0.08108642538320732,
+      "learning_rate": 4.3983391716411775e-06,
+      "loss": 0.3374,
+      "step": 4054
+    },
+    {
+      "epoch": 4.323027718550106,
+      "grad_norm": 0.08057675359517547,
+      "learning_rate": 4.384777011883343e-06,
+      "loss": 0.3327,
+      "step": 4055
+    },
+    {
+      "epoch": 4.32409381663113,
+      "grad_norm": 0.07748909745389584,
+      "learning_rate": 4.371234581108059e-06,
+      "loss": 0.3264,
+      "step": 4056
+    },
+    {
+      "epoch": 4.325159914712153,
+      "grad_norm": 0.0920305873286274,
+      "learning_rate": 4.3577118868171335e-06,
+      "loss": 0.3326,
+      "step": 4057
+    },
+    {
+      "epoch": 4.326226012793177,
+      "grad_norm": 0.07953455564884193,
+      "learning_rate": 4.344208936501449e-06,
+      "loss": 0.3327,
+      "step": 4058
+    },
+    {
+      "epoch": 4.3272921108742,
+      "grad_norm": 0.08324725165522431,
+      "learning_rate": 4.3307257376409155e-06,
+      "loss": 0.3329,
+      "step": 4059
+    },
+    {
+      "epoch": 4.3283582089552235,
+      "grad_norm": 0.08278692213524609,
+      "learning_rate": 4.317262297704541e-06,
+      "loss": 0.3374,
+      "step": 4060
+    },
+    {
+      "epoch": 4.3294243070362475,
+      "grad_norm": 0.07643673415451332,
+      "learning_rate": 4.3038186241503644e-06,
+      "loss": 0.338,
+      "step": 4061
+    },
+    {
+      "epoch": 4.330490405117271,
+      "grad_norm": 0.07581981080944176,
+      "learning_rate": 4.290394724425495e-06,
+      "loss": 0.3384,
+      "step": 4062
+    },
+    {
+      "epoch": 4.3315565031982945,
+      "grad_norm": 0.07771968698621509,
+      "learning_rate": 4.276990605966056e-06,
+      "loss": 0.3337,
+      "step": 4063
+    },
+    {
+      "epoch": 4.332622601279318,
+      "grad_norm": 0.07895244133624792,
+      "learning_rate": 4.2636062761972406e-06,
+      "loss": 0.3409,
+      "step": 4064
+    },
+    {
+      "epoch": 4.333688699360341,
+      "grad_norm": 0.07797323195669634,
+      "learning_rate": 4.2502417425332746e-06,
+      "loss": 0.3376,
+      "step": 4065
+    },
+    {
+      "epoch": 4.334754797441365,
+      "grad_norm": 0.07997960877114406,
+      "learning_rate": 4.236897012377421e-06,
+      "loss": 0.3371,
+      "step": 4066
+    },
+    {
+      "epoch": 4.335820895522388,
+      "grad_norm": 0.07600199349205194,
+      "learning_rate": 4.223572093121951e-06,
+      "loss": 0.3364,
+      "step": 4067
+    },
+    {
+      "epoch": 4.336886993603412,
+      "grad_norm": 0.08268817910100379,
+      "learning_rate": 4.210266992148188e-06,
+      "loss": 0.336,
+      "step": 4068
+    },
+    {
+      "epoch": 4.337953091684435,
+      "grad_norm": 0.0745603318232261,
+      "learning_rate": 4.196981716826471e-06,
+      "loss": 0.3323,
+      "step": 4069
+    },
+    {
+      "epoch": 4.339019189765459,
+      "grad_norm": 0.09208585615342714,
+      "learning_rate": 4.183716274516134e-06,
+      "loss": 0.3314,
+      "step": 4070
+    },
+    {
+      "epoch": 4.340085287846482,
+      "grad_norm": 0.07711887841230759,
+      "learning_rate": 4.170470672565557e-06,
+      "loss": 0.3331,
+      "step": 4071
+    },
+    {
+      "epoch": 4.341151385927505,
+      "grad_norm": 0.08621935992949319,
+      "learning_rate": 4.157244918312113e-06,
+      "loss": 0.3341,
+      "step": 4072
+    },
+    {
+      "epoch": 4.342217484008529,
+      "grad_norm": 0.08226222595272614,
+      "learning_rate": 4.144039019082184e-06,
+      "loss": 0.333,
+      "step": 4073
+    },
+    {
+      "epoch": 4.343283582089552,
+      "grad_norm": 0.07808395675180577,
+      "learning_rate": 4.1308529821911495e-06,
+      "loss": 0.3303,
+      "step": 4074
+    },
+    {
+      "epoch": 4.344349680170576,
+      "grad_norm": 0.0769818042274672,
+      "learning_rate": 4.117686814943382e-06,
+      "loss": 0.3417,
+      "step": 4075
+    },
+    {
+      "epoch": 4.345415778251599,
+      "grad_norm": 0.07879847363496587,
+      "learning_rate": 4.104540524632268e-06,
+      "loss": 0.3382,
+      "step": 4076
+    },
+    {
+      "epoch": 4.346481876332622,
+      "grad_norm": 0.08293035530571763,
+      "learning_rate": 4.091414118540158e-06,
+      "loss": 0.3331,
+      "step": 4077
+    },
+    {
+      "epoch": 4.347547974413646,
+      "grad_norm": 0.07727360738321705,
+      "learning_rate": 4.078307603938397e-06,
+      "loss": 0.337,
+      "step": 4078
+    },
+    {
+      "epoch": 4.348614072494669,
+      "grad_norm": 0.07551576922057891,
+      "learning_rate": 4.0652209880873214e-06,
+      "loss": 0.328,
+      "step": 4079
+    },
+    {
+      "epoch": 4.349680170575693,
+      "grad_norm": 0.07917186089359039,
+      "learning_rate": 4.052154278236242e-06,
+      "loss": 0.334,
+      "step": 4080
+    },
+    {
+      "epoch": 4.350746268656716,
+      "grad_norm": 0.07567017656306375,
+      "learning_rate": 4.039107481623417e-06,
+      "loss": 0.3345,
+      "step": 4081
+    },
+    {
+      "epoch": 4.3518123667377395,
+      "grad_norm": 0.07605848497463297,
+      "learning_rate": 4.026080605476104e-06,
+      "loss": 0.3311,
+      "step": 4082
+    },
+    {
+      "epoch": 4.3528784648187635,
+      "grad_norm": 0.08119621016465227,
+      "learning_rate": 4.013073657010518e-06,
+      "loss": 0.334,
+      "step": 4083
+    },
+    {
+      "epoch": 4.353944562899787,
+      "grad_norm": 0.07763766653584449,
+      "learning_rate": 4.000086643431838e-06,
+      "loss": 0.3346,
+      "step": 4084
+    },
+    {
+      "epoch": 4.355010660980811,
+      "grad_norm": 0.08680938596509069,
+      "learning_rate": 3.987119571934179e-06,
+      "loss": 0.3379,
+      "step": 4085
+    },
+    {
+      "epoch": 4.356076759061834,
+      "grad_norm": 0.07539433250460861,
+      "learning_rate": 3.974172449700633e-06,
+      "loss": 0.3312,
+      "step": 4086
+    },
+    {
+      "epoch": 4.357142857142857,
+      "grad_norm": 0.08358711815421738,
+      "learning_rate": 3.961245283903239e-06,
+      "loss": 0.3346,
+      "step": 4087
+    },
+    {
+      "epoch": 4.358208955223881,
+      "grad_norm": 0.07975462632947826,
+      "learning_rate": 3.948338081702958e-06,
+      "loss": 0.3278,
+      "step": 4088
+    },
+    {
+      "epoch": 4.359275053304904,
+      "grad_norm": 0.07783915298575807,
+      "learning_rate": 3.935450850249725e-06,
+      "loss": 0.3356,
+      "step": 4089
+    },
+    {
+      "epoch": 4.360341151385928,
+      "grad_norm": 0.07816454670420965,
+      "learning_rate": 3.9225835966823966e-06,
+      "loss": 0.3361,
+      "step": 4090
+    },
+    {
+      "epoch": 4.361407249466951,
+      "grad_norm": 0.08148461720782253,
+      "learning_rate": 3.909736328128748e-06,
+      "loss": 0.3333,
+      "step": 4091
+    },
+    {
+      "epoch": 4.362473347547974,
+      "grad_norm": 0.08042384846739183,
+      "learning_rate": 3.896909051705509e-06,
+      "loss": 0.3321,
+      "step": 4092
+    },
+    {
+      "epoch": 4.363539445628998,
+      "grad_norm": 0.08319011697671452,
+      "learning_rate": 3.884101774518327e-06,
+      "loss": 0.3354,
+      "step": 4093
+    },
+    {
+      "epoch": 4.364605543710021,
+      "grad_norm": 0.0782252777997426,
+      "learning_rate": 3.871314503661761e-06,
+      "loss": 0.3361,
+      "step": 4094
+    },
+    {
+      "epoch": 4.365671641791045,
+      "grad_norm": 0.07065927300340862,
+      "learning_rate": 3.858547246219293e-06,
+      "loss": 0.3312,
+      "step": 4095
+    },
+    {
+      "epoch": 4.366737739872068,
+      "grad_norm": 0.08468956849885999,
+      "learning_rate": 3.845800009263334e-06,
+      "loss": 0.3307,
+      "step": 4096
+    },
+    {
+      "epoch": 4.367803837953092,
+      "grad_norm": 0.08901806534396679,
+      "learning_rate": 3.833072799855173e-06,
+      "loss": 0.3315,
+      "step": 4097
+    },
+    {
+      "epoch": 4.368869936034115,
+      "grad_norm": 0.07301214410208626,
+      "learning_rate": 3.820365625045037e-06,
+      "loss": 0.3331,
+      "step": 4098
+    },
+    {
+      "epoch": 4.369936034115138,
+      "grad_norm": 0.08198685107248344,
+      "learning_rate": 3.8076784918720242e-06,
+      "loss": 0.3311,
+      "step": 4099
+    },
+    {
+      "epoch": 4.371002132196162,
+      "grad_norm": 0.08355479249429133,
+      "learning_rate": 3.7950114073641573e-06,
+      "loss": 0.3355,
+      "step": 4100
+    },
+    {
+      "epoch": 4.372068230277185,
+      "grad_norm": 0.07726712718675953,
+      "learning_rate": 3.7823643785383434e-06,
+      "loss": 0.333,
+      "step": 4101
+    },
+    {
+      "epoch": 4.373134328358209,
+      "grad_norm": 0.07279799524384821,
+      "learning_rate": 3.7697374124003872e-06,
+      "loss": 0.3303,
+      "step": 4102
+    },
+    {
+      "epoch": 4.3742004264392325,
+      "grad_norm": 0.08528796671091622,
+      "learning_rate": 3.757130515944951e-06,
+      "loss": 0.3342,
+      "step": 4103
+    },
+    {
+      "epoch": 4.3752665245202556,
+      "grad_norm": 0.08572585263620255,
+      "learning_rate": 3.7445436961556135e-06,
+      "loss": 0.3369,
+      "step": 4104
+    },
+    {
+      "epoch": 4.3763326226012795,
+      "grad_norm": 0.0748149784520539,
+      "learning_rate": 3.7319769600048237e-06,
+      "loss": 0.3387,
+      "step": 4105
+    },
+    {
+      "epoch": 4.377398720682303,
+      "grad_norm": 0.07657381538769502,
+      "learning_rate": 3.7194303144538847e-06,
+      "loss": 0.3327,
+      "step": 4106
+    },
+    {
+      "epoch": 4.378464818763327,
+      "grad_norm": 0.0807481452009072,
+      "learning_rate": 3.706903766452996e-06,
+      "loss": 0.3343,
+      "step": 4107
+    },
+    {
+      "epoch": 4.37953091684435,
+      "grad_norm": 0.08393196828122008,
+      "learning_rate": 3.6943973229412124e-06,
+      "loss": 0.3374,
+      "step": 4108
+    },
+    {
+      "epoch": 4.380597014925373,
+      "grad_norm": 0.0782549467161215,
+      "learning_rate": 3.681910990846462e-06,
+      "loss": 0.3287,
+      "step": 4109
+    },
+    {
+      "epoch": 4.381663113006397,
+      "grad_norm": 0.08070850418302111,
+      "learning_rate": 3.669444777085507e-06,
+      "loss": 0.337,
+      "step": 4110
+    },
+    {
+      "epoch": 4.38272921108742,
+      "grad_norm": 0.07534023101241813,
+      "learning_rate": 3.6569986885639954e-06,
+      "loss": 0.3341,
+      "step": 4111
+    },
+    {
+      "epoch": 4.383795309168444,
+      "grad_norm": 0.07838807058793515,
+      "learning_rate": 3.6445727321764035e-06,
+      "loss": 0.3353,
+      "step": 4112
+    },
+    {
+      "epoch": 4.384861407249467,
+      "grad_norm": 0.07822860155918042,
+      "learning_rate": 3.6321669148060833e-06,
+      "loss": 0.3342,
+      "step": 4113
+    },
+    {
+      "epoch": 4.38592750533049,
+      "grad_norm": 0.07732616158756847,
+      "learning_rate": 3.619781243325187e-06,
+      "loss": 0.3382,
+      "step": 4114
+    },
+    {
+      "epoch": 4.386993603411514,
+      "grad_norm": 0.0767992597048576,
+      "learning_rate": 3.6074157245947495e-06,
+      "loss": 0.3392,
+      "step": 4115
+    },
+    {
+      "epoch": 4.388059701492537,
+      "grad_norm": 0.08304386197535715,
+      "learning_rate": 3.5950703654646303e-06,
+      "loss": 0.3368,
+      "step": 4116
+    },
+    {
+      "epoch": 4.389125799573561,
+      "grad_norm": 0.07467146074464837,
+      "learning_rate": 3.5827451727735007e-06,
+      "loss": 0.333,
+      "step": 4117
+    },
+    {
+      "epoch": 4.390191897654584,
+      "grad_norm": 0.07659772726124718,
+      "learning_rate": 3.5704401533488865e-06,
+      "loss": 0.3332,
+      "step": 4118
+    },
+    {
+      "epoch": 4.391257995735607,
+      "grad_norm": 0.0779901058287516,
+      "learning_rate": 3.5581553140071256e-06,
+      "loss": 0.3353,
+      "step": 4119
+    },
+    {
+      "epoch": 4.392324093816631,
+      "grad_norm": 0.08271167127027722,
+      "learning_rate": 3.5458906615533883e-06,
+      "loss": 0.3325,
+      "step": 4120
+    },
+    {
+      "epoch": 4.393390191897654,
+      "grad_norm": 0.07640335606088489,
+      "learning_rate": 3.53364620278164e-06,
+      "loss": 0.3336,
+      "step": 4121
+    },
+    {
+      "epoch": 4.394456289978678,
+      "grad_norm": 0.08042193883025106,
+      "learning_rate": 3.5214219444746856e-06,
+      "loss": 0.3384,
+      "step": 4122
+    },
+    {
+      "epoch": 4.395522388059701,
+      "grad_norm": 0.07840556030820514,
+      "learning_rate": 3.5092178934041353e-06,
+      "loss": 0.3386,
+      "step": 4123
+    },
+    {
+      "epoch": 4.396588486140725,
+      "grad_norm": 0.08151804289922143,
+      "learning_rate": 3.497034056330382e-06,
+      "loss": 0.3317,
+      "step": 4124
+    },
+    {
+      "epoch": 4.3976545842217485,
+      "grad_norm": 0.08168021320582781,
+      "learning_rate": 3.4848704400026434e-06,
+      "loss": 0.3321,
+      "step": 4125
+    },
+    {
+      "epoch": 4.398720682302772,
+      "grad_norm": 0.07609160414759338,
+      "learning_rate": 3.4727270511589396e-06,
+      "loss": 0.3363,
+      "step": 4126
+    },
+    {
+      "epoch": 4.399786780383796,
+      "grad_norm": 0.07842449906210139,
+      "learning_rate": 3.4606038965260715e-06,
+      "loss": 0.3281,
+      "step": 4127
+    },
+    {
+      "epoch": 4.400852878464819,
+      "grad_norm": 0.07972055102539237,
+      "learning_rate": 3.4485009828196357e-06,
+      "loss": 0.3348,
+      "step": 4128
+    },
+    {
+      "epoch": 4.401918976545843,
+      "grad_norm": 0.07239605085558472,
+      "learning_rate": 3.4364183167440123e-06,
+      "loss": 0.3314,
+      "step": 4129
+    },
+    {
+      "epoch": 4.402985074626866,
+      "grad_norm": 0.07272119283107233,
+      "learning_rate": 3.4243559049923803e-06,
+      "loss": 0.3345,
+      "step": 4130
+    },
+    {
+      "epoch": 4.404051172707889,
+      "grad_norm": 0.07698965215086702,
+      "learning_rate": 3.412313754246688e-06,
+      "loss": 0.3364,
+      "step": 4131
+    },
+    {
+      "epoch": 4.405117270788913,
+      "grad_norm": 0.07835880219239048,
+      "learning_rate": 3.400291871177652e-06,
+      "loss": 0.3325,
+      "step": 4132
+    },
+    {
+      "epoch": 4.406183368869936,
+      "grad_norm": 0.07883352671429816,
+      "learning_rate": 3.3882902624447777e-06,
+      "loss": 0.3302,
+      "step": 4133
+    },
+    {
+      "epoch": 4.40724946695096,
+      "grad_norm": 0.07598330265933242,
+      "learning_rate": 3.3763089346963417e-06,
+      "loss": 0.3315,
+      "step": 4134
+    },
+    {
+      "epoch": 4.408315565031983,
+      "grad_norm": 0.07233756048331239,
+      "learning_rate": 3.3643478945693552e-06,
+      "loss": 0.3344,
+      "step": 4135
+    },
+    {
+      "epoch": 4.409381663113006,
+      "grad_norm": 0.07815412688938289,
+      "learning_rate": 3.352407148689625e-06,
+      "loss": 0.3355,
+      "step": 4136
+    },
+    {
+      "epoch": 4.41044776119403,
+      "grad_norm": 0.08194740814774278,
+      "learning_rate": 3.3404867036716994e-06,
+      "loss": 0.3353,
+      "step": 4137
+    },
+    {
+      "epoch": 4.411513859275053,
+      "grad_norm": 0.07605509113317131,
+      "learning_rate": 3.328586566118901e-06,
+      "loss": 0.3352,
+      "step": 4138
+    },
+    {
+      "epoch": 4.412579957356077,
+      "grad_norm": 0.07758256960053231,
+      "learning_rate": 3.316706742623268e-06,
+      "loss": 0.3391,
+      "step": 4139
+    },
+    {
+      "epoch": 4.4136460554371,
+      "grad_norm": 0.07526062814521468,
+      "learning_rate": 3.3048472397656115e-06,
+      "loss": 0.3331,
+      "step": 4140
+    },
+    {
+      "epoch": 4.414712153518123,
+      "grad_norm": 0.08486437341532607,
+      "learning_rate": 3.2930080641154816e-06,
+      "loss": 0.3323,
+      "step": 4141
+    },
+    {
+      "epoch": 4.415778251599147,
+      "grad_norm": 0.0830672341181274,
+      "learning_rate": 3.2811892222311694e-06,
+      "loss": 0.3319,
+      "step": 4142
+    },
+    {
+      "epoch": 4.41684434968017,
+      "grad_norm": 0.07212703483795091,
+      "learning_rate": 3.269390720659691e-06,
+      "loss": 0.3332,
+      "step": 4143
+    },
+    {
+      "epoch": 4.417910447761194,
+      "grad_norm": 0.0790520570244998,
+      "learning_rate": 3.257612565936805e-06,
+      "loss": 0.3359,
+      "step": 4144
+    },
+    {
+      "epoch": 4.418976545842217,
+      "grad_norm": 0.07333741287228444,
+      "learning_rate": 3.2458547645870086e-06,
+      "loss": 0.3285,
+      "step": 4145
+    },
+    {
+      "epoch": 4.4200426439232405,
+      "grad_norm": 0.07631456166587812,
+      "learning_rate": 3.2341173231234956e-06,
+      "loss": 0.3362,
+      "step": 4146
+    },
+    {
+      "epoch": 4.4211087420042645,
+      "grad_norm": 0.07268449281095922,
+      "learning_rate": 3.2224002480482075e-06,
+      "loss": 0.335,
+      "step": 4147
+    },
+    {
+      "epoch": 4.422174840085288,
+      "grad_norm": 0.07265611750504483,
+      "learning_rate": 3.210703545851792e-06,
+      "loss": 0.3323,
+      "step": 4148
+    },
+    {
+      "epoch": 4.423240938166312,
+      "grad_norm": 0.07887619669716188,
+      "learning_rate": 3.1990272230136266e-06,
+      "loss": 0.3305,
+      "step": 4149
+    },
+    {
+      "epoch": 4.424307036247335,
+      "grad_norm": 0.07492581724533613,
+      "learning_rate": 3.187371286001768e-06,
+      "loss": 0.3387,
+      "step": 4150
+    },
+    {
+      "epoch": 4.425373134328359,
+      "grad_norm": 0.07375810082278332,
+      "learning_rate": 3.175735741273007e-06,
+      "loss": 0.3323,
+      "step": 4151
+    },
+    {
+      "epoch": 4.426439232409382,
+      "grad_norm": 0.0756484197290268,
+      "learning_rate": 3.164120595272837e-06,
+      "loss": 0.3339,
+      "step": 4152
+    },
+    {
+      "epoch": 4.427505330490405,
+      "grad_norm": 0.07829165152170597,
+      "learning_rate": 3.1525258544354354e-06,
+      "loss": 0.3358,
+      "step": 4153
+    },
+    {
+      "epoch": 4.428571428571429,
+      "grad_norm": 0.07587711508081742,
+      "learning_rate": 3.140951525183691e-06,
+      "loss": 0.3389,
+      "step": 4154
+    },
+    {
+      "epoch": 4.429637526652452,
+      "grad_norm": 0.07159911504321198,
+      "learning_rate": 3.1293976139291814e-06,
+      "loss": 0.3293,
+      "step": 4155
+    },
+    {
+      "epoch": 4.430703624733475,
+      "grad_norm": 0.0699462370372574,
+      "learning_rate": 3.117864127072179e-06,
+      "loss": 0.3384,
+      "step": 4156
+    },
+    {
+      "epoch": 4.431769722814499,
+      "grad_norm": 0.07194403996523205,
+      "learning_rate": 3.106351071001621e-06,
+      "loss": 0.3298,
+      "step": 4157
+    },
+    {
+      "epoch": 4.432835820895522,
+      "grad_norm": 0.07348866543415371,
+      "learning_rate": 3.0948584520951488e-06,
+      "loss": 0.3327,
+      "step": 4158
+    },
+    {
+      "epoch": 4.433901918976546,
+      "grad_norm": 0.07901403560235466,
+      "learning_rate": 3.083386276719087e-06,
+      "loss": 0.3354,
+      "step": 4159
+    },
+    {
+      "epoch": 4.434968017057569,
+      "grad_norm": 0.07614682029865204,
+      "learning_rate": 3.071934551228406e-06,
+      "loss": 0.3399,
+      "step": 4160
+    },
+    {
+      "epoch": 4.436034115138593,
+      "grad_norm": 0.07095662489839566,
+      "learning_rate": 3.060503281966778e-06,
+      "loss": 0.3364,
+      "step": 4161
+    },
+    {
+      "epoch": 4.437100213219616,
+      "grad_norm": 0.07661517083301671,
+      "learning_rate": 3.049092475266533e-06,
+      "loss": 0.3407,
+      "step": 4162
+    },
+    {
+      "epoch": 4.438166311300639,
+      "grad_norm": 0.07612860870888066,
+      "learning_rate": 3.037702137448659e-06,
+      "loss": 0.336,
+      "step": 4163
+    },
+    {
+      "epoch": 4.439232409381663,
+      "grad_norm": 0.07257317111602125,
+      "learning_rate": 3.0263322748228117e-06,
+      "loss": 0.3347,
+      "step": 4164
+    },
+    {
+      "epoch": 4.440298507462686,
+      "grad_norm": 0.06942974184000834,
+      "learning_rate": 3.0149828936873084e-06,
+      "loss": 0.3379,
+      "step": 4165
+    },
+    {
+      "epoch": 4.44136460554371,
+      "grad_norm": 0.07367479114255061,
+      "learning_rate": 3.003654000329115e-06,
+      "loss": 0.3325,
+      "step": 4166
+    },
+    {
+      "epoch": 4.4424307036247335,
+      "grad_norm": 0.07405338625538695,
+      "learning_rate": 2.9923456010238426e-06,
+      "loss": 0.3377,
+      "step": 4167
+    },
+    {
+      "epoch": 4.443496801705757,
+      "grad_norm": 0.07391480199022413,
+      "learning_rate": 2.98105770203577e-06,
+      "loss": 0.3336,
+      "step": 4168
+    },
+    {
+      "epoch": 4.4445628997867805,
+      "grad_norm": 0.0697155481700767,
+      "learning_rate": 2.9697903096177973e-06,
+      "loss": 0.3323,
+      "step": 4169
+    },
+    {
+      "epoch": 4.445628997867804,
+      "grad_norm": 0.07479820610427203,
+      "learning_rate": 2.9585434300114734e-06,
+      "loss": 0.332,
+      "step": 4170
+    },
+    {
+      "epoch": 4.446695095948828,
+      "grad_norm": 0.07917174679867747,
+      "learning_rate": 2.9473170694469934e-06,
+      "loss": 0.3346,
+      "step": 4171
+    },
+    {
+      "epoch": 4.447761194029851,
+      "grad_norm": 0.07299094494689012,
+      "learning_rate": 2.9361112341431643e-06,
+      "loss": 0.3303,
+      "step": 4172
+    },
+    {
+      "epoch": 4.448827292110874,
+      "grad_norm": 0.07367637209674983,
+      "learning_rate": 2.924925930307447e-06,
+      "loss": 0.3346,
+      "step": 4173
+    },
+    {
+      "epoch": 4.449893390191898,
+      "grad_norm": 0.07495993556247976,
+      "learning_rate": 2.9137611641359222e-06,
+      "loss": 0.3321,
+      "step": 4174
+    },
+    {
+      "epoch": 4.450959488272921,
+      "grad_norm": 0.07267443806340775,
+      "learning_rate": 2.902616941813281e-06,
+      "loss": 0.3299,
+      "step": 4175
+    },
+    {
+      "epoch": 4.452025586353945,
+      "grad_norm": 0.0708925866597283,
+      "learning_rate": 2.8914932695128393e-06,
+      "loss": 0.3332,
+      "step": 4176
+    },
+    {
+      "epoch": 4.453091684434968,
+      "grad_norm": 0.07269636700113985,
+      "learning_rate": 2.880390153396544e-06,
+      "loss": 0.3303,
+      "step": 4177
+    },
+    {
+      "epoch": 4.454157782515992,
+      "grad_norm": 0.07332256346717313,
+      "learning_rate": 2.86930759961495e-06,
+      "loss": 0.3294,
+      "step": 4178
+    },
+    {
+      "epoch": 4.455223880597015,
+      "grad_norm": 0.07309550795550446,
+      "learning_rate": 2.8582456143071956e-06,
+      "loss": 0.3318,
+      "step": 4179
+    },
+    {
+      "epoch": 4.456289978678038,
+      "grad_norm": 0.07177315229711953,
+      "learning_rate": 2.8472042036010594e-06,
+      "loss": 0.3333,
+      "step": 4180
+    },
+    {
+      "epoch": 4.457356076759062,
+      "grad_norm": 0.0747294310777951,
+      "learning_rate": 2.8361833736129107e-06,
+      "loss": 0.3239,
+      "step": 4181
+    },
+    {
+      "epoch": 4.458422174840085,
+      "grad_norm": 0.07554108527569321,
+      "learning_rate": 2.8251831304477108e-06,
+      "loss": 0.3381,
+      "step": 4182
+    },
+    {
+      "epoch": 4.459488272921108,
+      "grad_norm": 0.07278816202889332,
+      "learning_rate": 2.81420348019902e-06,
+      "loss": 0.3347,
+      "step": 4183
+    },
+    {
+      "epoch": 4.460554371002132,
+      "grad_norm": 0.07323691646275544,
+      "learning_rate": 2.8032444289490012e-06,
+      "loss": 0.3337,
+      "step": 4184
+    },
+    {
+      "epoch": 4.461620469083155,
+      "grad_norm": 0.07761545782310925,
+      "learning_rate": 2.792305982768402e-06,
+      "loss": 0.3331,
+      "step": 4185
+    },
+    {
+      "epoch": 4.462686567164179,
+      "grad_norm": 0.0739374400916545,
+      "learning_rate": 2.7813881477165395e-06,
+      "loss": 0.3279,
+      "step": 4186
+    },
+    {
+      "epoch": 4.463752665245202,
+      "grad_norm": 0.07242003332809858,
+      "learning_rate": 2.7704909298413362e-06,
+      "loss": 0.3363,
+      "step": 4187
+    },
+    {
+      "epoch": 4.464818763326226,
+      "grad_norm": 0.07187298705284119,
+      "learning_rate": 2.7596143351792837e-06,
+      "loss": 0.3386,
+      "step": 4188
+    },
+    {
+      "epoch": 4.4658848614072495,
+      "grad_norm": 0.07239042923705526,
+      "learning_rate": 2.7487583697554555e-06,
+      "loss": 0.3323,
+      "step": 4189
+    },
+    {
+      "epoch": 4.466950959488273,
+      "grad_norm": 0.07678589531267026,
+      "learning_rate": 2.7379230395834764e-06,
+      "loss": 0.3329,
+      "step": 4190
+    },
+    {
+      "epoch": 4.468017057569297,
+      "grad_norm": 0.07189456346112492,
+      "learning_rate": 2.7271083506655728e-06,
+      "loss": 0.3382,
+      "step": 4191
+    },
+    {
+      "epoch": 4.46908315565032,
+      "grad_norm": 0.07515170520654751,
+      "learning_rate": 2.716314308992516e-06,
+      "loss": 0.3385,
+      "step": 4192
+    },
+    {
+      "epoch": 4.470149253731344,
+      "grad_norm": 0.07313542483059168,
+      "learning_rate": 2.7055409205436346e-06,
+      "loss": 0.3316,
+      "step": 4193
+    },
+    {
+      "epoch": 4.471215351812367,
+      "grad_norm": 0.07066165546162541,
+      "learning_rate": 2.6947881912868346e-06,
+      "loss": 0.3315,
+      "step": 4194
+    },
+    {
+      "epoch": 4.47228144989339,
+      "grad_norm": 0.07272630195981485,
+      "learning_rate": 2.6840561271785694e-06,
+      "loss": 0.3354,
+      "step": 4195
+    },
+    {
+      "epoch": 4.473347547974414,
+      "grad_norm": 0.07161497516187117,
+      "learning_rate": 2.6733447341638472e-06,
+      "loss": 0.3388,
+      "step": 4196
+    },
+    {
+      "epoch": 4.474413646055437,
+      "grad_norm": 0.07454202797819427,
+      "learning_rate": 2.662654018176212e-06,
+      "loss": 0.3333,
+      "step": 4197
+    },
+    {
+      "epoch": 4.475479744136461,
+      "grad_norm": 0.07508088009677,
+      "learning_rate": 2.6519839851377737e-06,
+      "loss": 0.3387,
+      "step": 4198
+    },
+    {
+      "epoch": 4.476545842217484,
+      "grad_norm": 0.07804787962517462,
+      "learning_rate": 2.6413346409591745e-06,
+      "loss": 0.3317,
+      "step": 4199
+    },
+    {
+      "epoch": 4.477611940298507,
+      "grad_norm": 0.07128589665687961,
+      "learning_rate": 2.630705991539602e-06,
+      "loss": 0.3251,
+      "step": 4200
+    },
+    {
+      "epoch": 4.478678038379531,
+      "grad_norm": 0.07355996310666547,
+      "learning_rate": 2.6200980427667635e-06,
+      "loss": 0.3297,
+      "step": 4201
+    },
+    {
+      "epoch": 4.479744136460554,
+      "grad_norm": 0.07500471190524303,
+      "learning_rate": 2.6095108005169188e-06,
+      "loss": 0.3292,
+      "step": 4202
+    },
+    {
+      "epoch": 4.480810234541578,
+      "grad_norm": 0.07946453180732672,
+      "learning_rate": 2.5989442706548574e-06,
+      "loss": 0.3352,
+      "step": 4203
+    },
+    {
+      "epoch": 4.481876332622601,
+      "grad_norm": 0.06970897348276822,
+      "learning_rate": 2.5883984590338738e-06,
+      "loss": 0.3279,
+      "step": 4204
+    },
+    {
+      "epoch": 4.482942430703625,
+      "grad_norm": 0.08238106944138114,
+      "learning_rate": 2.5778733714958027e-06,
+      "loss": 0.3324,
+      "step": 4205
+    },
+    {
+      "epoch": 4.484008528784648,
+      "grad_norm": 0.07235582940293811,
+      "learning_rate": 2.5673690138710018e-06,
+      "loss": 0.3365,
+      "step": 4206
+    },
+    {
+      "epoch": 4.485074626865671,
+      "grad_norm": 0.07346252026606727,
+      "learning_rate": 2.556885391978341e-06,
+      "loss": 0.3349,
+      "step": 4207
+    },
+    {
+      "epoch": 4.486140724946695,
+      "grad_norm": 0.07558966474067527,
+      "learning_rate": 2.5464225116251886e-06,
+      "loss": 0.3358,
+      "step": 4208
+    },
+    {
+      "epoch": 4.4872068230277184,
+      "grad_norm": 0.07040408366742829,
+      "learning_rate": 2.53598037860745e-06,
+      "loss": 0.331,
+      "step": 4209
+    },
+    {
+      "epoch": 4.4882729211087415,
+      "grad_norm": 0.07488263115221298,
+      "learning_rate": 2.5255589987095207e-06,
+      "loss": 0.3342,
+      "step": 4210
+    },
+    {
+      "epoch": 4.4893390191897655,
+      "grad_norm": 0.07051801023243459,
+      "learning_rate": 2.5151583777042988e-06,
+      "loss": 0.3306,
+      "step": 4211
+    },
+    {
+      "epoch": 4.490405117270789,
+      "grad_norm": 0.07182136451708207,
+      "learning_rate": 2.5047785213531882e-06,
+      "loss": 0.3272,
+      "step": 4212
+    },
+    {
+      "epoch": 4.491471215351813,
+      "grad_norm": 0.07727816037885224,
+      "learning_rate": 2.494419435406097e-06,
+      "loss": 0.3354,
+      "step": 4213
+    },
+    {
+      "epoch": 4.492537313432836,
+      "grad_norm": 0.07344914255236583,
+      "learning_rate": 2.4840811256014164e-06,
+      "loss": 0.3293,
+      "step": 4214
+    },
+    {
+      "epoch": 4.49360341151386,
+      "grad_norm": 0.07192568760462288,
+      "learning_rate": 2.4737635976660325e-06,
+      "loss": 0.3287,
+      "step": 4215
+    },
+    {
+      "epoch": 4.494669509594883,
+      "grad_norm": 0.0701357026382714,
+      "learning_rate": 2.4634668573153154e-06,
+      "loss": 0.3386,
+      "step": 4216
+    },
+    {
+      "epoch": 4.495735607675906,
+      "grad_norm": 0.06637179767672738,
+      "learning_rate": 2.4531909102531294e-06,
+      "loss": 0.3325,
+      "step": 4217
+    },
+    {
+      "epoch": 4.49680170575693,
+      "grad_norm": 0.07656271024622706,
+      "learning_rate": 2.442935762171819e-06,
+      "loss": 0.3394,
+      "step": 4218
+    },
+    {
+      "epoch": 4.497867803837953,
+      "grad_norm": 0.0714666290950529,
+      "learning_rate": 2.4327014187521948e-06,
+      "loss": 0.3367,
+      "step": 4219
+    },
+    {
+      "epoch": 4.498933901918977,
+      "grad_norm": 0.07606369364800027,
+      "learning_rate": 2.422487885663554e-06,
+      "loss": 0.3386,
+      "step": 4220
+    },
+    {
+      "epoch": 4.5,
+      "grad_norm": 0.06844926088782544,
+      "learning_rate": 2.4122951685636674e-06,
+      "loss": 0.3326,
+      "step": 4221
+    },
+    {
+      "epoch": 4.501066098081023,
+      "grad_norm": 0.07046377328122992,
+      "learning_rate": 2.4021232730987622e-06,
+      "loss": 0.3352,
+      "step": 4222
+    },
+    {
+      "epoch": 4.502132196162047,
+      "grad_norm": 0.07467074933049231,
+      "learning_rate": 2.3919722049035433e-06,
+      "loss": 0.3334,
+      "step": 4223
+    },
+    {
+      "epoch": 4.50319829424307,
+      "grad_norm": 0.07060359053074924,
+      "learning_rate": 2.38184196960118e-06,
+      "loss": 0.3362,
+      "step": 4224
+    },
+    {
+      "epoch": 4.504264392324094,
+      "grad_norm": 0.07379325677186419,
+      "learning_rate": 2.3717325728032935e-06,
+      "loss": 0.3429,
+      "step": 4225
+    },
+    {
+      "epoch": 4.505330490405117,
+      "grad_norm": 0.07310277967696212,
+      "learning_rate": 2.3616440201099567e-06,
+      "loss": 0.3388,
+      "step": 4226
+    },
+    {
+      "epoch": 4.50639658848614,
+      "grad_norm": 0.07006559210490895,
+      "learning_rate": 2.3515763171097115e-06,
+      "loss": 0.3355,
+      "step": 4227
+    },
+    {
+      "epoch": 4.507462686567164,
+      "grad_norm": 0.07133173745971073,
+      "learning_rate": 2.341529469379551e-06,
+      "loss": 0.3318,
+      "step": 4228
+    },
+    {
+      "epoch": 4.508528784648187,
+      "grad_norm": 0.07098735264122637,
+      "learning_rate": 2.3315034824848846e-06,
+      "loss": 0.3346,
+      "step": 4229
+    },
+    {
+      "epoch": 4.509594882729211,
+      "grad_norm": 0.0777511243612946,
+      "learning_rate": 2.3214983619795995e-06,
+      "loss": 0.3382,
+      "step": 4230
+    },
+    {
+      "epoch": 4.5106609808102345,
+      "grad_norm": 0.06908524988843785,
+      "learning_rate": 2.3115141134060215e-06,
+      "loss": 0.3351,
+      "step": 4231
+    },
+    {
+      "epoch": 4.5117270788912585,
+      "grad_norm": 0.07463758297713084,
+      "learning_rate": 2.301550742294887e-06,
+      "loss": 0.3385,
+      "step": 4232
+    },
+    {
+      "epoch": 4.5127931769722816,
+      "grad_norm": 0.07271121038748618,
+      "learning_rate": 2.2916082541653983e-06,
+      "loss": 0.3349,
+      "step": 4233
+    },
+    {
+      "epoch": 4.513859275053305,
+      "grad_norm": 0.07475883982198514,
+      "learning_rate": 2.281686654525177e-06,
+      "loss": 0.333,
+      "step": 4234
+    },
+    {
+      "epoch": 4.514925373134329,
+      "grad_norm": 0.07660331206011552,
+      "learning_rate": 2.2717859488702665e-06,
+      "loss": 0.3378,
+      "step": 4235
+    },
+    {
+      "epoch": 4.515991471215352,
+      "grad_norm": 0.07040590430776542,
+      "learning_rate": 2.2619061426851463e-06,
+      "loss": 0.3353,
+      "step": 4236
+    },
+    {
+      "epoch": 4.517057569296375,
+      "grad_norm": 0.07585806379146305,
+      "learning_rate": 2.252047241442723e-06,
+      "loss": 0.3395,
+      "step": 4237
+    },
+    {
+      "epoch": 4.518123667377399,
+      "grad_norm": 0.07561687413038395,
+      "learning_rate": 2.2422092506043036e-06,
+      "loss": 0.3367,
+      "step": 4238
+    },
+    {
+      "epoch": 4.519189765458422,
+      "grad_norm": 0.06884029998313776,
+      "learning_rate": 2.2323921756196263e-06,
+      "loss": 0.336,
+      "step": 4239
+    },
+    {
+      "epoch": 4.520255863539446,
+      "grad_norm": 0.0699098475783151,
+      "learning_rate": 2.2225960219268526e-06,
+      "loss": 0.3348,
+      "step": 4240
+    },
+    {
+      "epoch": 4.521321961620469,
+      "grad_norm": 0.07422236573639096,
+      "learning_rate": 2.212820794952526e-06,
+      "loss": 0.3304,
+      "step": 4241
+    },
+    {
+      "epoch": 4.522388059701493,
+      "grad_norm": 0.07002733474761201,
+      "learning_rate": 2.2030665001116213e-06,
+      "loss": 0.3297,
+      "step": 4242
+    },
+    {
+      "epoch": 4.523454157782516,
+      "grad_norm": 0.07060390805754775,
+      "learning_rate": 2.1933331428075146e-06,
+      "loss": 0.3305,
+      "step": 4243
+    },
+    {
+      "epoch": 4.524520255863539,
+      "grad_norm": 0.07131829668068262,
+      "learning_rate": 2.1836207284319724e-06,
+      "loss": 0.331,
+      "step": 4244
+    },
+    {
+      "epoch": 4.525586353944563,
+      "grad_norm": 0.07092568911405243,
+      "learning_rate": 2.1739292623651755e-06,
+      "loss": 0.3324,
+      "step": 4245
+    },
+    {
+      "epoch": 4.526652452025586,
+      "grad_norm": 0.07054221449199592,
+      "learning_rate": 2.164258749975683e-06,
+      "loss": 0.3366,
+      "step": 4246
+    },
+    {
+      "epoch": 4.52771855010661,
+      "grad_norm": 0.07706442180790107,
+      "learning_rate": 2.154609196620472e-06,
+      "loss": 0.3371,
+      "step": 4247
+    },
+    {
+      "epoch": 4.528784648187633,
+      "grad_norm": 0.07508166585212031,
+      "learning_rate": 2.144980607644871e-06,
+      "loss": 0.3321,
+      "step": 4248
+    },
+    {
+      "epoch": 4.529850746268656,
+      "grad_norm": 0.07179015508589877,
+      "learning_rate": 2.135372988382636e-06,
+      "loss": 0.3326,
+      "step": 4249
+    },
+    {
+      "epoch": 4.53091684434968,
+      "grad_norm": 0.07298872176356,
+      "learning_rate": 2.1257863441558867e-06,
+      "loss": 0.3336,
+      "step": 4250
+    },
+    {
+      "epoch": 4.531982942430703,
+      "grad_norm": 0.07322352264632,
+      "learning_rate": 2.116220680275114e-06,
+      "loss": 0.335,
+      "step": 4251
+    },
+    {
+      "epoch": 4.533049040511727,
+      "grad_norm": 0.07412360547888225,
+      "learning_rate": 2.1066760020392075e-06,
+      "loss": 0.333,
+      "step": 4252
+    },
+    {
+      "epoch": 4.5341151385927505,
+      "grad_norm": 0.07037038219362778,
+      "learning_rate": 2.0971523147354224e-06,
+      "loss": 0.3345,
+      "step": 4253
+    },
+    {
+      "epoch": 4.535181236673774,
+      "grad_norm": 0.07385081135383645,
+      "learning_rate": 2.0876496236393915e-06,
+      "loss": 0.3361,
+      "step": 4254
+    },
+    {
+      "epoch": 4.536247334754798,
+      "grad_norm": 0.07977224842487596,
+      "learning_rate": 2.0781679340151007e-06,
+      "loss": 0.3322,
+      "step": 4255
+    },
+    {
+      "epoch": 4.537313432835821,
+      "grad_norm": 0.07150280522896328,
+      "learning_rate": 2.0687072511149207e-06,
+      "loss": 0.3343,
+      "step": 4256
+    },
+    {
+      "epoch": 4.538379530916845,
+      "grad_norm": 0.07550230379695838,
+      "learning_rate": 2.0592675801795715e-06,
+      "loss": 0.3289,
+      "step": 4257
+    },
+    {
+      "epoch": 4.539445628997868,
+      "grad_norm": 0.07468614314295811,
+      "learning_rate": 2.0498489264381537e-06,
+      "loss": 0.3364,
+      "step": 4258
+    },
+    {
+      "epoch": 4.540511727078892,
+      "grad_norm": 0.07766324270385316,
+      "learning_rate": 2.040451295108099e-06,
+      "loss": 0.3332,
+      "step": 4259
+    },
+    {
+      "epoch": 4.541577825159915,
+      "grad_norm": 0.0736473092823455,
+      "learning_rate": 2.0310746913952075e-06,
+      "loss": 0.3307,
+      "step": 4260
+    },
+    {
+      "epoch": 4.542643923240938,
+      "grad_norm": 0.0734896770686822,
+      "learning_rate": 2.0217191204936393e-06,
+      "loss": 0.3396,
+      "step": 4261
+    },
+    {
+      "epoch": 4.543710021321962,
+      "grad_norm": 0.06952100029741023,
+      "learning_rate": 2.012384587585885e-06,
+      "loss": 0.3346,
+      "step": 4262
+    },
+    {
+      "epoch": 4.544776119402985,
+      "grad_norm": 0.07462393218580446,
+      "learning_rate": 2.003071097842795e-06,
+      "loss": 0.3356,
+      "step": 4263
+    },
+    {
+      "epoch": 4.545842217484008,
+      "grad_norm": 0.07095210368849357,
+      "learning_rate": 1.993778656423557e-06,
+      "loss": 0.3299,
+      "step": 4264
+    },
+    {
+      "epoch": 4.546908315565032,
+      "grad_norm": 0.07173596162673962,
+      "learning_rate": 1.9845072684757084e-06,
+      "loss": 0.3331,
+      "step": 4265
+    },
+    {
+      "epoch": 4.547974413646055,
+      "grad_norm": 0.07462211744625402,
+      "learning_rate": 1.975256939135104e-06,
+      "loss": 0.3331,
+      "step": 4266
+    },
+    {
+      "epoch": 4.549040511727079,
+      "grad_norm": 0.06797661761770143,
+      "learning_rate": 1.966027673525952e-06,
+      "loss": 0.3349,
+      "step": 4267
+    },
+    {
+      "epoch": 4.550106609808102,
+      "grad_norm": 0.06909061390055764,
+      "learning_rate": 1.9568194767607897e-06,
+      "loss": 0.333,
+      "step": 4268
+    },
+    {
+      "epoch": 4.551172707889126,
+      "grad_norm": 0.07230285207266503,
+      "learning_rate": 1.9476323539404697e-06,
+      "loss": 0.336,
+      "step": 4269
+    },
+    {
+      "epoch": 4.552238805970149,
+      "grad_norm": 0.06795386640951454,
+      "learning_rate": 1.9384663101541834e-06,
+      "loss": 0.3306,
+      "step": 4270
+    },
+    {
+      "epoch": 4.553304904051172,
+      "grad_norm": 0.07185284208858062,
+      "learning_rate": 1.9293213504794474e-06,
+      "loss": 0.3329,
+      "step": 4271
+    },
+    {
+      "epoch": 4.554371002132196,
+      "grad_norm": 0.06964066381645194,
+      "learning_rate": 1.9201974799820976e-06,
+      "loss": 0.3336,
+      "step": 4272
+    },
+    {
+      "epoch": 4.5554371002132195,
+      "grad_norm": 0.06841885093169003,
+      "learning_rate": 1.911094703716274e-06,
+      "loss": 0.3373,
+      "step": 4273
+    },
+    {
+      "epoch": 4.556503198294243,
+      "grad_norm": 0.07829671714309364,
+      "learning_rate": 1.9020130267244408e-06,
+      "loss": 0.3337,
+      "step": 4274
+    },
+    {
+      "epoch": 4.5575692963752665,
+      "grad_norm": 0.07118332494914492,
+      "learning_rate": 1.8929524540373868e-06,
+      "loss": 0.3337,
+      "step": 4275
+    },
+    {
+      "epoch": 4.55863539445629,
+      "grad_norm": 0.07251578667557201,
+      "learning_rate": 1.8839129906741903e-06,
+      "loss": 0.3374,
+      "step": 4276
+    },
+    {
+      "epoch": 4.559701492537314,
+      "grad_norm": 0.06723960489550845,
+      "learning_rate": 1.8748946416422464e-06,
+      "loss": 0.3306,
+      "step": 4277
+    },
+    {
+      "epoch": 4.560767590618337,
+      "grad_norm": 0.07111432053727844,
+      "learning_rate": 1.8658974119372475e-06,
+      "loss": 0.3356,
+      "step": 4278
+    },
+    {
+      "epoch": 4.561833688699361,
+      "grad_norm": 0.06575145639594351,
+      "learning_rate": 1.856921306543198e-06,
+      "loss": 0.3326,
+      "step": 4279
+    },
+    {
+      "epoch": 4.562899786780384,
+      "grad_norm": 0.07466010683175027,
+      "learning_rate": 1.847966330432387e-06,
+      "loss": 0.3331,
+      "step": 4280
+    },
+    {
+      "epoch": 4.563965884861407,
+      "grad_norm": 0.07631299648143247,
+      "learning_rate": 1.839032488565411e-06,
+      "loss": 0.3351,
+      "step": 4281
+    },
+    {
+      "epoch": 4.565031982942431,
+      "grad_norm": 0.06992148350886665,
+      "learning_rate": 1.8301197858911512e-06,
+      "loss": 0.3341,
+      "step": 4282
+    },
+    {
+      "epoch": 4.566098081023454,
+      "grad_norm": 0.06950406008223693,
+      "learning_rate": 1.8212282273467874e-06,
+      "loss": 0.3326,
+      "step": 4283
+    },
+    {
+      "epoch": 4.567164179104478,
+      "grad_norm": 0.0715513717486181,
+      "learning_rate": 1.8123578178577706e-06,
+      "loss": 0.3314,
+      "step": 4284
+    },
+    {
+      "epoch": 4.568230277185501,
+      "grad_norm": 0.07759734713431235,
+      "learning_rate": 1.8035085623378544e-06,
+      "loss": 0.3284,
+      "step": 4285
+    },
+    {
+      "epoch": 4.569296375266525,
+      "grad_norm": 0.07289404960905593,
+      "learning_rate": 1.7946804656890648e-06,
+      "loss": 0.3339,
+      "step": 4286
+    },
+    {
+      "epoch": 4.570362473347548,
+      "grad_norm": 0.07031085270085344,
+      "learning_rate": 1.7858735328017119e-06,
+      "loss": 0.3352,
+      "step": 4287
+    },
+    {
+      "epoch": 4.571428571428571,
+      "grad_norm": 0.0683840513250094,
+      "learning_rate": 1.7770877685543687e-06,
+      "loss": 0.3319,
+      "step": 4288
+    },
+    {
+      "epoch": 4.572494669509595,
+      "grad_norm": 0.07276081186481145,
+      "learning_rate": 1.768323177813902e-06,
+      "loss": 0.3314,
+      "step": 4289
+    },
+    {
+      "epoch": 4.573560767590618,
+      "grad_norm": 0.07091066982025185,
+      "learning_rate": 1.7595797654354374e-06,
+      "loss": 0.3327,
+      "step": 4290
+    },
+    {
+      "epoch": 4.574626865671641,
+      "grad_norm": 0.07166306063065245,
+      "learning_rate": 1.750857536262367e-06,
+      "loss": 0.3349,
+      "step": 4291
+    },
+    {
+      "epoch": 4.575692963752665,
+      "grad_norm": 0.07054961140916155,
+      "learning_rate": 1.7421564951263547e-06,
+      "loss": 0.332,
+      "step": 4292
+    },
+    {
+      "epoch": 4.576759061833688,
+      "grad_norm": 0.07361301895145772,
+      "learning_rate": 1.7334766468473275e-06,
+      "loss": 0.3337,
+      "step": 4293
+    },
+    {
+      "epoch": 4.577825159914712,
+      "grad_norm": 0.07045798280358322,
+      "learning_rate": 1.7248179962334699e-06,
+      "loss": 0.3359,
+      "step": 4294
+    },
+    {
+      "epoch": 4.5788912579957355,
+      "grad_norm": 0.06774482441192077,
+      "learning_rate": 1.7161805480812166e-06,
+      "loss": 0.3327,
+      "step": 4295
+    },
+    {
+      "epoch": 4.5799573560767595,
+      "grad_norm": 0.0718262080789473,
+      "learning_rate": 1.7075643071752735e-06,
+      "loss": 0.3345,
+      "step": 4296
+    },
+    {
+      "epoch": 4.581023454157783,
+      "grad_norm": 0.0691615331261111,
+      "learning_rate": 1.6989692782885914e-06,
+      "loss": 0.332,
+      "step": 4297
+    },
+    {
+      "epoch": 4.582089552238806,
+      "grad_norm": 0.0731179918926363,
+      "learning_rate": 1.6903954661823618e-06,
+      "loss": 0.3332,
+      "step": 4298
+    },
+    {
+      "epoch": 4.58315565031983,
+      "grad_norm": 0.07078605872262074,
+      "learning_rate": 1.6818428756060346e-06,
+      "loss": 0.3331,
+      "step": 4299
+    },
+    {
+      "epoch": 4.584221748400853,
+      "grad_norm": 0.06734701136745615,
+      "learning_rate": 1.6733115112973042e-06,
+      "loss": 0.3328,
+      "step": 4300
+    },
+    {
+      "epoch": 4.585287846481877,
+      "grad_norm": 0.06790749484835054,
+      "learning_rate": 1.6648013779820972e-06,
+      "loss": 0.3385,
+      "step": 4301
+    },
+    {
+      "epoch": 4.5863539445629,
+      "grad_norm": 0.07187466083950145,
+      "learning_rate": 1.656312480374589e-06,
+      "loss": 0.3356,
+      "step": 4302
+    },
+    {
+      "epoch": 4.587420042643923,
+      "grad_norm": 0.07459252483448003,
+      "learning_rate": 1.6478448231771914e-06,
+      "loss": 0.333,
+      "step": 4303
+    },
+    {
+      "epoch": 4.588486140724947,
+      "grad_norm": 0.0717675220591279,
+      "learning_rate": 1.639398411080535e-06,
+      "loss": 0.3319,
+      "step": 4304
+    },
+    {
+      "epoch": 4.58955223880597,
+      "grad_norm": 0.07118074349730492,
+      "learning_rate": 1.6309732487634989e-06,
+      "loss": 0.3318,
+      "step": 4305
+    },
+    {
+      "epoch": 4.590618336886994,
+      "grad_norm": 0.06815517038121584,
+      "learning_rate": 1.6225693408931898e-06,
+      "loss": 0.3395,
+      "step": 4306
+    },
+    {
+      "epoch": 4.591684434968017,
+      "grad_norm": 0.06997889027085387,
+      "learning_rate": 1.6141866921249282e-06,
+      "loss": 0.331,
+      "step": 4307
+    },
+    {
+      "epoch": 4.59275053304904,
+      "grad_norm": 0.07279040293236479,
+      "learning_rate": 1.6058253071022711e-06,
+      "loss": 0.3322,
+      "step": 4308
+    },
+    {
+      "epoch": 4.593816631130064,
+      "grad_norm": 0.0745884036380402,
+      "learning_rate": 1.5974851904569931e-06,
+      "loss": 0.3353,
+      "step": 4309
+    },
+    {
+      "epoch": 4.594882729211087,
+      "grad_norm": 0.06892857071091296,
+      "learning_rate": 1.589166346809079e-06,
+      "loss": 0.3336,
+      "step": 4310
+    },
+    {
+      "epoch": 4.595948827292111,
+      "grad_norm": 0.06646283141120592,
+      "learning_rate": 1.58086878076674e-06,
+      "loss": 0.3355,
+      "step": 4311
+    },
+    {
+      "epoch": 4.597014925373134,
+      "grad_norm": 0.06892657692261639,
+      "learning_rate": 1.5725924969263973e-06,
+      "loss": 0.3341,
+      "step": 4312
+    },
+    {
+      "epoch": 4.598081023454158,
+      "grad_norm": 0.07247620559120913,
+      "learning_rate": 1.5643374998726768e-06,
+      "loss": 0.3371,
+      "step": 4313
+    },
+    {
+      "epoch": 4.599147121535181,
+      "grad_norm": 0.0682871385404355,
+      "learning_rate": 1.5561037941784184e-06,
+      "loss": 0.337,
+      "step": 4314
+    },
+    {
+      "epoch": 4.600213219616204,
+      "grad_norm": 0.07060756058234041,
+      "learning_rate": 1.5478913844046716e-06,
+      "loss": 0.3356,
+      "step": 4315
+    },
+    {
+      "epoch": 4.601279317697228,
+      "grad_norm": 0.06829225466485105,
+      "learning_rate": 1.5397002751006863e-06,
+      "loss": 0.337,
+      "step": 4316
+    },
+    {
+      "epoch": 4.6023454157782515,
+      "grad_norm": 0.06766530449323094,
+      "learning_rate": 1.531530470803908e-06,
+      "loss": 0.3321,
+      "step": 4317
+    },
+    {
+      "epoch": 4.603411513859275,
+      "grad_norm": 0.0670931989643599,
+      "learning_rate": 1.5233819760399793e-06,
+      "loss": 0.335,
+      "step": 4318
+    },
+    {
+      "epoch": 4.604477611940299,
+      "grad_norm": 0.06938288654468625,
+      "learning_rate": 1.5152547953227515e-06,
+      "loss": 0.3318,
+      "step": 4319
+    },
+    {
+      "epoch": 4.605543710021322,
+      "grad_norm": 0.06867733948754896,
+      "learning_rate": 1.5071489331542543e-06,
+      "loss": 0.3314,
+      "step": 4320
+    },
+    {
+      "epoch": 4.606609808102346,
+      "grad_norm": 0.07777991942380198,
+      "learning_rate": 1.499064394024714e-06,
+      "loss": 0.3385,
+      "step": 4321
+    },
+    {
+      "epoch": 4.607675906183369,
+      "grad_norm": 0.06967592936560713,
+      "learning_rate": 1.4910011824125436e-06,
+      "loss": 0.3323,
+      "step": 4322
+    },
+    {
+      "epoch": 4.608742004264393,
+      "grad_norm": 0.06751157296924212,
+      "learning_rate": 1.482959302784357e-06,
+      "loss": 0.3393,
+      "step": 4323
+    },
+    {
+      "epoch": 4.609808102345416,
+      "grad_norm": 0.06979920693263406,
+      "learning_rate": 1.4749387595949195e-06,
+      "loss": 0.336,
+      "step": 4324
+    },
+    {
+      "epoch": 4.610874200426439,
+      "grad_norm": 0.07053350912123188,
+      "learning_rate": 1.4669395572872015e-06,
+      "loss": 0.3298,
+      "step": 4325
+    },
+    {
+      "epoch": 4.611940298507463,
+      "grad_norm": 0.07371928778646851,
+      "learning_rate": 1.4589617002923516e-06,
+      "loss": 0.3342,
+      "step": 4326
+    },
+    {
+      "epoch": 4.613006396588486,
+      "grad_norm": 0.07223172335830214,
+      "learning_rate": 1.4510051930296799e-06,
+      "loss": 0.3321,
+      "step": 4327
+    },
+    {
+      "epoch": 4.61407249466951,
+      "grad_norm": 0.066958488669665,
+      "learning_rate": 1.4430700399066821e-06,
+      "loss": 0.3312,
+      "step": 4328
+    },
+    {
+      "epoch": 4.615138592750533,
+      "grad_norm": 0.07053436931086877,
+      "learning_rate": 1.435156245319016e-06,
+      "loss": 0.3372,
+      "step": 4329
+    },
+    {
+      "epoch": 4.616204690831556,
+      "grad_norm": 0.06968744357208878,
+      "learning_rate": 1.4272638136505257e-06,
+      "loss": 0.3368,
+      "step": 4330
+    },
+    {
+      "epoch": 4.61727078891258,
+      "grad_norm": 0.06859881353886961,
+      "learning_rate": 1.4193927492731897e-06,
+      "loss": 0.3277,
+      "step": 4331
+    },
+    {
+      "epoch": 4.618336886993603,
+      "grad_norm": 0.07231488171540013,
+      "learning_rate": 1.4115430565471776e-06,
+      "loss": 0.3321,
+      "step": 4332
+    },
+    {
+      "epoch": 4.619402985074627,
+      "grad_norm": 0.0666791642191909,
+      "learning_rate": 1.4037147398208118e-06,
+      "loss": 0.3371,
+      "step": 4333
+    },
+    {
+      "epoch": 4.62046908315565,
+      "grad_norm": 0.0776009579849058,
+      "learning_rate": 1.3959078034305785e-06,
+      "loss": 0.3347,
+      "step": 4334
+    },
+    {
+      "epoch": 4.621535181236673,
+      "grad_norm": 0.06614100139929324,
+      "learning_rate": 1.3881222517010983e-06,
+      "loss": 0.3376,
+      "step": 4335
+    },
+    {
+      "epoch": 4.622601279317697,
+      "grad_norm": 0.06757259397296564,
+      "learning_rate": 1.3803580889451795e-06,
+      "loss": 0.3353,
+      "step": 4336
+    },
+    {
+      "epoch": 4.6236673773987205,
+      "grad_norm": 0.07131647416320316,
+      "learning_rate": 1.3726153194637548e-06,
+      "loss": 0.3441,
+      "step": 4337
+    },
+    {
+      "epoch": 4.6247334754797444,
+      "grad_norm": 0.06932530118685151,
+      "learning_rate": 1.3648939475459178e-06,
+      "loss": 0.329,
+      "step": 4338
+    },
+    {
+      "epoch": 4.6257995735607675,
+      "grad_norm": 0.07255409329631657,
+      "learning_rate": 1.3571939774689091e-06,
+      "loss": 0.3288,
+      "step": 4339
+    },
+    {
+      "epoch": 4.6268656716417915,
+      "grad_norm": 0.06673195482515673,
+      "learning_rate": 1.3495154134981126e-06,
+      "loss": 0.3354,
+      "step": 4340
+    },
+    {
+      "epoch": 4.627931769722815,
+      "grad_norm": 0.06572556100183624,
+      "learning_rate": 1.34185825988705e-06,
+      "loss": 0.3329,
+      "step": 4341
+    },
+    {
+      "epoch": 4.628997867803838,
+      "grad_norm": 0.06490987870262467,
+      "learning_rate": 1.3342225208773906e-06,
+      "loss": 0.3354,
+      "step": 4342
+    },
+    {
+      "epoch": 4.630063965884862,
+      "grad_norm": 0.07309588285106411,
+      "learning_rate": 1.3266082006989333e-06,
+      "loss": 0.331,
+      "step": 4343
+    },
+    {
+      "epoch": 4.631130063965885,
+      "grad_norm": 0.07038414592289387,
+      "learning_rate": 1.3190153035696196e-06,
+      "loss": 0.3435,
+      "step": 4344
+    },
+    {
+      "epoch": 4.632196162046908,
+      "grad_norm": 0.06928635687457206,
+      "learning_rate": 1.311443833695516e-06,
+      "loss": 0.3364,
+      "step": 4345
+    },
+    {
+      "epoch": 4.633262260127932,
+      "grad_norm": 0.06537020208079272,
+      "learning_rate": 1.303893795270823e-06,
+      "loss": 0.3316,
+      "step": 4346
+    },
+    {
+      "epoch": 4.634328358208955,
+      "grad_norm": 0.06896350457838996,
+      "learning_rate": 1.296365192477871e-06,
+      "loss": 0.3367,
+      "step": 4347
+    },
+    {
+      "epoch": 4.635394456289979,
+      "grad_norm": 0.06599969292032785,
+      "learning_rate": 1.2888580294871233e-06,
+      "loss": 0.3346,
+      "step": 4348
+    },
+    {
+      "epoch": 4.636460554371002,
+      "grad_norm": 0.06925949481192152,
+      "learning_rate": 1.281372310457143e-06,
+      "loss": 0.3349,
+      "step": 4349
+    },
+    {
+      "epoch": 4.637526652452026,
+      "grad_norm": 0.07032567037244174,
+      "learning_rate": 1.2739080395346347e-06,
+      "loss": 0.3324,
+      "step": 4350
+    },
+    {
+      "epoch": 4.638592750533049,
+      "grad_norm": 0.06808103931489896,
+      "learning_rate": 1.2664652208544205e-06,
+      "loss": 0.3364,
+      "step": 4351
+    },
+    {
+      "epoch": 4.639658848614072,
+      "grad_norm": 0.07415952469477234,
+      "learning_rate": 1.2590438585394372e-06,
+      "loss": 0.3368,
+      "step": 4352
+    },
+    {
+      "epoch": 4.640724946695096,
+      "grad_norm": 0.07472723302057231,
+      "learning_rate": 1.2516439567007254e-06,
+      "loss": 0.3312,
+      "step": 4353
+    },
+    {
+      "epoch": 4.641791044776119,
+      "grad_norm": 0.06723931963836592,
+      "learning_rate": 1.2442655194374464e-06,
+      "loss": 0.3363,
+      "step": 4354
+    },
+    {
+      "epoch": 4.642857142857143,
+      "grad_norm": 0.06731787952306181,
+      "learning_rate": 1.2369085508368862e-06,
+      "loss": 0.3384,
+      "step": 4355
+    },
+    {
+      "epoch": 4.643923240938166,
+      "grad_norm": 0.06995325462163068,
+      "learning_rate": 1.2295730549744023e-06,
+      "loss": 0.3331,
+      "step": 4356
+    },
+    {
+      "epoch": 4.644989339019189,
+      "grad_norm": 0.07264991945604682,
+      "learning_rate": 1.2222590359134868e-06,
+      "loss": 0.3335,
+      "step": 4357
+    },
+    {
+      "epoch": 4.646055437100213,
+      "grad_norm": 0.0664979493499962,
+      "learning_rate": 1.2149664977057296e-06,
+      "loss": 0.331,
+      "step": 4358
+    },
+    {
+      "epoch": 4.6471215351812365,
+      "grad_norm": 0.06852799522357257,
+      "learning_rate": 1.2076954443908152e-06,
+      "loss": 0.3342,
+      "step": 4359
+    },
+    {
+      "epoch": 4.6481876332622605,
+      "grad_norm": 0.0700385586316551,
+      "learning_rate": 1.20044587999653e-06,
+      "loss": 0.3324,
+      "step": 4360
+    },
+    {
+      "epoch": 4.649253731343284,
+      "grad_norm": 0.06744640423267867,
+      "learning_rate": 1.1932178085387514e-06,
+      "loss": 0.3336,
+      "step": 4361
+    },
+    {
+      "epoch": 4.650319829424307,
+      "grad_norm": 0.07133474365028102,
+      "learning_rate": 1.1860112340214624e-06,
+      "loss": 0.3357,
+      "step": 4362
+    },
+    {
+      "epoch": 4.651385927505331,
+      "grad_norm": 0.06780843638739921,
+      "learning_rate": 1.178826160436728e-06,
+      "loss": 0.3339,
+      "step": 4363
+    },
+    {
+      "epoch": 4.652452025586354,
+      "grad_norm": 0.06642970597324976,
+      "learning_rate": 1.1716625917647018e-06,
+      "loss": 0.3341,
+      "step": 4364
+    },
+    {
+      "epoch": 4.653518123667378,
+      "grad_norm": 0.07245838112455406,
+      "learning_rate": 1.1645205319736318e-06,
+      "loss": 0.3345,
+      "step": 4365
+    },
+    {
+      "epoch": 4.654584221748401,
+      "grad_norm": 0.07452837382246935,
+      "learning_rate": 1.1573999850198515e-06,
+      "loss": 0.3311,
+      "step": 4366
+    },
+    {
+      "epoch": 4.655650319829425,
+      "grad_norm": 0.07068223858842627,
+      "learning_rate": 1.1503009548477695e-06,
+      "loss": 0.3382,
+      "step": 4367
+    },
+    {
+      "epoch": 4.656716417910448,
+      "grad_norm": 0.06557927350005839,
+      "learning_rate": 1.143223445389876e-06,
+      "loss": 0.3304,
+      "step": 4368
+    },
+    {
+      "epoch": 4.657782515991471,
+      "grad_norm": 0.0663596931630506,
+      "learning_rate": 1.1361674605667505e-06,
+      "loss": 0.3306,
+      "step": 4369
+    },
+    {
+      "epoch": 4.658848614072495,
+      "grad_norm": 0.06703039652535904,
+      "learning_rate": 1.1291330042870396e-06,
+      "loss": 0.3375,
+      "step": 4370
+    },
+    {
+      "epoch": 4.659914712153518,
+      "grad_norm": 0.06939523423465807,
+      "learning_rate": 1.122120080447462e-06,
+      "loss": 0.3369,
+      "step": 4371
+    },
+    {
+      "epoch": 4.660980810234541,
+      "grad_norm": 0.07088459747200242,
+      "learning_rate": 1.115128692932821e-06,
+      "loss": 0.3397,
+      "step": 4372
+    },
+    {
+      "epoch": 4.662046908315565,
+      "grad_norm": 0.07022817101518151,
+      "learning_rate": 1.1081588456159786e-06,
+      "loss": 0.3359,
+      "step": 4373
+    },
+    {
+      "epoch": 4.663113006396588,
+      "grad_norm": 0.07123058169602929,
+      "learning_rate": 1.10121054235786e-06,
+      "loss": 0.3333,
+      "step": 4374
+    },
+    {
+      "epoch": 4.664179104477612,
+      "grad_norm": 0.0675246173798063,
+      "learning_rate": 1.0942837870074795e-06,
+      "loss": 0.3333,
+      "step": 4375
+    },
+    {
+      "epoch": 4.665245202558635,
+      "grad_norm": 0.06471932900262051,
+      "learning_rate": 1.087378583401888e-06,
+      "loss": 0.3321,
+      "step": 4376
+    },
+    {
+      "epoch": 4.666311300639659,
+      "grad_norm": 0.06686992898419163,
+      "learning_rate": 1.080494935366212e-06,
+      "loss": 0.3331,
+      "step": 4377
+    },
+    {
+      "epoch": 4.667377398720682,
+      "grad_norm": 0.06937875917999614,
+      "learning_rate": 1.073632846713637e-06,
+      "loss": 0.3362,
+      "step": 4378
+    },
+    {
+      "epoch": 4.6684434968017055,
+      "grad_norm": 0.06987137809858991,
+      "learning_rate": 1.0667923212454023e-06,
+      "loss": 0.3368,
+      "step": 4379
+    },
+    {
+      "epoch": 4.669509594882729,
+      "grad_norm": 0.06867240275257108,
+      "learning_rate": 1.0599733627508013e-06,
+      "loss": 0.3346,
+      "step": 4380
+    },
+    {
+      "epoch": 4.6705756929637525,
+      "grad_norm": 0.06826231476314719,
+      "learning_rate": 1.0531759750071857e-06,
+      "loss": 0.3313,
+      "step": 4381
+    },
+    {
+      "epoch": 4.6716417910447765,
+      "grad_norm": 0.06977322414302481,
+      "learning_rate": 1.0464001617799525e-06,
+      "loss": 0.3349,
+      "step": 4382
+    },
+    {
+      "epoch": 4.6727078891258,
+      "grad_norm": 0.07065917384047074,
+      "learning_rate": 1.0396459268225523e-06,
+      "loss": 0.3303,
+      "step": 4383
+    },
+    {
+      "epoch": 4.673773987206823,
+      "grad_norm": 0.06723899201810622,
+      "learning_rate": 1.0329132738764814e-06,
+      "loss": 0.3376,
+      "step": 4384
+    },
+    {
+      "epoch": 4.674840085287847,
+      "grad_norm": 0.06708033630230664,
+      "learning_rate": 1.0262022066712763e-06,
+      "loss": 0.3368,
+      "step": 4385
+    },
+    {
+      "epoch": 4.67590618336887,
+      "grad_norm": 0.06703109426648023,
+      "learning_rate": 1.0195127289245188e-06,
+      "loss": 0.3305,
+      "step": 4386
+    },
+    {
+      "epoch": 4.676972281449894,
+      "grad_norm": 0.07014172973051326,
+      "learning_rate": 1.0128448443418315e-06,
+      "loss": 0.3354,
+      "step": 4387
+    },
+    {
+      "epoch": 4.678038379530917,
+      "grad_norm": 0.07076031888009388,
+      "learning_rate": 1.0061985566168863e-06,
+      "loss": 0.3393,
+      "step": 4388
+    },
+    {
+      "epoch": 4.67910447761194,
+      "grad_norm": 0.06971401068803788,
+      "learning_rate": 9.99573869431365e-07,
+      "loss": 0.3405,
+      "step": 4389
+    },
+    {
+      "epoch": 4.680170575692964,
+      "grad_norm": 0.06595206438851949,
+      "learning_rate": 9.92970786455012e-07,
+      "loss": 0.3291,
+      "step": 4390
+    },
+    {
+      "epoch": 4.681236673773987,
+      "grad_norm": 0.07143631976414835,
+      "learning_rate": 9.863893113455857e-07,
+      "loss": 0.3294,
+      "step": 4391
+    },
+    {
+      "epoch": 4.682302771855011,
+      "grad_norm": 0.06704484678485922,
+      "learning_rate": 9.798294477488901e-07,
+      "loss": 0.3355,
+      "step": 4392
+    },
+    {
+      "epoch": 4.683368869936034,
+      "grad_norm": 0.06626390001420994,
+      "learning_rate": 9.732911992987382e-07,
+      "loss": 0.3352,
+      "step": 4393
+    },
+    {
+      "epoch": 4.684434968017058,
+      "grad_norm": 0.06399841711925291,
+      "learning_rate": 9.667745696169839e-07,
+      "loss": 0.3365,
+      "step": 4394
+    },
+    {
+      "epoch": 4.685501066098081,
+      "grad_norm": 0.06442430133602112,
+      "learning_rate": 9.602795623135042e-07,
+      "loss": 0.3306,
+      "step": 4395
+    },
+    {
+      "epoch": 4.686567164179104,
+      "grad_norm": 0.06647357574669362,
+      "learning_rate": 9.5380618098619e-07,
+      "loss": 0.3311,
+      "step": 4396
+    },
+    {
+      "epoch": 4.687633262260128,
+      "grad_norm": 0.06493583228185477,
+      "learning_rate": 9.473544292209591e-07,
+      "loss": 0.327,
+      "step": 4397
+    },
+    {
+      "epoch": 4.688699360341151,
+      "grad_norm": 0.06666580308482166,
+      "learning_rate": 9.409243105917532e-07,
+      "loss": 0.334,
+      "step": 4398
+    },
+    {
+      "epoch": 4.689765458422174,
+      "grad_norm": 0.06921334075814284,
+      "learning_rate": 9.345158286605182e-07,
+      "loss": 0.3363,
+      "step": 4399
+    },
+    {
+      "epoch": 4.690831556503198,
+      "grad_norm": 0.06673171229849625,
+      "learning_rate": 9.281289869772192e-07,
+      "loss": 0.328,
+      "step": 4400
+    },
+    {
+      "epoch": 4.6918976545842215,
+      "grad_norm": 0.06584271320243143,
+      "learning_rate": 9.217637890798348e-07,
+      "loss": 0.3394,
+      "step": 4401
+    },
+    {
+      "epoch": 4.6929637526652455,
+      "grad_norm": 0.0671007285287849,
+      "learning_rate": 9.154202384943622e-07,
+      "loss": 0.3318,
+      "step": 4402
+    },
+    {
+      "epoch": 4.6940298507462686,
+      "grad_norm": 0.06331709951463382,
+      "learning_rate": 9.090983387347863e-07,
+      "loss": 0.3347,
+      "step": 4403
+    },
+    {
+      "epoch": 4.6950959488272925,
+      "grad_norm": 0.06898127420033802,
+      "learning_rate": 9.027980933031188e-07,
+      "loss": 0.3374,
+      "step": 4404
+    },
+    {
+      "epoch": 4.696162046908316,
+      "grad_norm": 0.07051509740595827,
+      "learning_rate": 8.965195056893638e-07,
+      "loss": 0.3351,
+      "step": 4405
+    },
+    {
+      "epoch": 4.697228144989339,
+      "grad_norm": 0.06642503187647329,
+      "learning_rate": 8.902625793715391e-07,
+      "loss": 0.3289,
+      "step": 4406
+    },
+    {
+      "epoch": 4.698294243070363,
+      "grad_norm": 0.062149770848458644,
+      "learning_rate": 8.8402731781565e-07,
+      "loss": 0.3321,
+      "step": 4407
+    },
+    {
+      "epoch": 4.699360341151386,
+      "grad_norm": 0.06584174207599507,
+      "learning_rate": 8.778137244757068e-07,
+      "loss": 0.3303,
+      "step": 4408
+    },
+    {
+      "epoch": 4.70042643923241,
+      "grad_norm": 0.06691401972973397,
+      "learning_rate": 8.716218027937251e-07,
+      "loss": 0.3305,
+      "step": 4409
+    },
+    {
+      "epoch": 4.701492537313433,
+      "grad_norm": 0.0690126967878217,
+      "learning_rate": 8.654515561997034e-07,
+      "loss": 0.3354,
+      "step": 4410
+    },
+    {
+      "epoch": 4.702558635394456,
+      "grad_norm": 0.06740746680029383,
+      "learning_rate": 8.593029881116322e-07,
+      "loss": 0.3242,
+      "step": 4411
+    },
+    {
+      "epoch": 4.70362473347548,
+      "grad_norm": 0.06898874376741002,
+      "learning_rate": 8.531761019355067e-07,
+      "loss": 0.3387,
+      "step": 4412
+    },
+    {
+      "epoch": 4.704690831556503,
+      "grad_norm": 0.06501918561728087,
+      "learning_rate": 8.470709010653011e-07,
+      "loss": 0.3329,
+      "step": 4413
+    },
+    {
+      "epoch": 4.705756929637527,
+      "grad_norm": 0.06601690952715261,
+      "learning_rate": 8.409873888829768e-07,
+      "loss": 0.3378,
+      "step": 4414
+    },
+    {
+      "epoch": 4.70682302771855,
+      "grad_norm": 0.06584452917029096,
+      "learning_rate": 8.349255687584867e-07,
+      "loss": 0.3418,
+      "step": 4415
+    },
+    {
+      "epoch": 4.707889125799573,
+      "grad_norm": 0.07132046643473576,
+      "learning_rate": 8.288854440497629e-07,
+      "loss": 0.3356,
+      "step": 4416
+    },
+    {
+      "epoch": 4.708955223880597,
+      "grad_norm": 0.06688057560968289,
+      "learning_rate": 8.228670181027199e-07,
+      "loss": 0.3406,
+      "step": 4417
+    },
+    {
+      "epoch": 4.71002132196162,
+      "grad_norm": 0.06467424384256552,
+      "learning_rate": 8.168702942512507e-07,
+      "loss": 0.3404,
+      "step": 4418
+    },
+    {
+      "epoch": 4.711087420042644,
+      "grad_norm": 0.06658144857820635,
+      "learning_rate": 8.108952758172317e-07,
+      "loss": 0.3373,
+      "step": 4419
+    },
+    {
+      "epoch": 4.712153518123667,
+      "grad_norm": 0.06579984809073838,
+      "learning_rate": 8.049419661105129e-07,
+      "loss": 0.3327,
+      "step": 4420
+    },
+    {
+      "epoch": 4.713219616204691,
+      "grad_norm": 0.06783834549966072,
+      "learning_rate": 7.99010368428923e-07,
+      "loss": 0.3363,
+      "step": 4421
+    },
+    {
+      "epoch": 4.714285714285714,
+      "grad_norm": 0.065981531581556,
+      "learning_rate": 7.93100486058247e-07,
+      "loss": 0.3362,
+      "step": 4422
+    },
+    {
+      "epoch": 4.7153518123667375,
+      "grad_norm": 0.066232413118734,
+      "learning_rate": 7.872123222722617e-07,
+      "loss": 0.3285,
+      "step": 4423
+    },
+    {
+      "epoch": 4.7164179104477615,
+      "grad_norm": 0.06404021502472225,
+      "learning_rate": 7.813458803327001e-07,
+      "loss": 0.333,
+      "step": 4424
+    },
+    {
+      "epoch": 4.717484008528785,
+      "grad_norm": 0.06489096407537417,
+      "learning_rate": 7.755011634892695e-07,
+      "loss": 0.3336,
+      "step": 4425
+    },
+    {
+      "epoch": 4.718550106609808,
+      "grad_norm": 0.06614640728593563,
+      "learning_rate": 7.696781749796333e-07,
+      "loss": 0.3326,
+      "step": 4426
+    },
+    {
+      "epoch": 4.719616204690832,
+      "grad_norm": 0.06662313517511247,
+      "learning_rate": 7.638769180294292e-07,
+      "loss": 0.3354,
+      "step": 4427
+    },
+    {
+      "epoch": 4.720682302771855,
+      "grad_norm": 0.06630182982279928,
+      "learning_rate": 7.580973958522553e-07,
+      "loss": 0.3274,
+      "step": 4428
+    },
+    {
+      "epoch": 4.721748400852879,
+      "grad_norm": 0.06231304613251542,
+      "learning_rate": 7.523396116496573e-07,
+      "loss": 0.3351,
+      "step": 4429
+    },
+    {
+      "epoch": 4.722814498933902,
+      "grad_norm": 0.06561797780372054,
+      "learning_rate": 7.46603568611155e-07,
+      "loss": 0.3303,
+      "step": 4430
+    },
+    {
+      "epoch": 4.723880597014926,
+      "grad_norm": 0.06606066948383217,
+      "learning_rate": 7.408892699142156e-07,
+      "loss": 0.3337,
+      "step": 4431
+    },
+    {
+      "epoch": 4.724946695095949,
+      "grad_norm": 0.06532372136865669,
+      "learning_rate": 7.35196718724267e-07,
+      "loss": 0.3361,
+      "step": 4432
+    },
+    {
+      "epoch": 4.726012793176972,
+      "grad_norm": 0.07451746009716578,
+      "learning_rate": 7.295259181946801e-07,
+      "loss": 0.3353,
+      "step": 4433
+    },
+    {
+      "epoch": 4.727078891257996,
+      "grad_norm": 0.06275057701061912,
+      "learning_rate": 7.23876871466791e-07,
+      "loss": 0.3384,
+      "step": 4434
+    },
+    {
+      "epoch": 4.728144989339019,
+      "grad_norm": 0.06554740340061337,
+      "learning_rate": 7.182495816698787e-07,
+      "loss": 0.3302,
+      "step": 4435
+    },
+    {
+      "epoch": 4.729211087420042,
+      "grad_norm": 0.0690906174501306,
+      "learning_rate": 7.126440519211608e-07,
+      "loss": 0.3378,
+      "step": 4436
+    },
+    {
+      "epoch": 4.730277185501066,
+      "grad_norm": 0.06708686099386059,
+      "learning_rate": 7.070602853258112e-07,
+      "loss": 0.3358,
+      "step": 4437
+    },
+    {
+      "epoch": 4.731343283582089,
+      "grad_norm": 0.06891228927845029,
+      "learning_rate": 7.014982849769558e-07,
+      "loss": 0.3354,
+      "step": 4438
+    },
+    {
+      "epoch": 4.732409381663113,
+      "grad_norm": 0.06537602757352802,
+      "learning_rate": 6.95958053955641e-07,
+      "loss": 0.3371,
+      "step": 4439
+    },
+    {
+      "epoch": 4.733475479744136,
+      "grad_norm": 0.06604287763417778,
+      "learning_rate": 6.904395953308784e-07,
+      "loss": 0.328,
+      "step": 4440
+    },
+    {
+      "epoch": 4.73454157782516,
+      "grad_norm": 0.06405394013935223,
+      "learning_rate": 6.849429121596007e-07,
+      "loss": 0.3346,
+      "step": 4441
+    },
+    {
+      "epoch": 4.735607675906183,
+      "grad_norm": 0.06528070970380216,
+      "learning_rate": 6.794680074866833e-07,
+      "loss": 0.3419,
+      "step": 4442
+    },
+    {
+      "epoch": 4.7366737739872065,
+      "grad_norm": 0.06512815748529073,
+      "learning_rate": 6.740148843449401e-07,
+      "loss": 0.3331,
+      "step": 4443
+    },
+    {
+      "epoch": 4.73773987206823,
+      "grad_norm": 0.06646695491242108,
+      "learning_rate": 6.685835457551281e-07,
+      "loss": 0.3346,
+      "step": 4444
+    },
+    {
+      "epoch": 4.7388059701492535,
+      "grad_norm": 0.0672561173174234,
+      "learning_rate": 6.631739947259075e-07,
+      "loss": 0.3368,
+      "step": 4445
+    },
+    {
+      "epoch": 4.7398720682302775,
+      "grad_norm": 0.06724341190669111,
+      "learning_rate": 6.577862342539032e-07,
+      "loss": 0.3291,
+      "step": 4446
+    },
+    {
+      "epoch": 4.740938166311301,
+      "grad_norm": 0.06595018864181441,
+      "learning_rate": 6.524202673236524e-07,
+      "loss": 0.3335,
+      "step": 4447
+    },
+    {
+      "epoch": 4.742004264392325,
+      "grad_norm": 0.06439992810416004,
+      "learning_rate": 6.470760969076173e-07,
+      "loss": 0.3371,
+      "step": 4448
+    },
+    {
+      "epoch": 4.743070362473348,
+      "grad_norm": 0.06515103944575167,
+      "learning_rate": 6.417537259661899e-07,
+      "loss": 0.3392,
+      "step": 4449
+    },
+    {
+      "epoch": 4.744136460554371,
+      "grad_norm": 0.06518098968521178,
+      "learning_rate": 6.364531574476962e-07,
+      "loss": 0.3263,
+      "step": 4450
+    },
+    {
+      "epoch": 4.745202558635395,
+      "grad_norm": 0.06553602384394039,
+      "learning_rate": 6.311743942883652e-07,
+      "loss": 0.3319,
+      "step": 4451
+    },
+    {
+      "epoch": 4.746268656716418,
+      "grad_norm": 0.06646799447298098,
+      "learning_rate": 6.259174394123602e-07,
+      "loss": 0.3287,
+      "step": 4452
+    },
+    {
+      "epoch": 4.747334754797441,
+      "grad_norm": 0.06614233529232587,
+      "learning_rate": 6.206822957317693e-07,
+      "loss": 0.3309,
+      "step": 4453
+    },
+    {
+      "epoch": 4.748400852878465,
+      "grad_norm": 0.06540305031790555,
+      "learning_rate": 6.154689661465752e-07,
+      "loss": 0.3389,
+      "step": 4454
+    },
+    {
+      "epoch": 4.749466950959488,
+      "grad_norm": 0.06512815324295504,
+      "learning_rate": 6.102774535447031e-07,
+      "loss": 0.3341,
+      "step": 4455
+    },
+    {
+      "epoch": 4.750533049040512,
+      "grad_norm": 0.06377242318608925,
+      "learning_rate": 6.051077608019773e-07,
+      "loss": 0.3335,
+      "step": 4456
+    },
+    {
+      "epoch": 4.751599147121535,
+      "grad_norm": 0.06507854548241519,
+      "learning_rate": 5.99959890782138e-07,
+      "loss": 0.3292,
+      "step": 4457
+    },
+    {
+      "epoch": 4.752665245202559,
+      "grad_norm": 0.06569932677501229,
+      "learning_rate": 5.948338463368419e-07,
+      "loss": 0.3407,
+      "step": 4458
+    },
+    {
+      "epoch": 4.753731343283582,
+      "grad_norm": 0.06613270042766406,
+      "learning_rate": 5.897296303056444e-07,
+      "loss": 0.3418,
+      "step": 4459
+    },
+    {
+      "epoch": 4.754797441364605,
+      "grad_norm": 0.06517187632853641,
+      "learning_rate": 5.846472455160213e-07,
+      "loss": 0.3361,
+      "step": 4460
+    },
+    {
+      "epoch": 4.755863539445629,
+      "grad_norm": 0.06689709594911152,
+      "learning_rate": 5.795866947833472e-07,
+      "loss": 0.3365,
+      "step": 4461
+    },
+    {
+      "epoch": 4.756929637526652,
+      "grad_norm": 0.06666505765328634,
+      "learning_rate": 5.745479809109045e-07,
+      "loss": 0.3306,
+      "step": 4462
+    },
+    {
+      "epoch": 4.757995735607675,
+      "grad_norm": 0.06646312328818217,
+      "learning_rate": 5.695311066898779e-07,
+      "loss": 0.3335,
+      "step": 4463
+    },
+    {
+      "epoch": 4.759061833688699,
+      "grad_norm": 0.06785871586907363,
+      "learning_rate": 5.6453607489936e-07,
+      "loss": 0.3364,
+      "step": 4464
+    },
+    {
+      "epoch": 4.7601279317697225,
+      "grad_norm": 0.06542533321108031,
+      "learning_rate": 5.595628883063331e-07,
+      "loss": 0.3355,
+      "step": 4465
+    },
+    {
+      "epoch": 4.7611940298507465,
+      "grad_norm": 0.06297758066267968,
+      "learning_rate": 5.546115496656867e-07,
+      "loss": 0.3383,
+      "step": 4466
+    },
+    {
+      "epoch": 4.76226012793177,
+      "grad_norm": 0.065278353186685,
+      "learning_rate": 5.496820617202047e-07,
+      "loss": 0.3349,
+      "step": 4467
+    },
+    {
+      "epoch": 4.7633262260127935,
+      "grad_norm": 0.06538720202179056,
+      "learning_rate": 5.447744272005695e-07,
+      "loss": 0.3331,
+      "step": 4468
+    },
+    {
+      "epoch": 4.764392324093817,
+      "grad_norm": 0.06718790884542368,
+      "learning_rate": 5.398886488253485e-07,
+      "loss": 0.3379,
+      "step": 4469
+    },
+    {
+      "epoch": 4.76545842217484,
+      "grad_norm": 0.06449471524894189,
+      "learning_rate": 5.350247293010169e-07,
+      "loss": 0.3349,
+      "step": 4470
+    },
+    {
+      "epoch": 4.766524520255864,
+      "grad_norm": 0.06433921241405373,
+      "learning_rate": 5.301826713219305e-07,
+      "loss": 0.3319,
+      "step": 4471
+    },
+    {
+      "epoch": 4.767590618336887,
+      "grad_norm": 0.06817594418030926,
+      "learning_rate": 5.253624775703347e-07,
+      "loss": 0.3361,
+      "step": 4472
+    },
+    {
+      "epoch": 4.768656716417911,
+      "grad_norm": 0.06600839287904284,
+      "learning_rate": 5.205641507163694e-07,
+      "loss": 0.3343,
+      "step": 4473
+    },
+    {
+      "epoch": 4.769722814498934,
+      "grad_norm": 0.06616379620806581,
+      "learning_rate": 5.157876934180551e-07,
+      "loss": 0.3318,
+      "step": 4474
+    },
+    {
+      "epoch": 4.770788912579958,
+      "grad_norm": 0.06606131200753886,
+      "learning_rate": 5.110331083213105e-07,
+      "loss": 0.3376,
+      "step": 4475
+    },
+    {
+      "epoch": 4.771855010660981,
+      "grad_norm": 0.06329917860131302,
+      "learning_rate": 5.063003980599179e-07,
+      "loss": 0.3366,
+      "step": 4476
+    },
+    {
+      "epoch": 4.772921108742004,
+      "grad_norm": 0.06481792763454347,
+      "learning_rate": 5.01589565255558e-07,
+      "loss": 0.3394,
+      "step": 4477
+    },
+    {
+      "epoch": 4.773987206823028,
+      "grad_norm": 0.0687947416522106,
+      "learning_rate": 4.969006125177834e-07,
+      "loss": 0.3368,
+      "step": 4478
+    },
+    {
+      "epoch": 4.775053304904051,
+      "grad_norm": 0.06404270869284907,
+      "learning_rate": 4.922335424440361e-07,
+      "loss": 0.3323,
+      "step": 4479
+    },
+    {
+      "epoch": 4.776119402985074,
+      "grad_norm": 0.06757244081482576,
+      "learning_rate": 4.875883576196261e-07,
+      "loss": 0.3367,
+      "step": 4480
+    },
+    {
+      "epoch": 4.777185501066098,
+      "grad_norm": 0.06305226854333544,
+      "learning_rate": 4.829650606177438e-07,
+      "loss": 0.3345,
+      "step": 4481
+    },
+    {
+      "epoch": 4.778251599147121,
+      "grad_norm": 0.06783144251481625,
+      "learning_rate": 4.783636539994607e-07,
+      "loss": 0.3381,
+      "step": 4482
+    },
+    {
+      "epoch": 4.779317697228145,
+      "grad_norm": 0.07112258236166988,
+      "learning_rate": 4.737841403137067e-07,
+      "loss": 0.3319,
+      "step": 4483
+    },
+    {
+      "epoch": 4.780383795309168,
+      "grad_norm": 0.06778618377309222,
+      "learning_rate": 4.692265220973058e-07,
+      "loss": 0.3368,
+      "step": 4484
+    },
+    {
+      "epoch": 4.781449893390192,
+      "grad_norm": 0.061819307755361846,
+      "learning_rate": 4.6469080187493634e-07,
+      "loss": 0.3305,
+      "step": 4485
+    },
+    {
+      "epoch": 4.782515991471215,
+      "grad_norm": 0.06472343334221725,
+      "learning_rate": 4.601769821591529e-07,
+      "loss": 0.3301,
+      "step": 4486
+    },
+    {
+      "epoch": 4.7835820895522385,
+      "grad_norm": 0.06769820854581794,
+      "learning_rate": 4.5568506545037305e-07,
+      "loss": 0.3363,
+      "step": 4487
+    },
+    {
+      "epoch": 4.7846481876332625,
+      "grad_norm": 0.06599930627838296,
+      "learning_rate": 4.512150542368909e-07,
+      "loss": 0.3344,
+      "step": 4488
+    },
+    {
+      "epoch": 4.785714285714286,
+      "grad_norm": 0.06669160582080108,
+      "learning_rate": 4.467669509948591e-07,
+      "loss": 0.335,
+      "step": 4489
+    },
+    {
+      "epoch": 4.786780383795309,
+      "grad_norm": 0.06657029869145847,
+      "learning_rate": 4.423407581882932e-07,
+      "loss": 0.3318,
+      "step": 4490
+    },
+    {
+      "epoch": 4.787846481876333,
+      "grad_norm": 0.06875440174953255,
+      "learning_rate": 4.3793647826907203e-07,
+      "loss": 0.3301,
+      "step": 4491
+    },
+    {
+      "epoch": 4.788912579957356,
+      "grad_norm": 0.06843506786311358,
+      "learning_rate": 4.3355411367694164e-07,
+      "loss": 0.3377,
+      "step": 4492
+    },
+    {
+      "epoch": 4.78997867803838,
+      "grad_norm": 0.06459340857709406,
+      "learning_rate": 4.2919366683951135e-07,
+      "loss": 0.3336,
+      "step": 4493
+    },
+    {
+      "epoch": 4.791044776119403,
+      "grad_norm": 0.06386692441739224,
+      "learning_rate": 4.2485514017222674e-07,
+      "loss": 0.333,
+      "step": 4494
+    },
+    {
+      "epoch": 4.792110874200427,
+      "grad_norm": 0.06649809538469802,
+      "learning_rate": 4.205385360784142e-07,
+      "loss": 0.337,
+      "step": 4495
+    },
+    {
+      "epoch": 4.79317697228145,
+      "grad_norm": 0.06494463953807365,
+      "learning_rate": 4.162438569492455e-07,
+      "loss": 0.3381,
+      "step": 4496
+    },
+    {
+      "epoch": 4.794243070362473,
+      "grad_norm": 0.06443760229121664,
+      "learning_rate": 4.119711051637554e-07,
+      "loss": 0.331,
+      "step": 4497
+    },
+    {
+      "epoch": 4.795309168443497,
+      "grad_norm": 0.06694814400064329,
+      "learning_rate": 4.077202830888238e-07,
+      "loss": 0.3366,
+      "step": 4498
+    },
+    {
+      "epoch": 4.79637526652452,
+      "grad_norm": 0.061826241061163335,
+      "learning_rate": 4.0349139307918063e-07,
+      "loss": 0.3229,
+      "step": 4499
+    },
+    {
+      "epoch": 4.797441364605544,
+      "grad_norm": 0.06413189534977501,
+      "learning_rate": 3.992844374774141e-07,
+      "loss": 0.3357,
+      "step": 4500
+    },
+    {
+      "epoch": 4.798507462686567,
+      "grad_norm": 0.06301453498466782,
+      "learning_rate": 3.950994186139623e-07,
+      "loss": 0.3262,
+      "step": 4501
+    },
+    {
+      "epoch": 4.79957356076759,
+      "grad_norm": 0.0669025986464228,
+      "learning_rate": 3.909363388071041e-07,
+      "loss": 0.3333,
+      "step": 4502
+    },
+    {
+      "epoch": 4.800639658848614,
+      "grad_norm": 0.06727116449771486,
+      "learning_rate": 3.8679520036296823e-07,
+      "loss": 0.3285,
+      "step": 4503
+    },
+    {
+      "epoch": 4.801705756929637,
+      "grad_norm": 0.06423952902920181,
+      "learning_rate": 3.826760055755374e-07,
+      "loss": 0.3298,
+      "step": 4504
+    },
+    {
+      "epoch": 4.802771855010661,
+      "grad_norm": 0.06287554539652727,
+      "learning_rate": 3.785787567266219e-07,
+      "loss": 0.3314,
+      "step": 4505
+    },
+    {
+      "epoch": 4.803837953091684,
+      "grad_norm": 0.06428325628977336,
+      "learning_rate": 3.745034560858907e-07,
+      "loss": 0.3374,
+      "step": 4506
+    },
+    {
+      "epoch": 4.8049040511727075,
+      "grad_norm": 0.0615240827531194,
+      "learning_rate": 3.7045010591084893e-07,
+      "loss": 0.3346,
+      "step": 4507
+    },
+    {
+      "epoch": 4.8059701492537314,
+      "grad_norm": 0.06350727088776514,
+      "learning_rate": 3.66418708446834e-07,
+      "loss": 0.331,
+      "step": 4508
+    },
+    {
+      "epoch": 4.8070362473347545,
+      "grad_norm": 0.06394487503770653,
+      "learning_rate": 3.6240926592704173e-07,
+      "loss": 0.3281,
+      "step": 4509
+    },
+    {
+      "epoch": 4.8081023454157785,
+      "grad_norm": 0.06562664075180806,
+      "learning_rate": 3.58421780572491e-07,
+      "loss": 0.3364,
+      "step": 4510
+    },
+    {
+      "epoch": 4.809168443496802,
+      "grad_norm": 0.0641531508096308,
+      "learning_rate": 3.5445625459203715e-07,
+      "loss": 0.3311,
+      "step": 4511
+    },
+    {
+      "epoch": 4.810234541577826,
+      "grad_norm": 0.06317633180319338,
+      "learning_rate": 3.5051269018238075e-07,
+      "loss": 0.3327,
+      "step": 4512
+    },
+    {
+      "epoch": 4.811300639658849,
+      "grad_norm": 0.06640788788183018,
+      "learning_rate": 3.4659108952805e-07,
+      "loss": 0.3339,
+      "step": 4513
+    },
+    {
+      "epoch": 4.812366737739872,
+      "grad_norm": 0.06476401703214942,
+      "learning_rate": 3.4269145480140486e-07,
+      "loss": 0.3283,
+      "step": 4514
+    },
+    {
+      "epoch": 4.813432835820896,
+      "grad_norm": 0.06632322608724953,
+      "learning_rate": 3.388137881626419e-07,
+      "loss": 0.3333,
+      "step": 4515
+    },
+    {
+      "epoch": 4.814498933901919,
+      "grad_norm": 0.06584746139018403,
+      "learning_rate": 3.3495809175978944e-07,
+      "loss": 0.3341,
+      "step": 4516
+    },
+    {
+      "epoch": 4.815565031982942,
+      "grad_norm": 0.06225839649171285,
+      "learning_rate": 3.311243677287035e-07,
+      "loss": 0.339,
+      "step": 4517
+    },
+    {
+      "epoch": 4.816631130063966,
+      "grad_norm": 0.06266507769908893,
+      "learning_rate": 3.2731261819306306e-07,
+      "loss": 0.3338,
+      "step": 4518
+    },
+    {
+      "epoch": 4.817697228144989,
+      "grad_norm": 0.06361144124581149,
+      "learning_rate": 3.2352284526438347e-07,
+      "loss": 0.3262,
+      "step": 4519
+    },
+    {
+      "epoch": 4.818763326226013,
+      "grad_norm": 0.06657050940016936,
+      "learning_rate": 3.1975505104199446e-07,
+      "loss": 0.3347,
+      "step": 4520
+    },
+    {
+      "epoch": 4.819829424307036,
+      "grad_norm": 0.06336689637535396,
+      "learning_rate": 3.1600923761307077e-07,
+      "loss": 0.3314,
+      "step": 4521
+    },
+    {
+      "epoch": 4.82089552238806,
+      "grad_norm": 0.06697036655954806,
+      "learning_rate": 3.1228540705258827e-07,
+      "loss": 0.3373,
+      "step": 4522
+    },
+    {
+      "epoch": 4.821961620469083,
+      "grad_norm": 0.06669058010931536,
+      "learning_rate": 3.08583561423359e-07,
+      "loss": 0.3333,
+      "step": 4523
+    },
+    {
+      "epoch": 4.823027718550106,
+      "grad_norm": 0.06204857673664051,
+      "learning_rate": 3.0490370277601375e-07,
+      "loss": 0.3305,
+      "step": 4524
+    },
+    {
+      "epoch": 4.82409381663113,
+      "grad_norm": 0.06414424868604003,
+      "learning_rate": 3.0124583314900204e-07,
+      "loss": 0.3285,
+      "step": 4525
+    },
+    {
+      "epoch": 4.825159914712153,
+      "grad_norm": 0.06489342987918388,
+      "learning_rate": 2.976099545685962e-07,
+      "loss": 0.3389,
+      "step": 4526
+    },
+    {
+      "epoch": 4.826226012793177,
+      "grad_norm": 0.06392477993982196,
+      "learning_rate": 2.9399606904887856e-07,
+      "loss": 0.3301,
+      "step": 4527
+    },
+    {
+      "epoch": 4.8272921108742,
+      "grad_norm": 0.06216530530535471,
+      "learning_rate": 2.9040417859175884e-07,
+      "loss": 0.3278,
+      "step": 4528
+    },
+    {
+      "epoch": 4.8283582089552235,
+      "grad_norm": 0.06907666472180728,
+      "learning_rate": 2.8683428518695654e-07,
+      "loss": 0.3462,
+      "step": 4529
+    },
+    {
+      "epoch": 4.8294243070362475,
+      "grad_norm": 0.06389830168648562,
+      "learning_rate": 2.832863908120009e-07,
+      "loss": 0.3316,
+      "step": 4530
+    },
+    {
+      "epoch": 4.830490405117271,
+      "grad_norm": 0.0659969650902936,
+      "learning_rate": 2.7976049743224876e-07,
+      "loss": 0.334,
+      "step": 4531
+    },
+    {
+      "epoch": 4.8315565031982945,
+      "grad_norm": 0.06337199459457302,
+      "learning_rate": 2.762566070008621e-07,
+      "loss": 0.3379,
+      "step": 4532
+    },
+    {
+      "epoch": 4.832622601279318,
+      "grad_norm": 0.06123098557327691,
+      "learning_rate": 2.7277472145880837e-07,
+      "loss": 0.3337,
+      "step": 4533
+    },
+    {
+      "epoch": 4.833688699360341,
+      "grad_norm": 0.06421391981205138,
+      "learning_rate": 2.6931484273487796e-07,
+      "loss": 0.3351,
+      "step": 4534
+    },
+    {
+      "epoch": 4.834754797441365,
+      "grad_norm": 0.06631306221864534,
+      "learning_rate": 2.658769727456534e-07,
+      "loss": 0.3395,
+      "step": 4535
+    },
+    {
+      "epoch": 4.835820895522388,
+      "grad_norm": 0.06390409569316938,
+      "learning_rate": 2.6246111339554903e-07,
+      "loss": 0.335,
+      "step": 4536
+    },
+    {
+      "epoch": 4.836886993603412,
+      "grad_norm": 0.06255999806490187,
+      "learning_rate": 2.5906726657676685e-07,
+      "loss": 0.3371,
+      "step": 4537
+    },
+    {
+      "epoch": 4.837953091684435,
+      "grad_norm": 0.0638021517251054,
+      "learning_rate": 2.556954341693185e-07,
+      "loss": 0.3324,
+      "step": 4538
+    },
+    {
+      "epoch": 4.839019189765459,
+      "grad_norm": 0.06516423398474619,
+      "learning_rate": 2.5234561804102556e-07,
+      "loss": 0.3363,
+      "step": 4539
+    },
+    {
+      "epoch": 4.840085287846482,
+      "grad_norm": 0.06196390275858242,
+      "learning_rate": 2.4901782004751905e-07,
+      "loss": 0.3265,
+      "step": 4540
+    },
+    {
+      "epoch": 4.841151385927505,
+      "grad_norm": 0.06281847969805605,
+      "learning_rate": 2.457120420322179e-07,
+      "loss": 0.3373,
+      "step": 4541
+    },
+    {
+      "epoch": 4.842217484008529,
+      "grad_norm": 0.06560695196791491,
+      "learning_rate": 2.424282858263549e-07,
+      "loss": 0.3375,
+      "step": 4542
+    },
+    {
+      "epoch": 4.843283582089552,
+      "grad_norm": 0.06533330528579777,
+      "learning_rate": 2.3916655324895953e-07,
+      "loss": 0.3293,
+      "step": 4543
+    },
+    {
+      "epoch": 4.844349680170575,
+      "grad_norm": 0.06480021279499126,
+      "learning_rate": 2.35926846106862e-07,
+      "loss": 0.3355,
+      "step": 4544
+    },
+    {
+      "epoch": 4.845415778251599,
+      "grad_norm": 0.06287240019527855,
+      "learning_rate": 2.3270916619469342e-07,
+      "loss": 0.3364,
+      "step": 4545
+    },
+    {
+      "epoch": 4.846481876332622,
+      "grad_norm": 0.06700125191000252,
+      "learning_rate": 2.2951351529487685e-07,
+      "loss": 0.338,
+      "step": 4546
+    },
+    {
+      "epoch": 4.847547974413646,
+      "grad_norm": 0.06356697255958557,
+      "learning_rate": 2.2633989517764522e-07,
+      "loss": 0.335,
+      "step": 4547
+    },
+    {
+      "epoch": 4.848614072494669,
+      "grad_norm": 0.06622567910030012,
+      "learning_rate": 2.2318830760100995e-07,
+      "loss": 0.3398,
+      "step": 4548
+    },
+    {
+      "epoch": 4.849680170575693,
+      "grad_norm": 0.06340820773787822,
+      "learning_rate": 2.2005875431078794e-07,
+      "loss": 0.3361,
+      "step": 4549
+    },
+    {
+      "epoch": 4.850746268656716,
+      "grad_norm": 0.06467319645032904,
+      "learning_rate": 2.169512370405924e-07,
+      "loss": 0.3362,
+      "step": 4550
+    },
+    {
+      "epoch": 4.8518123667377395,
+      "grad_norm": 0.06494652144642511,
+      "learning_rate": 2.138657575118286e-07,
+      "loss": 0.334,
+      "step": 4551
+    },
+    {
+      "epoch": 4.8528784648187635,
+      "grad_norm": 0.06344916494239451,
+      "learning_rate": 2.1080231743368928e-07,
+      "loss": 0.3261,
+      "step": 4552
+    },
+    {
+      "epoch": 4.853944562899787,
+      "grad_norm": 0.06552581044666182,
+      "learning_rate": 2.0776091850315483e-07,
+      "loss": 0.3338,
+      "step": 4553
+    },
+    {
+      "epoch": 4.855010660980811,
+      "grad_norm": 0.06182298175401099,
+      "learning_rate": 2.0474156240501086e-07,
+      "loss": 0.3335,
+      "step": 4554
+    },
+    {
+      "epoch": 4.856076759061834,
+      "grad_norm": 0.06542556498649621,
+      "learning_rate": 2.0174425081182615e-07,
+      "loss": 0.3364,
+      "step": 4555
+    },
+    {
+      "epoch": 4.857142857142857,
+      "grad_norm": 0.06371796771439006,
+      "learning_rate": 1.9876898538394362e-07,
+      "loss": 0.3353,
+      "step": 4556
+    },
+    {
+      "epoch": 4.858208955223881,
+      "grad_norm": 0.06396421709422397,
+      "learning_rate": 1.9581576776951605e-07,
+      "loss": 0.3295,
+      "step": 4557
+    },
+    {
+      "epoch": 4.859275053304904,
+      "grad_norm": 0.06288608503660843,
+      "learning_rate": 1.9288459960446592e-07,
+      "loss": 0.3292,
+      "step": 4558
+    },
+    {
+      "epoch": 4.860341151385928,
+      "grad_norm": 0.06363937784876181,
+      "learning_rate": 1.899754825125122e-07,
+      "loss": 0.3366,
+      "step": 4559
+    },
+    {
+      "epoch": 4.861407249466951,
+      "grad_norm": 0.06336696226394795,
+      "learning_rate": 1.8708841810515244e-07,
+      "loss": 0.3274,
+      "step": 4560
+    },
+    {
+      "epoch": 4.862473347547974,
+      "grad_norm": 0.0635413388603882,
+      "learning_rate": 1.8422340798167181e-07,
+      "loss": 0.3359,
+      "step": 4561
+    },
+    {
+      "epoch": 4.863539445628998,
+      "grad_norm": 0.06482344768083516,
+      "learning_rate": 1.8138045372913414e-07,
+      "loss": 0.3387,
+      "step": 4562
+    },
+    {
+      "epoch": 4.864605543710021,
+      "grad_norm": 0.06543147651304172,
+      "learning_rate": 1.7855955692239525e-07,
+      "loss": 0.3369,
+      "step": 4563
+    },
+    {
+      "epoch": 4.865671641791045,
+      "grad_norm": 0.0634061615569733,
+      "learning_rate": 1.757607191240762e-07,
+      "loss": 0.3337,
+      "step": 4564
+    },
+    {
+      "epoch": 4.866737739872068,
+      "grad_norm": 0.06497196758674298,
+      "learning_rate": 1.7298394188459466e-07,
+      "loss": 0.3337,
+      "step": 4565
+    },
+    {
+      "epoch": 4.867803837953092,
+      "grad_norm": 0.0643026256592065,
+      "learning_rate": 1.7022922674213793e-07,
+      "loss": 0.3358,
+      "step": 4566
+    },
+    {
+      "epoch": 4.868869936034115,
+      "grad_norm": 0.06380088154716912,
+      "learning_rate": 1.6749657522267205e-07,
+      "loss": 0.3323,
+      "step": 4567
+    },
+    {
+      "epoch": 4.869936034115138,
+      "grad_norm": 0.062292059451068596,
+      "learning_rate": 1.6478598883995057e-07,
+      "loss": 0.3297,
+      "step": 4568
+    },
+    {
+      "epoch": 4.871002132196162,
+      "grad_norm": 0.06429186614170503,
+      "learning_rate": 1.620974690954924e-07,
+      "loss": 0.3318,
+      "step": 4569
+    },
+    {
+      "epoch": 4.872068230277185,
+      "grad_norm": 0.061982887675499386,
+      "learning_rate": 1.59431017478604e-07,
+      "loss": 0.3376,
+      "step": 4570
+    },
+    {
+      "epoch": 4.8731343283582085,
+      "grad_norm": 0.0636117943991852,
+      "learning_rate": 1.5678663546634832e-07,
+      "loss": 0.3342,
+      "step": 4571
+    },
+    {
+      "epoch": 4.8742004264392325,
+      "grad_norm": 0.06390306822281343,
+      "learning_rate": 1.541643245235891e-07,
+      "loss": 0.3332,
+      "step": 4572
+    },
+    {
+      "epoch": 4.8752665245202556,
+      "grad_norm": 0.06414487391507678,
+      "learning_rate": 1.515640861029466e-07,
+      "loss": 0.3319,
+      "step": 4573
+    },
+    {
+      "epoch": 4.8763326226012795,
+      "grad_norm": 0.06234839482693829,
+      "learning_rate": 1.4898592164481528e-07,
+      "loss": 0.3345,
+      "step": 4574
+    },
+    {
+      "epoch": 4.877398720682303,
+      "grad_norm": 0.06282579096563232,
+      "learning_rate": 1.4642983257736388e-07,
+      "loss": 0.3318,
+      "step": 4575
+    },
+    {
+      "epoch": 4.878464818763327,
+      "grad_norm": 0.06413128319015131,
+      "learning_rate": 1.4389582031653525e-07,
+      "loss": 0.3288,
+      "step": 4576
+    },
+    {
+      "epoch": 4.87953091684435,
+      "grad_norm": 0.061604350840344914,
+      "learning_rate": 1.4138388626603772e-07,
+      "loss": 0.3293,
+      "step": 4577
+    },
+    {
+      "epoch": 4.880597014925373,
+      "grad_norm": 0.06492239286709452,
+      "learning_rate": 1.388940318173537e-07,
+      "loss": 0.3359,
+      "step": 4578
+    },
+    {
+      "epoch": 4.881663113006397,
+      "grad_norm": 0.06750216444127496,
+      "learning_rate": 1.3642625834973555e-07,
+      "loss": 0.3319,
+      "step": 4579
+    },
+    {
+      "epoch": 4.88272921108742,
+      "grad_norm": 0.06551862088137983,
+      "learning_rate": 1.3398056723019638e-07,
+      "loss": 0.3371,
+      "step": 4580
+    },
+    {
+      "epoch": 4.883795309168444,
+      "grad_norm": 0.06303973420478017,
+      "learning_rate": 1.3155695981352356e-07,
+      "loss": 0.3339,
+      "step": 4581
+    },
+    {
+      "epoch": 4.884861407249467,
+      "grad_norm": 0.06431600254532914,
+      "learning_rate": 1.2915543744227433e-07,
+      "loss": 0.3387,
+      "step": 4582
+    },
+    {
+      "epoch": 4.88592750533049,
+      "grad_norm": 0.06224741808886467,
+      "learning_rate": 1.267760014467667e-07,
+      "loss": 0.3303,
+      "step": 4583
+    },
+    {
+      "epoch": 4.886993603411514,
+      "grad_norm": 0.0686255077912375,
+      "learning_rate": 1.2441865314507529e-07,
+      "loss": 0.3315,
+      "step": 4584
+    },
+    {
+      "epoch": 4.888059701492537,
+      "grad_norm": 0.06375747762463076,
+      "learning_rate": 1.220833938430621e-07,
+      "loss": 0.3329,
+      "step": 4585
+    },
+    {
+      "epoch": 4.889125799573561,
+      "grad_norm": 0.06360467146598325,
+      "learning_rate": 1.1977022483432355e-07,
+      "loss": 0.3304,
+      "step": 4586
+    },
+    {
+      "epoch": 4.890191897654584,
+      "grad_norm": 0.06371846108482049,
+      "learning_rate": 1.1747914740025235e-07,
+      "loss": 0.3316,
+      "step": 4587
+    },
+    {
+      "epoch": 4.891257995735607,
+      "grad_norm": 0.06123471557161623,
+      "learning_rate": 1.152101628099711e-07,
+      "loss": 0.333,
+      "step": 4588
+    },
+    {
+      "epoch": 4.892324093816631,
+      "grad_norm": 0.0653623572630271,
+      "learning_rate": 1.1296327232038995e-07,
+      "loss": 0.3383,
+      "step": 4589
+    },
+    {
+      "epoch": 4.893390191897654,
+      "grad_norm": 0.06317084300404086,
+      "learning_rate": 1.1073847717616659e-07,
+      "loss": 0.3355,
+      "step": 4590
+    },
+    {
+      "epoch": 4.894456289978678,
+      "grad_norm": 0.06436959516579645,
+      "learning_rate": 1.0853577860971965e-07,
+      "loss": 0.3375,
+      "step": 4591
+    },
+    {
+      "epoch": 4.895522388059701,
+      "grad_norm": 0.06436654116700972,
+      "learning_rate": 1.0635517784122862e-07,
+      "loss": 0.3361,
+      "step": 4592
+    },
+    {
+      "epoch": 4.896588486140725,
+      "grad_norm": 0.0638870216071855,
+      "learning_rate": 1.0419667607863393e-07,
+      "loss": 0.3359,
+      "step": 4593
+    },
+    {
+      "epoch": 4.8976545842217485,
+      "grad_norm": 0.06402271945130204,
+      "learning_rate": 1.0206027451764133e-07,
+      "loss": 0.3365,
+      "step": 4594
+    },
+    {
+      "epoch": 4.898720682302772,
+      "grad_norm": 0.06360664197572094,
+      "learning_rate": 9.994597434169529e-08,
+      "loss": 0.3333,
+      "step": 4595
+    },
+    {
+      "epoch": 4.899786780383796,
+      "grad_norm": 0.06298163310834395,
+      "learning_rate": 9.785377672201002e-08,
+      "loss": 0.3304,
+      "step": 4596
+    },
+    {
+      "epoch": 4.900852878464819,
+      "grad_norm": 0.06222784082540263,
+      "learning_rate": 9.578368281756068e-08,
+      "loss": 0.3307,
+      "step": 4597
+    },
+    {
+      "epoch": 4.901918976545842,
+      "grad_norm": 0.06603440394113538,
+      "learning_rate": 9.373569377506553e-08,
+      "loss": 0.3351,
+      "step": 4598
+    },
+    {
+      "epoch": 4.902985074626866,
+      "grad_norm": 0.061308075618675476,
+      "learning_rate": 9.17098107290082e-08,
+      "loss": 0.3373,
+      "step": 4599
+    },
+    {
+      "epoch": 4.904051172707889,
+      "grad_norm": 0.06278471458473436,
+      "learning_rate": 8.970603480161988e-08,
+      "loss": 0.3301,
+      "step": 4600
+    },
+    {
+      "epoch": 4.905117270788913,
+      "grad_norm": 0.0626450228202153,
+      "learning_rate": 8.772436710288822e-08,
+      "loss": 0.3327,
+      "step": 4601
+    },
+    {
+      "epoch": 4.906183368869936,
+      "grad_norm": 0.06235194061204951,
+      "learning_rate": 8.576480873055737e-08,
+      "loss": 0.3337,
+      "step": 4602
+    },
+    {
+      "epoch": 4.90724946695096,
+      "grad_norm": 0.06548793549407254,
+      "learning_rate": 8.382736077011899e-08,
+      "loss": 0.3319,
+      "step": 4603
+    },
+    {
+      "epoch": 4.908315565031983,
+      "grad_norm": 0.06161345936082664,
+      "learning_rate": 8.191202429481681e-08,
+      "loss": 0.3343,
+      "step": 4604
+    },
+    {
+      "epoch": 4.909381663113006,
+      "grad_norm": 0.06227151592088014,
+      "learning_rate": 8.001880036565102e-08,
+      "loss": 0.3299,
+      "step": 4605
+    },
+    {
+      "epoch": 4.91044776119403,
+      "grad_norm": 0.0628595755152068,
+      "learning_rate": 7.814769003136491e-08,
+      "loss": 0.335,
+      "step": 4606
+    },
+    {
+      "epoch": 4.911513859275053,
+      "grad_norm": 0.06133289029888643,
+      "learning_rate": 7.629869432845827e-08,
+      "loss": 0.3327,
+      "step": 4607
+    },
+    {
+      "epoch": 4.912579957356077,
+      "grad_norm": 0.06399713009061905,
+      "learning_rate": 7.447181428117844e-08,
+      "loss": 0.3368,
+      "step": 4608
+    },
+    {
+      "epoch": 4.9136460554371,
+      "grad_norm": 0.06471533837047432,
+      "learning_rate": 7.266705090152482e-08,
+      "loss": 0.3371,
+      "step": 4609
+    },
+    {
+      "epoch": 4.914712153518123,
+      "grad_norm": 0.0631935092408263,
+      "learning_rate": 7.08844051892399e-08,
+      "loss": 0.3346,
+      "step": 4610
+    },
+    {
+      "epoch": 4.915778251599147,
+      "grad_norm": 0.06449225156722846,
+      "learning_rate": 6.912387813181376e-08,
+      "loss": 0.3322,
+      "step": 4611
+    },
+    {
+      "epoch": 4.91684434968017,
+      "grad_norm": 0.06188700970750854,
+      "learning_rate": 6.738547070449297e-08,
+      "loss": 0.3325,
+      "step": 4612
+    },
+    {
+      "epoch": 4.917910447761194,
+      "grad_norm": 0.06443747535001423,
+      "learning_rate": 6.566918387026278e-08,
+      "loss": 0.335,
+      "step": 4613
+    },
+    {
+      "epoch": 4.918976545842217,
+      "grad_norm": 0.06527583600959296,
+      "learning_rate": 6.397501857985599e-08,
+      "loss": 0.3299,
+      "step": 4614
+    },
+    {
+      "epoch": 4.9200426439232405,
+      "grad_norm": 0.06254955888214284,
+      "learning_rate": 6.230297577175304e-08,
+      "loss": 0.3318,
+      "step": 4615
+    },
+    {
+      "epoch": 4.9211087420042645,
+      "grad_norm": 0.06654203572641401,
+      "learning_rate": 6.065305637218188e-08,
+      "loss": 0.3377,
+      "step": 4616
+    },
+    {
+      "epoch": 4.922174840085288,
+      "grad_norm": 0.06615365835912221,
+      "learning_rate": 5.902526129510477e-08,
+      "loss": 0.3324,
+      "step": 4617
+    },
+    {
+      "epoch": 4.923240938166312,
+      "grad_norm": 0.06358574578582833,
+      "learning_rate": 5.741959144223597e-08,
+      "loss": 0.3405,
+      "step": 4618
+    },
+    {
+      "epoch": 4.924307036247335,
+      "grad_norm": 0.06592424042080675,
+      "learning_rate": 5.583604770304174e-08,
+      "loss": 0.3329,
+      "step": 4619
+    },
+    {
+      "epoch": 4.925373134328359,
+      "grad_norm": 0.06489809675701481,
+      "learning_rate": 5.427463095471375e-08,
+      "loss": 0.3331,
+      "step": 4620
+    },
+    {
+      "epoch": 4.926439232409382,
+      "grad_norm": 0.06346784271705076,
+      "learning_rate": 5.27353420622001e-08,
+      "loss": 0.3378,
+      "step": 4621
+    },
+    {
+      "epoch": 4.927505330490405,
+      "grad_norm": 0.06492417082655748,
+      "learning_rate": 5.121818187818761e-08,
+      "loss": 0.3352,
+      "step": 4622
+    },
+    {
+      "epoch": 4.928571428571429,
+      "grad_norm": 0.06263743897374244,
+      "learning_rate": 4.9723151243106225e-08,
+      "loss": 0.3375,
+      "step": 4623
+    },
+    {
+      "epoch": 4.929637526652452,
+      "grad_norm": 0.06117815118487387,
+      "learning_rate": 4.825025098512015e-08,
+      "loss": 0.3407,
+      "step": 4624
+    },
+    {
+      "epoch": 4.930703624733475,
+      "grad_norm": 0.06245193201459552,
+      "learning_rate": 4.679948192013673e-08,
+      "loss": 0.3279,
+      "step": 4625
+    },
+    {
+      "epoch": 4.931769722814499,
+      "grad_norm": 0.06712207626671726,
+      "learning_rate": 4.537084485181531e-08,
+      "loss": 0.3347,
+      "step": 4626
+    },
+    {
+      "epoch": 4.932835820895522,
+      "grad_norm": 0.0637857168923382,
+      "learning_rate": 4.396434057154508e-08,
+      "loss": 0.3303,
+      "step": 4627
+    },
+    {
+      "epoch": 4.933901918976546,
+      "grad_norm": 0.06368341666107143,
+      "learning_rate": 4.257996985844948e-08,
+      "loss": 0.3293,
+      "step": 4628
+    },
+    {
+      "epoch": 4.934968017057569,
+      "grad_norm": 0.06745977801668324,
+      "learning_rate": 4.121773347940394e-08,
+      "loss": 0.3329,
+      "step": 4629
+    },
+    {
+      "epoch": 4.936034115138593,
+      "grad_norm": 0.06316814202138803,
+      "learning_rate": 3.987763218901375e-08,
+      "loss": 0.3323,
+      "step": 4630
+    },
+    {
+      "epoch": 4.937100213219616,
+      "grad_norm": 0.06238635312813817,
+      "learning_rate": 3.85596667296273e-08,
+      "loss": 0.3319,
+      "step": 4631
+    },
+    {
+      "epoch": 4.938166311300639,
+      "grad_norm": 0.06453862518488947,
+      "learning_rate": 3.7263837831327255e-08,
+      "loss": 0.3316,
+      "step": 4632
+    },
+    {
+      "epoch": 4.939232409381663,
+      "grad_norm": 0.06561075891909876,
+      "learning_rate": 3.5990146211939416e-08,
+      "loss": 0.3357,
+      "step": 4633
+    },
+    {
+      "epoch": 4.940298507462686,
+      "grad_norm": 0.06243805915307683,
+      "learning_rate": 3.473859257701495e-08,
+      "loss": 0.3357,
+      "step": 4634
+    },
+    {
+      "epoch": 4.94136460554371,
+      "grad_norm": 0.06580397289868427,
+      "learning_rate": 3.3509177619857056e-08,
+      "loss": 0.3354,
+      "step": 4635
+    },
+    {
+      "epoch": 4.9424307036247335,
+      "grad_norm": 0.06077030430444124,
+      "learning_rate": 3.2301902021494304e-08,
+      "loss": 0.3247,
+      "step": 4636
+    },
+    {
+      "epoch": 4.943496801705757,
+      "grad_norm": 0.06466626231200338,
+      "learning_rate": 3.111676645069839e-08,
+      "loss": 0.3321,
+      "step": 4637
+    },
+    {
+      "epoch": 4.9445628997867805,
+      "grad_norm": 0.06068940940138413,
+      "learning_rate": 2.9953771563966396e-08,
+      "loss": 0.3288,
+      "step": 4638
+    },
+    {
+      "epoch": 4.945628997867804,
+      "grad_norm": 0.06278652639592003,
+      "learning_rate": 2.881291800554298e-08,
+      "loss": 0.3372,
+      "step": 4639
+    },
+    {
+      "epoch": 4.946695095948828,
+      "grad_norm": 0.061309406774551704,
+      "learning_rate": 2.769420640739817e-08,
+      "loss": 0.3353,
+      "step": 4640
+    },
+    {
+      "epoch": 4.947761194029851,
+      "grad_norm": 0.06281507165537796,
+      "learning_rate": 2.6597637389240704e-08,
+      "loss": 0.3321,
+      "step": 4641
+    },
+    {
+      "epoch": 4.948827292110874,
+      "grad_norm": 0.061550308085559016,
+      "learning_rate": 2.552321155851356e-08,
+      "loss": 0.3325,
+      "step": 4642
+    },
+    {
+      "epoch": 4.949893390191898,
+      "grad_norm": 0.06375020106010763,
+      "learning_rate": 2.447092951038954e-08,
+      "loss": 0.3315,
+      "step": 4643
+    },
+    {
+      "epoch": 4.950959488272921,
+      "grad_norm": 0.06552775876706644,
+      "learning_rate": 2.344079182778458e-08,
+      "loss": 0.3445,
+      "step": 4644
+    },
+    {
+      "epoch": 4.952025586353945,
+      "grad_norm": 0.06809372509479077,
+      "learning_rate": 2.2432799081339997e-08,
+      "loss": 0.3325,
+      "step": 4645
+    },
+    {
+      "epoch": 4.953091684434968,
+      "grad_norm": 0.06352213483924132,
+      "learning_rate": 2.1446951829422468e-08,
+      "loss": 0.3363,
+      "step": 4646
+    },
+    {
+      "epoch": 4.954157782515992,
+      "grad_norm": 0.0628161589783622,
+      "learning_rate": 2.0483250618150708e-08,
+      "loss": 0.334,
+      "step": 4647
+    },
+    {
+      "epoch": 4.955223880597015,
+      "grad_norm": 0.06170202655763391,
+      "learning_rate": 1.954169598136435e-08,
+      "loss": 0.3379,
+      "step": 4648
+    },
+    {
+      "epoch": 4.956289978678038,
+      "grad_norm": 0.06284704035908611,
+      "learning_rate": 1.862228844062841e-08,
+      "loss": 0.334,
+      "step": 4649
+    },
+    {
+      "epoch": 4.957356076759062,
+      "grad_norm": 0.06343922078643444,
+      "learning_rate": 1.772502850525548e-08,
+      "loss": 0.3358,
+      "step": 4650
+    },
+    {
+      "epoch": 4.958422174840085,
+      "grad_norm": 0.06367103465728541,
+      "learning_rate": 1.6849916672270206e-08,
+      "loss": 0.3362,
+      "step": 4651
+    },
+    {
+      "epoch": 4.959488272921108,
+      "grad_norm": 0.06339160396502194,
+      "learning_rate": 1.5996953426449245e-08,
+      "loss": 0.3362,
+      "step": 4652
+    },
+    {
+      "epoch": 4.960554371002132,
+      "grad_norm": 0.06146054352349314,
+      "learning_rate": 1.51661392402902e-08,
+      "loss": 0.3327,
+      "step": 4653
+    },
+    {
+      "epoch": 4.961620469083155,
+      "grad_norm": 0.0644259680452347,
+      "learning_rate": 1.4357474574011598e-08,
+      "loss": 0.3275,
+      "step": 4654
+    },
+    {
+      "epoch": 4.962686567164179,
+      "grad_norm": 0.06489614895198179,
+      "learning_rate": 1.3570959875579547e-08,
+      "loss": 0.3354,
+      "step": 4655
+    },
+    {
+      "epoch": 4.963752665245202,
+      "grad_norm": 0.061701893851838184,
+      "learning_rate": 1.2806595580676651e-08,
+      "loss": 0.3296,
+      "step": 4656
+    },
+    {
+      "epoch": 4.964818763326226,
+      "grad_norm": 0.06305988708818967,
+      "learning_rate": 1.2064382112728646e-08,
+      "loss": 0.3271,
+      "step": 4657
+    },
+    {
+      "epoch": 4.9658848614072495,
+      "grad_norm": 0.06467949485812308,
+      "learning_rate": 1.1344319882873323e-08,
+      "loss": 0.3373,
+      "step": 4658
+    },
+    {
+      "epoch": 4.966950959488273,
+      "grad_norm": 0.0616143550364817,
+      "learning_rate": 1.064640928999605e-08,
+      "loss": 0.3352,
+      "step": 4659
+    },
+    {
+      "epoch": 4.968017057569297,
+      "grad_norm": 0.06426893511659539,
+      "learning_rate": 9.970650720703135e-09,
+      "loss": 0.3364,
+      "step": 4660
+    },
+    {
+      "epoch": 4.96908315565032,
+      "grad_norm": 0.06068623678209842,
+      "learning_rate": 9.317044549321808e-09,
+      "loss": 0.3317,
+      "step": 4661
+    },
+    {
+      "epoch": 4.970149253731344,
+      "grad_norm": 0.06197114220318593,
+      "learning_rate": 8.685591137922444e-09,
+      "loss": 0.3286,
+      "step": 4662
+    },
+    {
+      "epoch": 4.971215351812367,
+      "grad_norm": 0.06386456846237887,
+      "learning_rate": 8.076290836296352e-09,
+      "loss": 0.3352,
+      "step": 4663
+    },
+    {
+      "epoch": 4.97228144989339,
+      "grad_norm": 0.06241553173429086,
+      "learning_rate": 7.489143981960211e-09,
+      "loss": 0.3375,
+      "step": 4664
+    },
+    {
+      "epoch": 4.973347547974414,
+      "grad_norm": 0.06569632030334341,
+      "learning_rate": 6.924150900169402e-09,
+      "loss": 0.334,
+      "step": 4665
+    },
+    {
+      "epoch": 4.974413646055437,
+      "grad_norm": 0.06246704081644333,
+      "learning_rate": 6.381311903900234e-09,
+      "loss": 0.3328,
+      "step": 4666
+    },
+    {
+      "epoch": 4.975479744136461,
+      "grad_norm": 0.06349233758317888,
+      "learning_rate": 5.860627293849952e-09,
+      "loss": 0.3371,
+      "step": 4667
+    },
+    {
+      "epoch": 4.976545842217484,
+      "grad_norm": 0.061921416439339974,
+      "learning_rate": 5.36209735845894e-09,
+      "loss": 0.3386,
+      "step": 4668
+    },
+    {
+      "epoch": 4.977611940298507,
+      "grad_norm": 0.062197087906872645,
+      "learning_rate": 4.885722373879631e-09,
+      "loss": 0.3402,
+      "step": 4669
+    },
+    {
+      "epoch": 4.978678038379531,
+      "grad_norm": 0.06266534129824461,
+      "learning_rate": 4.431502604003157e-09,
+      "loss": 0.3249,
+      "step": 4670
+    },
+    {
+      "epoch": 4.979744136460554,
+      "grad_norm": 0.062785232042293,
+      "learning_rate": 3.999438300446024e-09,
+      "loss": 0.3328,
+      "step": 4671
+    },
+    {
+      "epoch": 4.980810234541578,
+      "grad_norm": 0.06336115371955366,
+      "learning_rate": 3.5895297025456686e-09,
+      "loss": 0.3298,
+      "step": 4672
+    },
+    {
+      "epoch": 4.981876332622601,
+      "grad_norm": 0.06207536666619536,
+      "learning_rate": 3.2017770373737877e-09,
+      "loss": 0.3376,
+      "step": 4673
+    },
+    {
+      "epoch": 4.982942430703625,
+      "grad_norm": 0.06040648920327252,
+      "learning_rate": 2.8361805197185678e-09,
+      "loss": 0.3313,
+      "step": 4674
+    },
+    {
+      "epoch": 4.984008528784648,
+      "grad_norm": 0.06070989404951965,
+      "learning_rate": 2.4927403521068926e-09,
+      "loss": 0.3359,
+      "step": 4675
+    },
+    {
+      "epoch": 4.985074626865671,
+      "grad_norm": 0.061482753937327835,
+      "learning_rate": 2.1714567247821393e-09,
+      "loss": 0.3363,
+      "step": 4676
+    },
+    {
+      "epoch": 4.986140724946695,
+      "grad_norm": 0.06142287227143837,
+      "learning_rate": 1.872329815726381e-09,
+      "loss": 0.3397,
+      "step": 4677
+    },
+    {
+      "epoch": 4.9872068230277184,
+      "grad_norm": 0.06141544242318959,
+      "learning_rate": 1.5953597906337437e-09,
+      "loss": 0.3335,
+      "step": 4678
+    },
+    {
+      "epoch": 4.9882729211087415,
+      "grad_norm": 0.06367883708761624,
+      "learning_rate": 1.3405468029370484e-09,
+      "loss": 0.3371,
+      "step": 4679
+    },
+    {
+      "epoch": 4.9893390191897655,
+      "grad_norm": 0.06116208249352169,
+      "learning_rate": 1.1078909937811688e-09,
+      "loss": 0.3295,
+      "step": 4680
+    },
+    {
+      "epoch": 4.990405117270789,
+      "grad_norm": 0.06315685975189177,
+      "learning_rate": 8.973924920541166e-10,
+      "loss": 0.3355,
+      "step": 4681
+    },
+    {
+      "epoch": 4.991471215351813,
+      "grad_norm": 0.06339091373459602,
+      "learning_rate": 7.090514143515137e-10,
+      "loss": 0.3304,
+      "step": 4682
+    },
+    {
+      "epoch": 4.992537313432836,
+      "grad_norm": 0.06281856911763233,
+      "learning_rate": 5.428678650165609e-10,
+      "loss": 0.3336,
+      "step": 4683
+    },
+    {
+      "epoch": 4.99360341151386,
+      "grad_norm": 0.062428937737431266,
+      "learning_rate": 3.988419360956286e-10,
+      "loss": 0.3368,
+      "step": 4684
+    },
+    {
+      "epoch": 4.994669509594883,
+      "grad_norm": 0.063210979241412,
+      "learning_rate": 2.769737073737844e-10,
+      "loss": 0.3338,
+      "step": 4685
+    },
+    {
+      "epoch": 4.995735607675906,
+      "grad_norm": 0.06278318354034783,
+      "learning_rate": 1.7726324636591075e-10,
+      "loss": 0.3374,
+      "step": 4686
+    },
+    {
+      "epoch": 4.99680170575693,
+      "grad_norm": 0.06485031870184303,
+      "learning_rate": 9.971060830338275e-11,
+      "loss": 0.3327,
+      "step": 4687
+    },
+    {
+      "epoch": 4.997867803837953,
+      "grad_norm": 0.062487731895542534,
+      "learning_rate": 4.431583613850876e-11,
+      "loss": 0.3312,
+      "step": 4688
+    },
+    {
+      "epoch": 4.998933901918977,
+      "grad_norm": 0.06242272501094243,
+      "learning_rate": 1.1078960571175856e-11,
+      "loss": 0.3316,
+      "step": 4689
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.08020564798786287,
+      "learning_rate": 0.0,
+      "loss": 0.3297,
+      "step": 4690
+    },
+    {
+      "epoch": 5.0,
+      "step": 4690,
+      "total_flos": 7.868010719084544e+16,
+      "train_loss": 0.37999461384724453,
+      "train_runtime": 78929.6999,
+      "train_samples_per_second": 30.397,
+      "train_steps_per_second": 0.059
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 4690,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 7.868010719084544e+16,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}